Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Crisis counselling online : building rapport with suicidal youth 2011

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


ubc_2011_spring_timm_maria.pdf [ 759.97kB ]
JSON: 1.0054473.json
JSON-LD: 1.0054473+ld.json
RDF/XML (Pretty): 1.0054473.xml
RDF/JSON: 1.0054473+rdf.json
Turtle: 1.0054473+rdf-turtle.txt
N-Triples: 1.0054473+rdf-ntriples.txt

Full Text

 CRISIS COUNSELLING ONLINE: BUILDING RAPPORT WITH SUICIDAL YOUTH  by   MARIA TIMM  B.A., The University of British Columbia, 2006    A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS OF THE DEGREE OF  MASTER OF ARTS  in  THE FACULTY OF GRADUATE STUDIES  (Counselling Psychology)    THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)     April 2011   © Maria Timm, 2011             ii              Abstract  Despite a documented increase in the use of online counselling services by youth, little research has been conducted on how counsellors establish rapport in the absence of voice tone and conventional spoken language.   As a result, no empirically validated guidelines exist for crisis counselling with youth online. Research shows that youth who access online services are often in extreme distress or suicidal. Youth‟s increasing affinity for online communication, coupled with a lack of research in this area, necessitate an examination of how rapport is built online. The current study was a qualitative exploration of client-counsellor interactions in online crisis counselling sessions with suicidal youth. Data sources consisted of transcripts obtained from an online crisis chat service for youth.  A collective case study was conducted, using content analysis of client-counsellor interactions followed by an examination of patterns across cases. It was found that tentative language, open-ended questions, and figurative language were used most in the Initial Contact phase and that interventions tended to be connection- building in nature. In the Risk Assessment phase, providing a context for questions, showing acceptance of coping methods, and statements of care were the most frequent interventions, and interventions tended to be connection- building in nature. In the Termination phase, summaries, questions about coping, expressions of care, and emoticons were used; connection-building and action- oriented types of interventions were both used. Overall, it was found that the counsellor tended to mirror the language patterns of the youth and that threats to rapport were handled with genuineness, often using informal language.   iii  Preface This research was conducted using information obtained in collaboration with the Vancouver Crisis Centre. Permission was granted by Vancouver Crisis Centre director Richard Kramer, MA, and Program Coordinator Jonathan Tanaka assisted with data collection. Ethics approval was granted from UBC Behavioural Research Ethics Board on June 15, 2010 (number H09-00974-A001); an amendment to the procedures was accepted on December 9, 2010.                   iv Table of Contents  Abstract ............................................................................................................................... ii Preface ................................................................................................................................ iii Table of contents ................................................................................................................ iv List of figures ..................................................................................................................... vi Acknowledgements ........................................................................................................... vii Dedication ........................................................................................................................ viii  CHAPTER 1: INTRODUCTION ....................................................................................... 1 1.1 Help-seeking trends ............................................................................................... 1 1.2 Advantages of online counselling ......................................................................... 2 1.3 Feedback from youth ............................................................................................. 4 1.4 Identified needs ..................................................................................................... 5 1.4.1 Importance of the therapeutic connection .......................................................... 6 1.4.2 Rapport or working alliance? ............................................................................. 7 1.4.3 Emotional connection and active engagement ................................................... 7 1.4.4 Real relationship ............................................................................................... 10 1.4.5 Trust .................................................................................................................. 11 1.5 Youth counselling: The role of rapport ............................................................... 11 1.6 Current research directions .................................................................................. 13 1.7 Language patterns ................................................................................................ 15 1.8 Use of emoticons ................................................................................................. 16 1.9 The current study ................................................................................................. 18 1.10 Tentative expectations: Rapport-building behaviours ....................................... 20  CHAPTER 2: METHOD .................................................................................................. 23 2.1 The collective case study ..................................................................................... 25 2.2 Content analysis ................................................................................................... 26 2.3 The data set .......................................................................................................... 27 2.3.1 Selection process .............................................................................................. 28 2.3.2  Inclusion criteria   ............................................................................................ 29 Chatter satisfaction ........................................................................................ 30 Defining rapport ............................................................................................ 31 2.3.3 Discarding of data ............................................................................................. 37 2.4 Data set for analysis ............................................................................................. 38 2.5 Delineation of phases .......................................................................................... 39 2.6 Content analysis ................................................................................................... 39 2.7 Analysis of themes .............................................................................................. 42 2.8 Dimensions of rapport ......................................................................................... 44 2.9 Trustworthiness ................................................................................................... 47 2.10 Ethical concerns ................................................................................................. 48  CHAPTER 3: FINDINGS................................................................................................. 51 3.1 Narrative descriptions .......................................................................................... 51 3.1.1 Case #1 ............................................................................................................. 51   v 3.1.2 Case #2 ............................................................................................................. 54 3.1.3 Case #3 ............................................................................................................. 59 3.1.4 Case #4 ............................................................................................................. 63 3.1.5 Case #5 ............................................................................................................. 69 3.1.6 Case #6 ............................................................................................................. 72 3.1.7 Case #7 ............................................................................................................. 75 3.1.8 Case #8 ............................................................................................................. 80 3.1.9 Case #9 ............................................................................................................. 85 3.2 Patterns and themes ............................................................................................. 88 3.2.1 Initial contact phase .......................................................................................... 89 3.2.2 Suicide assessment phase ................................................................................. 92 3.2.3 Termination phase ............................................................................................ 95  CHAPTER 4: DISCUSSION ............................................................................................ 98 4.1 Initial contact phase ............................................................................................. 98 4.1.1 Use of metaphor  .............................................................................................. 99 4.1.2 Reflection and clarification ............................................................................ 101 4.1.3 Dimensions of rapport .................................................................................... 102 4.2 Suicide assessment phase .................................................................................. 104 4.2.1 Respecting autonomy ..................................................................................... 105 4.2.2 Informal language ........................................................................................... 106 4.2.3 Real relationship ............................................................................................. 107 4.2.4 Dimensions of rapport .................................................................................... 107 4.3 Termination phase ............................................................................................. 110 4.3.1 Online linguistic conventions ......................................................................... 110 4.3.2 Dimensions of rapport .................................................................................... 112 4.4 Themes across phases ........................................................................................ 114 4.4.1 Mirroring language patterns ........................................................................... 114 4.4.2 Threats to rapport: the importance of the real relationship ............................ 116 4.5 Limitations and directions for future research ................................................... 119 4.5.1 Sampling procedures ...................................................................................... 119 4.5.2 Satisfaction scale ............................................................................................ 120 4.5.3 Methods .......................................................................................................... 121 4.5.4 Chat analysis ................................................................................................... 122 4.5.5 Definition of rapport ....................................................................................... 123 4.6 Directions for future research ............................................................................ 124 4.7 Implications for practice .................................................................................... 126 4.8 Conclusion ......................................................................................................... 128  REFERENCES ............................................................................................................... 130  APPENDICES ................................................................................................................ 137 Appendix A : Crisis centre satisfaction scale .......................................................... 137 Appendix B : YIBC disclaimer ............................................................................... 138 Appendix C : Volunteer consent form ..................................................................... 139 Appendix D: Coding of counsellor responses  ........................................................ 141   vi        List of Figures  Figure 2.1    Steps for analysis .................................................................................. 25 Figure 2.2    Selection process ................................................................................... 28                        vii  ACKNOWLEDGEMENTS  Thank you to my thesis supervisor Richard Young for his guidance and insights. Thank you to Richard Kramer, Director of the Vancouver Crisis Centre, for lending your support and your volunteers to this project.  Jonathan Tanaka, youthinbc coordinator for the Vancouver Crisis Centre, many thanks for your collaboration and your willingness to provide not only essential data but also intriguing conversations. I am grateful for the assistance and invaluable input of my research assistants, Nina Abizadeh and Merike Bruen.  Thank you also to Felipe Hirata and Dr. Johanna Timm, for lessons in formatting, courier service, and unwavering emotional support. Thank you to my sisters from Alabama for helping me live a balanced life and for commiserating when it was not so balanced. And to Sakae Alford for making me homemade ravioli while I was working on my formatting.  Lastly: Fox and Paris, thank you for your joyful presence and the comic relief while I was immersed in such a heavy subject.      viii  DEDICATION:   For the volunteers and staff of the Vancouver Crisis Centre and For  Florian          1 Chapter 1. Introduction  1.1 Help-seeking trends  Every 40 seconds, someone in the world takes their own life. Youth are particularly at risk: suicide is the second leading cause of death for Canadians between 10 and 24 (Canadian Mental Health Association, 2008). Similar statistics have been reported for the United States -  Bryan and Rudd (2006) have noted that suicide is the 11 th  leading cause of death for adults but the third leading cause of death for youth in America. Prior to taking action to end their lives, most people reach out to someone and inform them of their suicidal feelings. For example, 90% of suicidal youth give some form of warning sign prior to a suicide attempt (Granello, 2010b). Many people reach out to suicide prevention centres rather than family or friends (Mishara, 2007) this has found to be in part due to the stigma that surrounds the admission of suicidality (Evans, 2005). Suicidal ideation seldom comes “out of the blue;” it tends to be preceded by a crisis state that often involves intense depression or anxiety (Evans, 2005). If rapport is built with the distressed person before this crisis state escalates, the plunge into suicidal ideation is more likely to be averted (Evans, 2005). The Vancouver Crisis Centre, whose program this research examined, is a non- profit, community-based suicide prevention centre. It provides, in addition to a 24-hour crisis line and a school-based suicide prevention program, its newest service, “,” an online service that connects youth in crisis with trained volunteer counsellors in a one-to-one confidential chat.  This service was developed in 2004 in response to the observation that youth were not accessing the distress lines as frequently as they had in the past. Specifically, there was a 70% decline in calls to the distress line   2 from youth aged 13-24 in the period between 2001 and 2003 (Vancouver Crisis Centre, 2010). Based on a program evaluation including focus groups held with local youth, it was determined that an online modality would be preferable, as youth demonstrated decreased interest in using the telephone to obtain crisis support.  This preference of youth for online services is echoed by recent research in the field. A study examining the attitudes of youth towards telephone counselling found that, although over 98% of youth surveyed were aware of the existence of the local crisis hotline, only 1.7% had used a hotline in the past year and 2.1% reported ever using a crisis hotline. Researchers identified three factors that accounted for youths‟ non-use of these services: shame, self-reliance, and structure (referring to structural barriers such as lack of access to a private phone).  Alarmingly, researchers noted that “[…] objections to hotlines are strongest among individuals most in need of help” (Gould et al., 2006, p. 610).  Researchers also collected information regarding the use of alternate resources, and the Internet was cited by 18.2% of youth. Compared with the 1.7% who used the telephone option, this is significant. Gould and colleagues concluded their study by stating that “the Internet is one potential avenue for enhancing access to crisis services by youth […] it behooves hotline advocates to take advantage of  the Internet‟s growing accessibility and teenagers‟ propensity to use it as a means to obtain help” (Gould et al., p. 611). 1.2 Advantages of online counselling Research has found that online communication can play an important role in adolescent identity development (Beals, 2010) and peer interaction (Davis, 2010). It has also been found that the Internet has been used as a vehicle for working through grief (Williams &   3 Merten, 2009) and that it can strengthen peer relationships (Blas, Craig, Pepler & Connoly, 2008). Young people experiencing emotional distress are often hesitant to reach out for help, as they can feel embarrassment, shame, or that they are not “normal” (Martin, 2003). Increasing numbers of youth are using the Internet for counselling support when they are in emotional distress, possibly because they are drawn to the anonymity of this form of communication (Becker & Schmidt, 2006; Sumhramanyan, Smahel & Greenfield, 2006). Conducting crisis counselling online is a fairly new modality and has seen increasing use over the past few years (Whitlock, Powers & Eckenrode, 2006). Volume of chats on has also increased since its inception: the most recently reported statistics indicate a usage increase of 40% from 2006 to 2007 (Vancouver Crisis Centre, 2009).  This may reflect youth‟s increasing comfort with web-based media (Becker & Schmidt, 2006; Sumhramanyan, Smahel & Greenfield, 2006; Whitlock, Powers & Eckenrode, 2006) and their preference for this type of communication versus using the Crisis Centre‟s phone lines (Vancouver Crisis Centre, 2009). Online counselling has been recognized as having a number of distinct advantages over in-person counselling for both counsellor and client. For the client, anonymity is a key benefit. Adolescents particularly value anonymity and confidentiality in the counselling relationship (Geldard & Patton, 2007; Heath, Crow & Wiles, 2007; Martin, 2003).  In fact, it has been suggested that adolescents are more likely to be truthful in an online environment than in other types of exchanges (Whitlock, Powers, & Eckenrode, 2006). It has been hypothesized that the lack of identifying characteristics of the crisis counsellor may enable the youth to create the “ideal chat partner” (Becker & Schmidt,   4 2006, p. 238). For the counsellor, advantages include the ability to reach a more diverse population, access to full transcripts of sessions, low cost, and ease of standardization of services (King, Bambling, Reid & Thomas, 2006). Despite these positive factors and the documented increase in use of online counselling services, no empirically derived guidelines exist for conducting crisis counselling with youth online. The need for such guidelines is heightened with the knowledge that the presenting concerns of youth accessing these services are often of a serious nature: 22% of youth that log onto are suicidal, compared to only 10% of people who use the phone lines (Vancouver Crisis Centre, 2009).  1.3 Feedback from youth The Vancouver Crisis Centre‟s site has an option for youth to provide their input with respect to the services: there is a tab titled “Feedback;” clicking on this directs youth to a survey. This survey contains ten questions, two of which are qualitative:  “If you had to give a volunteer counsellor a couple tips on how to chat with youth in crisis, what would you say?” and “What was the hardest/most annoying thing about logging in, chatting and/or sign out?”  Based on a yearly summary of youth responses for 2009-2010 conducted by the Vancouver Crisis Centre, one of the most frequent complaints from youth regarding the counsellor‟s skills is a difficulty in building a therapeutic connection without sounding “like a robot;” youth frequently cited difficulties in building a connection with their counsellor online using only text. Some samples of youth responses reflecting this are: “Try not to be so automated, monotone. Just talk like a regular person with emotion,” “Don‟t repeat. Try not to talk like a robot,” and “Don‟t just be like „whatever‟ or those annoying chat robots on MSN ” (Vancouver   5 Crisis Centre, 2009).  These concerns have been echoed by providers and consumers of online youth crisis services in the UK (Tomlin, 2009) and Australia (King, Bambling, Lloyd, Gomurra & Smith, 2006), and take on increased importance when the chatter is displaying suicidal ideation.  The technological competence of today‟s youth and their increasing affinity for communication via the Internet necessitate a reexamination of our modes of delivering counselling services. 1.4 Identified needs The online crisis counsellor has only one tool with which to build rapport, and this is language.  Vancouver Crisis Centre volunteer counsellors are trained in traditional Rogerian client-centered techniques: reflection of feelings and content, as well as crisis intervention skills such as risk assessment and exploration of coping skills and resources. This is consistent with most other online crisis services (Mishara, 2007; Tomlin, 2009), and reflects the general understanding that individuals in crisis greatly benefit from empathic support and validation of their experience (Granello, 2010b; Hoff, 2009). However, as mentioned, both youth in crisis and crisis line volunteer counsellors have noted that simple reflection of feelings and content may appear “robot-like” when written out in the chat; without voice tone it is difficult to convey nuances and depth of emotion. The Vancouver Crisis Centre currently conducts on-going evaluations with regard to demographics, types of issues, chat volume based on time of day, and many other service delivery aspects. However, no evaluation has been conducted that looks specifically at the language and techniques used by counsellors in the actual crisis chat. Given the high usage of this service by suicidal youth and the importance of early intervention with regard to suicide, it seems advisable to focus on what is working in how   6 crisis counsellors interact with suicidal youth online. Specifically, the online nature of the service requires the counsellor to use language in a way that builds rapport in the absence of voice tone.  Gaps in the existing research on online crisis chats preclude the development of a training program supported by empirical evidence.  1.4.1 Importance of the therapeutic connection Numerous studies have shown that rapport-building is a vital component of the counselling relationship or “working alliance” and has been linked to positive therapeutic outcomes (Hackney & Cormier, 2009; Sandhu et al., 1993). Several studies have found that working alliance is the single most significant predictor of successful therapy, accounting for between 22% (Knaevelsrud & Maercker, 2007) and 30% (Fletcher- Tomenius, 2009) of variance in therapeutic outcome, regardless of therapeutic orientation. Groth-Marnat (1997), in his review of factors affecting therapy, concluded: “the overall quality of the therapeutic relationship accounts for at least as much of the outcome variance as specific techniques” (p. 596).  Working alliance has been found to be a robust predictor of therapeutic outcome regardless of whether this outcome was assessed by the client, therapist, or an independent observer (Hilsenroth & Cromer, 2007).  The important emotional connection between client and counsellor is often referred to as “rapport.” Researchers have frequently acknowledged the reciprocal nature of rapport: rapport is defined by Hackney and Cormier (2009) as “the psychological climate that emerges from interpersonal contact between you and the client” (p. 43). The nomological network of rapport is explored below.     7  1.4.2 Rapport or working alliance?  Many researchers consider rapport and working alliance to be interchangeable concepts; researchers frequently use measures of working alliance such as the Working Alliance Inventory (WAI) to assess rapport (Sharpley & Ridgway, 1992). Others (Auerbach et al., 2008) have found that rapport is most closely correlated with the bond dimension of working alliance as conceptualized by Bordin (1979).  Factor analysis has found high correlations between rapport and the bond subscale of the Working Alliance Inventory (WAI) developed by Horvath and Greenberg in 1989 (Efstation, Patton & Kardash, 1990).  In a study that aimed to assess the degree of rapport between client and counsellor, the WAI was used as a measure of rapport and researchers observed that the bond dimension of the WAI is “very similar to rapport” […] (Sharpley & Ridgway, 1992, p. 1).  In fact, one version of the Working Alliance Inventory (the AWAI-A, Schlosser & Gelso, 2005) has a subscale titled Rapport. Hence, for the purposes of this study, the terms rapport and working alliance were used interchangeably, with the understanding that working alliance is a broader construct which subsumes rapport. 1.4.3 Emotional connection and active engagement Working alliance, also referred to as therapeutic alliance, was initially conceptualized by Bordin (1979) as containing three discrete components: bond, (defined as the degree to which there is an atmosphere of mutual trust, acceptance and confidence in the therapeutic relationship), goals (the degree to which client and counsellor agree on and are engaged in the goals of treatment), and tasks (the degree to which counsellor and client reach agreement on how to reach treatment goals). Fletcher-Tomenius and Vossler (2009) summarized Bordin‟s original conception of working alliance as containing both   8 connection-oriented and collaborative components they defined it as: “the extent to which both [client and counsellor] work collaboratively and purposefully and connect emotionally (p.24).” Bachelor (1995) conducted a qualitative assessment of therapeutic interventions associated with the development of working alliance. Using a phenomenological approach, she found three distinct types of interventions associated with client-reported perceptions of working alliance: nurturant, collaborative, and insight- oriented. Nurturant-type interventions involved behaviours such as listening non- judgmentally, providing empathy, and being attentive. Collaborative behaviours included problem-solving, making suggestions, and reacting constructively to criticisms by the client. Insight-oriented interventions included a focus on patient distress and clarification and support of emotional disclosures, with the counsellor often providing a new way of looking at the current situation. Building rapport with the client who is suicidal is of especial importance considering the potentially life-threatening nature of suicidality. Despite the fact that most research shows rapport as being most closely linked the bond dimension of working alliance as per Bordin (1979) or the nurturant type of working alliance as per Bachelor (1995), when working with the suicidal client, most researchers consider rapport as containing an action-oriented dimension as well (e.g. Granello, 2010a). This is likely because of the nature of suicide assessment, which by necessity involves the use of directed, collaborative action on the part of client and counsellor. Neimeyer and Pfeiffer (1994) determined that there are four “factors” that help to build rapport when conducting a suicide assessment: reflection of negative feelings, elaboration of the complaint, exploration of suicidality, and involvement. These factors were used in the creation of   9 their instrument, the Suicide Intervention Response Inventory (SIRI), which has been subject to multiple revisions since its inception in 1989 (Neimeyer & Bonnelle, 1997). Thus, although there are differences in how rapport is conceptualized by different theorists, there appears to be an understanding in the literature that rapport is composed of both action-oriented and connection-building components and that both are important in building a therapeutic relationship with the client.  With respect to working with the suicidal client, it has been found that different types of rapport-building interventions are used depending on the “phase” of suicide assessment (i.e. the beginning, middle, or end of the session). Granello (2010a) created a 7-step model of suicide assessment in which he outlined 25 practical strategies for working with suicidal clients.  When initial contact is made with the client (after any immediate lethality has been assessed), he advised to focus on building the emotional connection with the client (Granello, 2010a). In the middle phase (in which the “formal assessment” occurs), Granello suggested that the counsellor use connection-building interventions such as “encouraging emotional ventilation,” and “acknowledging psyche- ache;” he urged the counsellor to “listen, understand, and validate” (p. 220). However, in this assessment phase he also advocated use of action-oriented interventions such as “engage social support,” and “establish a problem-solving framework” (p.220). In the final (termination) phase, Granello advised a focus on action-oriented interventions (such as creating a safety plan and connecting the client to resources). These recommendations are made for in-person counselling – as previously mentioned, no empirically validated guidelines exist for conducting online counselling for youth.    10 1.4.4   “Real relationship” A construct that seems to belong in the nomological network of rapport is what Gelso (2005) referred to as the “real relationship.” This has been defined as “the personal relationship that exists between therapist and client from the first moment of contact” (p.640). Gelso distinguished notion of the real relationship from the concept of working alliance on the grounds that the real relationship is considered to exist independently of the working alliance that is formed over time and is “more basic than the alliance and existing apart from the work of therapy” (Gelso, 2005, p. 640). The real relationship is said to be composed of two main features: genuineness and realism. Realism refers to the extent to which client and therapist experience each other “in ways that fit him or her rather than projections based on fears and wishes related to significant others from the past” (Gelso, 2009, p. 254) and genuineness is described as “the ability to be who one truly is, as opposed to being phony or inauthentic” (Gelso, 2009, p. 254).  Gelso et al. (2005) developed a measure of this construct termed the Real Relationship Inventory- Therapist form (RRI-T). The RRI-T has two subscales entitled Genuineness and Realism. The items on this scale appear to be related nomologically to Rogers‟ (1961) concept of therapeutic congruence, considered to be one of the three main components of working alliance. For example, one item from the Inventory reads, “I am able to respond genuinely to my client.”  Those involved with youth in a counselling capacity frequently note the importance for youth of someone who is not “fake;” who genuinely expresses his or her feelings as they arise (Martin, 2003).     11 1.4.5 Trust It has also been posited that the construct of trust is strongly related to the working alliance, and that development of trust is particularly important in online counselling. Trust emerged as a key factor in the development of working alliance when conducting online counselling in a study conducted by Fletcher-Tomenius and Vossler (2009). One commonly used measure of (in-person) counselling outcomes, the Counsellor Rating Effectiveness Scale (CERS, Strong, 1968, cited in Horvath, 1989), is designed to measure four types of interventions related to therapeutic outcomes; one of these is therapist trustworthiness. The CERS has been found to be highly correlated with the Working Alliance Inventory (WAI, Horvath, 1989), in a study that found that the establishment of trust was a key factor in the forming of a therapeutic alliance (Farsimadan, Draghi- Lorenz, & Ellis, 2007).  The development of trust is particularly important when counselling adolescents: Martin (2003) noted that the development of trust is a key step in the establishment of working alliance when counselling youth.  1.5 Youth counselling: The role of rapport Building a strong working alliance, and hence building rapport, is especially important yet challenging with adolescent clients (Martin, 2003), and is of crucial importance when the adolescent client is suicidal (Evans, 2005). In fact, it has been noted that “the therapeutic relationship plays a major role in determining a [suicidal] client‟s willingness to seek help, and clients indicate that a strong alliance with a helping professional has a significant impact on helping them through a serious emotional crisis” (Granello, 2010b, p. 367).  Suicidal adolescents are commonly viewed as a challenging population to work with. It often takes time to build rapport and to navigate around the various defenses and   12 resistance that youth may display. Romig and Gruenke (2001), who worked with inmates at a correctional facility, list several common resistance strategies displayed by clients. These include giving very little information, being silent or aimlessly chatting, and disrespecting the boundaries of the counselling relationship. When working with suicidal clients, it has been found that these particular clients may present as hostile, sarcastic, or withdrawn (Granello, 2010b). Thus, working with suicidal clients who are adolescents presents particular challenges to building rapport. There are several possible reasons for client resistance. Fear of losing control, self-protection, and doubt about confidentiality and the intentions of the counsellor can prevent the suicidal adolescent from fully participating in the counselling process (Romig & Gruenke, 2001; Martin, 2003).   Recognizing this, Paulson and Everall (2003) conducted a qualitative examination of counselling behaviours perceived as helpful among adolescents aged 13-18. It was found that two main type of interventions of positive factors emerged; these were termed “self-expression” and “therapeutic interaction.” Thus, it appears clear that the nature of the interaction between client and therapist is of vital importance when counselling adolescents, and this importance is increased when the adolescent is in suicidal crisis. Several key counsellor behaviours have been identified as particularly effective in establishing rapport. Many of these are non-verbal: the counsellor‟s posture, eye movements, gestures, and facial expressions have all been shown to contribute to client perceptions of counsellor empathy (Sandhu et al., 1993), which is considered an important component of working alliance (Horvath & Greenberg, 1989).  In a recent study, Bedi (2007) found that the counsellor‟s presentation   13 and body language and the counsellor‟s nonverbal gestures were two of eleven key factors considered by clients to contribute to working alliance. Online crisis counsellors must work without these important visual cues in their therapeutic arsenal; they do not even have the benefit of voice tone, which telephonic crisis counsellors have as a resource in the absence of visual cues. Shao-Kang (2008), who conducted a study on online language use, noted that “the goal of nonverbal cues is to convey emotion; the goal of verbal cues is to communicate ideas” (p. 595).  In counselling, conveying emotion is a crucial component of the therapeutic process. The research shows that difficulties conveying emotion online may have detrimental effects on the counselling process.  For example, Cook and Doyle (2002) found lower therapeutic alliance scores in online compared with face-to-face counselling (FTF) in some key aspects of the therapeutic alliance, although other components of the therapeutic alliance were found to be comparable in online and face-to-face counselling. Hanley (2009) conducted an assessment of the quality of the working alliance using a mixed-method approach by combining a quantitative self-report measure (the Therapeutic Alliance Quality scale) and qualitative interviews with users of an online counselling service in the UK. Although they noted that the establishment of rapport was more challenging in online counselling, they found a medium to high quality working alliance, which underscores the potential usefulness of this medium as a counselling tool.  1.6 Current research directions Recently, Williams, Bambling, King, and Abbott (2009) examined transcripts of online counselling sessions with youth using content analysis.  They sought to identify how counsellors communicate in online sessions and how various counsellor behaviours had a   14 positive effect on the session as operationalized by a list of positive therapeutic outcomes; for example, “encouragement by therapist empowered client” (Williams et al., 2009, p. 97).   It was found that counsellor behaviours that could be misinterpreted due to lack of voice tone had decreased effectiveness in the sessions. It was also noted that counsellors used more connection-building interventions than task-focused ones, and attributed this to a difficulty in communicating due to lack of voice-tone. This somewhat paradoxical finding highlights the importance of building a working alliance in online counselling. An Australian team of researchers (King, Bambling, Lloyd, Gomurra & Smith, 2006) conducted a qualitative assessment of youth‟s experience with the online crisis counselling service provided by the local Kids Help Phone centre. Using interview data, they noted that a major disadvantage cited by youth was that online counsellors were not able to sufficiently recognize the depth or subtleties of their expressed emotions. In addition, youth reported that building a connection with their counsellor online was impeded by the lack of voice tone and visual input, causing difficulty in gauging the counsellor‟s appraisal of their emotional state (King et al., 2006). In a follow-up study that aimed to expand on these findings, Bambling, King, Reid and Wegner (2008) examined the experience of counsellors on the Kids Help Line using a focus group methodology. Areas of concern from the counsellors‟ point of view included lack of non- verbal cues, potential misunderstanding of the severity of problems such as suicidal ideation, and clients‟ misunderstanding of the counsellors‟ empathic statements. Specific skills that counsellors found to be helpful in establishing rapport included use of emoticons (such as a “smiley face” or ), directly addressing potential   15 miscommunications, and use of icons and scales to help the client identify the intensity of their emotional experience. The researchers summarized their findings by stating that “methods [that] continue to enhance both the process and the effectiveness of online counselling are needed” (Bambling, King, Reid & Wegner, 2008, p. 115). Several ways of using language have been found be effective when working with adolescents and may lend themselves particularly well to online counselling.  An overview of this online- specific language is presented below, with suggestions for its applicability to online counselling.  1.7 Language patterns It has been found that “normal adolescent conversational behaviours” (Geldard & Patton, 2007, p. 28) were seen as more useful than traditional counselling responses in a review of a peer counselling program.  Researchers noted that the adolescent participants, when confronted with counselling techniques designed for adults, indicated discomfort, stating that because these were perceived as so different from normal adolescent verbal interaction, they implied a status difference between helper and helpee (Geldard & Patton, 2007). It was found that “conversational behaviours,” (Geldard & Patton, 2007, p. 33) which involved more informal language, were viewed as a positive part of the counselling relationship. Since research shows that adolescents are increasingly connecting with peers online, it seems reasonable to design a framework for counselling online that has some features of adolescent language use, to better facilitate rapport. An important aspect of the technique known as non-verbal communications (NVC), along with kinesics and proxemics (which are primarily visual aspects of communication), is the use of paralinguistics. These are the elements of human speech   16 that are non-verbal, such as “mmhmm.” Youth feedback from the youthinbc site indicates that these non-verbal attenders have decreased effectiveness in an online context; in the words of one youth: “Stop using mmhm in a chat! It doesn‟t come across the same way (empathically) as it does when you hear a voice. There are a lot of different ways to enunciate an “mmhm” and a person in crisis is not going to „hear‟ an empathic tone” (Vancouver Crisis Centre, 2009).   Research with adolescents has shown that they feel more comfortable engaging in a therapeutic encounter if they feel the person they are talking with is similar to them (Geldard & Patton, 2007), and mirroring language patterns, including paralinguistics, has been shown to help facilitate this; it can also help the counsellor to feel closer to the client, as research shows that mirroring or “matching” behaviour has a reciprocal effect (Bower, 2010; Gonzales, Hancock, & Pennebaker, 2010).  Adolescents engage in unique language patterns when online; most notably, this includes the use of emoticons and abbreviations.  1.8 Use of emoticons   Recently, research has focused specifically on how adolescents‟ web-based communication differs from “normal” speech or writing. Merchant (2001) describes the “emergence of new linguistic conventions” (Merchant, 2001. p. 6), and points out that adolescents, when engaging in online chats, tend to use emoticons (characters representing emotions that are formed by using punctuation such as brackets) to convey a sense of emotion and perhaps serve to simulate voice tone. The word “emoticon” is a merging of the words “emotion” and “icon” (Goldsborough, 2008). For example “:D” is an emoticon representing a “big smile.”  Merchant refers to these as “paralinguistic conventions” (Merchant, 2001,   17 p. 9) and suggests that they are a vital component of online communication.  Derks, Bos, and von Grumbkow (2008) found that the primary uses of emoticons are to convey emotion, reinforce the content of the message, and express humor. They also found that emoticons were used in a way that resembled facial expressions in normal, “live” conversations. They suggested that these non-verbal linguistic elements serve to supplement the verbal material in the message, lending it greater emotional intensity and “providing social cues beyond what is found in the verbal text of a message” (2008, p. 99). Shao-Kang (2008) found that when faced with pure text in the absence of emoticons, most internet users “cannot perceive the correct emotion, attitude, and attention intents” (2008, p. 597). It was further found that emoticons “allow receivers to correctly understand the level and direction of emotion, attitude, and attention expression” (Shao- Kang, 2008 p. 597). It was concluded that emoticons can function as “quasi- nonverbal cues” (p. 597). Martin (2003) found that adolescents communicate their emotions less clearly than do adults, partly because their capacity to think abstractly is still developing, and partly because they find it hard to open up to a counsellor who, as an adult, may be seen as an authority figure and therefore perceived as less likely to understand the adolescents‟ concerns. Martin gave the example of an adolescent saying to a therapist, “English sucks.” He explained that this phrase might have a number of different meanings, depending on the tone in which it is said. For example, it could mean the adolescent feels hopeless about school, angry at the teacher, or frustrated by the subject matter (Martin, 2003). Because   18 adolescents sometimes fail to elaborate on such statements, this can make therapy a frustrating experience. These difficulties are magnified when one considers the lack of voice tone in online chatting. Emoticons are thus a vital way in which the online counsellor can gauge what the adolescent is trying to say without asking too many questions, as adolescents tend to place great importance on whether the counsellor  “gets” them (Martin, 2003). It has been shown that adolescents initially have trouble connecting with a counsellor because the counsellor may, as an authority figure, evoke the same negative feelings as their parents do. Adolescents are often in the process of separating from their parents, and frequently have strong and conflicting feelings about this (Martin, 2003). Using emoticons can help to establish rapport by mimicking the types of exchanges youth have with their peers.  1.9 The current study Both current research and anecdotal observations seem to indicate a need for the crisis counsellor to use language that builds rapport in a manner that is not artificial and seeks to convey an accurate understanding of expressed emotions. As evidenced in the above summary of pertinent literature, preliminary work in the field has focused on gathering young people‟s and counsellors‟ perspectives on what is effective and what is hindering in online counselling with youth. However, most of this research has employed after-the-fact interviews, questionnaires, or focus-group designs. Although studies such as the one conducted by Williams et al. (2009) have examined how various counsellor responses can have a positive impact on the online counselling session, there is a   19 lack of research specifically examining how counsellor responses facilitate the development of rapport when working with suicidal youth online. In addition, little research has been conducted using transcripts of the actual crisis chats, which have the benefit of being untainted by the research context, and most research has focused on general counselling behaviours and outcomes across different sessions with multiple concerns. The main body of research on this topic focuses on the use of interventions designed for in-person counselling, such as use of empathy and open-ended questions. However, little work has been done that examines how online-specific language, such as use of emoticons, is used in online chats in a counselling context. It has been noted that current evaluations of suicide helplines suffer from methodological limitations: one common method of evaluating the success of suicide prevention services is through satisfaction surveys filled out by clients. One problem with this is response bias: inevitably, those clients who fill out the surveys are likely to differ from those who have not (for example, have strong positive or negative views of the service). Furthermore, as Mishara points out  “[…] because there are no empirical data linking satisfaction to improvements in clients […] satisfaction studies appear to have little value as indicators of the success of helplines” (Mishara, Chagnon, Daigle, Balan, Campbell et al.,2007, p. 310). The current study aimed to build on the research cited above in an effort to fill this gap. It did so by using transcripts from the Vancouver Crisis Centre to explore how specific counsellor responses, including use of emoticons and abbreviations, were used to build rapport in chats when the youth was suicidal. The study focused on chats containing solely suicidal ideation because suicidal ideation is both life-   20 threatening and more common in online chats than in phone calls to the Crisis Centre - as mentioned previously, 22% of youth that log on are suicidal (Vancouver Crisis Centre, 2009).  In addition, there is a scarcity of research on how suicidality is dealt with in online counselling with youth.  It was hoped that this research would contribute to a greater understanding of the counsellor behaviours associated with building an effective relationship with suicidal youth online. The following research questions were put forward: 1)  How is rapport built in the initial stages of the chat? 2) How is rapport maintained during and following disclosure of suicidality? 3) How does the counsellor manage chat termination while maintaining rapport? 4) If rapport is threatened, how does the counsellor re-establish or maintain rapport? 1.10 Tentative expectations: Rapport-building behaviours The exploratory, qualitative nature of this research and the lack of existing research on this topic precluded the proposition of firm predictions with respect to which behaviours would be found in specific phases of rapport-containing chats. However, based on the existing research on both youth counselling and online counselling which has been outlined above, tentative research expectations existed; it was noted that these had the potential to bias any interpretation of the results. These tentative expectations are outlined below. It is notoriously difficult to build rapport with adolescent clients. The impersonal nature of online communication adds to this difficulty. Since, unlike in face-to-face and telephonic interactions, client and counsellor did not have voice tone available with which to build rapport, it was posited that online-specific linguistic conventions such as   21 emoticons and abbreviations might be used to facilitate rapport in the chats. Existing research and theory pertaining to both online counselling and counselling adolescents as cited above indicated that this would be an effective strategy. In addition, based on previous research on online services such as YIBC (Bambling, King, Reid & Wegner; Williams, Bambling, King and Abbott, 2009), it was posited that use of empathy, clarifying questions, and scaling questions would assist in building rapport, although, since no research yet exists on chats in which the youth was suicidal, these ideas were mainly speculative as the research was exploratory. Researchers‟ tentative expectations with respect to use of rapport-building interventions in the different phases of the chat are discussed below. It was considered plausible that the initial phase of the chat would be mostly concerned with building rapport. This expectation was supported by the existing literature on suicide intervention: it has been frequently noted that a hierarchical approach is beneficial when conducting a risk assessment.  For example, clinicians working with suicidal clients recommend beginning with more innocuous questions (for example, asking about current functioning) and then moving towards more direct risk assessment questions; this is also seen as a way to reduce resistance (Bryan & Rudd, 2006). Considering that the focus in the beginning of the counselling relationship is most often the time where trust is established and an interpersonal connection is initially formed (Bedi, 2007) and considering the research on suicide assessment that points to the importance of early connection-building (Granello, 2010b), it was tentatively expected that interventions in this beginning phase would be more focused building an emotional connection than on the other, more action-oriented components of rapport. Since online-   22 specific language such as emoticons and abbreviations (for example “lol” for “laughing out loud” or “pos” for “parents over shoulder”) seem to be ways in which emotion is conveyed online (Shao-Kang, 2009), it was thought that these would be most present in this phase of the chat.  It was also expected that boundary-setting and clarification of the chatter‟s expectations of the service would likely take place in this phase. The second phase of the chat was proposed to involve more action-oriented types of interventions considering the nature of risk assessment, which usually involves a focus on collaborative behaviours  - for example, encouraging the youth to put the suicidal means away (Granello, 2010a). Given the serious nature of suicidal ideation, it was considered likely that online-specific language would be less present in this phase of the chat, as it may be perceived as flippant. Granello (2010a) has noted that the conclusion of an (in-person) suicide assessment is usually concerned with more behaviourally-oriented interventions, such as facilitating a referral to other resources or creating a safety plan or no-harm agreement (Granello, 2010a). It was considered possible that this trend would also be observed in online counselling. However, as noted, all of the above speculations were tentatively held considering the lack of research on the topic of online suicide counselling with youth.         23  Chapter 2. Method As stated previously, the research questions this study hoped to address were as follows: 1) How is rapport built in the initial stages of the chat? 2) How is rapport maintained during and following disclosure of suicidality? 3) How does the counsellor manage chat termination while maintaining rapport? 4) If rapport is threatened, how does the counsellor re-establish or maintain rapport? In this chapter, I outline the methods used to answer the research questions proposed below and explain how they relate to the nature of the research questions. Following this, I describe the selection process and inclusion criteria as well as make note of ethical concerns that arose and how they were addressed by the study design. It was decided that the exploratory, open-ended nature of the research questions fit with a qualitative mode of inquiry. In addition, the length, depth, and complexity of the chats received warranted a more in-depth exploration than could be achieved by quantitative means. Therefore, in order to answer the above research questions, a qualitative, exploratory content analysis was performed on the data set.  Individual chats were treated as “cases” and were examined as such, following the suggestions for case study research described by Stake (1995).  As stated earlier, researchers have called for “methods [that] continue to enhance both the process [italics mine] and the effectiveness of online counselling […]” (Bambling, King, Reid & Wegner, 2008, p. 115).  Qualitative research tends to be more oriented towards process than content, although it may employ methods that focus on content, such as content analysis. Curtis (2004) conducted a content analysis of student-to-student interactions in an online learning forum. Curtis distinguished between manifest (what the chatter actually says) and latent content (goals   24 implicit in the chatter‟s response). The analysis of data in the present study focused first on the types of interventions used by the counsellor (by using content analysis), then on how the crisis counsellor met the goal of establishing rapport (by examining patterns across cases). Stake (1995) noted that combining different approaches to studying the same phenomenon increases the validity of a qualitative inquiry.  Combining content analysis with the thematic exploration characteristic of case studies was thought to facilitate the development of a richer picture of what occurred in the rapport-containing chats. By identifying relevant themes across cases, it was hoped that the counsellor behaviours found in rapport-containing chats could be identified and tentative recommendations for service or for future research be put forward.  This research was conducted using an inductive rather than deductive approach; starting from more detailed observations, then examining patterns, and lastly putting forward tentative hypotheses about the observations. Therefore, to summarize, chats were examined first on a micro- level (coding responses), then on a macro-level (examination of themes). The aim of both these approaches was two-fold: first, to identify if it was possible to observe the components of rapport commonly proposed to exist in in-person counselling in an online context. Secondly, the study aimed to explore how the crisis counselor was able to build rapport in an online environment. The ways in which the cases were examined is described in Figure 2.1, followed by a more detailed description of this process.    25  Figure 2.1 Steps of Analysis 2.1 The collective case study Data were examined using the collective case study approach identified by Stake (1995). The collective case study approach is one in which multiple cases are examined for the existence of themes or patterns.  Stake (1995) described two possible ways in which researchers make sense of case studies: “through direct interpretation of the individual instance and through aggregation of instances until something can be said about them as a class (p. 74).” This study used the latter approach, following the assertion by Stake that although it may be possible to find something of relevance in a single case, it is more often the case that findings of importance make themselves known as repeated patterns (1995). Each chat between volunteer counsellor and youth was referred to as a “case.” The case study approach was considered particularly appropriate for both the type of data used in this study (chat transcripts, each forming a “case”) and the exploratory, process- oriented nature of the research questions. Content Analysis (Crisis Centre categories  dimensions of rapport) Narrative Descriptions Themes and Patterns   26  The suggestions given by Stake (1995) for analyzing case studies were used by the researchers in the examination of the transcripts after content analysis had been conducted. Stake noted that “All researchers have great privilege and obligation: the privilege to pay attention to what they consider worthy of attention and the obligation to make conclusions drawn from those choices meaningful to colleagues and clients (p. 49).” With respect to how the data were transformed into case studies, this phase of the analysis involved describing what was observed as occurring in the chat by asking guiding questions (this procedure is described below), using as much detail in the description as possible to facilitate a rich picture for the reader while balancing this with a need for confidentiality. Creswell (2009) observed that one of the ways to add rigour to qualitative research is to “provide detailed descriptions of the setting […]” since this can help the ensuing results “become more realistic and richer” (Creswell, 2009, p.192). The cases were first described in narrative fashion; following this, patterns were identified using research assistants to improve reliability; lastly, interpretations (referred to as “assertions” as per Stake) were made.  2.2 Content analysis Content analysis has been defined succinctly as “a method of analyzing written, verbal, or visual communication messages” (Elo & Kyngas, 2007, p. 107). It was first used to analyze articles in magazines and newspapers in the 1900s.  It involves classifying certain words or word combinations into categories with the assumption that they share the same meaning. Several criticisms of content analysis have been cited, including, for example, that it is too simplistic of a method or that it is not truly qualitative or quantitative in nature (Elo & Kyngas, 2007). However, many other benefits of this method have been   27 enumerated. These include the flexibility of content analysis, its ability to analyze both manifest and latent content, and the fact that it does not impact the participants being studied as it is employed after the fact, thus eliminating a large potential source of bias. However, of course, it is not immune to bias: researcher bias is a threat inherent in any form of data analysis involving human interpretation.   Williams et al. (2009) have used content analysis to investigate the language patterns of youth and counsellor responses in an online counselling session; therefore, content analysis seemed congruent with the aims of the current research. In addition, the flexibility of this method allows for open-ended inquiry, which suited the exploratory nature of this study. Stake (1995) noted that a qualitative approach is recommended when undertaking an initial review of a given program, so that cases can be thoroughly examined for the presence of patterns and themes; initial findings from this type of analysis can assist the researcher with program development. 2.3 The data set In order to clarify the procedures used, a table is presented below to illustrate the steps taken when selecting the data for analysis; following this, the steps are described in more detail. Below is the selection process:     28   Figure 2.2. Selection Process 2.3.1 Selection process  Fifty chats with identifying information removed were provided by Jonathan Tanaka, training coordinator of the program at the Vancouver Crisis Centre. Transcripts were selected by Mr. Tanaka using the inclusion criteria with respect to chat demographics described below, after obtaining volunteer consent. Due to concerns about volunteer counsellor anonymity, the Vancouver Crisis Centre had requested that data be collected in this manner, as allowing me to select the cases would put me in contact with volunteer counsellor aliases. Since every staff member is provided a list of aliases, this would mean that I would potentially be able to match a given alias with the corresponding volunteer counsellor name. 50 chats meeting initial inclusion criteria for demographic information 16 eliminated as did not contain rapport or chatter did not present with suicidality 17 contained rapport as per operational definiton 8 were eliminated due to difficulty obtaining volunteer consent and other issues 9 chats remained for analysis   29  2.3.2 Inclusion criteria The inclusion criteria for this study were as follows: only chats with Contemplating Suicide as a problem code (chats are identified in this way if there is a suicide assessment undertaken by the volunteer counsellor) that were rated 5 on the Crisis Centre Satisfaction Scale (see Appendix A and description below) were included in the analysis. On December 24, 2009, the Vancouver Crisis Centre placed a disclaimer on the website indicating to youth that their (non-identifying) information may be used for research purposes (see Appendix B). Only chats received after this date were included in the analysis. The exclusion criteria were as follows: chats which did not contain Contemplating Suicide as a problem code, were rated 1, 2, 3, or 4,  on the Satisfaction Scale, and which had been received prior to December 24, 2009 were excluded from the analysis. As well, chats received from individuals younger than 14 or older than 18 were excluded since the Vancouver Crisis Centre‟s services were developed for high school students, and youth within this age range are the most frequent users of the service (Vancouver Crisis Centre, 2009).  Demographic information was collected as well with respect to the age, gender, location, and presenting problem (in addition to suicide); volunteer counsellor demographics included gender and time spent on the lines (plus or minus 100 hours). With respect to chat duration, initially, a timeframe of 40-60 minutes was proposed. Since the average chat received by the Crisis Centre is 50 minutes long (Vancouver Crisis Centre, 2010), this was considered a timeframe that represented an average chat and was thought to be long enough to capture the start and resolution of a suicidal crisis. An additional criterion for inclusion was the presence of chatter-perceived   30 rapport, the definition of which is outlined below, following a description of the Satisfaction Scale. Chatter satisfaction  Hackney and Cormier (2009) noted that one of the most reliable predictors of a strong counselling alliance – which includes rapport - is client-expressed satisfaction with the relationship. Thus, chatter satisfaction was used as a criterion of inclusion in the present evaluation.  “Chatter satisfaction” was quantified in the following manner: All sessions with clients are documented by volunteer counsellors as per Crisis Centre protocol. At the conclusion of each chat, crisis counsellors are required to assess the satisfaction of the client on a 5-point scale, with 5 indicating the highest level of satisfaction (see Appendix A). This scale has been developed by the Vancouver Crisis Centre based on the seminal work of Hoff (1975, 2009) and has been subject to multiple revisions based on staff and volunteer counsellor input and pilot studies. It fulfills the quality control standards set by the American Association of Suicidology, the accrediting body of the Vancouver Crisis Centre (Vancouver Crisis Centre, 2010). Use of a Likert-type scale such as the Satisfaction Scale was considered appropriate in light of the nature of the variable being studied (satisfaction, which can be reasonably quantified in interval terms), and research shows that most people are able to place themselves on a continuum with regard to this variable (Palys & Atchison, 2008). This type of scale is frequently used in the area of program development (Gould, Kalafat, Harris & Kleinman, 2007). To avoid ambiguity and increase objectivity, short descriptions have been developed by the Vancouver Crisis Centre to describe each level of the scale (see Appendix A).  Only chats rated 5 on this scale were included in the analysis. In order to increase inter-rater reliability and avoid   31 researcher bias (Hodges & Videto, 2003), two trained research assistants who were uninformed regarding the hypotheses of the study assessed satisfaction ratings independently, based on the chat transcript, as it was thought that some volunteer counsellors might over or underreport satisfaction based on individual volunteer counsellor characteristics (for example, newer volunteer counsellors may  have been more concerned with being perceived as effective and therefore may have assigned themselves higher satisfaction ratings).  However, the possibility of this rating inflation is lower than when rating crisis phone-calls, since volunteer counsellors on YIBC are aware that staff have access to full transcripts of their sessions; hence, they would be less likely to rate their chats in a way that was inconsistent with actual content of the chat.  When there was disagreement between satisfaction ratings, as was the case in two of the chats, this disagreement was discussed and the statements re-rated by all raters. If agreement was not reached, then the chat in question was discarded. One chat was discarded as a result of this process. Defining rapport Chats meeting the inclusion criteria defined above were further examined for the presence of chatter-perceived rapport, which would point to the existence of a positive working alliance (Auerbach, May, Stevens, & Kiesler, 2008; Sharpley & Ridgway, 1992). However, before chats containing rapport could be identified, a definition of rapport that was amenable to the aims of the study needed to be developed.  Researchers have found that rapport, while clearly affecting positive changes in the counselling relationship, has been an elusive concept to define. The existing definitions of rapport cited above were discarded for use in this study because they lacked specificity, which   32 made it difficult to apply them to behaviours observed in the chat transcripts. In addition, it was noted that most definitions of rapport spoke to only one or two dimensions of the construct; as noted, the nomological network of rapport encompasses a wide variety of different but related constructs.  Bachelor (1995) noted that “the accumulated body of data in the area also attests to the diversity of definitional and measurement approaches to the construct” (p. 323). Existing research on both rapport and working alliance was examined, as were the  nomological networks of these constructs (as noted in the preceding chapter), with the goal of developing a definition of rapport that was both parsimonious and amenable to online counselling. Externally validated measures of working alliance for in-person counselling were initially considered as operationalizations of rapport for this study. For example, the Observer form of the Working Alliance Inventory (WAI; Horvath, 1989) and the Session Evaluation Questionnaire (SEQ; Stiles, Gordon & Lani, 2002) were evaluated with regard to their applicability for this study.  Both these measures have observer forms which would have permitted the researcher to evaluate the session without requiring input from the crisis volunteer counsellor or the youth. However, it was noted that the these instruments were designed with the implication that they are measuring not one isolated session as is the case in crisis counselling, but are assessing longer-term development of rapport in the counselling process. For example, one item from the WAI states: “There is a mutual perception that the goals of the sessions are important for the client” (Horvath & Greenberg, 1989).  Questions such as this were not considered appropriate for chats on, whose short-term nature essentially provides the reviewer with a “snapshot” of the counselling relationship; the fact that the item in question refers to   33 “sessions” in the plural renders it inappropriate for use when analyzing a single crisis chat.  In addition, measures such as the WAI and SEQ have not been evaluated with regard to their applicability for online counselling or for counselling with youth. Hence, an investigation was conducted into the nomological networks of the construct commonly known as “rapport,” with the primary objective being to determine an operational definition of this construct that was appropriate for both online counselling and relevant for counselling with youth (for example, as noted, the construct of trust appears to be particularly relevant when attempting to build a relationship with youth). Considering the nomological network of working rapport, it is reasonable to assume that when assessing the convergent validity of a given measure of this construct, moderate correlations would exist between measures of rapport and measures of trust, real relationship, and working alliance, and this seems to be the case based on the research cited above. In addition, these components of rapport appear to be especially relevant for counselling adolescents, as previously noted. Based on the existing research and nomological networks of rapport, the following definition of rapport was developed:  “The existence of a genuine emotional connection between client and counselor in which a climate of mutual trust, collaboration, and understanding is created, as operationalized by verbal statements or actions made by the client that imply the existence of such a connection.”  Two trained research assistants, who both had experience on other crisis lines (CHIMO Crisis line and Women Against Violence Against Women crisis line located in Richmond and Vancouver, BC, respectively) but who were unaffiliated with the   34 Vancouver Crisis Centre and who had never participated in online counselling, were provided with the preceding definition of rapport and with 50 chats matching the inclusion criteria.  They were asked to highlight portions of youth responses that appeared to be congruent with this definition. Youth responses were then categorized based on content. Youth-perceived rapport was examined rather than counsellor- perceived rapport since this was in keeping with the goals of the study (i.e., it was considered more important that youth perceived a connection than counsellor, since the service is tailored towards youth). In addition, most of the research on rapport indicates that if one party perceives rapport to exist, the other does as well, most likely due to the reciprocal nature of rapport and rapport-building behaviours (Hackney & Cormier, 2009). Meetings were held with the research assistants to define which chatter responses were indicative of rapport. These were then grouped into themes based on common characteristics. It was found that four clear, discrete categories of youth responses indicating rapport were consistently highlighted by the primary researchers and the research assistants. These were: 1) Expresses clear agreement with volunteer counsellor‟s statements 2) Collaborates with risk assessment and safety planning 3) Openly discloses suicidality/sensitive information 4) Expresses liking of/connection with the volunteer counsellor After discussions between the research assistants and me, amendments to these categories were made. The first category, Expresses Clear Agreement with Volunteer Counsellor Statements, was highlighted consistently because it implied that the chatter felt understood, which is considered an important component of rapport (Hackney &   35 Cormier, 2009). However, it was noted that the youth simply saying “yeah” in response to volunteer counsellor statements could not necessarily be considered a sign of rapport, since this could be viewed as passively contributing to the conversation and is not even necessarily indicative of agreement, since this response could be sarcastic in nature. However, instances where the youth clearly or emphatically expressed agreement (such as “exactly!” or “totally!”) were included. The context of this response was examined as well to rule out possible sarcasm. It was noted that in cases where the youth felt understood, emotional disclosure usually followed, which indicated that the volunteer counsellor‟s initial understanding of the youth‟s concerns possibly allowed a climate of trust to flourish which created a safe space for future disclosures. Therefore, use of statements in this category appeared to contribute to rapport by facilitating both trust and understanding, which are considered in the literature to be strongly associated with rapport (Fletcher-Tomenius & Vossler, 2009). The second category, Collaborates with Risk Assessment, was more difficult to examine in terms of how it contributed to rapport. Initially, a decrease in suicide risk as defined by compliance with the volunteer counsellor‟s risk assessment requests (e.g. “yes I will place the pills in another room, yes I will stop cutting,”) was considered indicative of rapport; the category was initially termed “Compliance with Risk Assessment.” However, after discussions between myself and the two RA‟s, it was decided that overt compliance was not necessarily reflective of rapport since first of all, there was no way to tell whether the youth was actually complying and secondly, honest non-compliance with risk assessment (e.g. “I don‟t know if I can put the pills in another room;” “I don‟t know if I can stop cutting”) was considered more indicative of rapport, since it implied that   36 trust had been established to such a degree that the youth was willing to honestly disclose information that may be negatively received. However, willingness to engage with the volunteer counsellor with regard to the risk assessment process (for example, by following volunteer counsellor suggestions or by asking the volunteer counsellor for advice) was considered indicative of rapport as it seemed to contribute to the collaborative aspect of rapport. Hence, the word Compliance was replaced by Collaboration in this category. An example of collaboration with risk assessment would be a spontaneously generated statement such as “I will call you guys if things get worse,” or “I will try to keep myself safe.” A related aspect of this category was youth receptivity to the provision of additional resources, such as suicide counselling centres (e.g. “I will call that number you gave me).” The third category, Willingness to Disclose, was considered to contribute to rapport by indicating that trust has been established. Examples of this category included disclosure of suicidal plans, previous attempts, cutting, and other sensitive information (disclosure of suicidality by itself was not considered indicative of rapport since as per the inclusion criteria, all chats contained suicidality). However, the timing of a given disclosure also determined whether it was considered evidence of rapport. For example, if a youth logged on and immediately disclosed sensitive information (e.g. “I am cutting right now”), this disclosure could not be considered as a result of pre-existing rapport. Thus, disclosures made prior to volunteer counsellor involvement were not considered as evidence of rapport. Clarifying statements made by the youth (e.g. “No, I mean xxxx”) were also included in this category, as it indicated that the youth cared enough to make sure the volunteer counsellor understood what they were trying to convey.   37 The fourth category, Expresses Liking of/Connection with the volunteer counsellor, appeared to indicate rapport by implying the genuine emotional connection or “real relationship” as per Gelso et al. (2005) that has been considered a vital part of rapport and seems particularly important when building a therapeutic connection with youth. Examples of this category included asking “Can I chat with you again?”, statements such as “You really understand me,” as well as spontaneous provision of identifying information such as an email address (which also signifies trust), which occurred in two instances. Only chats which contained at least one instance of each type of youth statement were included in the analysis. 2.3.3 Discarding of data Of the 50 chats initially obtained, 17 of these contained rapport as per the operational definition. Of these, eight chats were eliminated either because the volunteer in question had left the Centre (three instances) or refused to sign the consent form (one instance). One chat was also eliminated because, although it contained evidence of rapport, the volunteer in question did not follow the Crisis Centre model, and after a discussion with the program coordinator, it was decided to eliminate this chat from the analysis. Two chats were eliminated because they ended abruptly, either because of a problem with the internet connection or because the chatter suddenly had to terminate the chat. One chat was discarded in the analysis phase as it was found that the chatter, who had entered “16” in the age box, later in the chat disclosed her actual age as 23. Because this age was above the inclusion criteria age range of 15-19 (high school age), it was eliminated from the sample.    38 2.4 Data set for analysis  Nine chats obtained from the Vancouver Crisis Centre‟s online database formed the data used in this study. The ages of chatters ranged from 15 to 18 with a mean age of 16.88. In terms of gender, the chats consisted of three male and six female chatters. The volunteer counsellors consisted of three male and six female volunteer counsellors. With respect to chatter/volunteer counsellor gender, three of the chats consisted of a same- gender dyad (female chatter and female volunteer counsellor). Three chats consisted of a female chatter and a male volunteer counsellor, and three chats consisted of a male chatter with a female volunteer counsellor. The volunteer counsellors‟ level of experience was as follows: three of the chats were taken by volunteer counsellors with less than 100 hours of experience on the lines; six were taken by volunteer counsellors with 100 hours or more of experience. The youth gave their location as a Canadian city (seven instances) an American city (one instance) and a town in the United Kingdom (one instance). The presenting concerns of the chatters (in addition to suicide) were as follows: Suicidal friend (three instances), eating disorder (two instances); depression (two instances); childhood abuse (one instance); and self-harm (one instance). Chats selected ranged between 35 and 87 minutes, with a mean time of 50.22 minutes.  As mentioned, initially, a range of 40-60 minutes was considered for the inclusion criteria. However, upon closer examination of the chats, it was found that measuring the duration of the chat in minutes did not accurately reflect the actual amount of interaction in a given chat. For example, in some chats there were long pauses between responses because the chatter either left the room or was otherwise occupied while chatting, while in other “shorter” chats, more   39 actual interaction was observed. Therefore, chats lasting between 35 and 87 minutes formed this data set. 2.5 Delineation of phases Each chat was separated into three phases: Initial Contact, Suicide Assessment, and Termination. These are phases recognized by the Vancouver Crisis Centre as occurring in the course of a suicide-containing call or chat; they are also similar to categories found in research on the topic of suicide risk assessment (Bryan & Rudd, 2006; Granello, 2010a; Gould et al., 2007; Halderman, Eyman, Kerner, & Schlacks, 2009).  Based on the existing research on suicide assessment, it was expected that each “phase” would likely involve different goals and needs for both the volunteer counsellor and the chatter, and thus different phases were expected to involve different types of interventions. Using research assistants to ensure inter-rater reliability, these three phases were delineated and operationally defined.  The phases and what they entail were delineated by the research assistants and me following the examination of the 50 chats initially provided. The Initial Contact phase was determined to start from the volunteer counsellor greeting the chatter and to end with the disclosure of suicidality. The Suicide Assessment phase was determined to begin with the volunteer counsellor‟s asking of the question “are you suicidal?” or its equivalent and to end with the chatter or volunteer counsellor indicating that the chat will end soon (e.g., by referring to activities to be done after the chat); the Termination phase was determined to begin at this point. 2.6 Content analysis In the first phase of the study, a detailed content analysis was conducted on the existing client-counsellor interactions observed in transcripts obtained from the Vancouver Crisis   40 Centre‟s database. Crisis counsellor responses were coded based on categories developed by the Vancouver Crisis Centre as part of its training materials (see Appendix D). These categories were developed for training purposes by the Vancouver Crisis Centre and represented the range of responses volunteer counsellors were trained to provide on the phone lines; hence, they were considered most likely to capture the different types of responses given by volunteer counsellors. Also, given that this research was in part a program evaluation regarding the in-session behaviours of counsellors working at the Vancouver Crisis Centre, using the Crisis Centre codes of volunteer counsellor responses seemed appropriate.  Creswell (2009) stated that when coding in qualitative research, it is advisable to keep the several “categories” of codes in mind when engaging in the process of coding. The first “category” cited refers to codes reflecting topics that would be anticipated based on the literature reviewed and on “common sense” (Creswell, 2009, p. 187).  The Crisis Centre codes would appear to fall under this category, as common sense dictates that volunteers trained by the Vancouver Crisis Centre would engage in behaviours that fit these parameters. The second category of codes cited by Creswell refers to those that are “surprising and not anticipated at the beginning of the study” (p.187); lastly, Creswell cited codes that “address a larger theoretical perspective in research” (p. 187). These three categories were kept in mind by researchers both in the initial coding phase and in the second phase of the analysis, in which themes and patterns were examined. Codes were developed in the following manner: a guiding question was asked (“How is the counsellor attempting to communicate with the youth?”) and categories were generated based on the answer to this question. For example, the crisis counsellor‟s statement “Sounds like you‟re feeling   41 really sad” was coded as  “counsellor reflects client‟s sadness” by one research assistant, as “reflecting feelings” by myself, and as “shows empathy” by the second research assistant. It was decided to code this response as “reflection of emotion” as per the Crisis Centre category of that name, since “empathy” was determined to be an ambiguous code that could encompass a variety of behaviours.     Creswell (2009) noted that in order to increase the trustworthiness of the results when coding, the researcher should pay close attention to a possible “drift” in codes as the research progresses; namely, a shift in the meaning of various codes. In the current study, this was accomplished by following Creswell‟s suggestion of repeatedly comparing data with the codes, as well as by using written documentation and continued checking of codes with the research assistants.   Given the online nature of youthinbc and the fact that the crisis centre responses were designed for phone counselling, additional categories emerged and these were agreed upon by multiple raters prior to being included in the coding scheme. These included use of emoticons, use of figurative language (including metaphor, simile and personification), and informal language (see Appendix D). Multiple raters, consisting of trained research assistants, were used to ensure reliability of the categories.  According to Stake (1995), this investigator triangulation is recommended to increase the validity of the information examined.  If agreement regarding a particular code was not reached, then raters engaged in conversation with the aim of reaching consensus. In some cases, consensus regarding the applicability of crisis centre categories for a given code was not reached. In this case, new codes were generated, again using multiple raters. Following   42 data analysis, the coded chats were sent to my research supervisor for verification and, if needed, revision.  A sample application of these categories is demonstrated below: Crisis Counsellor 6:25:06 PM it sounds like having not heard for your boyfriend for so long has got you feeling really terrible and has triggered you to want to injure yourself and has also triggered some feelings of suicide  (reflection of content + emotion) Youth 6:25:17 PM yes Youth 6:26:16 PM idk what to do. except die Crisis Counsellor 6:26:49 PM It looks like you are going through a lot of pain, unbearable pain and you are feeling like suicide is the only option right now, (reflection of content + emotion) Crisis Counsellor 6:27:12 PM I am worried about you (self-disclosure/immediacy), I'm just wondering, are you attempting suicide right now? (close-ended question to explore safety) Youth 6:27:37 PM no not yet. Crisis Counsellor 6:28:29 PM so, you are feeling like you might attempt suicide tonight? (close-ended question to explore safety) Youth 6:28:48 PM or in the morning when no one is here. i dont want to be saved Crisis Counsellor 6:30:03 PM I am really glad you logged on tonight, (expression of encouragement) i can tell you are going through some really heavy stuff.(empathy) It sounds like you are planning on killing yourself either tonight or in the morning - have you thought about how you might do it?  (Close-ended question to explore safety)  (Obtained from Vancouver Crisis Centre YIBC database, January 17, 2010).  2.7 Analysis of themes  Stake (1995) identified that “most case study reports present both coded data and direct interpretation […] (p. 29)” since this is considered to provide the researcher with a richer data set than either would produce alone.  The second phase of the analysis consisted of a more in-depth thematic exploration of the chats. It was built on the information obtained in the primary analysis and aimed to use the data obtained to construct a thematic   43 “picture” of what occurred in the separate stages of the chat.  As mentioned, each chat was separated into three discrete stages: Initial Contact, Suicide Assessment, and Termination. Again, multiple raters (myself and trained research assistants unaffiliated with the Vancouver Crisis Centre) were utilized and inter-rater agreement sought regarding the definition of each stage. The coding system in the current study used Curtis‟ method of beginning with guiding questions. In this case, the guiding questions were the research questions outlined above and were addressed in the second phase when examining chats for the existence of patterns and themes: 1) How is rapport built in the initial stages of the chat? 2) How is rapport maintained during and following disclosure of suicidality? 3) How does the counsellor manage chat termination while maintaining rapport? 4) If threats to rapport occur, how does the counsellor re-establish or maintain rapport?   Threats to rapport were defined as any behaviour of client or counsellor that threatened existing rapport or hindered the development of rapport, as identified above. For example, the collaborative aspect of rapport could be threatened by the youth‟s resistance to safety-planning, and the client‟s trust in the therapeutic process could be threatened by the youth‟s questioning of the crisis counsellors‟ training, motives, or ability to understand his or her experience. Before patterns and themes could be examined, counsellor interventions which had been coded in the primary analysis had to be examined with respect to how they corresponded to dimensions of rapport. Since researchers in the field of suicide assessment (Granello, 2010a; Hoff, 2009) cite different “types” of interventions as   44 occurring in different phases of a given session with a suicidal client, the current study involved an examination of how dimensions of rapport were used online in the various phases of the chat. 2.8 Dimensions of rapport The counsellor responses obtained through content analysis in the first phase of the analysis were examined to determine whether they fit with a particular “dimension” of rapport, with the intention of determining whether dimensions of rapport for in-person counselling are also used in online counselling. In addition, researchers were interested in determining whether different dimensions of rapport were present in different phases of the chat. These dimensions of rapport and how they were arrived at are now outlined. Initially, Bachelor‟s (1995) categories of nurturant, collaborative, and insight- oriented dimensions of rapport were considered to provide a guiding framework for identifying different dimensions of rapport rather than the more traditional bond, task, and goal dimensions of working alliance identified by Bordin (1979).  These categories seemed both amenable to online counselling and were  considered to be more specific to the  bond dimension of working alliance, which was expected to be more salient due to the  expected lack of task and goal elements in a “one-off” crisis chat. However, as Hilsenroth (2007) noted, “fostering these distinct three forms of alliance […] may be difficult to universally achieve in patients as each may perceive and prefer different forms” (p. 208). Since both cited theorists had developed their definitions for use in in- person counselling, it was decided to determine whether dimensions of rapport were perhaps different in online counselling.   45 In this study, the two research assistants and I began examining possible dimensions of rapport by asking the guiding question, “How might this response build the connection between client and counsellor?” Similar counsellor responses were grouped together on a thematic basis, ensuring that multiple raters reached agreement prior to a response being assigned a category. For example, the response coded as “open-ended question to explore safety” was categorized by myself and the research assistants as [the volunteer counsellor builds the relationship by] „engaging chatter in safety planning;‟ „encouraging action,‟ and „encouraging collaborative action‟ by myself and the two research assistants.  Categories were assigned labels that best fit the proposed therapeutic goals of the counsellor‟s response (see Appendix D). As mentioned in the literature review, most definitions of rapport appear to reference both action-oriented and connection-building dimensions regardless of which theorist or measure is used or cited (Bordin, 1979; Bachelor, 1995; Pfeiffer, 1995). Upon examination of the counsellor responses used in the chats, it was decided to “collapse” the categories created by Bachelor (1995) and those created by Bordin (1979) into two categories since it was found that two categories emerged consistently: these were given the labels “action- oriented” (defined as engaging the chatter in goal or task-directed action) and “connection-building” (defined as creating an atmosphere of mutual trust, understanding and acceptance). These categories were generated with the understanding that they represented the intention of the volunteer counsellor (to build a connection or to engage the youth in action planning), not its effect on the youth. Therefore, they were not seen as causing rapport, since inferring causation from a mere existence of two phenomena together is inadvisable in research. Of course, some counsellor responses were not   46 considered as attempts to build the relationship and were not coded as such; for example, responding to questions about the service and so forth (see Appendix D). However, every counsellor response which could be classified as an attempt to build the relationship could be placed in one of these two categories. Other ways of using language were coded as well, although these could not be classified as “types of counsellor responses.” These included use of emoticons, use of abbreviations, use of figurative language, and use of informal language (see Appendix D). The number of response types in each given phase was counted; however, specific figures were not cited in the analysis unless there was a significantly higher number of a given response (more than one-quarter of total responses), since this study was qualitative in nature. Frequently the more salient features of a given chat were with respect to themes and patterns of youth/volunteer counsellor interaction rather than specific numbers of instances a particular response was made.  Action-oriented or connection-building responses were considered to “dominate” if they were present more than 60% in a given phase of a chat.  Youth responses were not coded in the same manner since the goals of the study were to examine interventions used by the counsellor, not the youth; however, the interaction between youth and counselor was qualitatively examined in each phase of the chat.    Following the coding of the types of counsellor responses, the interaction between client and counsellor was transformed into narrative descriptions which attempted to capture the interaction between client and counsellor in as richly descriptive a manner as possible.  When the narrative descriptions were written, each sentence or group of related sentences was examined with the guiding question: “What is happening in the chat at this   47 moment in the relationship between chatter and volunteer counsellor, how is it capturing aspects of the relationship between youth and volunteer counsellor?” 2.9 Trustworthiness It has been noted that “the challenge for those seeking to synthesize from case studies is to develop respect for the individual case context while permitting some blurring of unique features to occur” (Rossman, 1993, p. 42). To prevent confidentiality being breached, identifying features were obscured by eliminating identifying details. For example, one youth‟s detailed description of sexual abuse removed and only general features were reported, and another‟s detailed description of her activities with her friends was edited for references to locations. Questions or uncertainties regarding this process were directed to my research supervisor.  It was recognized that that this blending of cases needed to be balanced by the researchers‟ desire to describe unique features of individual chats. These features included illustrations of a particular pattern or noteworthy interactions worthy of further study. Using Stake‟s (1995) suggestions regarding case study construction, an effort was made to balance detailed narrative descriptions with appropriate vignettes that appeared to be particularly relevant to the research question or were considered potent “examples” of a particular pattern.   It was recognized that the cases were being viewed through the subjective lens of me as a researcher; therefore, the picture that emerged of each case would inevitably differ from that recorded by another researcher. Research assistants were used to ensure that the written reports conformed to the chat transcripts, and completed case studies were sent to my supervisor for feedback. In qualitative research, “trustworthiness” refers to the reliability of the findings (Tobin & Begley, 2003). Trustworthiness is generally broken   48 down into four categories: credibility, transferability, dependability, and confirmability (Tobin & Begley, 2003). To address credibility (the degree to which the research findings were a „credible‟ representation of the raw data), I ensured that the data were viewed independently through the lenses of multiple trained individuals (two research assistants as well as my research supervisor). To address transferability (the ease with which other researchers would be able to replicate or extend this study), I included appendices detailing my coding system and attempted to describe the procedures used to analyze the data in as much detail as possible. With respect to dependability and confirmability, I directed questions regarding the study‟s design and the processes of data collection to my research supervisor to ensure that the quality of the processes of data collection was sound and to ensure that the study‟s findings were representing the actual data set. 2.10 Ethical concerns  Informed consent was an obvious ethical concern regarding data collection. Volunteer counsellors were recruited from the Vancouver Crisis Centre and provided with a consent form indicating that their information would remain anonymous and that this research was not evaluative in nature (see Appendix C). For youth, however, the issue of consent was more challenging, since, being anonymous, they could not be contacted directly.  As mentioned in the Inclusion Criteria section, a statement was entered on the website prior to access to the chat indicating that chatter information may be used anonymously to help improve the service (see Appendix B). Only chats received after this change to the website were examined. The Crisis Centre‟s policies prevented the use of an electronic “consent form” or check box indicating assent. The reasoning behind this was that the ticking of a box or other form of assent may have prevented youth from fully   49 participating in the counselling process due to concerns about their anonymity, which would have compromised the ethical principles of beneficence and non-maleficence. This procedure is in keeping with policies of similar anonymous services for youth. For example, researchers of an online crisis service in Australia stated that “a core element of the ethos of the service is the provision of confidential services […]” (King, Bambling, Reid & Thomas, 2006, p. 77). Palys and Atchison (2008) noted that in some cases in which archival data are coded, it may not be practical or necessary to obtain informed consent, particularly when participants are, as was the case in this study, by definition anonymous. They pointed out that “when coding archival records […] lack of identification may not be a problem since it implies anonymity […]” (p. 75).   In fact, in the previously cited study by Williams et al. (2009), researchers were able to obtain ethics approval to use transcripts of chats from the online service provided by the Kids Help Line in Australia. In order to protect the participants‟ confidentiality, the Crisis Centre removed any identifying information prior to researchers gaining access to the chats.  Although IP addresses are theoretically traceable, tracing is a lengthy process that involves the police; hence, for the purposes of the study, they were considered non- identifying information. The Vancouver Crisis Centre removed uniquely identifying chatter information from the transcript, such as IP address, email addresses, schools‟ names and so forth before providing me with the transcripts.  The Crisis Centre provided a signed statement indicating that this procedure had been done. Only I, my research supervisor, and trained research assistants with no Crisis Centre affiliation had access to the chats. No identifying portions of transcripts were included in the final thesis to protect the anonymity of the participants.   50 An additional ethical consideration was the fact that, at the time this research was proposed, I was casually employed by the Vancouver Crisis Centre in a supervisory capacity. My role at the Vancouver Crisis Centre initially involved supervising volunteer counsellors on both the phone lines and, as well as providing authorization for sending emergency services to youth assessed as at high risk of ending their lives.  My employment there had the potential to create a conflict of interest as it was essentially a dual relationship. To ensure no duality of relationship for the duration of the research, the Crisis Centre offered me only shifts supervising volunteer counsellors in the phone lines in order to eliminate my access to volunteer counsellors working on  I did not have access to volunteer counsellor names and volunteer counsellor pseudonyms were numerically coded to prevent this. The Crisis Centre provided a signed letter to my research supervisor and me indicating its consent to my use of data for research purposes. Ethics approval was received from the UBC Ethics Board.  It was initially proposed that no transcripts or parts thereof would be included in the final thesis. However, after an examination of the data, it was found that a richer and more nuanced picture of the manner in which rapport was built could be obtained by using direct quotes from the youth and the volunteer counsellor. An amendment was submitted to the UBC ethics board requesting that we be permitted to use short, non-identifying quotes from the youth and volunteer counsellor. This amendment was approved by UBC on December 9, 2010.  Chats were transferred from the Vancouver Crisis Centre and stored on the hard drive of a password-protected PC.    51 Chapter 3. Findings In this chapter, the findings as they relate to each individual case study are presented using narrative descriptions of each case. These descriptions were structured to reflect the different types of responses used by the counsellor (obtained through the content analysis conducted in the earlier phase of the study) and the response they engendered in the youth, with the aim of capturing how rapport between client and counsellor appeared to be maintained and enhanced by counsellor responses. An effort was made to supplement descriptions of client-counsellor interaction with verbatim quotes from both client and counsellor, in order to provide a rich picture of the interaction in each given chat or part thereof. In addition, particularly noteworthy “vignettes” were selected for inclusion in selected segments. General themes and patterns were then extrapolated from the content analysis.  Following this, assertions were made with respect to these narrative descriptions. Stake (1995) noted that “Case researchers must make choices about the amount of analysis and interpretation in order to create assertions, which are the researcher's proposed generalizations […] (p. 85).”  3.1 Narrative descriptions  3.1.1 Case #1 This case is a crisis chat between an 18-year-old female from a Canadian city who is struggling with an eating disorder and a male volunteer counsellor who has less than 100 hours of experience on the crisis line. The chatter is classified as a “returning chatter” meaning she has accessed the crisis chat more than 15 times in the past six months. The chat duration is 37 minutes in total: nine minutes for the termination phase, 27 minutes for the risk assessment phase, and one minute for the termination phase.   52 Initial contact phase  The volunteer counsellor opens this chat with an informal greeting (“Hi, chatter 1”), then asks an open-ended question about the chatter‟s current emotional state. This chat has a very short rapport-building phase, as the chatter jumps into suicidal statements in the first minute of the chat: she discloses that she is concerned over her weight (she explains she has an eating disorder) and then states “now I want to kill myself.” Seven seconds later, she amends this: “I don‟t think I actually will.” About 17 seconds later, she again amends this with “But I really want to.”  The volunteer counsellor does not speak during these pauses in the chat, which allows the chatter space to clarify and amend her statements, which she does, frequently. This chatter, based on the IP address, has contacted the Crisis Centre before, and therefore her level of comfort with the service may have contributed to this very rapid disclosure of sensitive information. The volunteer counsellor responds to this not by immediately focusing on the overt suicidal statement, but by acknowledging the difficulty of the chatter‟s situation and  making educated guesses regarding the chatter‟s emotional state. The two most frequent volunteer counsellor responses in this phase are acknowledgements of the difficulty of the chatter‟s situation (considered a connecting-building intervention) and questions regarding coping (considered an action-oriented intervention). The volunteer counsellor expresses encouragement of the chatter‟s choice of coping mechanism, which is singing: “Well, it‟s good that you have something you love and you know will make better” – but also acknowledges the temporary nature of the effects of the coping: “at least for the time being.” He uses a number of summarizing statements and further expressions of encouragement: “This whole experience sounds really painful for you and I‟m really glad   53 you came on here and can talk to me about all of this.”  The volunteer counsellor waits for nine minutes before going back to explore the chatter‟s initial suicidal statement, using this time to build a connection and gain an understanding of the chatter‟s emotional experience. Thus, the main skills used to build a connection are reflections of the hard time the chatter is going through and an exploration of how the chatter is currently coping with her difficult emotions, setting the stage for the risk assessment which will follow. Connection-building responses are the most frequently used in this phase. Suicide assessment phase  As the risk assessment section of the chat begins, the volunteer counsellor provides a context for asking the question about suicide: he states, “[Chatter x], I know you said you don‟t think you‟ll do it, but you did mention at the beginning of the chat that you were thinking of suicide.” He waits a few seconds, then asks “Is that something you are thinking seriously about?” The chatter appears relieved that the issues have been brought up once more, as she openly discloses not only that she does have suicidal thoughts but that “the urge is kinda strong right now.” However, about 20 seconds later, she spontaneously counters her own suicidal feelings by stating “But I don‟t want to hurt my family.”  In between the chatter‟s admission of suicidal feelings and the chatter‟s own response to them, the volunteer counsellor is silent. The volunteer counsellor reflects the chatter‟s content and emotion and makes sure he is understanding the chatter by use of clarifying statements such as “Am I understanding that right?” The volunteer counsellor then uses reflective statements and multiple statements acknowledging the difficulty of the chatter‟s situation to explore current supports. The chatter discloses additional information about her difficulty in talking to her parents,   54 which the volunteer counsellor reframes as her need to feel understood. The volunteer counsellor uses this information as a bridge towards brainstorming other sources of support with the chatter, such as counselling offered by the local hospital. The volunteer counsellor phrases his suggestions tentatively and checks in with the chatter‟s perceived ability to follow-through with the suggestions; for example, “Would that be possible/manageable?” The chat then moves into the Termination phase. The salient of type of response is the action-oriented type; however, the connection-building type of response is also very present; the volunteer counsellor tends to use connection-building responses interspersed with action-oriented ones, particularly when asking questions about the chatter‟s suicidality. Termination phase This termination phase is very short (lasting only one minute); this seems due to the rapid typing of both the chatter and the volunteer counsellor and the lack of pauses between responses. The volunteer counsellor initiates the termination by asking an open-ended question regarding chatter‟s plans for after the chat. The chatter states she will go sing; the volunteer counsellor expresses approval regarding this (“oh good!”) and an emoticon ( ). The volunteer counsellor also advises the chatter to log on if she needs to, and advises the chatter of the availability of the services. The chatter responds by using an emoticon (), which the volunteer counsellor mirrors (“bye  ”). Action-oriented types of responses are used most frequently in this phase. 3.1.2 Case #2 This case is a crisis chat between a 15-year-old female chatter from a Canadian city and a female volunteer who has worked on the lines for more than 100 hours. Her presenting   55 concerns are depression, self-harm, and past sexual harassment. She is classified as a returning chatter. The chat duration is 41 minutes in total; the initial contact is 17 minutes in length, the suicide assessment phase is 18 minutes, and the termination phase is six minutes in length. Initial contact phase This chat begins with an informal greeting (“Hi, Chatter 2”) followed by an open-ended question regarding the chatter‟s current emotional state (“What‟s going on for you today?”) The chatter replies:  “ not at my best I must say but not at my worst either.” The use of the word “sigh” invokes the voice tone that is absent in face-to-face communication and may provide the volunteer counsellor with a sense of the emotion behind the words. The volunteer counsellor uses reflective statements as well as open- ended questions asking the chatter to elaborate: “How would it be for you to give me some more information about that?”  The volunteer counsellor then uses reflection of the chatter‟s emotion and content, as well as open-ended questions about the situation which seem to facilitate a greater understanding and help build rapport by showing interest in the chatter‟s experience. It appears the counsellor‟s reflections and open-ended questions are helping to build rapport, as the chatter begins to disclose increasingly more emotionally-charged information; she states that “becuz...suddenly it appears that the dam has broken and I can‟t take it anymore I feel detached and stupid for trying to put the crap that happened behind me…for example I was sexually harassed…” Thus, the chatter uses metaphor to describe her current emotional experience; in the absence of voice tone, this may be one way to communicate. The volunteer counsellor not only reflects the emotion expressed   56 but also mirrors and extends the dam metaphor when discussing possible coping strategies: “Oh, so the downside of writing is that it's very raw and real for you, and you're scared that you might not be able to handle the results of writing your true feelings down. Also, earlier you mentioned that suppressing your feelings is something that keeps you from cutting, but eventually the dam burst and the feelings get through, and I can imagine how difficult it must be trying to decide between the one option where you have to suppress your feelings and the other where the rawness of them is something that scares you.” Additional skills used in this segment included open-ended questions to explore perceived supports in the chatter‟s life as well as possible coping methods. Also, the volunteer counsellor explores deterrents to self-harm, creating a positive focus: she asks “[…] you mentioned earlier that it has been a while since you cut yourself, so I‟m wondering what has been working in your life that has stopped you from cutting yourself?” To summarize, the most frequent volunteer counsellor intervention in this section is reflection of emotion (nine responses) followed closely by reflection of content (eight responses). As in Chat # 1, connection-building responses are the most common in this phase. Suicide assessment phase In this phase, the volunteer counsellor introduces the question about suicide by providing a summary of the chatter‟s situation, thus acknowledging the difficulty of the chatter‟s situation and normalizing potential suicidal thoughts.  Questions about suicidal plan, intent, and previous history are then asked, again interspersed with reflective statements. The volunteer counsellor uses what was coded as “educated guesses” regarding the   57 chatter‟s emotions and help-seeking behaviours, using tentative language to allow the chatter room to disagree. “Educated guesses” refer to statements by the volunteer counsellor that are not direct reflections of the chatter‟s words but infer a given experience (see Appendix D). An example in this chat would be “I‟m guessing you don‟t feel comfortable with your therapist; is that why you didn‟t tell him you were feeling suicidal?” The chatter agrees with this statement (“yeah”). The volunteer counsellor accepts this response by using a reflective statement and checks in on the chatter‟s medication use. The chatter discloses that she does not like the side-effects of the medication but admits: “i dont think i'll ever want to be on them again...but i might need them.” The volunteer counsellor asks a number of closed questions in a fairly short period of time (about 15 minutes), but seems to maintain rapport through the use of empathy, mostly either reflecting the chatter‟s  stated emotions or through educated guesses regarding emotions when the chatter is not forthcoming. In this portion of the chat, the most frequent volunteer counsellor responses are in fact closed questions and reflection of chatter emotions (tied at six responses for each intervention).  For example, volunteer counsellor #2 asks about the specificity of the chatter‟s plan in the following manner: “I'm getting the sense that you feel exasperated by not knowing what the best course of action for you is, and you mentioned that overdosing is something that you‟re considering now, so I'm wondering if you've planned a specific date and time during which you will carry through with your plan?” The volunteer uses empathy and reflective statements to capture the chatter‟s feelings of ambivalence, then leads into problem-   58 solving and open-ended questions regarding the possibility of YIBC as a resource and an alternative to self-harm: “Okay, so it seems like you're experiencing a struggle not only about whether or not you should be taking medication from your psychiatrist, but you're also conflicted about whether suppressing your feelings or writing them down would be a better way to cope, and sometimes when the struggle gets to much, I'm sensing that that's when self- harming becomes an option. I'm wondering how it would be for you to come back and chat with us, or call our distress line, if you're ever feeling like things are getting to be too much for you and you're wanting to harm yourself?” The volunteer counsellor asks open-ended question regarding the chatter‟s support network as well as methods of coping other than self-harm. The chatter openly expresses ambivalence in response to the volunteer counsellor‟s statements, stating that: “sigh...i could try that...but nothing seems to work.” The volunteer counsellor acknowledges the chatter‟s hesitance and the difficulty of the situation. She then asks a number of risk assessment questions regarding plan, timing, and previous history of suicide. She also makes sure the chatter agrees to come on to chat if the risk does become more imminent.  In this phase, action-oriented and connection-building responses are used an equal number of times. Termination phase  This chat comes to an organic conclusion with the chatter stating “I feel better now” and thanking the volunteer counsellor. She also gives the volunteer counsellor an indication that she will follow-through with the suggestions for help, stating that she will ask the hospital about support groups for her eating disorder.  The chatter also uses an emoticon   59 (smiley face) at this point, which seems to indicate a lighter mood and a connection with the volunteer counsellor. The volunteer counsellor concludes the chat with a reflective statement regarding the utility of the youth‟s current coping methods as well as a disclosure of volunteer counsellor emotions, (“I‟m glad you feel better after talking”). The most frequent volunteer counsellor action is reflection of content; this termination phase also includes an expression of care by the volunteer counsellor and acknowledgment of difficulty of chatter‟s situation. Thus, connection-building responses are the most frequently used in this chat. 3.1.3 Case #3 This case is a crisis chat between a 16-year-old male chatter from a city in the United Kingdom who is experiencing depression and social isolation; he is also seeking advice with respect to his suicidal friend. He is speaking with a female volunteer who has more than 100 hours of experience on the lines. He is classified as a new chatter (less than 15 chats in a six-month period). The duration of the chat is a total of 40 minutes, with five minutes for the initial contact phase, 27 for the suicide risk assessment phase, and eight minutes for the termination phase. Initial contact phase This chat is the continuation of a previous chat, which the volunteer counsellor acknowledges immediately by saying “Hello again.” The youth responds to this with an emoticon (smiley face). The volunteer counsellor mirrors this non-verbal response by countering with a smiley face of her own. The volunteer counsellor then makes a number of educated guesses with regard to the chatter‟s current situation and reframes the chatter‟s values of friendship and caring. In addition, the volunteer counsellor reflects the   60 content of the chatter‟s experience and makes educated guesses regarding his emotional experience. The Initial Contact phase with this chatter only lasts five minutes; however, it can be assumed that rapport was built in the previous chat (which was not included as it did not meet inclusion criteria). The most frequent response in this chat was reframing of chatter values, which could serve to build rapport by allowing the chatter to feel a sense of empowerment and self-worth, something he appears to struggle with as evidenced by various statements along the lines of “I feel disgusted with myself…I screw it up[…]” It appears that the volunteer counsellor was able to counter these negative feelings and build a connection by reaching between the lines and pulling out positive values of caring for his suicidal friend and how important the friendship is to him. The connection- building type of response is the most common in this chat. Suicide assessment phase Volunteer counsellor # 3 leads into the risk assessment questions by both providing a rationale for the questions as well as providing a “window” for the chatter to disclose suicidality; it is phrased in an informal, conversational manner that seems to open the door for the chatter to respond affirmatively, which he does: “I meant to ask you earlier – I know the stress level has been high for you lately, working through this [waits 7 seconds] “have you yourself been feeling suicidal?” The chatter appears relieved to disclose his feelings, stating simply “Yes, actually.” The volunteer counsellor uses the reflective skill of summarizing as well as open-ended questions regarding protective factors (e.g. “what do you think made things a little better for you?”) to help get a sense of the situation. This invested interest in the chatter‟s experience appears to foster rapport by allowing the chatter to feel that he   61 matters: he states that he feels more understood on YIBC than when meeting with his psychiatrist: “I find it very difficult to open up […], sometimes it feels like I‟m wasting their time because I hardly talk.” The volunteer counsellor frequently acknowledges the difficulty of the chatter‟s situation by using statements such as “I imagine it would be really difficult to go through feeling like you‟re personally responsible for this?” Additional skills used throughout the risk assessment include numerous reframes of the chatter‟s values, as well as some perspective-taking to help facilitate brainstorming (e.g. “If you had a friend in this situation, what would you tell them?”). Non-verbal attenders (e.g. “mmhm”) and summarizing statements are also used frequently in this phase. The chatter responds to the volunteer counsellor‟s statements in a manner that suggests rapport has been built; he frequently replies “exactly” to the volunteer counsellor‟s reflections, and he uses emoticons (most notably, a “smiley” face) to denote positive emotion. Towards the end of the risk assessment, when the chatter has agreed to contact YIBC if he has thoughts of acting on his suicidal feelings, the chatter and volunteer counsellor both express satisfaction with the relationship by mirroring each other‟s emoticons: the chatter states that coming onto YIBC would be easier than calling in, and uses a smiley face () to reinforce the positive emotion, to which the volunteer counsellor responds not with words but with another smiley face ().  This exchange seems to open the doors for more positive exchange, as the chatter then overtly thanks the volunteer counsellor: “Thank you so much for being here […] I‟ve been on here before, and what you do is really good.” The volunteer counsellor responds to the chatter‟s   62 thanks and normalizes the experience: “I know some people find it weird to talk on the phone […] lots of people prefer this chat.” The chatter agrees: “Exactly…I wouldn‟t feel comfortable at all talking on the phone about stuff like this.” In summary, in this risk assessment, the volunteer counsellor uses the skills of normalization, acknowledging the difficulty of the chatter‟s situation, and emoticons to build rapport in the absence of voice tone. Notably, the volunteer counsellor also acknowledges potential resistance on the part of the chatter, which appears to help him feel more understood and validated. Connection-building responses are the most frequently used in this chat. Termination phase This chat begins its termination phase with the youth initiating its‟ end; the youth indicates that he will be going to bed soon. The volunteer counsellor asks a closed question regarding the chatter‟s proposed actions and an open-ended question regarding coping. The chatter responds to the coping question by opening up still further: after about a minute of silence, he responds with “that‟s a whole other thing, really…I self harm sometimes.”  The volunteer counsellor acknowledges this in a non-judgmental manner: he simply reflects this and acknowledges its function: “Okay, so sometimes to deal with stress you try self-harm…” The use of three dots indicates a natural pause or an indication that the chatter may continue. When the chatter simply responds with “yes‟” the volunteer counsellor asks “Does it help you?” to which the chatter responds affirmatively (“yeah, it does”).  However, interestingly, when asked if he is thinking of self-harming after the chat, the chatter states that he is not going to cut tonight. He discloses that he will simply go to bed since the chat has helped him, and punctuates this   63 statement with an emoticon (). The volunteer counsellor tells the chatter he is glad he logged on, and mentions the possibility of an outreach email to check in on the chatter who responds by declining. However, he states that he will likely log back on in the near future, thus implicitly referencing the usefulness of the service. The chat concludes with the chatter saying goodbye with a smiley face and the volunteer counsellor mirroring this emoticon. In this chat, the most common response made by the volunteer counsellor is with respect to coping, which was considered an action-oriented response. 3.1.4 Case #4 This case consists of a 15-year-old female chatter from an American city who is accessing the service for advice regarding her suicidal friend. She is classified as a new chatter and she is chatting with a female volunteer who has less than 100 hours of service on the lines. The chat lasts 69 minutes in total, with the initial contact lasting 24 minutes, the suicide assessment lasting 40 minutes, and the termination phase lasting 15 minutes. Initial contact phase The volunteer counsellor begins by informally greeting the chatter (“Hi there”) and then asking an open-ended question regarding her current state: “How are you doing this evening?”). She tries to establish the chatter‟s needs by asking close-ended questions: “Do you need some support?” The chatter responds affirmatively, and the volunteer counsellor expresses appreciation for her decision to log on; she also makes an educated guess regarding the current situation and a further probing question: “I‟m glad you felt comfortable coming to us. It sounds like this is a rough time for you - has something happened recently?” The chatter responds by asking a number of questions about the service, which the volunteer counsellor responds to by clarifying her role in an informal   64 manner: “I'm sensing you're concerned about our level of training, but we are trained and if you and I come upon an issue I'm not familiar with or that a pro should be dealing, with then I would ask if you'd like me to find you a resource to somewhere like that. But I'll do my best :) ” This chatter‟s conversational style could be characterized as fairly terse and reticent to start with, as she gives one-word answers and waits a significant period of time before replying to the volunteer counsellor (at one point, she waits three minutes before responding, which is a fairly long time in comparison to other responses she has made). The volunteer counsellor uses the skills of reframing the chatter‟s values (“caring for your friend is important for you”) and asks a number of closed questions regarding actions taken so far. Interspersed with these are reflective statements. The volunteer counsellor also uses educated guesses regarding the goals of the chatter, using tentative language which appears to allow the chatter the space to disagree with the volunteer counsellor‟s reflections. She uses clarifying questions when the chatter is not forthcoming or vague, which results in the chatter elaborating on her concerns. In this chat, which lasts 24 minutes, a total of 11 questions are asked, five of which are closed questions. A number of questions were centered around the chatter‟s needs, goals, and values; in this chat, the action-oriented dimension of rapport is the most salient in the counsellor‟s responses. Suicide assessment phase Similar to volunteer counsellor # 1, volunteer counsellor # 4 frames her risk assessment question in the context of the chat by referring to the chatter‟s initial question about her training: she states, using informal language, “Hey, umm…earlier in the chat you asked   65 me if I was trained in suicide prevention…I‟m just wondering if you‟re thinking about killing yourself?” She softens the potential harshness of the question and provides a rationale for the question by immediately typing, “Sorry to spring the question on you like that I am just worried about you.” The chatter responds to this question honestly: “Half thinkin‟ bout it but no plan.” Following this, risk assessment questions regarding plan, means, intent, and previous history are asked, interspersed with reflective statements as well as clarifying questions. The chatter responds affirmatively to the volunteer counsellor reflections, and is able to successfully convey emotion using only text; for example she replies “yeeesssssss” to the volunteer counsellor‟s reflection “Ahh so it sounds like you‟d feel pretty crappy about leaving people behind to deal with it”; the drawn-out word seems to imply a more intense affirmation than a simple “yes,” indicating that the volunteer counsellor was successful in understanding her experience. It appears that rapport may have been established in this stage since the chatter openly discloses self-harming behaviours when asked about her coping strategies. The volunteer counsellor responds by expressing acceptance of this coping method and open- ended questions, encouraging the chatter to explore alternative methods of coping. The chatter appears to feel comfortable enough to dispute these suggestions; stating that “No I want to cut.” She waits a few seconds, then elaborates on the utility of this method of coping: “To get pain out and numb.” The volunteer counsellor accepts this explanation and asks further questions regarding her current state (“Oh you want to cut NOW?”  ) The volunteer counsellor then checks in with the chatter‟s safety, using informal language: “Mhmm so there are a couple of self-harming behaviors you do for that, you're   66 tellin‟ me. I just want to make sure though, that when you do cut (cuz hey, you say it helps you feel better), that you're not doing it too deep or anything....or if you did, do you think you'd be willing to call for help? ” The chatter responds by clarifying the nature of her self-harming: “I don‟t do it deeply just on the surface not deep.”  The volunteer counsellor then checks in regarding external referrals; the chatter, however, declines, saying that although she has a therapist, she does not feel comfortable discussing  things “on a deep level” with them. The chatter also expresses that she feels more comfortable using YIBC, implying once again that rapport has been established. The chatter asks a number of questions about the service and the volunteer counsellor‟s availability, which also implies that a personal connection has been established. The chatter then explains the rationale for her questions: “This helps lot havin‟ someone to talk to.” The volunteer counsellor responds to the thanks with a reflective statement:  “I‟m glad you find it helpful” as well as an emoticon (:D). The chat then takes an interesting turn: the chatter states “yeah I understand;” 20 seconds later, she asks “Can I cut now?” The volunteer counsellor seems understandably taken aback; this is shown in the text in the form of an informal expression of confusion and a clarifying question “Um, are you asking me if you can cut?” When the chatter responds affirmatively, the volunteer counsellor openly states her role and the limitations thereof and moves the focus back onto the chatter:  “Well honestly I can't stop you by reaching through the monitor, but if your friend asked you the same question what do you think you would say?” The volunteer counsellor waits about 20 seconds, then asks a closed question asking the chatter to brainstorm other options for cutting. The chatter responds only to the first part of the message:   67 “well i would not let them, I would tell the friend i care about you alot and am worried that you could really cut and injure deeply so please don‟t do it its not worth it...... By over-exercising that helps stop urges although hungry.” The volunteer counsellor expresses appreciation and agreement with this response, then expresses her rationale for suggesting alternate methods of coping: she states her concern for the chatter‟s physical safety. She also clarifies that it is her role as a volunteer counsellor to check on the chatter‟s safety. The volunteer counsellor also makes a number of statements that show acceptance of the chatter‟s motivation to engage in cutting and not eating as coping strategies, which may have helped foster rapport since the chatter implies she feels understood;  for example: “mmhm on the one hand you're saying eating would help; it sounds like you're sick of feeling weak and hungry, but that you don't plan on eating cuz I guess that's another way you distract yourself from the emotional pain- by focusing on the pain of hunger.” The chatter responds affirmatively: “Yes exactly.” Eventually, the chatter is able to identify other methods of coping for the evening: reading, yoga, and deep breathing are things she will engage in. The volunteer counsellor lightens the mood towards the end of the risk assessment by calling attention to her spelling mistakes “Pardon my typos, haha.”  The most frequent volunteer counsellor responses in this chat are regarding coping: a total of 12; these were considered to represent the action-oriented type of response. Termination phase The volunteer counsellor begins this phase by asking the chatter if she will feel safe tonight. The chatter states openly “to be honist [sic]...I don't know.” The volunteer counsellor attempts to brainstorm options with the chatter with respect to people who can   68 support her for the night. The chatter identifies that she can call a friend and that talking to someone before she goes to bed will be helpful. The volunteer counsellor then attempts to find the chatter resources in her area (local crisis centres) since this chatter is located in the United States. The chatter expresses appreciation for the usefulness of this information: she states “ok that sounds great. Sorry I can‟t do it for myself” The volunteer counsellor normalizes this: “That‟s not a problem at all – most people don‟t know that kind of stuff.” The chatter then expresses her appreciation for the volunteer counsellor: “But you people are the nicest and most nonjudgmental people I have found to talk to.” The volunteer counsellor responds by using an emoticon () and clarifying the reason for referring the chatter to external resources, explaining that while YIBC is only open until 1am PST, the local resources are open 24/7. The volunteer counsellor frequently uses emoticons to punctuate her statements which seems to lighten the mood. She then leads into making a no-harm agreement with the chatter, asking her if she would promise to contact YIBC if she feels like hurting herself. The chatter answers honestly: “I‟ll try to do that.” The chatter then self-identifies things she will do after the chat to help herself cope: she states she will “find a friend to chat with.” The chatter concludes the chat with the words:   “Thank you So Much for all you help” (waits 20 seconds) “You help make me feel worth it.” The volunteer counsellor responds to the thanks (“Well I'm really glad that I can be helpful, and you definitely are worth it”). She then once more asks the chatter about making a no-harm agreement, and this time the chatter assents more firmly than the last time: (“I promise”). The volunteer counsellor responds with an emoticon () and an expression of positive emotion (“I'm sooo glad to hear it”) and   69 expresses appreciation of the chatter‟s decision to log on (“I'm really really glad that you decided to chat today [15/f]. I think you made a really healthy, positive choice for yourself by deciding to talk when you're feeling this way”).  The chatter thanks the volunteer counsellor, uses an emoticon (); the volunteer counsellor mirrors this emoticon () and ends the chat. The action-oriented type of response is the most frequently used in this phase. 3.1.5 Case #5 This case describes a 15-year-old female chatter from a Canadian city who is accessing the crisis chat regarding self-harm. She is classified as a returning chatter. She is chatting with a female counsellor who has less than 100 hours of experience on the lines. The chat lasts a total of 63 minutes, with the initial contact phase lasting two minutes, the suicide assessment phase lasting 55 minutes, and the termination phase lasting five minutes. Initial contact phase  This phase is very short in length, since the chatter immediately discloses her suicidal feelings (within two minutes).The most prevalent response in this section is reflection of emotion: the volunteer counsellor‟s use of this intervention seems to allow the chatter to feel understood, as she responds affirmatively (“Yes, Exactly”) to the statement made by the volunteer counsellor: “I can imagine it‟s very frightening to want to live but not be able to come up with a reason to live.” The chat then moves in to the Risk Assessment phase. The most prevalent type of response is the connection-building one.      70 Suicide assessment phase The volunteer counsellor begins the risk assessment by asking the chatter how she has dealt with the suicidal feelings in the past. The chatter openly states that she has “not dealt very smartly” with these feelings; specifically, that she has self-harmed and attempted suicide in the past.  She honestly states her current situation: “And I am afraid I will do something stupid again.” The volunteer counsellor reframes this as an implied goal of the chatter:  “Sounds like you want to find new ways of dealing with these intense feelings you have.” The chatter responds affirmatively, implying that a good working alliance has been established: “My ears are wide open for ideas.” The volunteer counsellor then attempts to engage the chatter in brainstorming options for “less destructive things.” The chatter actively participates in brainstorming by identifying options she has tried in the past: (exercising, watching movies, etc); however, she shares that these methods are not working for her anymore. The volunteer counsellor acknowledges the difficulty of the chatter‟s situation and makes an educated guess with respect to the chatter‟s emotional state: “So you‟re tried several different things (exercising, watching movies, hiding) and found they didn‟t help. So I guess you‟re feeling stuck.” The volunteer counsellor also asks the chatter questions with respect to suicidal means, plan, and intent; the chatter responds honestly by disclosing her plan (to take her mother‟s prescription medication), although she appears to want to lighten the mood by exclaiming “gosh I hate being so morbid” immediately after disclosing this. The volunteer counsellor frequently reflects the difficulty of the chatter‟ situation and makes a number of educated guesses on her emotional state. The chatter appears to be emotionally   71 impacted by the volunteer counsellor‟s reflections: she states “Thanks for not shooting me down or saying I am an idiot. I Realy [sic] apreceat [sic] it.”  About 20 seconds later, she adds “I need the thought that some one cares.” The volunteer counsellor responds to this by reframing the chatter‟s values: “I'm getting that it's something really important to you, to have caring support that doesn't judge you.” She also makes a connection between the chatter‟s feelings and her actions, tentatively framed as a question: “Seems that you haven't told anyone about your thoughts because, even though you'd really want the support, you don't want to take the risk that they'll think badly about you, (like maybe by thinking that you're morbid?)” The chatter agrees, and the volunteer counsellor reflects this, then gently explores the chatter‟s help-seeking patterns (she admits she is also afraid her friends will think she is morbid). The chatter then corrects the volunteer counsellor, stating that she is more worried about bothering her friends than she is about what they think of her. The volunteer counsellor responds by correcting herself and by reframing the chatter‟s disclosure in terms of her values: “It seems like you feel guilty for telling people about your painful feelings. Sounds like you care a lot about your friends and want to protect them from your emotions- so they don't worry about you, maybe?” The volunteer counsellor also gently confronts the chatter on her reticence to discuss emotionally charged topics such as suicide with her counsellor: “Sounds like you feel like you're burdening your counsellor with your problems, even though that's kind of what he or she is there for... have you ever talked to him or her about feeling not important enough to worry about?” She allows the chatter to imagine what it might be like to share some of these feelings with her counsellor. The chat then   72 moves to the Termination phase. The action-oriented dimension of rapport is most salient in this chat. Termination phase The volunteer counsellor initiates the termination of this chat by expressing her appreciation for the chatter‟s decision to log on. She uses an emoticon () to simulate a smile. The chatter responds by expressing her own appreciation for the service: “Me too. Thanks for understanding everything. You are a really nice person to talk to.” The volunteer counsellor acknowledges the thanks and uses an emoticon () to express her happiness with this.  She asks the chatter what she will do after the chat to take care of herself and reminds the chatter of the nature of the service and suggests that chatter contact the Crisis Centre if she needs to. The chatter thanks the volunteer counsellor by name and wishes her a good night.  Expression of care is the most frequent statement in this phase; hence the most salient type of response is the connection-building one. 3.1.6 Case #6 This case describes a 16-year-old female chatter from a Canadian city who is chatting with a male volunteer who has more than 100 hours of experience on the lines. She is classified as a returning chatter and her presenting concern is depression. The chat duration is 35 minutes in total; the initial contact lasts two minutes, the suicide assessment lasts 27, and the termination phases lasts six minutes. Initial contact phase Following the pattern seen in previous chats, this chat opens with an informal greeting by the volunteer counsellor as well as an open-ended question regarding the chatter‟s current emotional state. As with Chat #5, this chat moves fairly quickly into the risk assessment   73 phase: it takes the chatter only three minutes to disclose that she would like to take all the pills she has. Precipitated by this disclosure, the volunteer counsellor asks more specific questions regarding triggers (“what‟s goin‟ on for you today?”), and the chat then moves into the risk assessment phase. The most salient type of response in this phase is the action-oriented type. Suicide assessment phase This assessment starts with the chatter initiating the conversation on suicide rather than the volunteer counsellor: “I sometimes feel like if I can‟t take it I will end my life.” The volunteer counsellor acknowledges the chatter‟s suicidal state, then asks a number of questions regarding the length of the chatter‟s feelings and about triggers for the suicidal feelings. The volunteer counsellor makes an educated guess regarding the chatter‟s emotions based on the content, and the chatter responds by elaborating on her emotional experience. The chatter initially gives simply short, one-word answers but eventually elaborates and shares the thoughts that precipitate the feelings of wanting to end her life: “I got this general feeling and it kept building on it's self. I keep feeling like I am not worth it like the people in my life would like it better if I wasn't around.” The volunteer counsellor uses educated guesses to infer the emotions likely to accompany this: “I would imagine that feeling like the people around you don't want you around must leave you feeling really hurt and lonely.”  The chatter agrees with the volunteer counsellor‟s statements and the volunteer counsellor continues to offer empathic statements. The volunteer counsellor then moves into exploring supports with the chatter, as well as providing reassurance regarding the importance of the chatter‟s concerns: “You know, I can tell that you really don‟t feel like you‟re worth anything right now, but you‟re   74 definitely not wasting your counsellor‟s time.” The volunteer counsellor then references the chatter‟s previous disclosure of availability of pills and uses this as a rationale for asking about suicide: “I‟m wondering, since you mentioned having a bottle of pills in front of you, have you taken any?” The volunteer counsellor expresses happiness (“I‟m glad to hear that”) when the chatter states she has only taken prescribed medications. The volunteer counsellor asks further risk assessment questions and eventually asks the chatter if she would be comfortable placing the pills in another room. The chatter complies using the abbreviations kk and brb (Okay and Be Right Back, respectively). The volunteer counsellor mirrors this online-specific language use by responding with a smiley face () when the chatter states she has put the pills in another room. The volunteer counsellor also expresses understanding in an informal manner, using online- specific abbreviations (“icic”, meaning “I see, I see”). He attempts to establish the chatter‟s reasons of living or deterrents to suicide by asking open-ended questions such as “What has kept you going to this point?” The volunteer counsellor also uses the skills of reframing to identify the chatter‟s goals: in response to the chatter‟s comment that she feels “so alone,” the volunteer counsellor responds with: “ic, so it‟s important for you to be connected to someone.” As with other volunteer counsellors, he provides the chatter with contact information and general information regarding the Crisis Centre‟s availability and other services. He inserts an emoticon () into this sentence. The chatter provides her location and thanks the volunteer counsellor, then asks a question about YIBC‟s hours. The chatter once more thanks the volunteer counsellor. The volunteer counsellor responds by using an emoticon () and shares “No problem, it‟s what I‟m here for.”    In this chat, informal language is used seven times in 27 minutes, which is   75 with more frequency than any other counsellor behaviour, and the most frequently seen dimension of rapport this phase is the connection-building one. Termination phase The volunteer counsellor leads into the termination phase by reflecting the chatter‟s earlier expressed need to eat, asking the chatter if this is something she plans to do after the chat. The chatter appears to consider this and recognizes that this is an adaptive option for her; she acknowledges that she “probably should […] or I would be a mess physically.”  The volunteer counsellor then proceeds with a summary of the chat, coupled with an expression of appreciation for chatter‟s decision to log on.  He then indicates the impending end of the chat, giving an expression of encouragement (“I do have to go soon and wish you the best of luck!)” The chatter responds with a powerful expression of appreciation for the chat: she tells the volunteer counsellor “this chat saved my life” to which the volunteer counsellor responds with a “big smiley face” emoticon (:D). The chatter ends the chat by thanking the volunteer counsellor again and stating that she will go and eat now. Connection-building and action-oriented responses are used with equal frequency in this chat. 3.1.7 Case #7 This case describes a 16-year- old female from a Canadian city whose presenting concern is an eating disorder. She is classified as a returning chatter. She is chatting with a male volunteer who has more than 100 hours of experience on the crisis line. The chat lasts for a total of 51 minutes; 21 for the initial contact, 26 for the suicide assessment, and four minutes for the termination phase.    76 Initial contact phase   The volunteer counsellor opens this chat informally by apologizing for the delay in responding and asking an open-ended question regarding the chatter‟s current emotional state. The chatter responds openly by stating “soso,” pausing about 30 seconds, and then typing “I want to cut.” The volunteer counsellor responds by expressing appreciation for the chatter‟s decision to log on, coupled with an educated guess regarding the chatter‟s situation: “You must have a lot going on for you tonight?”  The chatter agrees and elaborates regarding her current concern (her eating disorder). The volunteer counsellor moves into assessment of self-harm based on the chatter‟s statements that she “hates” her body: he states: “Sounds like you‟re really pretty angry at your body right now. I‟m wondering if you‟re cutting right now?” The chatter responds honestly, stating that she is not cutting yet but that she could “get a scissors soon here.”  The volunteer counsellor responds by clarifying his intentions and his goals for the chat, and asks an open-ended question with respect to possible options of keeping safe: “I would like to help you sort through these feelings tonight, and also want to help you stay safe, how would it be for you to move away from the scissors, or move the scissors to another room.” The chatter, implying that she feels safe enough to disclose her experience honestly, tells the volunteer counsellor that she would rather keep the scissors close to her. The volunteer counsellor accepts this and simply asks the chatter to let him know if she starts cutting during the chat, which the chatter agrees to. The volunteer counsellor then offers the chatter the idea of exploring (together) other options for dealing with the anger toward her body that the chatter is experiencing.  The chatter agrees with this. The volunteer counsellor then provides a rationale for his suggestion: “Thanks for that. I just   77 want to work with you to keep you safe.” He also uses a number of empathic statements (e.g. “Sounds like you are feeling really overwhelmed with your frustrations towards your body”). He also frequently acknowledges the difficulty of the chatter‟s situation and shows acceptance of the chatter‟s preferred method of coping and her frequent expressions of desire to continue to cope by cutting, taking diet pills, and exercising. The chatter responds briefly and affirmatively to the volunteer counsellor‟s statements and reflections (e.g. “yes” or “exactly”). The volunteer counsellor then explores possible avenues of professional support (the chatter states she is receiving treatment for her eating disorder and that it is going “ok”). The chatter appears open and honest with her responses, elaborating more as the chat progresses. Towards the end of this phase, she explains to the volunteer counsellor the function of her self-harming behavior: “I try to distract so I don‟t kill me.” The most frequently used counsellor responses in this chat are collaborative in nature; however, these are interspersed with connection-building responses, although these are fewer in number. The chat then moves into the Suicide Risk Assessment phase. Suicide assessment phase The volunteer counsellor leads into this phase with a clarifying statement: he asks the chatter if the cutting and pill-taking is suicidal in intent or for coping purposes. The chatter expresses that she is afraid that her coping behaviours may end up killing her. The volunteer counsellor responds in an action-oriented manner by stating “Ok, so let‟s work together to keep you safe tonight […].” He then attempts to turn the focus to brainstorming possible coping activities. However, the chatter effectively draws attention to her ambivalence regarding suicide: she asks the volunteer counsellor “Why should I   78 stay safe?” The volunteer counsellor waits about 30 seconds before responding, and his response is simply a reflective/clarifying statement of the chatter‟s previous disclosure as well as a statement regarding his intention. He gently points out an inconsistency in the chatter‟s communication: “When I say stay safe I am talking about keeping you alive, it sounded like that is something you wanted as well.” The chatter acknowledges her own mixed feelings: she states that she “halfway” wants to stay safe. The volunteer counsellor reflects this back to the chatter and thanks her for her honesty. The chatter responds with brief affirmative statements (“yup”), as seems to be her pattern. The volunteer counsellor normalizes the youth‟s ambivalence (“mixed feelings are common”) and focuses on exploring options for coping. He allows the chatter to take initiative by stating “you mentioned distraction” and leaving this as an open statement for the chatter to respond to. The chatter agrees but once more attempts to engage conversation around her questions about whether it is “worth it” to continue living (“really with all the suffering and pain is it worth it?”). The volunteer counsellor responds quite directly, making a general statement about human experience and attempting to flush out this chatter‟s reasons for living: “it is interesting how life can be so full of pain, but still be balanced out to make it worthwhile. What are some of the things that make life worthwhile for you.”  The chatter takes two minutes to respond, and when she does, she is the most loquacious she has been in the chat thus far: she shares that she values being around her loved ones and “living out [her] dreams and passions.” The volunteer counsellor reflects and validates this disclosure asks the chatter to further elaborate on what those passions may be. He expresses interest in knowing more about what is important to the chatter. The chatter responds by sharing that helping others is something she enjoys. The volunteer counsellor   79 reflects his statement (“So helping others is something you really value”) and asks the chatter once more to elaborate by asking the chatter if she has an outlet for doing so in her life. The chatter states that she does, and eventually even discloses that her suicidal ideation seems to be subsiding: “So maybe life is worth it.”  The volunteer counsellor builds on these statements and attempts to have the chatter reflect more on her reasons for living by suggesting that she journal or create something that reflects these reasons. The chatter decides to journal and listen to music. She then, without prompting, promises the volunteer counsellor that she will not kill herself, emphasizing the word “promise” by writing in capital letters. The chat then moves into the Termination phase.  Action- oriented and connection-building responses are equally present in this chat. Termination phase The chatter begins this phase by stating that she needs to go and rest. She thanks the volunteer counsellor for her help. The volunteer counsellor expresses appreciation for the chatter‟s decision to log on in an informal manner (“I‟m glad we got to chat”) and asks the chatter if she would log back on if she feels like harming herself (effectively acquiring an informal no-harm agreement). The chatter promises to do so, and the volunteer counsellor thanks the chatter and tells her to take care. The chatter responds by thanking the volunteer counsellor; she elaborates by stating “you guys have helped me more than you know.” The chatter concludes the chat by saying “goodnight.”  The connection-building dimension of rapport is most salient in this chat.      80  3.1.8 Case #8 This case describes an interaction between an 18-year-old male chatter from a Canadian city and a female volunteer counsellor who has more than 100 hours of experience on the lines.  He is classified as a new chatter and he is accessing the chat service for concerns related to childhood abuse. The chat lasts a total of 87 minutes: 16 for the initial contact, 40 for the risk assessment, and 31 minutes for the termination phase. Initial contact phase  The volunteer counsellor begins this chat with an open-ended question to explore the chatter‟s current emotional state. As the chatter opens up about the fact that he is having a difficult day, the volunteer counsellor frequently uses phrases indicating acknowledgement of difficulty of chatter‟s situation. The volunteer counsellor initially leaves pauses of 10-15 seconds between comments; however, the chatter draws attention to his dislike of this style, stating that he is not a talkative person and that it would be helpful if the volunteer counsellor spoke more in order for him to feel comfortable. The volunteer counsellor responds by sharing that she had thought the chatter was typing; hence the pauses between statements. The chatter clarifies that he was in fact typing but erasing his words. The volunteer counsellor asks the chatter if there is something specific he would like to address. The chatter states that yes, there is, but expresses his difficulty with articulating it. The volunteer counsellor normalizes these feelings “ it can be really difficult to know how to express what you're feeling to someone you don't know”  to which the chatter responds that he does trust the volunteer counsellor but that it is hard to get started. The volunteer counsellor expresses understanding of this difficulty and lets the chatter know he can take his time. She also attempts to help the chatter open up by   81 asking prompting questions such as “Did something happen today to prompt you to come on the chat?” The chatter then shares that he was watching TV, which triggered him.  The volunteer counsellor reflects the chatter‟s emotions and uses educated guesses to gauge the chatter‟s feelings. The chatter agrees with the volunteer counsellor‟s reflections and elaborates by using metaphorical speech: he states he feels “like a turtle without it‟s shell” and shares that English is not his first language. The volunteer counsellor responds with a statement of acceptance and reassurance, coupled with a clarifying question regarding the chatter‟s emotions: “no that makes perfect sense - you feel exposed? Vulnerable?”   The chatter appears to feel understood by the volunteer counsellor‟s statements (based on his answers of “yes” or “exactly”), which in this section consist of reflections of content and emotion, coupled with clarifying questions. Many of the reflections are phrased tentatively (for example, the volunteer counsellor puts a question mark after a feeling word, or states “I‟m wondering if you‟re feeling […]”).The chat then moves to the risk assessment phase. The connection-building type of response is the most frequently used in this phase. Suicide assessment phase The volunteer counsellor begins the suicide assessment by reflecting the current situation, thereby providing a context for the question “are you suicidal.” The chatter seems taken aback by the question, though - he responds with “I wasn‟t expecting that question.” The volunteer counsellor responds to this statement with an acknowledgment of difficulty of the chatter‟s situation as a rationale for her questions. The chatter then refers the volunteer counsellor to his previous chats, stating that this may help her understand his   82 experience. The volunteer counsellor states she will read them but notes that this may take some time.  The volunteer counsellor acknowledges that she has read some information referring to a difficult childhood and then brings the chatter back to the present by referring to his past thoughts of suicide and asking how these match up with how he is currently feeling.  The chatter replies that he still feels uncertain about suicide as an answer to his problems. The volunteer counsellor reflects this ambivalence and the chatter clarifies that he “won‟t do it” [end his life]. The volunteer counsellor then checks in on the chatter‟s current safety, responding to his earlier statement in which he shared he was feeling cold and having trouble breathing. There are a few minutes of confusion as both chatter and volunteer counsellor enter text at once, leading to crossed messages. However, humour is used by the counsellor in this instance (“Oops, crossed messages! Guess I was typing too fast!”), which seems to dispel any potential awkwardness.   The volunteer counsellor then asks about the chatter‟s existing supports (his therapist) and his difficulty opening up to him about his suicidal thoughts, despite the fact that he “knows he should.” The chatter acknowledges that he is looking for help, but that he finds it intimidating to open up to someone in-person. The volunteer counsellor validates the chatter‟s experience by using empathic statements and normalizing his experience; for example: “Some people do find it a bit easier to just write stuff down anonymously . . . I guess it can feel like there's less at stake if you don't know the person and can't see them.” The volunteer counsellor also reframes the chatter‟s needs for control over bringing up difficult issues, and the conversation turns to an exploration of what a session with the therapist might look like. The chatter opens up with respect to his fears around discussing his suicidal feelings and the volunteer counsellor responds with empathy and validation.   83 The chatter then shares with the volunteer counsellor that even thinking about disclosing sensitive information to his therapist is causing him to feel physically ill. He keeps the volunteer counsellor informed of his actions as he goes to the washroom to throw up. The volunteer counsellor shows acceptance and normalizes the chatter‟s reactions. She also frequently checks in on the chatter‟s current emotional and physical well-being.  The volunteer counsellor reframes the chatter‟s current distress by focusing on the importance of this meeting with his psychologist and reframing the distress as an indication of its value. The chatter appears to agree with the volunteer counsellor‟s statements and acknowledges that he is putting pressure on himself to “get better” which may be interfering with his progress in therapy. The volunteer counsellor engages in some task- oriented conversation by attempting to focus the chatter on what he perceives as the most important goal of therapy. The chatter replies fairly succinctly: “Stop thinking about killing myself every day.” This leads back to the volunteer counsellor checking in on the current status of the suicidal thoughts. At this point, the chatter makes a process-oriented comment that changes the dynamic of the chat: he asks the volunteer counsellor why she is not wrapping up the chat at this point. The volunteer counsellor responds by making an educated guess regarding chatter needs -  (“sounds like you need to talk”) -   and a statement regarding the volunteer counsellor‟s emotions (“I am happy to talk to you”) . She also asks the chatter if the feeling of not being worth the volunteer counsellor‟s time is a feeling he has with his therapist.  He states that he and his therapist have talked about this pattern that he has of wanting help but simultaneously resisting it. The volunteer counsellor acknowledges the difficulty of this situation and normalizes the chatter‟s experience: “It‟s funny how the body can get in the way, even when another part of you   84 really wants to be able to talk and explain things.” The chatter seems to react very emotionally to these statements, using the punctuation […] to indicate a silence and then disclosing that he feels like crying. After a pause, he then expresses his emotion by using capitol letters: “ARRRR”! The volunteer counsellor reassures the chatter of his freedom to express himself with a reference to the mode of communication they are using: “you can [cry] if you want to…I can‟t see you!”  The chatter opens up about feeling “sad and small and scared‟ and explains his pauses by telling the volunteer counsellor “...i‟m crying now (just so you know if I‟m not typing).” The volunteer counsellor then expresses her appreciation of the chatter‟s decision to log on and allows the chatter the space to express his emotions (“its ok ... you can just take your time”).  At several points, the youth draws attention to the way in which he communicates, checking for clarification:  “I bet you can hear more than just the words I type.”  The counsellor replies: “When you capitalize and so on it helps me get some idea of the emotions, and how out of control things feel for you” to which the chatter responds “that helps to know that.” The chatter makes several remarks indicating suicidal ideation, but afterwards explains to the volunteer counsellor that he is not planning on acting on them - that it helps him feel better just to express these thoughts. In this phase, the most common volunteer counsellor skill used is reflection of emotion (18 times), followed by clarifying questions and statements (12) and educated guesses (10). Informal language is also used fairly frequently (six times). Thus, the predominant theme in this chat appears to be focused on the volunteer counsellor attempting to understand and reflect the emotional experience of the chatter, which was considered part of the connection- building dimension of rapport.   85 Termination phase This termination phase is initiated by the chatter, who asks the volunteer counsellor for suggestions with respect to things he could do to cope after the chat. The volunteer counsellor points the chatter towards stress management exercises on the YIBC site. The volunteer counsellor and the chatter go through these together, with the volunteer counsellor providing information and guidance regarding the breathing exercises on the site. The chatter expresses that he finds this exercise helpful and that he has printed off a copy to look at “when things get ugly.” The volunteer counsellor then begins to summarize the chat, and the chatter then jumps in with “I‟ll be ok. If you‟re looking for a way to end the chat I‟ll be ok.”  The volunteer counsellor responds to this by asking the chatter to think of some possible ways to cope after the chat, and this is explored with the chatter providing suggestions, with the volunteer‟s prompting. The volunteer counsellor lets the chatter know he is able to access the services if he needs to and advises him of the availability of the service as well as the option of calling in to the Crisis Centre if he needs to as YIBC shuts down at 1am. The chatter indicates that a personal connection has been made by asking “if i call and you are the one answering, can you tell me we`ve spoken?” The volunteer counsellor assents and the chat concludes with the chatter thanking the volunteer counsellor. The action-oriented type of response is the most salient in this chat.  3.1.9 Case #9 This case consists of an 18-year-old male from a Canadian city who is accessing the services for his suicidal friend. He is chatting with a female counsellor who has had more than 100 hours of experience on the crisis line. He is classified as a returning chatter. The   86 chat lasts a total of 66 minutes with 26 minutes for the initial contact phase, 42 minutes for the suicide assessment phase, and two minutes for the termination phase. Initial contact phase  This chat begins with the volunteer counsellor asking an open-ended question with respect to the chatter‟s current emotional state. The volunteer counsellor uses reflective statements to respond to the chatter‟s disclosed emotions. The chatter also lets the volunteer counsellor know what prompted him to log onto the site: he had a previous conversation with a volunteer counsellor, who told him he could log on if he needed support.  This opens the conversation up and allows the volunteer counsellor to inquire about what type of support the chatter is searching for. The chatter discloses that he is worried about his friend who is suicidal. The volunteer counsellor reframes the chatter‟s emotions by emphasizing the importance of his friend to him and intersperses reflective statements with questions about what the chatter has done about the situation thus far (e.g. talked openly with his friend about suicide). The volunteer counsellor also makes an educated guess with respect to the chatter‟s goals. The chatter then discloses that he believes if he loses his friend, he will not be able to live without her. The volunteer counsellor reflects this and clarifies that this is what the chatter means. However, she stays away from going into a suicide assessment; instead, she asks the chatter what an ideal situation would look like. The chatter states that a perfect situation would be his friend not being suicidal. The chat then moves into the Risk Assessment phase. The most frequently used type of response in this phase is the action-oriented one.     87 Suicide assessment phase This phase is initiated by the volunteer counsellor who provides a context for the question by referring back to a statement that the volunteer counsellor made earlier: “You said earlier that you will kill yourself if your friend does. I'm wondering how certain you are about that? For example, if your friend kills herself, will you definitely kill yourself?” She also checks in with respect to the chatter‟s ambivalence: “Or are you unsure?”  The chatter responds by stating quite honestly that at the moment, he is experiencing suicidal thoughts and that he “knows he will” end his life if his friend does.  The volunteer counsellor explore supports and coping and expresses concern for the chatter: “ I‟m worried about you.” The chatter further expands on his feelings, stating that he cannot imagine life without his friend. He shares that he himself was suicidal in the past and that his friend was the one who got him through it.   The volunteer counsellor responds by reflecting this back and asking the chatter if he would consider logging on to YIBC if he thinks he may act on those thoughts.  She also lets the chatter know she is happy he feels comfortable using the service. The chatter responds with “Ya. I‟ll prob come on and chat.” The volunteer counsellor expresses her happiness at hearing this, and also advises the chatter to let his suicidal friend know about the service. The chatter states he has already told her about YIBC and that he has suggested that they both go online and chat with a volunteer counsellor simultaneously. The volunteer counsellor responds by explaining that the nature of the service prevents volunteer counsellor simultaneous conversations but states that both the chatter and his friend are welcome to log on separately. The volunteer counsellor also lets the chatter know that in the case of an imminent risk, 911 is an option for both himself and his friend. The chat then moves into   88 the Termination stage. Action-oriented responses are the most frequently used in this phase; however, connection-building ones are also very much present. Termination phase  The volunteer counsellor initiates the end of this chat by stating that he will have to end the chat in a few minutes. She asks if the chatter has plans for the rest of the evening. The chatter states that he will “just be chillin‟ in.” The volunteer counsellor reflects this and reassures the chatter that both he and his friend are welcome to come and chat if they need to. The chatter thanks the volunteer counsellor for listening and the volunteer counsellor responds with “No problem at all, that's why we're here. Take care now, bye!” This concludes the chat. Connection-building responses are the most frequently used in this phase. 3.2 Observed patterns and themes In this section, themes and patterns arising from the narrative descriptions above are examined. Following the method used by Stake (1995), the findings in this chapter are presented below as they relate to the research questions proposed by the study. It was found that certain specific patterns with respect to counsellor responses and salient dimensions of rapport existed in each phase, and it was hoped that an exploration of these patterns would add a broader context to the narrative descriptions. Patterns common to each phase are outlined below.  Research question #4 (“How does the counsellor maintain rapport when threats to rapport occur?”) is applied to all three phases separately. Following this, themes and patterns common to all phases are described.     89 Research Question # 1: How is rapport built in the initial stages of the chat? 3.2.1 Initial contact phase The timing of the initial contact phase ranged between two and 26 minutes, with a mean time of 13.55 minutes. Every volunteer counsellor began the chat with an open-ended question regarding the chatter‟s current emotional state. An open-ended question was defined in this study as a question to which the answer cannot be “yes” or “no.” An example of an open-ended question would be “What have you done in the past to cope with these feelings?”  Open-ended questions and tentative language (e.g. “I‟m guessing you‟re feeling scared”) were used frequently in this phase, their purpose appearing to be to show understanding and build rapport. The tentative language seemed to allow the chatter space to elaborate or clarify, which generally occurred after their use.  Clarifying questions and educated guesses were used mostly in cases where the chatter was reticent or responded with one-word answers. They appeared to help build rapport; for example, in Chat # 2, the chatter initially responded to the volunteer counsellor with terse, one- word answers; however, she gradually elaborated and began to discharge more emotionally-charged information as the volunteer counsellor continued to reflect and ask open-ended questions. As well, the volunteer counsellor frequently paused after making a statement or question to allow the chatter time to answer; in addition, he or she often used tentative language or phrased her reflections as “guesses.”  In instances where the chatter was new to the service, discussions around confidentiality occurred. In several of the chats (for example, chat #5), the chatter disclosed or hinted at suicidality, and the volunteer counsellor, rather than moving into a full risk assessment,  tended to simply use empathy and reflective statements in this stage.   90 Reflection of emotion was the most common counsellor response in the initial contact phase. Acknowledgment of the difficulty of the chatter‟s situation was also a common theme; for example, it was the most frequent response in Chats #6 and # 5. These two responses were considered to represent the connection-building type of response.  Connection-building responses were the most common type (dominating five out of the nine chats). Reframing of chatter values and goals (considered to represent the action-oriented type of response) was also frequently done in this phase and tended to occur just prior to the start of the Risk Assessment phase. Action-oriented responses were dominant in four out of the nine chats. Most notably, the volunteer counsellor frequently reframed the chatter‟s statements as goals, and in several chats, youth and volunteer counsellor were mutually engaged in working to find options to help the chatter cope without cutting. The volunteer counsellor often made an educated guess with respect to the chatter‟s goals in this phase, which appeared to have the effect of focusing and containing the chat and providing a framework within which to work, particularly if the chatter presented as reticent. Connection-building responses were not only used in the form of reflective statements: in several chats, the volunteer counsellor frequently reframed the chatter‟s statements as values (e.g. “sounds like it‟s really important to you to have someone respect you”), which appeared to have the effect of allowing the youth to see their statements in a new way, perhaps reflecting the insight-oriented dimension of rapport as per Bachelor (1995).      91  Threats to rapport: Initial contact phase When the volunteer counsellor used closed questions (questions to which there could only be a “yes” or “no” answer such as “did something happen today to cause you to feel suicidal?”), this appeared to stall the progress of the chat as the chatter tended to simply answer “yes” or “no,” which in turn seemed to prompt the volunteer counsellor to ask more questions; thus, this portion of the chat ended up looking more like an interrogation. In contrast, using open-ended questions appeared to facilitate dialogue on the part of the chatter, as noted above. A common counselling intervention (the miracle question) seemed to fall flat in chat # 5. The volunteer counsellor asked the chatter to imagine what a perfect situation would look like; the chatter responded that he would like his friend to not be suicidal. The volunteer counsellor reflected this and the chatter appeared to feel understood; however, she expressed frustration at not being able to change the situation. The volunteer counsellor simply reflected this frustration, which led the chatter to revert back to one-word answers. This appears to reflect some criticisms voiced by youth regarding the YIBC site: one comment reads “I think it‟s great you guys try to figure out how we are feeling…but telling us how we feel doesn‟t help…for example saying „so you‟re feeling lonely‟ ” (Vancouver Crisis Centre, 2009). Another way in which rapport was challenged in the beginning of one chat was when the youth asked the volunteer counsellor about her training. The volunteer counsellor responded,  “I‟m sensing you‟re concerned about our level of training, but we are trained and if you and I come upon an issue I‟m not familiar with or that a pro should be dealing, with then I would ask if you‟d like me to find you a resource to somewhere like that. But I‟ll do my best ”. The youth responded to this statement by opening up   92 about her concerns (“ok well i feel so helpless and lost right now”), indicating that rapport had been re-established.  Research question #2:  How is rapport maintained during and after a disclosure of suicidality?    3.2.2 Suicide assessment phase  With respect to the timing of the risk assessment phase, the range was 18-55 minutes, with a mean time of 29.40 minutes. It was the longest of the three phases. In every chat but one (chat #6), the conversation on suicide was initiated by the volunteer counsellor. This phase, due to the nature of its content (suicidal ideation) contained many questions; this was the most common type of counsellor response in this phase across all chats. A common theme with respect to how these questions were asked in these rapport- containing chats concerns both the timing and content of the questions. Often, the chatter would disclose suicidal thoughts in the Initial Contact phase of the chat; however, the volunteer counsellor would not immediately begin the risk assessment and would instead use empathy and reflective statements. He or she then tended re-introduce the issue in the Risk Assessment portion of the chat, usually providing a context for asking the question by referring to an earlier statement made by the youth. In addition, the volunteer counsellor often paused between questions for several seconds to allow the youth to respond. Risk assessment questions were frequently interspersed with both reflective and normalizing statements.   In almost half the chats (four out of nine), the chatter was engaging in self- harming behavior as a coping mechanism. In all of these chats, the volunteer counsellor showed acceptance of cutting as a coping mechanism and attempted to engage the chatter   93 in exploring other options for coping. The crisis volunteer counsellor frequently utilized the counselling intervention of normalization in the chat, both with respect to coping and when faced with chatter reticence to disclose information. External supports such as counsellors or family members were commonly explored by the volunteer counsellor in this phase. Both goal and task components of working alliance as identified by Bordin (1979) seem salient here; the volunteer counsellor frequently engaged the youth in safety planning (task) to keep the chatter from acting on the suicidal feelings (goal). Hence, the action-oriented dimension of rapport was most salient in this phase. Specific ways in which the volunteer counsellor engaged the chatter included providing a rationale for asking the question (e.g., “I am concerned for you […]”). Also, when the volunteer counsellor made direct requests of the chatter (e.g., “Would you put the pills in another room?”), the question tended to be asked in a manner that allowed the chatter the autonomy to disagree (for example, “How might it be for you to put the pills away?”). Although action-oriented responses dominated in four out of the nine chats, connection- building statements were used often in conjunction with the action-oriented responses; they also dominated three out of the nine chats.  In two of the chats, action-oriented and connection-building responses were used an equal number of times. Interestingly, many informal statements were made in this phase - so many that it was determined that a separate category be created to capture their use. Informal statements were defined for this study as language not normally used in formal written communication, including expressions such as “kinda,” “sorta,” “umm,” and “hey.” Agreement was sought between raters as to the inclusion of statements in this category. Informal language was considered a separate category from online-specific language such as emoticons and abbreviations,   94 since it is not specific to online communication (see Appendix D). More informal statements occurred in the suicide assessment phase than in either of the other two phases, although it is of course possible that this was due to the Risk Assessment phase being longer in average duration than the other two phases.  Threats to rapport: Risk assessment phase A potential threat to rapport occurred in Chat #1 when the chatter challenged the volunteer counsellor with respect to her cutting behavior. The chatter had shared that cutting was her preferred method of coping and that she had tried to cope in other ways but been unsuccessful. The volunteer counsellor acknowledged and reflected this statement back to the chatter.  It seemed that rapport was being established, with the chatter spontaneously stating “This helps a lot, having someone to talk to.” It appeared jarring, then, when the chatter suddenly asked the volunteer counsellor “Can I cut now?” The volunteer counsellor, as noted in the narrative description of the chat, first clarified the question (“Um, are you asking me if you can cut?”), then answered by honestly stating his limitations “Well honestly I can‟t stop you by reaching through the monitor” and further asked: “What would you tell a friend in this situation?” The youth seemed actively engaged judging by her response to this question: “I would tell the friend I am worried […] that you could really cut and injure deeply so please don‟t do it it‟s not worth it.” Another instance in which rapport was threatened was in Chat # 4, when the volunteer counsellor attempted to engage the chatter in brainstorming coping options: she suggested: “Okay, let‟s work together to keep you safe tonight.” The chatter countered with “Why should I stay safe?” The volunteer counsellor responded by clarifying what   95 she meant by “staying safe”; she also pointed out an inconsistency in the chatter‟s communication: “By staying safe I mean keeping you alive, it sounded like that is something you wanted as well.”  From here, as noted in the analysis of Chat 4, the chatter disclosed her ambivalence: “I halfway want to stay safe,” which the volunteer counsellor normalized: (“Mixed feelings are common”); hence, it appeared that rapport had been reestablished. Furthermore, questions asked following this exchange seemed to allow the youth to exercise her autonomy; for example, when asking the chatter to brainstorm options for coping, the volunteer referred to a coping strategy mentioned by the youth (“you mentioned distraction”). Later, she prompted the youth to identify things that made living worthwhile for her. Thus, it appeared that the volunteer counsellor had responded to the cues of the chatter by normalizing her mixed feelings; she then engaged the chatter in action-planning on the chatter‟s terms. Research Question #3: How does the counsellor manage chat termination while maintaining rapport?  3.2.3 Termination phase The timing of the Termination phase varied greatly: A range of 1-31 minutes was found, with a mean time of 9.33 minutes. This was the shortest of all three phases. In terms of who initiated the termination, in four out of the nine chats it was the youth; in five chats it was the volunteer counsellor; therefore, it was not consistently one or the other who decided when the chat would end, and who ended the chat did not appear to impact rapport. In each Termination phase, the volunteer counsellor asked how the chatter would cope; in addition, emoticons, most notably, smiley faces () were used by both chatter and volunteer counsellor. The pattern appeared to be that the chatter would use an   96 emoticon and the volunteer counsellor would mirror it. Expressions of care and support were also very common and occurred in all of the chats, as did references to the availability of both and the Crisis Centre‟s 24 hour distress line.  Most of the chats also included a reference to a no-harm agreement (i.e., the chatter agreed to contact the Crisis Centre if they felt like acting on their suicidal thoughts); the no-harm agreement was frequently re-iterated by the volunteer counsellor in this phase. Frequently, the volunteer counsellor summarized what had transpired in the body of the chat prior to making a statement that the chat was about to end. Connection-building and action-oriented responses were used with equal frequency. Threats to rapport: Termination phase In one of the chats (Chat # 8), there was an observed difficulty in maintaining rapport during the Termination phase. The Termination phase lasted 31 minutes as a result of a back-and-forth exchange between the chatter and the volunteer counsellor where the chatter repeatedly asked the volunteer counsellor to stay on the chat. In fact, earlier during the Risk Assessment phase of this chat, he openly challenged the volunteer counsellor by stating: “How come you‟re letting me talk? How come you‟re not talking about ending the chat?” He also used what could be perceived as sarcasm: “Don‟t worry, with my psychiatrist it‟s 59 minutes 59 seconds, I‟m used to it.”  The volunteer counsellor expressed her willingness to stay on the chat with the youth and her concern for him.  Interestingly, the termination phase of this chat had initially begun with the chatter initiating termination; he asked the volunteer counsellor for suggestions for coping after the chat, and the volunteer counsellor referred him to breathing exercises on the YIBC website. The chatter appeared appreciative of this, but when the volunteer   97 counsellor began to summarize the chat, the chatter then jumped in, interrupting the volunteer counsellor by stating “I‟ll be ok. If you‟re looking for a way to end the chat, I‟ll be ok.” Rather than taking this at face value, the volunteer counsellor asked the chatter to brainstorm some options for what to do after the chat. Hence, she did not allow the chatter to alienate her through his use of sarcasm; rather, she seemed to maintain a connection by engaging the chatter in a joint goal, which he cooperated with. The chat ended with the chatter sharing his ideas for how he planned to cope after the chat and letting the volunteer counsellor know he would reach out through either YIBC or the crisis centre phone line. It appeared that a personal connection had been made between chatter and volunteer counsellor, since at the very end of the chat, the chatter asked the volunteer counsellor if he would be the one answering the phone if he were to call.               98  Chapter  4.   Discussion In this chapter, the patterns observed in the preceding chapter are examined with respect to how they relate to existing research. Patterns across cases are examined here, as are patterns across phases. Explanations for the observed patterns are put forward; following this, limitations of the study as well as implications for practice and suggestions for future research are discussed.  4.1 Initial contact phase Interestingly, in three of the nine chats, the chatter initially stated that he or she was looking for support for a suicidal friend; later in the chat, the chatter themselves opened up about their own suicidal feelings. This finding that youth initially framed their purpose for contacting YIBC as seeking help for a third party is understandable in light of the sensitive nature of suicidality and the stigma that often exists surrounding its admission (Evans, 2006). The initial contact phase takes on additional importance considering that it is in this phase that the volunteer counsellor must build sufficient trust for the youth to feel safe opening up about their own suicidality. It may be that the youth initially disclosed thoughts regarding their friend‟s suicidality to “test out” the volunteer counsellor‟s response.   A pattern was observed that later in the chat (usually after 10-15 minutes), the chatter would disclose their own suicidality, perhaps once they perceived it as “safe.” One way in which safety appeared to be created by the volunteer counsellor was through transforming the emotional content that the youth brought into less threatening “pictures;” namely, through use of metaphor and simile.     99  4.1.1 Use of metaphor Interestingly, it was found that in a number of chats during the initial contact phase (most notably, chats # 2 and #8), metaphor and simile, both forms of figurative language, were used by both the volunteer counsellor and the youth.  The initial contact was the only phase in which use of figurative language occurred. The pattern appeared to be that the youth would put forward the metaphor or simile (“I feel like a turtle without its shell”), which was reflected and expanded upon by the volunteer counsellor (“So you‟re feeling pretty boxed-in and vulnerable… bet you wish you could just hide under that shell”). Adolescents are notoriously resistant to counselling (Martin, 2003). Developmentally, they are at a crossroads between childhood and adulthood. During this time, conflicts about freedom and choice are often at their most extreme (Martin, 2003). Hence, adolescents may display resistance towards the counselling process. Romig and Gruenke (2001) found that one technique that seemed helpful in dispelling resistant behaviours was to approach the issue in an indirect manner; specifically, by using metaphorical language. In their study of prison inmates, use of metaphor was found to be an effective tool for establishing rapport. One example of this linguistic device that received a positive response in their study was the question, “What day of the week will it be when you put down that rock you‟re carrying on your head?” (Romig & Gruenke, 2001, p. 416). Martin (2003) suggested that when working with adolescents, listening to the verbal and non-verbal communication of the client is of particular importance. He stated that “[…] the bulk of the meaning is carried through tone, expression, context, and, most important, what is implied by the words used” (Martin, 2003, p. 2). Martin further stated   100 that approaching the adolescent client‟s emotional concerns indirectly may feel safer to the adolescent and can help facilitate rapport. He specifically recommended the use of metaphor as an effective example of this indirect approach, explaining that, developmentally, adolescents have an easier time understanding concrete images than abstract psychological concepts (Martin, 2003). Several additional reasons for the effectiveness of metaphors have been cited, which seem to be directly applicable to adolescent counselling. For example, it has been demonstrated that metaphoric language, by virtue of its indirect nature, enables the client to feel more in control of the therapeutic exchange. Thus, using metaphor may address the issues of power and control that are often so salient in adolescence (Martin, 2003), as the client does not feel coerced or pressured into sharing intense personal information. Romig and Gruenke (2001) found that the use of metaphor is particularly useful in helping people express “volatile or highly painful emotional issues” (p. 416). This finding would indicate that using metaphorical language may be especially relevant for youth accessing online crisis counselling resources. These youth usually spontaneously log on, as opposed to having a scheduled meeting as is normal in face-to-face counselling, and often present as in crisis or emotionally distraught.  In fact, two of the most thoroughly explored metaphors in this study are images of vulnerability: one chatter expressed that he felt “like a turtle without its shell” and another stated that she was afraid to talk about her painful emotions in case “the dam burst[s] through” and she was “flooded.” The Initial Contact phase is arguably the phase in which the youth is the most vulnerable, as it is the phase when they are most likely deciding whether the volunteer counsellor is to be trusted. Hence, in the chats examined, the youth were likely making decisions regarding   101 how much or how little to share with the crisis counsellor in this phase. This process of trust-building is likely present throughout all phases of the chat; however, it is likely that the newness of the relationship between chatter and volunteer in the initial contact phase would render trust-building particularly important. The online nature of the communication increases the likelihood of misunderstandings arising between youth and volunteer counsellor.   Expanding the expressed metaphor and reflecting it back to the youth, which the volunteer counsellors did in the two cases described, may have  helped to create a common ground for the adolescent and the counsellor, by meeting them “within their frame of reference” (Romig & Gruenke, 2001, p. 414). 4.1.2 Reflection and clarification Asking questions with respect to the chatter‟s experience and asking for clarification occurred frequently in the initial contact stage.  Reflection of emotion and reframing of chatter values were also often used in this phase. Both these skills appeared to facilitate rapport through allowing the chatter to feel heard and understood, as evidenced by chatter statements such as “exactly” or “thanks for understanding.” It appeared that tentatively phrased reflections or those framed as educated guesses were best received by the youth; simple reflections that echoed the youth‟s words appeared to have an adverse effect on rapport as evidenced by short, one-word answers or silence.  Reflection and clarification- seeking may have contributed to rapport by allowing the chatter to experience a feeling that he or she was important; that his or her concerns were valid, since feeling “misunderstood” is a very common experience for youth (Martin, 2003).  In addition, it has been frequently noted that one of the drawbacks of online communication is a difficulty in getting one‟s message across in the absence of voice tone; it has been found   102 that this can impact the encoding and decoding of the message (Weiten, 2009). Clarification and using tentative language may have helped to facilitate rapport in this initial phase of the chat by allowing space for the chatter to clarify or correct the counsellor‟s statements, thus engaging them more actively in the chat. The importance of using tentative language and asking for clarification when communicating online has been highlighted by Williams, Bambling, King and Abbott (2009) who noted that counsellors‟ behaviours can often be misinterpreted in an online environment, and by Bambling, King, Reid and Wegner. (2008) who found that misunderstanding a counsellor‟s empathic statements was one of the main sources of dissatisfaction among youth who were consumers of an online crisis service. In this study, it was noted in several chats that the volunteer counsellor would not immediately respond to the chatter‟s suicidal disclosure during the initial contact phase. Bryan and Rudd (2006) noted that when conducting a suicide risk assessment, it is recommended that the clinician begin by asking more innocuous questions first (such as questions regarding current functioning), gradually leading into the more “serious” questions such as those regarding timing of suicide plan, method, and previous attempts. This seems even more relevant when considering that the lack of voice-tone can impede the progress of rapport-building in online counselling. 4.1.3 Dimensions of rapport  The volunteer counsellor‟s use of more connection-building responses than action- oriented ones in the initial contact phase is consistent with the research on in-person counselling for suicidal clients (Granello, 2010a; Granello, 2010b). This finding is also consistent with the findings of Williams, Bambling, King and Abbott (2009), who found   103 that online crisis counsellors used more rapport-building responses than task-focused ones; however, it must be noted that Williams et al. examined the whole body of the chat rather than breaking it down into discrete components; also, they did not focus solely on clients presenting with suicidal ideation. Therefore, their findings and how they relate to those of the current study must be viewed in light of this fact. As previously mentioned, building rapport with youth is a notoriously difficult undertaking even in a “normal” counselling context.  Attempting to build trust online and when sensitive subjects such as suicidality are addressed adds to this challenge. Thus, it makes sense that the first part of the chat would be taken up with building a connection prior to engaging the youth in action-planning. In addition, suggesting or attempting to imply “what to do” early on in the chat may have a negative impact on the youth‟s sense of autonomy which is an important developmental task in adolescence (Martin, 2003). However, there were Initial Contact phases in which action-oriented responses were more dominantly used. These tended to be instances where the chatter was engaged in potentially dangerous coping behavior such as cutting or contemplating taking pills, or instances where the chatter was reticent and was having a hard time getting started.  In the former case, determining the chatter‟s current safety took precedence over building rapport; in the latter, it appeared that focusing on goals for the chat was a “safer” way to begin than discussing feelings. When action-oriented responses were used in this phase as a reaction to chatter‟s presentation as reticent; the volunteer counsellor tended to use open and closed questions to draw out the most important things to the chatter and build a foundation for the work that would follow.   104   As stated, the timing of the initial contact phase was between two and 26 minutes, with a mean time of 13.55 minutes. A number of explanations are possible for this large range.  The Initial Contact phase may be shorter for some chatters who are familiar with the YIBC service. It is also possible that the two chats with a very short Initial Contact phase are examples of the anonymous nature of online contact, which appears to motivate chatters to be more truthful in a shorter period of time, as research has indeed found to be the case (Whitlock, Powers, & Eckenrode, 2006). 4.2 Suicide assessment phase The suicide assessment phase was the longest phase of the three examined, which makes sense in light of the detailed nature of a suicide risk assessment as well as the sensitive nature of suicidality – in most of the chats, there was some hesitance with respect to disclosing information in this phase. In every chat but one, the conversation about suicide was initiated by the volunteer counsellor. This finding is perhaps a reflection of the stigma around mental health in general and suicide in particular (Evans, 2005). Since in this is phase, many sensitive questions (regarding suicidal plan, intent, previous history, and so on) are asked as part of the risk assessment process, the way in which these questions were asked appeared to impact rapport. In the risk assessment phase, asking questions, especially sensitive ones, in rapid succession may have negatively impacted rapport by making the chatter feel bombarded, causing them to shut down, as evidenced by short, one-word replies to the risk assessment questions in chats where this occurred.   With respect to in-person counselling, Granello (2010a) created a list of 25 strategies for clinicians working with suicidal clients. He noted that in the first few stages of suicide assessment, it is advisable to “slow things down,” “create a therapeutic   105 window” and “stay with the client” (p.220). These three suggestions all seem to contraindicate the asking of questions in rapid succession. With respect to online counselling, “softening” the asking of questions by interspersing them with caring statements  becomes even more important, since the potential for miscommunication or the counsellor sounding overly “clinical” is greater considering the absence of voice tone, which has been observed by researchers in the field as well as by youth (Williams, Bambling, King & Abbott, 2009). A common theme in this phase was for sensitive questions to be interspersed with reflective statements. It appeared that in this phase, reflection of the chatter‟s emotional experience may have contributed to rapport by “taking the edge off” the somewhat impersonal, clinical-sounding questions. In addition, these reflections implicitly provided a rationale for the volunteer counsellor‟s questions. The volunteer counsellor frequently used normalizing statements interspersed with questions in this phase.  It is likely that this would have contributed to developing rapport, in particular considering a developmental perspective since much research indicates that many young people, who are in the process of forming an identity, may feel they are “abnormal” when experiencing suicidal thoughts (Evans, 2005). 4.2.1 Respecting autonomy In a number of chats, the volunteer counsellor prefaced risk assessment questions with the phrase “if you feel comfortable,” which likely enabled the youth to feel more comfortable disclosing personal or sensitive information by allowing them to experience a sense of autonomy. Much research on adolescent development stresses the importance of perceived autonomy and self-determination for youth. Martin (2003) noted the   106 particular importance of respecting autonomy when counselling adolescents, and it appears that this is relevant when counsellors are attempting to maintain rapport online. As mentioned earlier, any perceived threat to the client‟s autonomy can impact the therapeutic alliance, particularly if that client is an adolescent. There are many instances in the Suicide Assessment phase in which the crisis counsellor appeared to actively promote the youth‟s autonomy. For example, engaging the youth in safety planning by asking them to brainstorm options, showing acceptance of coping methods (including those considered “risky” such as cutting); and asking their opinion on matters (for example, asking the chatter what they might tell a friend in their situation) could all be viewed as autonomy-enhancing interventions. Using tentative language and phrasing suggestions as questions is another way the youth‟s autonomy appeared to have been encouraged by the volunteer counsellor. 4.2.2 Informal language Despite the serious nature of the content of suicide assessment, many informal statements were made in this phase.  In fact, they were used so frequently that a separate category was created to capture their use. This finding is interesting in light of the prediction that these would be most present at the beginning (initial contact) and end (termination) of the chat, given the serious nature of suicide. Informal statements may have helped to build rapport by mimicking the types of exchanges youth have with their peers. Geldard and Patton (2007) found similar results when they examined adolescents‟ perspectives on helping/hindering behaviours of their counsellors: “conversational” behaviours were viewed by adolescents as helpful in the counselling relationship.  Abbreviations were also used frequently in the Suicide Assessment phase. The use of abbreviations may indicate a   107 level of comfort and an assumption that the volunteer counsellor is sufficiently similar to the youth that they would use the same linguistic conventions. Use of informal language as well as computer-specific language such as abbreviations and emoticons, seemed to help facilitate rapport, perhaps by adding an element of the “real relationship,” as it may have seemed that the counsellor presented  him or herself as on the same level as the youth. Interestingly, despite the finding of Bambling et al. (2008) regarding use of icons and scales to help the client identify the intensity of their emotional experience, there were no scaling questions asked in this phase or in the other two phases. Since Bambling et al. (2008) examined a broader range of chats in which suicide was not the only presenting concern, it is possible that scaling questions were more commonly used in chats that were not as high risk, since scaling questions (e.g. “on a scale of 1-10, how much emotional pain are you in?”) may be perceived as overly clinical when interacting with a suicidal adolescent, especially given the impersonal nature of online communication.  4.2.3 Real relationship When threats to rapport occurred in the suicide assessment phase, such as the instance where the youth challenged the volunteer counsellor by asking “Can I cut now?”, the volunteer counsellor first responded with “Um, are you asking me if you can cut?” and followed this with “Well honestly I can‟t stop you by reaching through the monitor.” The response above appears to reflect elements of the Real Relationship as characterized by Gelso (2005): the volunteer counsellor was expressing genuine surprise as well as an informal, honest admission of his inability to physically restrain the chatter. In their analysis of online crisis chats received by the Kids Helpline in Australia, Bambling et al.   108 (2008) found that directly addressing miscommunications was cited by the crisis counsellors as an important skill to develop. It appears that this skill was valuable with respect to the findings of the current study, particularly in instances where rapport was threatened. In addition to honestly expressing his limitations, in this same chat, the volunteer counsellor also deflected the question “Can I cut now?” by asking the chatter what they would tell a friend. This strategy would likely be useful with youth given that friends are an important part of a young person‟s identity (Martin, 2003). In addition, this response stayed away from moralistic or judgmental responses on the part of the volunteer counsellor, which would likely have alienated the youth (Martin, 2003). The research on youth counselling indicates that approaching potentially sensitive topics from an emotional distance; for example, asking the youth to share how they think their family dog might feel about their parents‟ divorce, can help them to respond without feeling vulnerable (Goldenberg & Goldenberg, 2008). In this chat, asking the youth what she might tell a friend in this situation also appeared to empower the youth to be proactive in the counselling process. One interesting technique used by the volunteer counsellor in Chat # 3 was an informal acknowledgement of possible chatter resistance to the volunteer counsellor‟s suggestion of finding a support group. Specifically, the volunteer counsellor prefaced his suggestion with “this might sound kind of lame, but some people find it helpful […].” This response appears once more to be an example of the volunteer counsellor showing genuineness or “being real” in the relationship. In addition to acknowledging possible resistance, this response also provided a rationale for the volunteer counsellor‟s   109 suggestion (“some people find it helpful”) and normalized the chatter‟s experience as well by implying that he was not the first person to show resistance to advice. 4.2.4 Dimensions of rapport Considering that the suicide assessment phase involved the disclosure and subsequent assessment of the chatter‟s suicidality, it is not surprising that the action-oriented type of intervention was most salient in this portion of the chat. It has been noted in the research on suicide counselling that “[suicidal] clients require not only empathic attention, but also directive management” (Neimeyer & Pfeiffer, 1994, p. 138).  As mentioned earlier, questions were frequently asked in this phase.  Not surprisingly, a majority of these questions were concerned with suicide (plan, intent, means, and previous history of suicidality). Clarifying questions were also frequently used. These questions may have allowed the youth to feel that they were important, that the volunteer counsellor truly cared about understanding what they had to say.  Clarifying questions appeared to help engage the youth in the counselling process. However, it is noted that in this phase, connection-building interventions were used with almost as much frequency as action- oriented ones (they dominated three out of nine chats, and in two chats, their use was equal with action-oriented interventions). The trend appeared to be to “sandwich”  action- oriented items (“how would it be for you to put the pills away?”) with connection- building interventions such as reflective statements (“sounds like this is a really hard night for you”). This finding is consistent with research on suicide assessment which indicates that the role of emotion-focused empathic reflections should not be underestimated, and that crisis line volunteer counsellors should first be trained in basic   110 empathy before any formal risk assessment is taught (e.g., Granello, 2010a; Hoff, 2009; Tomlin, 2009). 4.3 Termination phase 4.3.1 Online linguistic conventions Emoticons and abbreviations were used more often in the termination phase than in the other two phases. The manner in which they were used seems to be consistent with literature on rapport-building in counselling:  as stated earlier, mimicking usage of nonverbal as well as verbal client language can build rapport (Trout & Rosenfeld, 1980, cited in Sandhu et al., 1993), and emoticons are considered to be a form of non-verbal communication (Shao-Kang, 2008).  In the termination phase, the volunteer counsellor tended to only use an emoticon or abbreviation after a youth had used it first, hence “mirroring” the youth‟s language pattern. Goldsborough (2008) noted that in online communication, people tend to use emoticons to “show that they are part of an in-group” ( p. 16). Since people tend to like those who are similar to them (Weiten et al., 2009), it is likely that mirroring emoticon use would serve to enhance rapport. It was initially expected that emoticon use would be found more frequently in the initial contact phase; however, this was not found to be the case. One possible explanation for this is the observation by analysts of online communication that emoticons may be ineffectively used “[...] as with jargon to show people they are part of the in-group […] which just labels them as „newbies‟ ” (Goldsborough 2009, p. 16). Using emoticons early on in the chat, without waiting to observe the youth‟s patterns, may actually adversely affect rapport, as the youth may perceive the counsellor as “trying too hard” which would   111 compromise the genuineness and realism that are part of the Real Relationship (Gelso, 2009). The way in which emoticons were used in the chats examined is also consistent with the research on their use: they appeared to be used towards the end of the chat to emphasize the content of the message or to convey a lightening of mood (such as when the volunteer counsellor apologized for her spelling mistakes). This finding echoes that of Derks, Bos, and von Grumbkow (2008) who noted that the primary use of emoticons in online communication appears to be to convey emotion, reinforce the sender‟s message, and express humour. Interestingly, emoticons appeared to be used to express only positive emotions in these chats; there were no instances of a “sad face” emoticon () or other negative emotion in any of the chats received. This is an interesting finding.  It may be that this use of emoticons reflects the way in which the “sad face” emoticon tends to be used in current online communication. Goldsborough (2008) noted that the most common manner in which emoticons are used is to express humour; he elaborated that negative emoticons are most frequently used to express empathy or disappointment. Thus, it seems that in online communication, negative emoticons tend to be used to describe less intense emotions and are often used to convey a lightening of mood even when discussing an event that would normally generate negative emotion (e.g. “I have a math test tomorrow ”).  Since suicidal youth are often expressing intense negative emotions, using the “unhappy face” emoticon to express this may not be the most appropriate medium. However, youth did seem to use language in a specific way when they wanted to express negative emotion in the chats examined in this study: capitalizing words and overuse of punctuation (e.g., “AHHHHHH!!!!”) seemed to substitute for voice   112 tone when they youth was trying to communicate the intensity of their emotional experience to the counsellor. 4.3.2   Dimensions of rapport In the termination phase, an equal balance of action-oriented and connection-building responses was used. In four of the chats, action-oriented responses were most commonly used; in four chats, connection-building responses were most commonly used, and in one chat, their use was equal. Upon closer examination, it was discovered that even in those chats where one type of response “dominated,” the difference between the use of both types of responses was negligible – they were both strongly represented even in chats where one type of response “dominated.” This finding is interesting in light of the function of the termination phase in crisis counselling. Ending a chat with a youth who is suicidal involves a fine balance between assessing risk and ensuring that the youth possesses the internal and external resources to cope on their own following the chat.  In his guide to working with (in-person) suicidal clients, Granello (2010a) recommended that the final phase of counselling (steps six and seven of his seven-step model) be taken up with what he terms “behavioural strategies” (p. 220). These include suggestions such as “draft [ing] a short-term, positive action plan” and “ensur [ing] follow-up care” (p. 220). Connection-building interventions such as “listen, understand, and validate” (p.220) were suggested for the earlier stages of the therapeutic interaction.   Therefore, the presence of both types of responses in almost equal measure in the Termination phase seems to contradict the research on suicide counselling for in-person clients. There are several possible reasons for this finding, which will now be discussed.   113  When ending a chat online, the volunteer counsellor risks alienating the youth or reinforcing their sense of isolation by initiating the end of a chat.  Youth are often extremely vulnerable to rejection (Martin, 2003) and may perceive the volunteer counsellor‟s attempts to end the chat as such (in fact, this occurred in two of the chats in this data set). Because the interaction is taking place online and because of the short-term, anonymous nature of the service, it is unlikely that the volunteer and youth will “meet” again. The challenge the volunteer counsellor is presented with is to leave the youth with the feeling that he or she has been involved in a meaningful and supportive conversation with the crisis line volunteer, and that he or she possesses the resources to safely cope with his or her intense emotions following the termination of the chat. Hence, the frequent use of connection-building responses could have served to act as reassurance to the chatter that they were important and that the volunteer counsellor was glad they logged on (in fact, these were the two most frequent types of connection-building interventions in this phase). The use of action-oriented responses also followed a specific pattern in the termination phase: responses in this category tended to focus on the youth‟s immediate plans following the chat, as well as on safety planning in the event of imminent suicidality - no-harm agreements and other safety measures were most often set up in this phase. With respect to who took the initiative to end the chat, there was no trend in one direction; the youth initiated the termination four out of nine times, and the volunteer counselor in initiated the remaining five chats. What appears salient here is not who initiated the termination but how it was handled. Threats to rapport in this phase generally involved the youth resisting in some way to the volunteer counsellor‟s decision   114 to end the chat. This was done sometimes with sarcasm (“Don‟t worry if you‟re trying to find a way to end the chat [...] I‟m used to it”) or with anxiety (“Please stay on with me […] for just a few more minutes”). This threat to rapport was generally resolved by the volunteer counsellor engaging the chatter in joint goals (for example, in one chat, the volunteer counsellor asked “What can you do after this chat to relax?”, which countered the youth‟s implied statements that she did not care or was only trying to end the chat to “get rid of him”). In addition, the volunteer counsellor appeared to maintain rapport by offering statements of care or reassurance (“I‟m so glad you decided to reach out and chat with me today”). This balance of action-oriented and connection-building responses appeared to both empower the client to take action and reassure them that the end of the chat did not signal the end of the connection with the service. On this note, in every termination phase, the volunteer counsellor mentioned the availability of the YIBC service and encouraged the chatter to re-access as needed. In instances where the chatter was not from BC, local resources were provided. This may have had the effect of providing a “bridge” for the chatter to plan for the time between the current point of contact and the next couple of days. 4.4 Themes across phases 4.4.1 Mirroring language patterns A theme that was prevalent across phases was the tendency of the volunteer counsellor to mimic the youth‟s conversational patterns, whether this was by reflecting and extending a metaphor put forward by the youth in the initial contact phase, mirroring the informal language of the youth in the risk assessment phase, or echoing the use of emoticons in the termination phase.   115 The findings of the present study seem consistent with research on building rapport in traditional in-person counselling with respect to mirroring language patterns r “matching” (Hackney & Cormier, 2009).  The “chameleon effect” refers to the tendency of humans to mimic the non-verbal behaviour of interaction partners, without being aware of it (Lakin, Jefferis, Cheng, & Chatrand, 2003).  It has been suggested that this behaviour evolved because it helped early humans forge social ties and has been linked to rapport, liking, and feelings of affiliation (Lakin et al., 2003).  It has been found that using similar language patterns can facilitate a stronger bond in romantic relationships (Bower, 2010) and in small groups working together (Gonzales, Hancock, & Pennebaker, 2010).   With respect to the field of counselling, research on this topic for the past 30 years has suggested that mirroring the body language used by the client (also referred to as “matching” (Hackney & Cormier, 2009) can have a strong positive impact on the client/counsellor relationship (Trout & Rosenfeld, 1980, cited in Sandhu et al., 1993). Mirroring clients‟ spoken language has been found to be an effective tool as well – preliminary research by Bandler and Grindler (1979) showed that mirroring not only voice tone, inflection, and speed of speech but also the client‟s “predicates and syntax” (1979, cited in Sandhu et al, 1993, p. 3) can facilitate rapport in the relationship. A later study echoed the finding that mirroring verbal and non-verbal behaviour has a positive effect on the therapeutic relationship. (Storms, 1982, cited in Sandhu et al., 1993).The findings of the present study suggest that the practice of mirroring language patterns to help facilitate rapport may be applicable to online counselling as well. Rather than mirroring tone or inflection, the volunteer counsellor seemed to “match” the youth‟s use   116 of language in the chat; for example, by reflecting the youth‟s figurative language, use of informal language, or use of emoticons. 4.4.2 Threats to rapport: The importance of the real relationship As mentioned, in six of the chats, chatter and volunteer counsellor were of opposite sexes. Interestingly, at no point was the gender difference mentioned in any of the chats. In fact, one of the chats consisted of a female chatter and male volunteer counsellor and the presenting concern was the youth‟s eating disorder. Despite the fact that eating disorders are considered a mainly “female” concern both statistically and anecdotally, there was no mention of the gender difference in this chat, and rapport appeared strong throughout. This is interesting in light of the fact that most research suggests that clients prefer a counsellor who is similar to them on key characteristics including gender (Hackney & Cormier, 2009). It is possible that the acute crisis state made identifying characteristics of the counsellor less relevant; it is also possible that the online nature of the communication rendered such information less important due to the lack of visual cues present; as noted by Becker and Schmidt (2006), a common phenomenon in online chatting is the tendency to imbue one‟s chat partner with personally desirable characteristics. With respect to difficulties establishing or maintaining rapport, the lack of voice tone in online communication seemed to make it challenging for volunteer counsellors to use their usual skills of empathic reflecting in the same way that they would in a phone call. One of the most frequent criticisms of YIBC by youth is the lack of a personal connection due to the communication style of the volunteer. For example, one youth provided the following feedback on the site: “I find the scripted type responses quite   117 frustrating sometimes. I‟d love someone with a bit of personality;” another wrote: “Don‟t sound as…textbook. Tell me what you honestly think of the situation…within reason of course.”  (Vancouver Crisis Centre, 2009). Examples of this experience of youth were found in a number of chats in which the volunteer counsellor simply repeated the content put forward by the youth, changing only a few words.  In the Initial Contact phase, it was noted that this strategy resulted in a breakdown in communication, with the youth responding either with silence or with short, one-word answers. In contrast, directly addressing threats to rapport using informal, colloquial language appeared to be a helpful strategy in maintaining it. The clearest instance of this occurred in response to a youth‟s challenging a volunteer counsellor‟s training and competence. When this occurred, the volunteer counsellor responded with: “I'm sensing you're concerned about our level of training, but we are trained and if you and I come upon an issue I'm not familiar with or that a pro should be dealing, with then I would ask if you'd like me to find you a resource to somewhere like that. But I'll do my best :)”.  This informal, honest, and collaborative response appeared to engage the chatter in the process of counselling as it was action- oriented in nature, appearing to demonstrate the collaborative dimension of rapport as identified by Bachelor (1995), or the task dimension of rapport as identified by Bordin (1979).   Across all phases of the chats, it was found that when the connection between client and counsellor was on shaky ground (for example when the client questioned the service or the counsellor, or expressed themselves using anger or sarcasm), there was a distinct pattern with respect to how the counsellor responded. In the Initial Contact phase, threats to rapport appeared to be mostly expressed as questions about the volunteer   118 counsellor‟s role or capacity (“Are you, like, trained or just a kid like me?”). In the Risk Assessment phase, threats were mostly in the form of reactions to the assessment questions (“Why should I tell you my suicide plan?”) or as resistance towards safety planning (“Why should I stay safe?”) In the termination phase, the most common threat to rapport was chatter resistance to ending the chat (“Can you stay for a few more minutes?”). Regardless of what the threat to rapport entailed, it was found that the counsellor tended to respond using similar types of interventions. Specifically, each time that rapport was threatened, the counsellor responded to this threat in an informal, honest, and collaborative manner. Informal language was used by the volunteer counsellor often in response to the youth‟s use of such language, but not always.  As mentioned previously, the mirroring of youth language appeared to build rapport. It is possible that use of informal language helped defuse potentially conflictual situations by lightening the mood and suggesting a similarity between helper and helpee. The frank, honest nature of the responses appears to mirror the Real Relationship as identified by Gelso (2005). As mentioned, the Real Relationship appeared to make an important contribution to rapport and the research indicates that this type of genuine relationship is especially important when building a connection with youth (Martin, 2003). Feedback from youth on the YIBC site seems to underscore the importance of the online counsellor “being real.” For example, one youth wrote: “Less deflecting on questions, especially when the answer is important […] be more connecting and don‟t ask generic questions”, another wrote: “Give me your opinion on the situation and act like you give a shit. Thanks.” Still another implored the volunteers to “Give honest personal answers please” (Vancouver Crisis Centre, 2009).  In addition to providing a genuine, honest response, focusing on   119 collaboration (by identifying goals such as coping for after the chat) may have alleviated threats to rapport by changing the focus and allowing the chatter to develop a sense of agency. 4.5 Limitations and directions for future research  4.5.1 Sampling procedures The present study used archival methods (existing chat transcripts selected by the Vancouver Crisis Centre program coordinator who was provided with a list of inclusion criteria). While my position at the Crisis Centre enabled me to have access to data that would normally not be accessible to an “outsider,” my role there also prevented me from hand-selecting chats for analysis. Thus, although the possibility of researcher bias was circumvented by this method, the presenting concerns, gender, location, and other information was not controlled by the researchers during the initial selection of 50 chats or the 17 rapport-containing chats, of which eight were eliminated due to the various issues cited above.  The nine chats that were eventually selected from the initial selection of 50 inherently contained evidence of rapport as per the operational definitions agreed upon by the research assistants and me. Thus, they did not contain the challenges to rapport that are representative of those faced by volunteer counsellors on the lines.  On a related note, researchers were cognizant of the fact that what might be perceived as “rapport” could simply be familiarity with the services, and that in the future, distinctions between new and existing users could be made. However it was noted that even chatters who are frequent users can have chats in which no rapport is built; hence, the volunteer   120 counsellors‟ actions appear to have greater bearing on whether rapport is experienced than does the frequency with which the youth access the service.  At this time, the Vancouver Crisis Centre does not distinguish between “new” and “returning” chatters; they only classify a “returning” chatter as such if they have accessed 15 times or more in the last six months. Therefore, even a “new” chatter may have accessed up to 14 times in the last month. The Crisis Centre is in the process of installing a new callsheet recording system that will likely allow changes in the recording process that may enable staff to delineate first-time chatters. 4.5.2 Satisfaction scale Another potential limitation concerns the materials used to determine the inclusion criteria of the study. Specifically, the satisfaction scale used in this evaluation was developed for crisis counselling by phone. Because of the newness of the online service, no research exists with regard to whether the criteria used to assess satisfaction online should differ from those used to assess phone counselling. In addition, the construct validity of the satisfaction scale is not well-established, although it has good face validity and research shows that people are generally able to respond to Likert-type scales in a consistent manner (Palys & Atchison, 2008).  Future studies using alternate measures of satisfaction could be conducted. For example, King et al. (2003), in their evaluation of an Australian suicide helpline for youth, developed a reliable and valid assessment of caller satisfaction/call effectiveness, since they found existing measures to be lacking in validity. This measure, based on the MINI (Mini International Neuropsychiatric Interview), has demonstrated reliability and validity for evaluating crisis phone calls; however, King et al. advise that lay people (e.g. program evaluators, researchers) would   121 require advanced training to be able to utilize this measure; in addition, its use has not been evaluated for online counseling. Hence, it was not deemed realistic for the current evaluation. 4.5.3 Methods As is the case with any research method, the ones chosen for this study (content analysis and the collective case study approach) come with some limitations. Inherent in both of these is the threat of researcher bias: as a result of existing literature, the researchers of the present study (my research supervisor and me) inevitably had expectations with respect to what might be found in the results. These expectations may have impacted the manner in which the data was analyzed, although steps were taken to minimize researcher bias, for example, having two research assistants uninformed of the goals of the study concurrently analyze much of the data. Content analysis, the method used to initially analyze the transcripts, has some limitations. Content analysis by its nature tends to disregard the context of the words that are being analyzed; hence, it is up to the researcher to create meaning from these words. In this study, this was done by creating categories for specific counsellor interventions. The categories used for counsellor interventions in this study were internally used by the Vancouver Crisis Centre to describe volunteer counsellor behaviour on the phone lines. There are two potential problems with this categorization. First, it is not clear whether these categories are also amenable to online counselling; second, choosing categories that were created by the Crisis Centre increased the possibility of finding interventions that fit the parameters of crisis centre training, thereby possibly leaving out other relevant categories of interventions (as mentioned, one chat was eliminated by the program   122 coordinator because it did not conform to Crisis Centre training). However, it was hoped that the use of research assistants and the researchers‟ own expectation that there would likely be alternate types of interventions used would help mitigate this threat of researcher bias. The case study approach, by virtue of its focus on a smaller number of participants, is limited in scope and hence in generalizability to contexts outside the one studied (youth exhibiting suicidal behavior in a synchronous online forum). Also, the case study approach is inherently subjective in nature; since no person viewing the chats would write about them in quite the same way, it is probable that additional unique features or salient themes existed which were not included in the final analysis.  My position at the Vancouver Crisis Centre as a past reviewer of chats may have biased my view of the current chats since I normally viewed them through the lens of a supervisor; i.e., my role was to check for accuracy and ensure that proper risk assessment procedures had been followed. Thus, the examination of the data may have been affected by my past role as a supervisor. However, it was noted that had I not attained this position at the Centre, I would not have had access to this rich data set; hence, this limitation is tempered with gratitude. 4.5.4 Chat analysis  Another limitation concerns the manner in which the chats were examined. The chats were examined by breaking them down into three discrete phases (Initial Contact, Disclosure of Suicidality, and Termination). The reasoning behind this decision was that each “phase” was thought to include different goals and needs for both the volunteer counsellor and the chatter, and thus considered to involve different interventions, which   123 has been found to be the case according the literature on this topic (e.g. Granello, 2010a; Kalafat, 2002; Halderman, Eyman, Kerner, & Schlacks, 2009).  The research questions were concerned with examining how rapport is built in the different phases.  However, we did not look at the chats as a whole, which may have eliminated potentially useful information about patterns and types of rapport occurring throughout the chat from beginning to end. For the purposes and scope of this study, it seemed most feasible to break up the chats in this manner. 4.5.5 Definition of rapport  A potential limitation of this study concerns the manner in which rapport was operationalized. As mentioned earlier, the focus in this study was on youth-perceived rapport, since this was in keeping with the goals of the study; it was considered more important that youth perceived a connection than counsellor, since the service is tailored towards youth. As well, Gould et al. (2007) note that a large body of research indicates that client‟s own ratings have greater validity than those of a third party such as a counsellor.   It has also been noted that when one party in a counselling relationship perceives rapport to exist, the other is likely to perceive this as well (Hackney & Cormier, 2009).  However, in this study, youth were not actually asked what they thought constituted rapport; given the archival nature of this research, what constituted “rapport” had to be inferred from statements made by the youth using pre-determined categories generated by the researchers. This is a limitation that could not be overcome due to the lack of access to the youth themselves as a result of the anonymous nature of the service. The fact that counsellor interventions were grouped into the categories Connection-Building and Action-Oriented may have prevented other types of potentially   124 relevant interventions from being recognized; any time that specific criteria are used to define concepts, the possibility exists that important elements may be neglected due to the narrowness of focus. Thus, future research could examine other ways of delineating the different types of counsellor interventions. On a related note, a threat to the validity of the findings is the fact that this study relied on counsellor evaluations of satisfaction with respect to the satisfaction scale, which was used an one of the inclusion criteria for the study.  As stated previously, the Crisis Centre‟s policies prevent youth responses from being linked to actual chats, so this prevents the use of a satisfaction scale that can be filled out by youth themselves. This prevents a triangulation of data sources which Stake (1995) noted is a helpful process in increasing the validity of case study research.  However, it was anticipated that the requirement of inter-rater agreement (multiple reviewers assessing chatter satisfaction based on the transcript) would help mitigate this threat. 4.6 Directions for future research Future research could examine rapport-containing chats involving youth with other presenting concerns (for example anxiety, depression, bullying) to examine whether the types of interventions used by counsellors are similar to those used in chats where the youth is suicidal. In addition, non-rapport containing chats could be compared with the rapport-containing chats used in the present analysis and a comparative study could be done to observe the differences between counsellor interventions used. Another avenue of comparative exploration could be to conduct a comparison between transcripts of in-person counselling sessions with suicidal youth and the transcripts obtained in this study; this would likely provide   125 researchers with valuable information regarding differences in communication and counselling interventions. Follow-up studies could examine the generalizability of these findings to other presenting concerns; for example, for non-suicidal youth in crisis. In addition, studies could examine the applicability of the findings to adult e-counselling clients, who are an increasing population. Many rural communities with mental health needs, for example the Haida Gwaii Nation, are now connected to the Internet (Telus, 2006) and thus could make good use of an online crisis counselling service.   In the future, given the Crisis Centre‟s new documenting system, chats could be selected in which the chatter is “new” to the service in order to eliminate the possibility that familiarity with the services rather than counsellor interventions is contributing to rapport. On a related note, differences between volunteer counsellors with more than 100 hours verses “newer” volunteers with fewer than 100 hours could be examined, to see whether volunteer counsellor experience on the chat has an effect on the types of responses found. Future research could also examine each chat as a whole rather than focusing on “phases” in order to observe what occurs in a given chat in a temporal manner. One clear limitation of the current study is the lack of input from both volunteers and youth with respect to how rapport is built in online chats. In order to include the views of youth and volunteers, focus groups could be held with youth and with volunteers, the results of which could be analyzed thematically for recurrent patterns or themes.  As mentioned, the youthinbc site has a tab for feedback from youth. This   126 feedback could also be analyzed with respect to how closely it relates to the findings of this current study. 4.7 Implications for practice Although the intention of qualitative research is usually not to generalize the results to a broader context, Creswell (2009) noted that when examining several cases, as was done in this study, generalizations with respect to the context in which the study was conducted may be cautiously made. Stakeholders potentially impacted by this study include paid staff and volunteer counsellors of the Vancouver Crisis Centre, the Board of Directors, people in crisis, other community organizations, and donors. For those involved in providing support to suicidal adolescent clients online, the results of the present study can be used to inform and enhance their practice, keeping in mind the limitations of the study as defined above. Specifically, when first engaging the client in the initial contact phase, it appears helpful to frame statements tentatively, especially when attempting to reflect the emotional experience of the youth. Staying away from “closed questions” (e.g., questions that require a “yes” or “no” answer) also seems advisable.  Due to the synchronous nature of online communication, it appears helpful to allow for time for the youth to respond.  Online counsellors should also be aware that youth who initially state that they are accessing services “for a friend” may actually be in need of assistance themselves. Use of metaphorical language and imagery may also be useful tools to build a connection in the absence of voice tone by helping the crisis counsellor and youth visually conceptualize the issue or emotion at hand. However, in the current study it was   127 found that metaphors and other imagery were first used by the youth; the counsellor would then reflect or extend them. Therefore, the crisis counsellor would be advised to be conscious of any verbal imagery used by the youth in the counselling session. With respect to the risk assessment portion of the chat, it appears that providing a rationale for the risk assessment questions asked during this phase is helpful. Practitioners should also be aware that despite the inevitably serious nature of the content discussed in this phase, using informal language can be helpful to build a connection, particularly when the youth is using it as well. Crisis counsellors are trained to help those in distress and to decrease risk factors as much as possible. Therefore, when confronted with a youth who is self- harming, the instinctive response may be to attempt to dissuade the youth from this potentially dangerous coping method. However, based on the results of this study, practitioners are encouraged to not immediately show disapproval; showing acceptance while engaging the youth in brainstorming other options appears to be a more effective strategy, considering the stigma associated with self-harm; this has been found to the case in studies examining adolescents who self-harmed (Fortune, Sinclair & Hawton, 2008).  Statements of care appear to be important to use throughout the chat, and, when coupled with summary statements, seem particularly important when ending the chat. With respect to computer- specific language, it appears that using emoticons and abbreviations can be helpful; this strategy seems most effective when the youth initiates their use and towards the end of the chat, once rapport has been established. When the client-counsellor   128 working relationship is threatened, addressing the issue openly, directly and informally can help the crisis counsellor maintain the trust and collaboration necessary to maintain the helping relationship. 4.8 Conclusion The initial goal of this study was to determine what themes and patterns emerged in online crisis counselling chats where the youth was suicidal. It was found that different types of counsellor responses were present depending on the phase of the chat. Specifically, in the Initial Contact phase, the youth tended to be hesitant to disclose their suicidality; in fact, they were more likely to initially state that they were seeking support for a friend. In this phase, counsellor interventions tended to be tentative and connection-building in nature; in several instances, metaphors used by the youth were taken up by the volunteer counsellor. This was thought to substitute for voice tone in this initial trust-building phase of the interaction. When the youth displayed challenging behaviours in the chat, the volunteer counsellor responded openly and directly, often using informal language and sometimes using humour. This strategy was used throughout the phases when threats to rapport occurred.  With respect to the Suicide Assessment phase, it was found that questions regarding suicide were most often interspersed with empathic statements and a rationale was given for asking the question. Contrary to the researchers‟ expectations, informal language was used frequently in this phase; in fact, it was used more frequently than in the other two phases.  It was hypothesized that this may have helped the youth to feel more comfortable while discussing the emotionally charged topic of suicide. Counsellor responses tended   129 to be action-oriented in the suicide assessment phase; this was thought to be due to the nature of suicide intervention, which by necessity involves collaborative action on the part of client and counsellor. However, it was noted that connection- building responses were also frequently used in this phase; in fact, they dominated in three out of the nine chats and in two of the chats were used with equal frequency as the action-oriented responses.  It was noted that this is consistent with literature on suicide assessment which consistently indicates that classical Rogerian skills such as empathic reflection should form the basis of every conversation on suicidality (Halderman et al., 2010; Neimeyer & Pfeiffer, 1994; Paulson & Everall, 2003). It was also found that in almost half of the chats (four out of nine), the youth was either engaging in or contemplating self-harm, and that in instances where this occurred, the counsellor showed acceptance while engaging the chatter in action-oriented strategies to explore alternative coping methods. When it became time to end the chat, it was found that counsellors tended to use summaries, provide statements of care, and use emoticons frequently. The findings of this study, while providing limited generalizability due to the small size and specificity of the sample and the qualitative nature of the research as well as the limitations outlined above, nevertheless are hoped to provide researchers and clinicians with new information regarding the development of rapport in online counselling with suicidal youth. It is hoped that this study has contributed a greater understanding of what occurs in these chats, so that youth in crisis can be served appropriately and effectively using an online medium.   130 References Auerbach, S., May, J., Stevens, M ., & Kiesler, J.  (2008). The interactive role of working  alliance and  client-counsellor interpersonal behaviours in adolescent substance  abuse  treatment. International Journal of Clinical Health and Psychology, 8 (3),  617-629. Bachelor, A. (1995). Clients‟ perceptions of the therapeutic alliance: A qualitative   analysis. Journal of Counseling Psychology, 42 (3), 323-337. Bambling, King, Reid & Wegner (2008). Online counselling: The experience of   counsellors providing synchronous single-session counselling to young people.  Counselling and Psychotherapy Research, 8, (2), 110-116. Blais, J., Craig, W., Pepler, D. & Connolly, J. (2008). Adolescents online: The  importance of internet activity choices to salient relationships. Journal of Youth &  Adolescence, 37 (5), 522-536. Beals, L. (2010). Content creation in virtual worlds to support adolescent identity  development. New Directions for Youth Development, 128, 45-53. Bedi, R. (2007). Concept mapping the client‟s perspective on counselling alliance  formation. Journal of Counseling Psychology, 53 (1), 26-35. Becker, K., & Schmidt, M.H. (2005). When kids seek help online: Internet chat rooms  and suicide. Reclaiming Children and Youth, 13, (4), 229-230.  Bower, B. (2010). Talking alike cements relationships. Science News, 178 (13),  8-9. Bryan, C.J. & Rudd, M.D. (2006). Advances in the assessment of suicide risk. Journal of  Clinical Psychology in Session, 62, (2), 185-200.   131 Canadian Mental Health Association: Statistics (n.d.). Retrieved Sept. 14, 2009 -1258-1404&lang=1 Creswell, J.W. (2009). Research Design: Qualitative and Quantitative Methods. New   York: Sage Publications.  Cook, J., & Doyle, C. (2002). Working alliance in online therapy as         compared to face-to-face therapy: Preliminary results. Cyberpsychology         and Behavior, 5, 95-105. Curtis, R. (2004). Analysing students‟conversations in chat room discussion groups.   College Teaching, 52 (4), 143-149. Davies, K. (2010). Coming of age online: The developmental underpinnings of girls‟  blogs. Journal of Adolescent research, 25 (1), 145-171. Derk, D, Bos, A., & von Grumbkow, J. (2008). Emoticons in computer-mediated   communication: Social motives and social context. CyberPsychology and  Behavior, 11 (1), 99-101. Efstation, J.F., Patton, M.J., & Kardash, C.M. (1990). Measuring the working alliance in  counsellor supervision. Journal of Counselling Psychology,37 (3), 322-329. Elo, S., & Kyngas, H. (2008). The qualitative content analysis process. Journal of   Advanced Nursing, 62, (1), 107-115. Evans, D. L.  (2005). Treating and preventing adolescent mental health disorders: What    we know and what we don’t know. New York: Oxford University Press. Farsimadan, F., Draghi-Lorenz, R.; Ellis, J. (2007).  Process and outcome of therapy in        ethnically similar and dissimilar therapeutic dyads. Psychotherapy Research   7, (5), 567-  575.   132 Fletcher-Tomenius, L. J., & Vossler, A. (2009). Trust in online therapeutic relationships:  The therapist‟s experience. Counselling Psychology Review, 24, (2), 24-33.   Gelso, C.J., Kelley, F.A., Fuertes J.N., Marmarosh, C.,  Holmes, S.E. , Costa, C. &   Hancock,  G.R.,  (2005). Measuring the real relationship in psychotherapy: Initial   validation of the therapist form. Journal of Counselling Psychology, 52, (4), 640-  649. Gelso, C.J. (2009). The real relationship in a postmodern world: Theoretical and  empirical exploration. Psychotherapy Research, 19 (3), 253-264. Goldsborough, R. (2009). Technology today. Community College Week, 22, 16-16. Goldenberg, H., & Goldenberg, I. (2008). Family therapy:  An overview (7th Ed.)   Belmont, CA:  Thomson-Brooks/Cole. Gonzales, A., Hancock, J.T., & Pennebaker, J. (2010). Language style matching as a   predictor of social dynamics in small groups. Communication Research, 37 (1),  3-17. Gould, M.S., Greenberg, T., Harris-Munkfakh, J., Kleinman, M., & Lubell, K. (2006).  Teenagers‟ attitudes about seeking help from telephone crisis services (hotlines).   Suicide and Life-threatening Behaviour, 36 (6), 601-615. Gould, M.S., Kalafat, J., Harris-Munfakh, J.L., & Kleinman, M. (2007).  An evaluation   of crisis hotline outcomes part 2: Suicidal callers. Suicide and Life-Threatening   Behaviour, 37 (3), 338-352. Granello, D.H. (2010a). A suicide crisis intervention model with 25 practical strategies  for implementation. Journal of Mental Health Counseling, 32 (3), 218-235.   133 Granello, D.H. (2010b). The process of suicide assessment: Twelve core principles.  Journal of Counseling and Development, 88, 363-370. Groth-Marnat, G. (1997). Handbook of Psychological Assessment. New York: John  Wiley & Sons. Fortune, S., Sinclair, J., & Hawton, K. (2008). Adolescents views on preventing self-  harm: A large community study. Social Psychiatry and Psychiatric  Epidemiology, 43, 96-100. Hackney, H.L. & Cormier, S. (2009). The professional counsellor. Columbus, Ohio:   Pearson Education. Halderman, B. L., Eyman, J.R., Kerner, L., & Schlacks, B. (2010). A paradigm for  the telephonic assessment of suicidal ideation. Suicide & Life-Threatening  Behavior, 39 (6), 639-647. Hanley, T. (2009). The working alliance in online therapy with young people:  preliminary findings. British Journal of Guidance and Counselling, 37,   (3), 257-269. Heath, S., Charles, V., Crow, G., & Wiles, R. (2007). Informed consent, gatekeepers, and   go-betweens: negotiating consent in child-and-youth-orientated institutions.   British Educational Research Journal, 33 (3), 403-417. Horvath, A. O. & Greenberg, L.  (1989). The development and validation of the working  alliance inventory. Journal of Counselling Psychology, 36, 223-233. Interlock Employee and Family Assistance Programs (n.d.) Retrieved September 17,  2009 from   134 King, R., Bambling, M., Lloyd, C., & Gomurra, R. Smith, G.; Reid, W., & Wegner, K.   (2006). Online counselling: The motives and experiences of young people              who choose the Internet instead of face to face or telephone counselling.   Counselling and Psychotherapy Research, 6 (3): 169- 174. Knaevelsrud, C., & Maerker, A. (2007). Internet-based treatment for PTSD reduces   distress and facilitates the development of a strong therapeutic alliance: a   randomized controlled clinical trial. BMC Psychiatry, 7 (13), 1-10. Martin, D.G. (2003) Clinical practice with adolescents. Pacific Grove, CA: Brooks/Cole             Thompson. Merchant, G. (2001). Teenagers in Cyberspace: an investigation of language use and   language change in cyberspace. Journal of Research in Reading, 24, 1-14. Mishara, B.L., Chagnon, F., Daigle, M., Balan, C., Campbell, J., & Berman, A. (2007).  Which behaviors and intervention styles are related to better short-term outcomes   in telephone crisis intervention? Results from a silent monitoring study of calls to  the U.S. 1-800-SUICIDE network. Suicide and Life-Threatening Behavior, 37 (3),  308-321. Neimeyer, R. A, & Bonelle, K. (1997).  The suicide intervention response inventory: A  revision and validation. Death Studies, 21 (1), 59-81. Neimeyer, R. A. & Pfeiffer, A. M. (1994). Evaluation of suicide intervention  effectiveness. Death Studies, 18 (2), 131-166. Lakin, J.L., Jefferis, V.E., Cheng, C.M. & Chartrand, T.L. (2003). The chameleon effect   as social glue: Evidence of the evolutionary significance of non-conscious   mimicry. Journal of Nonverbal Behavior, 27 (3), 145-162.   135 Paulson, B.L. & Everall, R.D. (2003). Suicidal adolescents: Helpful aspects of  psychotherapy. Archives of Suicide Research, 7 (4), 309-321. Pennebaker, J.W., Francis, M.E., & Booth, R.J. (2001) Linguistic inquiry and   word count.  Mahwah, NJ: Erlbaum. Rogers, C. R. (1961).  On becoming a person: A therapist's view of psychotherapy.   Boston: Houghton Mifflin. Romig, C & Gruenke, C. (1991). The use of metaphor to overcome inmate resistance to       mental health counseling. Journal of Counseling & Development, 69, 414-418. Rossman, G. (1993). Building explanations across case studies: A framework for   synthesis. Boulder, CO: Colorado University School of Education. Sandhu, D.S., Reeves, T.G. & Portes, P.R (1993). Cross-cultural counselling and   neurolinguistic mirroring with Native American adolescents. Journal of   Multicultural Counselling & Development, 21 (2), 106-118. Sharpley, C.F., Fairne, E., Tabray-Collins, E., Bates, R., & Lee, P. (2000). The use of  counsellor and client-perceived rapport. Counselling Psychology Quarterly, 13  (1), 99-116. Shao-Kang, L. (2008). The nonverbal communication function of emoticons   Cyberpsychology and behavior, 11 (5), 595-597. Schlosser, L.Z. & Gelso, C.J. (2005). The advisory working alliance inventory – Advisor   version: Scale development and validation. Journal of Counselling Psychology,   52 (4), 650-654.   136 Sharpley, C.F., Ridgeway, & I. R.  (1992). Development and field-testing of a procedure   for coached clients to asses rapport during trainees‟ counselling interviews.  Counselling Psychology Quarterly 5 (2), 149-160. Stiles, W.B., Reynolds, S., Hardy, G.E., Rees, A., Barkham, M., & Shapiro, D.A. (1994).  Evaluation and description of psychotherapy sessions by clients using the session  evaluation questionnaire and the session impacts scale. Journal of Counselling  Psychology, 41, (2), 175-185. Subramanyam, K., Smahel, D., & Greenfield, P. (2006). Connecting developmental   constructions to the internet: Identity presentation and sexual exploration in   online teen chat rooms. Developmental Psychology, 42, (3), 395-406. Telus News Releases (2006).  Longest over-water radio Internet transmission to connect  Haida Gwaii to Broadband. Retrieved March 26, 2009 from  bin/news_viewer.cgi?news_id=770&mode=2&news_keywords=2006 Tobin, G.A. & Begley, C.M. (2004). Methodological rigour within a qualitative  framework. Journal of Advanced Nursing 48 (4), 388-396. Tomlin, A. (2009). SMS Training Module. Samaritans, 1-37. Weiten, W., Lloyd, M. A.. Dunn, D.S. & Hammer, E. Y.  (2009). Psychology Applied to   Modern Life: Adjustment in the 21st Century. Belmont, CA: Wadsworth   Learning. Whitlock, J.L., Powers, J.L., & Eckenrode, J. (2006). The virtual cutting edge: The   internet and adolescent self-injury. Developmental Psychology, 42 (3), 407-417.   137 Williams, A. I. & Merten, M.J. (2009). Adolescents‟ online social networking following  the death of a peer. Journal of Adolescent Research, 24 (1), 67-90. Williams, R., Bambling, M., King, R., & Abbott, Q. (2009). In-session processes in   online counselling with young people: An exploratory approach. Counselling and  Psychotherapy Research, 9 (2), 93-100. Vancouver Crisis Intervention and Suicide Prevention Centre (2009). Statistics. Vancouver Crisis Intervention and Suicide Prevention Centre (2010). History. Retrieved   April 7, 2010 from                                138     Appendix A Satisfaction Scale, Vancouver Crisis Centre                   1 2 3 4 5 N/A Caller Says Not Helpful            Caller Implies    Not Helpful Caller Gives No Indication          Caller Implies Helpful/Thanks Caller Says Helpful/Than ks Not Applicable "You're not helpful." "I'll phone someone else.  Sorry to bother you." "I'm no further ahead. This isn't helping."  "You don't understand." Caller becomes angry, annoyed, irritated with your responses. Caller is sarcastic to your responses. Caller hangs up abruptly.  Caller makes no comment about accuracy of your empathic statement s. Caller sounds same through the call. Caller just responds to you.   „That‟s right” “That‟s it.”  Caller   sounds calmer with responses Caller responds positively to your statements.   "Thank you." "I feel better." "You've been helpful."   Prank Call Incomplete Call Sex Caller Caller just swears and hangs up     139  Appendix B: YIBC Disclaimer (Additions to the website on December 24 th 2009 are noted in bold)  PRIVACY & LEGAL Youth in BC volunteers are trained in active listening skills, support and crisis intervention by professional staff members of the Crisis Intervention and Suicide Prevention Centre of BC. Youth in BC volunteers are not licensed professionals; therefore are not trained to give advice. As such, we may encourage you to consult with an appropriate professional about your problems and concerns. The Youth in BC online chat is designed for private communication between you and a Youth in BC volunteer. To ensure you receive the best quality of support, supervisors will occasionally monitor chat sessions and chat logs. If we are concerned that you or someone else is in immediate danger, we are required by law to contact the authorities. This may include reporting pertinent identifying information to appropriate agencies that can intervene on behalf of the person in danger. Youth In BC values and respects user‟s rights to confidentiality. Users are not required to provide any personal identifying information while using our services. However, we encourage users to input their age and location to ensure accurate statistical analysis of our user base. Youth In BC uses this statistical information in order to provide effective, ongoing and relevant service to youth in need.  Information provided us is used to evaluate and improve our existing services and to design further programs to support youth like yourselves. Youth in BC is not responsible for contents of any off-site pages or services provided by other organizations. We do our best to ensure that all links on our site are accurate and working. Please report any broken links or links which may have become inappropriate to the Youth in BC Coordinator at Thank you!           140 Appendix C: Consent Form                                                                                                                                   CONCCONONSON Volunteer Consent Form  Crisis Counselling Online: Building Rapport with Suicidal Youth  Principal Investigator:  Dr. Richard Young, UBC Department of Educational and Counselling Psychology and Special Education  Co-Investigators: Maria Timm, Graduate Student, MA Program, UBC Department of Educational and Counselling Psychology and Special Education  Purpose:  You have been invited to take part in this research because we are interested in examining if certain ways of communicating are more effective than others in helping youth who use our youthinbc.  The results of this study will be used to improve the existing youthinbc chat to better reflect the concerns and counselling needs of youth, as well as improve other online programs like youthinbc. In partnership with the Vancouver Crisis Centre and UBC, and on behalf of youth in distress, youthinbc appreciates your help!  Study Procedures:  If you decide to volunteer for this study, we will be reviewing transcripts of your chats with youth. The chat will be stripped of all information that identifies you as the volunteer counsellor before it is seem by the researchers. You will remain anonymous and this research will not impact your volunteer counsellor duties at the Vancouver Crisis Centre. This research is not evaluative in nature. Potential Risks: Knowing your transcripts are being examined may produce some anxiety. There are no other known risks to participation in this project.    141 Potential Benefits: By contributing to a fuller understanding of what constitutes effective chats, your participation in this study will help the many youth who access youthinbc receive better service and support. You will have the opportunity to request a copy of the completed study and/or the study results.  Confidentiality: Your identity will be kept strictly confidential. All data will be identified only by code number and kept in a locked cabinet. You, or the youth chatters, will not be identified by name or your volunteer counsellor identity in any reports of the study. Contact for information about the study: If you have any questions or desire further information with respect to this study, you may contact Richard Young at [604-822-6380].  Contact for concerns about the rights of research subjects: If you have any concerns about your treatment or rights as a research subject, you may contact the Research Subject Information Line in the UBC Office of Research Services at 604-822-8598, or, if long distance, e-mail to  Consent: Your consent is entirely voluntary and you may refuse to participate or withdraw from the study at any time without negative consequences. Your signature indicates that you have received a copy of this consent form for your own records.  Signed:  Print name:       Date:     142 Appendix D: Coding of Counsellor Responses Perception checking/Clarifying; e.g. “Do I understand you correctly?” (Connection- Building) Focusing the caller on the present (Action-Oriented) Open-ended question used to explore coping (Action-Oriented) Closed-ended used to explore coping (Action-Oriented) Boundary-setting (asking caller to modify a recurring behaviour that is interfering with volunteer counsellor‟s ability to provide service) (Action-Oriented) Tentative lead (e.g. “sounds like you feel”) (Connection-Building) Reflection of Content (Connection-Building) Reflection of Feelings (Connection-Building) Close-ended question to explore safety (Action-Oriented) Open-ended question to explore safety (Action-Oriented) Close-ended question to explore support (Action-Oriented) Open-ended question to explore support (Action-Oriented) Showing Acceptance for caller‟s situation/coping methods (Connection-Building) Open-ended question to explore options (Action-Oriented) Referral to social service (Action-Oriented) Referral to emergency service (Action-Oriented) Referral to friend or family (Action-Oriented) Statement of caring or concern (Connection-Building)   143 Reframing of needs, goals, or values (e.g. “Sounds like you really want to be respected in this relationship.” (Connection-Building) Responses Not Part of Vancouver Crisis Centre Manual Expression of encouragement (e.g. “I wish you all the best”) (Connection-Building) Educated guess on emotion (when chatter does not overtly state emotion) (Connection- Building) Educated guess on chatter‟s situation (when chatter does not overtly describe situation) (Connection-Building) Reflection of implied goal (e.g. “Sounds like you‟re looking for another way to cope with these feelings”) (Action-Oriented) Additional online-specific responses Emoticons e.g. () Abbreviations (e.g. “lol”) Responses not coded as Acton-Oriented or Connection-Building: Provision of information about Crisis Centre services Summary statements Clarification of volunteer counsellor role Informal language (defined as language not normally used in formal written communication, including expressions such as “kinda”, “sorta,” “umm”, “hey”). Non-verbal response (e.g. “Mmhhmm”) Use of figurative language (defined as the use of words or expressions in a way that is different from the literal meaning, including metaphor, simile, personification, and onomatopoeia).


Citation Scheme:


Usage Statistics

Country Views Downloads
United States 21 25
South Africa 10 0
Canada 3 3
China 2 3
Japan 2 0
City Views Downloads
Washington 16 0
Unknown 13 40
Buffalo 2 0
Beijing 2 0
Kinston 2 0
Tokyo 2 0
Mountain View 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}


Share to:


Related Items