UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A comparison of criticism received face-to-face or via text message among young adults : does mode of… DeClerck, Drew 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2016_february_DeClerck_Drew.pdf [ 709.07kB ]
Metadata
JSON: 24-1.0223353.json
JSON-LD: 24-1.0223353-ld.json
RDF/XML (Pretty): 24-1.0223353-rdf.xml
RDF/JSON: 24-1.0223353-rdf.json
Turtle: 24-1.0223353-turtle.txt
N-Triples: 24-1.0223353-rdf-ntriples.txt
Original Record: 24-1.0223353-source.json
Full Text
24-1.0223353-fulltext.txt
Citation
24-1.0223353.ris

Full Text

     A COMPARISON OF CRITICISM RECEIVED FACE-TO-FACE OR VIA TEXT MESSAGE AMONG YOUNG ADULTS: DOES MODE OF COMMUNICATION MATTER? by  Drew DeClerck   B.A., The University of Calgary, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF ARTS  in  THE COLLEGE OF GRADUATE STUDIES  (Psychology)   THE UNIVERSITY OF BRITISH COLUMBIA  (Okanagan)   January 2016   © Drew DeClerck, 2015  ii Abstract The popularity of text messaging has increased dramatically in the last decade, such that most young adults use this form of communication daily.  The extent to which negative exchanges over text messaging can impact wellbeing in the same way as face-to-face communication remains unclear.  In the computer-mediated communication literature, cues-filtered-out theories predict that the impact of text messaging would be weaker, given its lack of tone and non-verbal cues.  However, proponents of the social information processing theory argue that strategies are developed when using a new communication technology resulting in an increased ability to have meaningful communications.  According to adaptive structuration theories, individual psychological characteristics also likely play an important role in determining one’s reaction to text messaging, but this has also received little empirical attention.  The present study aimed to address the current gaps in the literature by using a laboratory-based experimental paradigm to compare the effects of criticism provided face-to-face and via text message to a no-criticism control group, and to examine the moderating effects of rejection sensitivity and rumination.  A total of 170 emerging adults took part in an acute laboratory stressor followed by either: 1) text message criticism (n = 53), 2) face-to-face criticism (n = 58), or 3) no feedback (control; n = 59).  Levels of self-reported stress and positive and negative affect were measured at baseline, post-stress-task, and post-feedback.  Trait levels of rejection sensitivity and rumination were also assessed post-feedback.  Using ANCOVA procedures, it was determined that critical feedback via text message and face-to-face were both associated with significantly lower levels of positive affect compared to the control group (findings related to stress and negative affect were in the expected directions, but were non-significant).  No significant moderation effects were found for rejection sensitivity or rumination on the relationship between criticism and subsequent levels of stress or affect.  The results suggest that criticism expressed using text messaging can lead to negative emotional outcomes, and at levels similar to face-to-face communication.  Future  iii studies would benefit from a more comprehensive assessment of various modes of communication, as well as naturalistic assessment methods, such as intensive longitudinal designs.    iv Preface The research described in this document received approval from the Okanagan Research Ethics Board – Office of Research Services and was assigned the tracking number H13-02483.  The results of this project have been accepted for presentation at the 17th annual Convention of the Society for Personality and Social Psychology: DeClerck, D., Tracey, A., Legg, N., Woodworth, M., & Holtzman, S. (2016, January).  Text message vs. face-to-face criticism following a stressful task: Does mode of communication matter? Poster accepted for presentation at the Annual Convention of the Society for Personality and Social Psychology, San Diego, CA.   v Table of Contents Abstract ........................................................................................................................................... ii Preface ........................................................................................................................................... iv Table of Contents ............................................................................................................................ v List of Tables ............................................................................................................................... viii Acknowledgements........................................................................................................................ ix Dedication ....................................................................................................................................... x Chapter 1: Introduction ................................................................................................................... 1 Negative Social Interactions and Wellbeing ............................................................................... 3 Negative Interactions and Computer Mediated Communication ................................................ 6 Chapter 2: Theories of Computer-Mediated Communication ...................................................... 10 Cues-Filtered-Out Theories....................................................................................................... 10 Electronic Propinquity Theory .................................................................................................. 12 Social Information Processing Theory ...................................................................................... 14 Adaptive Structuration and Related Theories ........................................................................... 15 Chapter 3: Psychological Factors that Influence the Effect of Negative Interactions .................. 18 Rejection Sensitivity ................................................................................................................. 18 Trait Rumination ....................................................................................................................... 19 Chapter 4: Current Study .............................................................................................................. 21 Chapter 5: Research Questions and Hypotheses .......................................................................... 24 Research Question 1 ................................................................................................................. 24  vi Research Question 2: ................................................................................................................ 24 Chapter 6: Methods....................................................................................................................... 26 Participants ................................................................................................................................ 26 Measures ................................................................................................................................... 28 Statistical Analysis .................................................................................................................... 29 Chapter 7: Results ......................................................................................................................... 32 Data Screening/Cleaning .......................................................................................................... 32 Preliminary Descriptive Analysis ............................................................................................. 34 Main Analysis ........................................................................................................................... 39 Research Question 1: Effect of Criticism on Mood and Stress................................................. 39 Research Question 2: Moderating Role of Rejection Sensitivity and Rumination. .................. 41 Rejection Sensitivity ................................................................................................................. 41 Rumination ................................................................................................................................ 42 Chapter 8: Discussion ................................................................................................................... 46 Summary of Results .................................................................................................................. 46 Research Question 1: Effect of Criticism on Mood and Stress................................................. 47 Research Question 2: Moderation role of Rejection Sensitivity and Rumination .................... 50 Limitations and Future Directions ............................................................................................ 51 Future Directions ...................................................................................................................... 53 Conclusions ............................................................................................................................... 55 References..................................................................................................................................... 57  vii Appendix A - Timeline of the 75-minute Protocol ....................................................................... 69   viii List of Tables Table 1    Means, Standard Deviations, and Ranges for Positive Affect ....................................... 35 Table 2    Means, Standard Deviations, and Ranges for Negative Affect ................................... 366 Table 3    Means, Standard Deviations, and Ranges for Stress ................................................... 377 Table 4    Moderating Role of Rejection Sensitivity on the Relationship Between        Experimental Condition and Positive Affect, Negative Affect, and Stress ................... 44 Table 5   Moderating Role of Rumination on the Relationship Between Experimental      Condition and Positive Affect, Negative Affect, and Stress ......................................... 45  ix Acknowledgements I would like to thank those who helped me throughout my education and career – First, Dr. Scott Oddie, Dr. Greg Wells, and Dr. Anomi Bearden: I could not have made it to where I am without your mentorship and support during my time at Red Deer College.  More recently, I have received unwavering support from my supervisor Dr. Susan Holtzman and the members of the UBCO health psychology lab.  None of this would be possible without you.  Finally, I am grateful for the guidance provided by my committee members Dr. Brian O’Connor and Dr. Michael Woodworth. Words cannot explain the amount of gratitude that I have for all of the support I have received from you throughout my academic career.       x Dedication I would like to dedicate this work to my parents, my brother, and my girlfriend, Brittney.  You always believed in me and supported me along the way.  I could not have done this without your love and support. 1 Chapter 1: Introduction An association between social relationships and wellbeing has been well established (Cohen & Wills, 1985; Cohen et al., 1998; Cohen, 2004; Gilbert, Quinn, Goodman, Butler, & Wallace, 2013; Rosengren, Orth-Gomér, Wedel, & Wilhelmsen, 1993).  Individuals who feel satisfied with their support networks and believe they will have access to support when needed have better health outcomes (Gilbert et al., 2013).  On the other hand, negative aspects of relationships such as conflict (Bolger, DeLongis, Kessler, & Schilling, 1989; Cohen et al., 1998; Uchino, 2009), rejection (Twenge, Catanese, & Baumeister, 2002; Gerber & Wheeler, 2009), and loneliness (Cacioppo et al., 2002) have been shown to have a detrimental impact on wellbeing.  In recent years, there has been a dramatic rise in the use of mobile technologies to interact with our social networks (Choi & Toma, 2014; Coyne, Stockdale, Busby, Iverson, & Grant, 2011; Lenhart, 2010; Smith, 2011; Smith, 2014).  In Canada, approximately 78% of all residents own cell phones and an average of 250 million text messages are sent each day (Canadian Wireless Telecommunications Association, 2013).  A recent report by the Pew Research Center highlighted that emerging adults (18-29 years old) are among the highest users of text messaging (Smith, 2011).  Of this age group, 95% use their phone for text messaging and they typically send or receive at least 85-100 messages each day (Skierkowski & Wood, 2012; Smith, 2011).  With such widespread adoption of text messaging, a number of questions have arisen regarding the effectiveness of this mode of communication in providing support in relationships, and the impact of negative text messaging interactions on relationships and wellbeing.   Recent research has produced results that indicate there are both positive and negative aspects of text messaging.  For example, advantages of text messaging include its ability to allow people to remain connected throughout the day and hold private conversations while in the company of others, contributing to increased emotional bonding (Davie, Panting, & Charlton,  2 2004; Pettigrew, 2009).  Text messaging has additionally been identified as a channel to meet new individuals and get to know them better while avoiding awkward initial interactions (Cupples & Thompson, 2010).  Similarly, anxious individuals may be able to use this mode of communication to reduce the stress associated with face-to-face interactions (Lee, Chang, Lin, & Cheng, 2014) and there is a growing body of evidence that those more experienced with text messaging are able to engage in increasingly rich interactions (Ogara, Koh, & Prybutok, 2014; Reid & Reid, 2010; Walther, 2011).   However, not only is there evidence that texting may be a less effective means of providing support than face-to-face communication (Murdock, 2013; Seltzer, Prososkia, Zieglerc, & Pollak, 2012; Sherman, Michikyan, & Greenfield, 2013), there is emerging evidence that texting may actually contribute to negative relational and emotional outcomes.  For example, because text messaging allows for perpetual contact, it may result in undesirable expectations of relational maintenance or feelings of entrapment (Hall & Baym, 2012; Thomée, Härenstam, & Hagberg, 2011).  Text messaging also creates potential for miscommunications as a result of the lack of tone or brevity (Kamibeppu & Sugiura, 2005) and can cause feelings of anxiety when waiting for a reply to a sent message (Cupples & Thompson, 2010; Igarashi, Motoyoshi, Takai, & Yoshida, 2008; Kamibeppu & Sugiura, 2005), which can lead to relational conflict.  For example, in a recent study of post-secondary students, more than 60% of participants reported that texting contributed to at least some degree of relationship conflict (Ogletree, Facher, & Gill, 2014).  Additional research has found that people are willing to say negative things via text message that they would not otherwise say in person (Allen, 2012; Patchin & Hinduja, 2006; Reid & Reid, 2010).   Emerging work by Holtzman and colleagues has yielded similar findings (DeClerck, Meisner, & Holtzman, 2014; Shallow, DeClerck, Hudson, Rempel, & Holtzman, 2014).  For example, in an online study of 185 emerging adults, 79% reported being criticized via text  3 message and over 50% stated that waiting for a response generates stress and uncertainty.  Similarly, an analysis of qualitative interviews with 40 emerging adults identified waiting for a response as the most stressful or bothersome aspect of text messaging, followed by miscommunications.  Despite the growing evidence to suggest that text messaging may be a source of interpersonal stress for emerging adults, there have been no studies (to the author’s knowledge) that have compared how the effects of negative interactions that occur via text messaging may differ from those that occur face-to-face or who may be more or less affected by either communication modality.  Given that text messaging may be replacing face-to-face communication as a form of social interaction for many emerging adults (Pierce, 2009) this question warrants investigation.   Negative Social Interactions and Wellbeing A focus on the negative aspects of text messaging is warranted given the large body of literature that has demonstrated both emotional and physical health consequences of negative social interactions (Bolger et al., 1989; Cohen et al., 1998; Rook, 2001).  Social interaction allows for interpersonal conflict, bad influences, and can lead loss (Cohen, 2004).  In fact, studies of daily stress and mood suggest that interpersonal stress has the strongest impact on wellbeing over and above other types of daily stress, including financial problems and feeling overwhelmed at work or at home (Bolger et al., 1989).  An intriguing example of how conflict can lead to worsened health can be seen in a study conducted by Cohen and colleagues (1998) in which participants were exposed to a virus known to cause the common cold.  Those who were engaged in an ongoing interpersonal conflict (of at least a month) were more than twice as likely to develop a cold than participants without any major conflicts.  This was found to be true for those engaged in a conflict within relationships of an intimate, friendly, or familial nature.  Additionally, lower relationship quality (Salmela-Aro, Aunola, & Nurmi, 2008), and high levels  4 of negative social exchanges (Rook, 2001) have been linked with increased depressive symptoms.    A number of laboratory based studies have additionally found a link between conflict and wellbeing.  For example, Ewart et al. (1991) asked married couples with hypertension to talk about a topic that was considered to be a current source of significant conflict between them for 10 minutes while monitoring their blood pressure.  The conversations were then coded for hostile, supportive, and neutral remarks.  It was found that negative (hostile) interactions were related to increases in blood pressure.  In a similar fashion, Kiecolt-Glaser and colleagues (1993) assessed autonomic, endocrine, and immune functioning in newlywed couples over a 24-hour period.  During this period, couples were asked to spend 30 minutes discussing a current marital problem.  Those who displayed more negativity and hostility during the conflict had increased levels of epinephrine, norepinephrine, adrenocorticotropic hormone (ACTH) and growth hormone and decreased levels of prolactin and reduced immune functioning.  Interestingly, the changes to cardiovascular, endocrine, and immune functioning outlined in these studies were tied only to negative interactions and not to supportive, neutral, or avoidant comments.  Additionally, these findings were more pronounced for females than males (for a more extensive review of how negative interactions affect physiological processes see Robles & Kiecolt-Glaser, 2003).   Social media has recently received a fair amount of attention related to how its use may impact social relationships.  Contrary to popular belief, online social networking may not lead to expanded social groups or emotional connectedness offline.  A study by Pollet, Roberts, and Dunbar (2011) found that although increased social media use was associated with a larger online network of “friends” there was no association with larger social groups offline, or with increased feelings of emotional connectedness with offline friends.  However, an interesting study by Coviello and colleagues (2014) found that emotional expression on Facebook was “contagious.” Using data from millions of Facebook users, they determined that rainfall in a  5 given city was correlated with negative emotional posts by its citizens as well as people in their social network who lived in other cities that it was not raining.  It may be that people are directly affected by the emotions of people on social media sites with no real avenue to seek social support when they are negatively affected (or when they have been exposed to “contagious” negative posts).  Other literature has also found relationships between increased Facebook use and declines in subjective wellbeing (Kross et al., 2013), and increases in loneliness (Song et al., 2014).    Other research on the negative aspects of social relationships has highlighted the particularly detrimental impact of social rejection on wellbeing.  Many of an individual’s thoughts and behaviours are at least partially rooted in efforts to build a strong sense of belonging within their social environment (Baumeister & Leary, 1995).  Such feelings are needed in order to feel secure, have meaningful relationships, and maintain a strong psychological foundation (Baumeister & Leary, 1995).  When one is rejected, the impact on wellbeing may be dramatic.  In our evolutionary history rejection would likely have resulted in banishment from clans or groups and often meant death.  Therefore, it is likely that humans evolved an overly sensitive detection system that results in more false alarms than misses (Wesselmann, Nairne, & Williams, 2012; Williams, 2007).  Zadro, Williams, and Richardson (2004) provided evidence for this in a series of experiments.  Participants were asked to play a video game called Cyberball which is a simulated ball tossing game that has been used in a number of experiments consistently leading to feelings of rejection (e.g., Masten, Telzer, Fuligni, Liberman, & Eisenberger, 2012; Williams et al., 2000; Williams et al., 2002).  The game is rigged to intentionally pass the ball to the participants in the rejection conditions less than the others.  Those in the rejection conditions reported lower levels of belonging, control, self-esteem, and meaningful existence regardless of whether they were told that the other members in the game were controlled by people, computer-generated, or even when they were told that the  6 game was rigged to not pass to them.  These findings were taken as strong evidence that humans operate under a very primitive and sensitive rejection detection system that is activated with even the slightest hint of exclusion, regardless of context.  In line with these results, other laboratory-based studies have shown increases in aggression and hostility (Twenge, Baumeister, Tice, & Stucke, 2001; Williams, 2007), negative mood (Gerber & Wheeler, 2009), risk taking (Twenge, Cantanese, & Baumeister, 2002), rumination (Smith & Williams, 2004; Wirth & Williams, 2009), procrastination (Twenge et al., 2002), and decreases in positive mood and self-esteem (Gerber & Wheeler, 2009) following social rejection.  Additionally, structured interviews conducted by Faulkner and Williams (1995) about the long-term effects of ostracism have revealed that it may lead to depression, lowered self-esteem and self-worth, learned helplessness, and even increased suicide attempts (as cited in Williams & Zadro, 2001 pp.  21-53). Negative Interactions and Computer Mediated Communication There is emerging evidence to suggest that the effects of rejection may be as strong, if not stronger, when it comes in the form of digital communication.  In fact, digital interactions may even create scenarios in which individuals falsely perceive that they have been rejected.  Unlike face-to-face interactions, text messaging typically involves waiting for a reply for an indeterminate amount of time, and this period of waiting has been cited as a major stressor associated with the use of text messaging in adolescents (Cupples & Thompson, 2010; Igarashi et al., 2008; Kamibeppu & Sugiura, 2005).  For example, it is possible that someone who is waiting for a reply to a text message will mistakenly interpret the lack of response as an intentional act of ignoring or rejection.  While waiting for a reply to a text message, those with text message dependency (defined as ‘‘text-messaging-related compulsive behavior that causes psychological/behavioral symptoms resulting in negative social outcomes,” p.  2313) report particularly high levels of anxiety related to the possibility that they are being ignored or ostracized (Igarashi et al., 2008).  The sometimes ambiguous nature of the content of text  7 messages may also create situations in which individuals become concerned that they are being rejected, and often expend significant effort in an attempt to reduce the uncertainty.  This was confirmed in a study by Smith and Williams (2004) in which participants who were ostracized via text message (after a brief period being included in a text message conversation) reported worse mood and lowered feelings of belonging, control, self-esteem, and meaningful existence compared to those who remained included in the conversation.  Those who did not receive a reply for an extended period of time attempted to clarify that there was in fact an intentional act of rejection and used provocation (identified as “any comments that participants may make in an effort to provoke someone to text message them” p.  297) in attempts to regain control and obtain more information (Smith and Williams, 2004).   There is additional research showing that other forms of CMC can lead to feelings of rejection.  For example, Tobin et al., (2014) examined the effects of ostracism on Facebook.  They invited small groups of undergraduates to their lab and set up each participant at a computer with a Facebook account that was linked with the other participants accounts (i.e., all of the accounts were “friends”).  Participants were then asked to post a status update using these accounts and to respond to the other participants’ posts.  Participants in the feedback (inclusion) condition were able to post and comment freely with each other while participants in the no feedback (exclusion) condition were blocked so that the other participants couldn’t see their posts or respond to them.  A confederate was additionally present to ensure that all participants in the feedback conditions received at least one response to their post.  Participants in the no feedback condition felt that the other participants were less interested in their posts and scored significantly lower on measures of sense of inclusion, belonging, self-esteem, control, and meaningful existence similar to findings of rejection leading to reduced satisfaction with basic social needs.    8 Additionally, Choi and Toma (2014) employed a daily diary method to determine patterns in undergraduates’ sharing of emotional events.  At the end of each day participants were asked to report on the most important emotional event of the day (either positive or negative depending on group assignment).  It was determined that positive events were more likely to be shared using “non-intrusive” media (e.g., text messaging and social media) while negative events were more likely to be shared over a voice call or face-to-face.  However, negative affect was increased following a negative interaction regardless of the avenue through which the interaction took place.  Some reasons that people may prefer to discuss negative topics in person could be the lack of control, inability to talk things out, or potential for miscommunications via text.  In fact, a recent study by Murdock (2013) found that as frequency of text messaging increased in a first-year-college-student sample, so too did psychological vulnerability to interpersonal stress.  It was argued from these findings that text messaging may be a poor method of communication for dealing with interpersonal stress in close relationships. Finally, in a recent study by Kothgassner and colleagues (2014) the effects of rejection face-to-face versus a computer generated environment were compared using the aforementioned Cyberball game.  For this study, 48 young adult females were assigned to an inclusion or exclusion condition as well as one of three game types: 1) Cyberball in the same room as the two other people they were playing with, 2) Cyberball with other players in a separate room (called the avatar condition), or 3) Cyberball against computer-generated players (called the agent condition).  For conditions two and three, the non-participant players in the game were actually computer-generated.  However, the participant was told that they were playing with other people in the avatar condition and with computer-generated characters in the agent condition.  Participants then completed a number of measures assessing satisfaction with basic social needs (belonging, self-esteem, control, and meaningful existence), aggression, and social presence.  Heart rate was additionally measured throughout the Cyberball task and for five minutes after.   9 Participants in the exclusion conditions showed lower levels of needs satisfaction independent of which Cyberball condition they were in.  Moreover, for those who were excluded, aggression was increased to the same amount for the face-to-face and avatar conditions (and more so for the agent condition).  Additionally, there was an increased stress response in all exclusion conditions as indicated by increases in heart rate, although it should be noted that the greatest increase for heart rate was found in the face-to-face condition.  This study highlights how although the physiological response to rejection may not be as dramatic in digital rejection it appears as though perceptions of rejection are similar regardless of whether it is done face-to-face or through a digital environment.  Similar results were reported by Zadro et al. (2004) who found that rejection via Cyberball was able to elicit a similarly negative response when comparing their results to previous research on ostracism.      10 Chapter 2: Theories of Computer-Mediated Communication Although the empirical literature remains sparse regarding negative social interactions via text messaging, there are a number of computer-mediated communication theories that can inform the question of whether these interactions may influence wellbeing in the same way as face-to-face interactions.  Of particular relevance are the Cues-Filtered-Out theories, Electronic Propinquity Theory, Social Information Processing theory, and Adaptive Structuration theory.  These theories each contribute to our understanding of how digital communication plays a role in its users’ lives within the context of text messaging in interpersonal relationships. Cues-Filtered-Out Theories The term cues-filtered-out (CFO) was coined by Culnan and Markus (1987) and can be used to explain a group of theories (including social presence theory, lack of social context cues hypothesis, and media richness theory) that claim that as the number of verbal (e.g., tone, volume, inflection) and non-verbal (e.g., body language, facial expressions, demeanor) cues are reduced in communications, so too is the quality of the interaction.  The lack of nonverbal cues is therefore thought to reduce the likelihood that any important socially-oriented communication will result.  This would suggest that negative interactions via texting would not have the same impact as those that occur face-to-face.  It could be argued that the lack of hostile tone, and missing facial expressions accompanying the criticism or conflict would therefore not be received as negatively by recipient.  However, others have argued that the nature of text messaging itself (e.g., lack of tone, brevity etc.) may actually lead to more conflict than other forms of communication (Kamibeppu & Sugiura, 2005).  This may be a result of the lack of cues leading to uncertainty which creates the possibility of misinterpretations and/or negative communication, especially for those higher in rejection sensitivity (more on this later). In line with CFO theory, studies of positive social interactions via different communication modalities have shown that as the number of verbal and non-verbal cues  11 decreases, so too does the ability of interactions to buffer stress or lead to feelings of connectedness.  For example, Seltzer et al. (2012) randomly assigned children to receive support from their mothers face-to-face, via phone, or using instant messaging via a computer following an acute laboratory stress task.  Oxytocin (a stress-buffering hormone) and cortisol (a stress hormone) levels were then measured and compared to controls (who received no support).  Interestingly, those who received support face-to-face or over the phone showed increased levels of oxytocin while children in the instant messaging and control groups did not.  Similarly, cortisol levels following support were similar for the instant messaging group and the control group indicating that text based communication may not be ideal for providing support when one is stressed.  They concluded that auditory cues are more important than the content of the exchange in providing support.  Similarly, a recent study by Sherman et al. (2013) compared the effectiveness of face-to-face, video, audio, and text-based conversations in promoting bonding (identified as feelings of connectedness and affection) between pairs of close female friends.  For this study, pairs of friends were recruited to engage in four conversations each: 1) face-to-face, 2) video-chat, 3) phone call, and 4) text based.  Levels of bonding were determined using self-report measures and affiliation cues such as facial expressions/reactions.  Participants reported feeling connected in all conditions; however, in-person interactions were the most effective at promoting bonding, followed by video, audio, and text in that order.   Although the CFO theories have received a fair amount of attention, a meta-analysis done by Walther et al. (1994) highlights how studies that sought to prove or disprove it have come up with mixed results.  To compound this, many of these studies were conducted during the beginning stages of CMC adoption when there may not yet have been time to develop more positive habits for its use (Walther et al., 1994; for a detailed critique of these theories see Walther, 2011) However, as text messaging has become more common, users have developed strategies to attempt to overcome some of the deficits in verbal and non-verbal cues, such as  12 textisms (e.g., emoticons, letter repetition, capital letters for emphasis).  Indeed, in the same study by Sherman and colleagues described above, the self-reported frequency of use in each of the computer mediated conditions (video, audio, or text) was associated with increased bonding (i.e., those who used the different modes of communication more frequently in their everyday lives had higher levels of bonding while using that form of communication).  Similarly, the use of textisms was associated with greater bonding in the text-based communications, however this was not enough to improve the levels of bonding to that of the other modes of communication.  In summary, text based communication led to the weakest feelings of connection regardless of textism use and experience with this type of CMC.   Taken together, these studies indicate that there is at least some credibility to the CFO theories and that interactions that occur via text are not equivalent to those done face-to-face.  The absence of visual and auditory information present in text messaging may cause certain users to misread the tone of a message or create hypothetical reasons as to why the other person is not responding.  On the other hand, some users may be missing cues that were meant to relay information.  These misinterpretations may lead to increased feelings of rejection or frustration when discussing an important or difficult topic.  While these studies allude to the potential deficits of text messaging in positive interactions, there are (to the author’s knowledge) no equivalent studies looking at how negative interactions via text may differ from those done face-to-face.   Electronic Propinquity Theory  Electronic Propinquity Theory (EPT) claims that psychological closeness can be achieved using CMC but is dependent on a number of factors.  These factors include: 1) its bandwidth, or the capability of the CMC to relay communicative cues; 2) its capacity to allow immediate exchanges between communicators; 3) the communication skills of its users; 4) the level of complexity of the task; and 5) the number of choices of communication mediums  13 available (Korzenny, 1978).  Using these criteria, it can be expected that psychological closeness and satisfaction with communication using a CMC is increased as the bandwidth, immediacy of exchanges, and communication skills of its users are increased.  Additionally, as the level of complexity of the task (or communication) is increased, higher bandwidth communications are preferred, and as the number of available mediums for communication are increased, the lower bandwidth communications are seen as less favorable.  For example, if someone has the option to use text-based communication, voice chat, or video chat, they will be least satisfied with the text-based option.  However, if the only option is for text based chat, then it is likely that they will be as satisfied as someone who only has the option for video chat.  Although this theory was created before many of the currently popular CMC technologies were available, it was presented in a way that allows for the integration of new technologies (Walther, 2011).    This theory had received little attention since its conception until Walther and Bazarova (2008) revisited it in a study that set out to determine if the number of options of CMC available affected user satisfaction with each type of communication medium.  For this study, they created groups that differed in the number of available communication mediums (face-to-face, video chat, voice chat, and text-based chat) with some groups having access to multiple options, and others having access to only one type of communication.  Interestingly, those who had access to only one medium (e.g., text-based chat) were more satisfied with this medium than when they used the same medium but had the option to use other higher bandwidth mediums (e.g., using text-based chat when also having the option to use video chat).  Another important finding of this study was that there were no differences found between any of the mediums when they were the only available option.  The findings that communication options interact with satisfaction may be behind some of the inconsistencies in findings from qualitative designs and field studies when many communication options are available, to those found in experimental studies where only one option is available (Walther & Bazarova, 2008).  14 This theory indicates that when the only available mode of communication is text messaging (for example, when someone is otherwise occupied and unable to use any other type of communication) it should have the same ability to create psychological closeness and have similar user satisfaction as other types of communication.  Conversely, this implies that when multiple options are available (as is often the case), or when the level of complexity of the communication is increased (e.g., during an argument or important conversation) text messaging may not be the best option for communication and may lead to dissatisfaction from its users.  To add to this, the time required for communication exchanges also plays a role in satisfaction which indicates that waiting for a reply may cause distress (Cupples & Thompson, 2010; Igarashi et al.  2008; Kamibeppu & Sugiura, 2005) over and above other types of communication where exchanges happen at a quicker rate (e.g., in face to face communication).   Social Information Processing Theory Social information processing (SIP) theory acknowledges the lack of nonverbal cues in CMC but purports that people are motivated to improve communication and are able to develop strategies that allow maximum use of the available cues in whichever form of communication they are using (Walther, 1992, 2011).  Due to the lack of verbal and nonverbal cues present in text based communication the content comes with much less information than face-to-face or spoken exchanges.  However, according to this theory the users of this technology are likely to attempt to make up for this deficit in order to improve its effectiveness.  Another important feature of this theory is the understanding that CMC runs at a much slower pace than face-to-face communication and therefore may not be appropriate for all types of communication (e.g., when discussing important or distressing topics).  This feature has been supported by research on text messaging (Murdock, 2013) and stresses the need to take extra time when communicating in order to appropriately process meaning (Walther, 2011).    15 Using SIP theory, it could be argued that negative interactions should not be conducted via text message due to the potential for miscommunication (Kamibeppu & Sugiura, 2005) and the anxiety associated with waiting for a reply to a text (Cupples & Thompson, 2010; Igarashi et al. 2008; Kamibeppu & Sugiura, 2005).  This anxiety is likely increased during a conflict and may lead to even more negative feelings as a result of the interaction.  Given that people try to maximize their understanding of the content of a text message there is heightened attention paid to the available cues.  Certain people may overanalyze messages or impose meaning on the cues that may, or may not, have been intended.  For example, a delayed reply or a short message could be a result of someone being busy or in a rush.  However, it could also be an intentional way of showing the recipient that they are upset with them (e.g., the silent treatment).  Unfortunately, using these strategies to relay information makes it more difficult to know if there is an intentional act of rejection and certain people (e.g., those higher in rejection sensitivity and rumination) are more likely to misread, or spend additional time ruminating over, the content of the message (or lack thereof).  Given this, it could be argued that text messaging should not be used for negative interactions.  However, it is apparent that these interactions are taking place via text (Ogletree er al., 2014) and understanding how they impact those engaging in them is increasingly important as text messaging becomes more common. Adaptive Structuration and Related Theories One of the more comprehensive theories related to technology use and adoption is adaptive structuration theory (AST; Desanctis and Scott, 1994).  This theory is an integrative approach to information technology adoption which states that as new technologies are developed there is an interaction between the technology and the social environment in which it is used.  AST claims that technology use evolves in concert with those who use it resulting in changes to social rules and norms.  Desanctis and Scott (1994) assert that the impact of new technologies depends on: 1) the nature of the technology (its structure, potential uses etc.); 2)  16 how the technology is adopted or used by group members in relation to structures already in place; and 3) the development of new structures over time while the technology is present.  This theory is related to, although more complex than, Social Influence Theory (SIT) which focuses on user perceptions of CMC rather than simply focusing on the properties of the technology itself (Fulk, Steinfield, Schmidtz, & Power, 1987).  SIT theory claims that the social network in which the CMC technology is being used affects, or shapes, the perceived richness or utility of the technology.  A user will alter their perception of a CMC based on how they view it, as well as how they believe the members of their social network view it. AST is supported by studies that have investigated how text messaging is used and how users may adapt their interactions using this technology.  For example, it has been found that more experienced users are able to have increasingly rich interactions via text message indicating that as one becomes more familiar with texting they may adapt how they use it (Ogara, et al., 2014; Reid & Reid, 2010; Sherman et al., 2013).  This finding is also supported by Channel Expansion Theory (CET) whose central component is that as one becomes more familiar with a communication technology they increase their ability to have “rich” communications with it (see Carlson & Zmud, 1999; D’Urso & Rains, 2008).  While some users may adopt positive habits related to cell phone use such as increased connectedness within relationships (Davie et al., 2004; Pettigrew, 2009), there is evidence that this technology may produce negative changes in relational maintenance.  For example, text messaging has become so common that there is an expectation to always be available, resulting in feelings of entrapment (Hall & Baym, 2012; Thomée, et al., 2011).  Furthermore, studies are producing findings that people are willing to say things through text that they wouldn’t say face-to-face (Allen, 2012; Patchin & Hinduja, 2006; Reid & Reid, 2010).   These findings highlight that there may be something about text messaging that leads to increased feelings of stress, or that it may not be the best method to discuss potentially important  17 or stressful topics (Murdock, 2013).  As some users and groups are clearly developing negative habits related to text message use (i.e., adopting negative uses of the structures of this technology), it is important to determine the extent to which negative interactions via text impact its users, and who is most dramatically impacted by these interactions.  Because text messaging has become such a central part of emerging adults’ communicative practices it could be argued that negative interactions that occur via this modality have the same impact as those done face-to-face.  Furthermore, the nature of text messaging may create an environment that leads to increased distress for certain individuals (e.g., those higher in certain risk factors such as rejection sensitivity and rumination).      18 Chapter 3: Psychological Factors that Influence the Effect of Negative Interactions AST highlights the ways in which different people and groups tend to interact with technology in different ways.  It therefore makes sense to assume that not everyone will be impacted by negative social interactions via text messaging in the same way.  What follows is an overview of two key variables that have been identified in the broader social support literature to be moderators of the impact of criticism on wellbeing.  Although the majority of studies have not examined these individual difference variables in the context of CMC interactions, it is reasonable to expect that these factors also play a role in moderating the impact of criticism that occurs via text messaging with those higher in these traits being more negatively impacted.   Rejection Sensitivity  Studies have shown that being accepted by peers in childhood and adolescence leads to better health outcomes (Butler, Doherty, and Potter, 2007; White and Kistner, 2011; Zimmer-Gembeck et al., 2013).  However, the mere perception of peer acceptance may be enough to affect wellbeing regardless of actual levels of acceptance (White and Kistner, 2011; Zimmer-Gembeck et al., 2013).  When someone is low in peer acceptance, or has a biased self-perception of peer acceptance (e.g., when someone perceives that they are less accepted than they are by their peers), they are likely to develop rejection sensitivity later in life.  Rejection sensitivity has been conceptualized as a “disposition to anxiously expect, readily perceive, and overreact to rejection” (Downey, Freitas, Michaelis, & Khouri, 1998, p 545) and has been associated with increased anxiety, depression, loneliness, and aggression, as well as lowered interpersonal competence (Butler, Doherty, and Potter, 2007; White and Kistner, 2011; Zimmer-Gembeck et al., 2013). In a study by Downey & Feldman (1996) using an undergraduate student sample, it was found that people higher in rejection sensitivity were more likely to experience feelings of rejection following ambiguous interactions with romantic partners, even after controlling for  19 possible baseline differences in negative mood.  Those higher in rejection sensitivity were also more likely to perceive that insensitivity or a lack of empathy from their romantic partners was intentionally hurtful, more likely to be concerned about their partners leaving them, and were in less satisfying relationships.  This dissatisfaction likely stems from behaviours that those with high rejection sensitivity engage in that potentially put the relationship under duress.  This claim was supported by a follow up study by Downey et al. (1998) which found that women higher in rejection sensitivity were more likely to engage in behaviors that increased the likelihood of rejection from their partner during naturally occurring relationship conflicts. The ambiguity associated with text messaging may therefore create a potentially damaging environment (in a variety of interpersonal relationships) for those high in rejection sensitivity.  Waiting for a reply might be an overly stressful situation for these individuals.  Additionally, the lack of tone and visual cues allow for an increased likelihood of misinterpreting neutral statements as negative for those high in rejection sensitivity.  Furthermore, negative interactions, such as criticism, are likely to be perceived as more negative than they are intended to be via text.  The reaction could be to respond with aggression, or withdraw and avoid further communications.  It may therefore be that those higher in rejection sensitivity will be more adversely affected by criticism via text message than criticism face-to-face.  However, these ideas have yet to be tested and warrant investigation.   Trait Rumination  Another personality factor that is likely to lead to an adverse reaction to text messaging is a tendency to ruminate.  Rumination has been conceptualized a number of different ways (Smith & Alloy, 2009) mostly related to repetitive and intrusive negative thoughts.  Rumination is different than reflection in that the former is related to more maladaptive and neurotic thoughts while the latter can be adaptive (Trapnell & Campbell, 1999).  A ruminative thought pattern has been identified as a contributor to increased distress following a difficult event (Nolen- 20 Hoeksema & Davis, 1999), difficulty with time management and problem solving, tendencies to dwell on past failures (Lyubomirsky Kasri, & Zehm, 2003), as well as the development and maintenance of depressed mood (Smith & Alloy, 2009) and negative affect (Feldner, Leen-Feldner, Zvolensky & Lejuez, 2006).  Additionally those higher in rumination tend to spend more time focusing on negative stimuli (Joorman, Dkane, & Gotlib, 2006) and may benefit less from social support (Afifi, Afifi, Merrill, Denes, & Davis, 2013). The tendency of those higher in rumination to focus on negative stimuli (Joorman et al., 2006) adds an increased likelihood that they will be adversely affected by negative interactions regardless of the avenue through which they take place (e.g., face-to-face or through text messaging).  Additionally, given that text message conversations are likely to take place over a longer period of time (as evident in SIP theory), those higher in trait rumination have a period of increased vulnerability where they are likely to spend additional time brooding and questioning what was meant by brief or vague comments.  Therefore, it is possible that those higher in rumination will be affected more so by negative interactions that take place through text messaging than those done face-to-face.  No known studies have focused on this and an understanding of how text messaging affects these individuals is needed given their heightened vulnerability to depression (Smith & Alloy, 2009).      21 Chapter 4: Current Study A number of aspects of CMC make it difficult to answer questions about its impact or effectiveness compared to face-to-face interactions.  First, there are numerous different types/avenues of CMC currently available, including text messaging, social media, email, online video games, and chat rooms.  While there is a growing body of literature for each of these modalities, the aforementioned theories would lead one to expect that different groups, settings, expectations, and experiences would lead to these technologies having different outcomes on their users.  It is additionally common for many relationships to make use of multiple modes of communication (Walther, 2011) making it even more difficult to isolate the effects that certain types of communication may have on relationships.  Furthermore, many of the CMC theories were developed in the 80’s and 90’s when many of the current technologies were not available or did not have such a central role in our everyday lives.   The majority of studies that have investigated the stressful aspects of text messaging have been conducted on adolescent populations (Cupples & Thompson, 2010; Davie, et al., 2004; Hofferth & Moon, 2012; Kamibeppu & Sugiura, 2005; Leatherdale, 2010; Lenhart, Ling, Campbell, & Purcell, 2010; Pierce, 2009; Thompson & Cupples, 2008; Underwood, Rosen, More, Ehrenreich, & Gentsch, 2012).  However, research on the role of texting in social relationships is important in emerging adulthood for a number of reasons.  First, as mentioned earlier, this population is among the highest users of text messaging (Smith, 2011).  Emerging adulthood is additionally an integral period in development when one’s identity is explored and a number of friendly and romantic relationships are typically developed (Arnett, 2000).  This is also a period during which communication technologies are increasingly used to maintain relationships as many social groups are geographically separated due to educational or vocational endeavours (Manago, Taylor, & Greenfield, 2012).  Finally, poor relationship quality during this period has been linked to depression, burnout, lower income, and dysfunctional  22 coping strategies in adulthood (Salmela-Aro et al., 2008).  To date, much of the research on the role of social support in wellbeing among emerging adults has focused mainly on face-to-face interactions. Further issues with past research in this area are related to the tendency for published studies to rely on self-reported perceptions of, and attitudes towards, text messaging, as well as cross sectional research designs (Angster, Frank, & Lester, 2010; Cupples & Thompson, 2010; Davie et al., 2004; Faulkner & Culwin, 2005; Hofferth & Moon, 2012; Holtgraves, 2011; Kamibeppu & Sugiura, 2005; Lee, et al., 2014; Lenhart et al., 2010; Lepp, Barkley, & Karpinski, 2014; Ogara et al., 2014; Pettigrew, 2009; Pierce, 2009; Reid & Reid, 2010; Sultan, 2014; Takao, Takahashi, & Kitamura, 2009; Thompson & Cupples, 2008; Walsh, White, & McD Young, 2010).  While these methodologies are an important first step in understanding how text messaging is used they are subject to response biases (e.g., difficulty with recall) and make interpretation and generalization of findings difficult.   The majority of research on text messaging has focused largely on how and why it is used and the types of people that use it, leaving questions about how it compares to face-to-face interactions.  Additionally, there are no known randomized control studies that have investigated how receiving critical feedback via text message following a stressful event impacts emerging adults, compared to criticism face-to-face.  The extent to which text messaging interactions, particularly those that are critical in nature, can impact wellbeing remains unclear.  Obtaining a more comprehensive understanding of the impact of text messaging as a form of communication, and determining who may be more at risk of reacting negatively to criticism through digital communications are the main focus of this study.  In order to achieve the precision required to answer these questions and ensure consistency between participants, the current study employed a randomized, controlled, experimental paradigm.  Specifically, participants were asked to engage in an acute laboratory stressor and then received negative feedback face-to-face, or via  23 text message, or no feedback at all, in order to compare the effects of these different types of communication.    24 Chapter 5: Research Questions and Hypotheses Research Question 1: Does criticism received via text message or face-to-face following an acute laboratory stressor have a greater effect on levels of perceived stress and positive and negative affect compared to those who do not receive criticism? Hypothesis 1: It was hypothesized that receiving criticism (either via text message or face-to-face) following the stress task would be associated with higher levels of perceived stress and negative affect and lower levels of positive affect, compared to controls who received no feedback. Research Question 2: Are participants who score higher on rejection sensitivity or trait rumination more likely to be adversely affected by criticism via text message than those who receive criticism face-to-face or no criticism following an acute laboratory stressor? Hypothesis 2a: It was hypothesized that participants who reported higher rejection sensitivity would report higher levels of perceived stress, and negative affect, as well as lower levels of positive affect following feedback than those who reported lower rejection sensitivity. Hypothesis 2b: Furthermore, participants in the text message criticism condition who reported higher rejection sensitivity were hypothesized to report greater increases in perceived stress and negative affect, as well as lower positive affect than participants who were in the face-to-face criticism and no feedback conditions.   Hypothesis 2c: It was hypothesized that participants who reported higher trait rumination would report greater perceived stress, and negative affect, as well as lower positive affect following feedback than those who reported lower trait rumination. Hypothesis 2d: Furthermore, participants who were in the text message criticism condition who reported higher trait rumination were hypothesized to report higher levels of  25 perceived stress and negative affect, as well as lower levels of positive affect than participants who were in the face-to-face criticism and no feedback conditions.      26 Chapter 6: Methods Participants A sample of 198 emerging adults was recruited through the UBC Okanagan psychology undergraduate research subject pool and the community to take part in this lab-based experiment.  All participants were required to be 18-25 years old and speak fluent English.  Participants were not included if they indicated that they have been diagnosed with a psychological disorder.  Of this sample, there were 18 participants who dropped out and therefore did not complete the protocol and/or fill out all forms, two participants were found to lie outside the allowed age range, and protocol issues required seven additional participants to be removed.  The resulting number of retained participants before data cleaning was 171. Procedure Participants were randomly assigned to one of the three study conditions:  1) negative feedback via text messaging, (2) in-person negative feedback, or (3) no contact (control group).  Upon arrival to our research lab, participants were greeted by a similar-aged female research assistant who took them to a private room to obtain informed consent (for a detailed timeline see appendix A).  This initial period was also used to build rapport between the RA and participant.  Specifically, the RA was instructed to smile often, behave in a warm and friendly manner, and to engage the participant in small talk (in an attempt to make them feel more comfortable and to help establish some commonalities).  Following this, the participant took part in a 15-minute acclimation period to reduce any stress associated with the unfamiliar environment.  The participant then completed baseline measures.  After this, they were run through the Trier Social Stress Task (TSST).   The TSST is a well validated, reliable acute lab stressor known to raise both self-reported and biological markers of stress (Kirschbaum, Pirke, & Hellhammer, 1993; Kudielka, Hellhammer, & Kirschbaum, 2007).  This protocol has been used in hundreds of published  27 articles (see Kudielka et al., 2007 for a review) as a means to induce stress in participants.  Subjects were introduced to the task (a 5-minute speech and a 5-minute verbal math task) and then given 5 minutes to prepare a speech about why they deserve a vocational position on campus.  During this period, there was no note taking allowed.  They were then asked to give this speech to a committee comprised of three volunteers (one male and two female) that were trained to remain neutral and refrain from providing positive feedback (e.g., no smiling, nodding, leaning forward etc.).  Following this they were prompted to complete a math task in which they were asked to serially subtract 17 from 2023.  Throughout the speech and math task, committee members followed a standardized protocol.  They were also asked to take notes related to strengths and weaknesses of the participant’s presentation.   After the TSST, subjects were taken back to the private room to fill out the post stress-task measures.  For the control group, participants remained in this room for 10 minutes (which is the approximate time that was allowed in the feedback conditions).  For the feedback conditions, the research assistant provided standardized negative feedback (i.e., criticism) using the committee members notes to help provide context.  The committee members were asked to provide negative comments for every participant regardless of their performance so that the research assistant would have relevant material for the feedback sessions.  The standardized feedback was comprised of example statements to be used based on the participant’s performance.  For example, if the participant was confident that they had done well on the task the research assistant was instructed to reply with a statement such as “I see…The committee thought you did okay.”  For those who reported being nervous or stressed: “yeah, some people do pretty well in those situations, but it seems like you struggled.”  For those who thought they did poorly, the research assistant may have responded with “they felt like you could have said more” or something similar.  The criticism was intended to be approximately 10 minutes long.  However, due to the different pace at which face-to-face and text based interactions occur, the  28 text message condition was typically longer (approximately 12 minutes, while the face-to-face criticism was approximately 8 minutes).  Following the criticism, the participant was asked to complete the final measures.   Measures Participants completed self-report measures of stress and mood (positive and negative affect) at (a) baseline (upon arrival, following a brief ‘acclimation’ period), (b) post stress induction, and (c) post feedback (or control) condition.  Stress was measured using a 10cm visual analogue scale ranging from “Not at all Stressed” to “Extremely Stressed” (Kirschbaum et al., 1993) which asked participants to mark an X on the line that best represented how stressed they felt at that moment.  Positive and negative affect was measured using the 5-item joy, contentment, anxiety, and hostility subscales of the Derogatis Affects Balance Scale (DABS; Derogatis, 1996).  Participants were asked to indicate the extent to which they were experiencing each feeling in the current moment using a 5-point Likert scale ranging from 0 (Very Slightly or Not at all) to 4 (Extremely).  The joy (e.g., joyous, delighted) and contentment (e.g., calm, relaxed) scales were combined to measure positive affect, while the anxiety (e.g., anxious, timid) and hostility (e.g., enraged, angry) subscales were combined to measure negative affect.  The DABS has been shown to have adequate reliability with α coefficients ranging from .79-.94 and adequate validity has been documented (Derogatis, 1996).  In the present study, Cronbach’s alpha for the positive affect scale at T1 (α = .91), T2 (α = .94), and T3 (α = .94) was excellent.  Related to the negative affect scale, Cronbach’s alpha was also strong for T1 (α = .78), T2 (α = .87), and T3 (α = 90). Following the feedback condition, participants were also asked to complete a series of standardized measures, including measures of trait rumination (the rumination scale of the Rumination Reflection Questionnaire; Trapnell & Campbell, 1999), rejection sensitivity (Rejection Sensitivity Questionnaire-Personal; Downey & Feldman, 1996), perceptions of  29 support from the research assistant, and demographics.  The rumination scale of the Rumination Reflection Questionnaire (RRQ) is a 12-item measure that asks respondents to indicate how much they agree with a statement (1: strongly disagree – 5: strongly agree) related to their rumination tendencies.  This scale has been shown to have high internal consistency in undergraduate populations (Trapnell & Campbell, 1999).  In the present study, Cronbach’s alpha for the rumination scale was .75.  The Rejection Sensitivity Questionnaire-Personal (RSQ-P) assesses perceptions of anxiety related to how others would respond to a number of scenarios common to college students through eight two-part questions.  This is measured using six-point Likert scales.  The RSQ-P has been shown to have adequate internal consistency (Downey & Feldman, 1996).  Cronbach’s alpha in the current study was .79.  Participants in the two criticism conditions were also asked to provide feedback related to the research assistant.  Feedback was provided using a seven-point Likert scale to determine how hurt they felt (not at all – extremely) by the research assistant after the stress task.   Statistical Analysis Initially, data were analyzed to identify problems with missing data, univariate and multivariate outliers, normality, and multicollinearity.  Missing data were examined to ensure that the pattern was random so as to not affect generalizability.  Data were then analyzed to ensure randomization was effective by testing for between-group baseline differences on demographics and other variables.  Following this, a manipulation check of the stress task was conducted.  Independent t-tests were used to compare levels of stress at baseline and immediately following the stress task to ensure that perceptions of stress increased.  Similar analyses were run to test for changes in positive and negative affect from baseline to post stress task.  Perceptions of support from the confederate were analyzed to ensure that participants felt criticized.  30 To address hypothesis 1, three one-way analyses of covariance (ANCOVA) were run to compare levels of 1) perceived stress, 2) positive affect, and 3) negative affect after participants either received negative feedback via text messaging, in-person negative feedback, or no contact (control group).  Levels of stress and affect immediately following the stress task were included as control variables in each respective model.  Planned contrast tests comparing face-to-face and text message criticism to the control condition were then conducted.  A final comparison between the text message criticism and face-to-face criticism groups was then conducted to determine if these groups differed significantly from one another. To address hypotheses 2a and 2b, three moderated regression analyses were run (one for the dependent variable of stress, one for the dependent variable of negative affect, and one for the dependent variable of positive affect).  In order to account for the three-category variable, two dummy variables were created following recommendations by Field (2009).  For the first variable, the text messaging criticism group was coded as one while the other two groups were coded as zero.  For the second variable, the face-to-face criticism group was coded as one while the other two groups were coded as zero.  The first analysis was conducted to determine if rejection sensitivity moderated the relationship between the three conditions (text message criticism, face-to-face criticism, and control) and perceived stress.  To examine this, a multiple linear regression was conducted.  The dependent variable was the difference score for T2 and T3 perceived stress.  The independent variables were post-stress task levels of perceived stress, the two dummy coded variables for the criticism conditions, rejection sensitivity, and the interaction between each dummy variable and rejection sensitivity.  The second analysis was conducted to determine if rejection sensitivity moderated the relationship between the conditions and negative affect.  To examine this, another multiple linear regression was conducted.  The dependent variable was the difference score for T2 and T3 negative affect.  The independent variables in this regression model were post-stress task levels of negative affect, the two dummy coded  31 variables for the criticism conditions, rejection sensitivity, and the interaction between each dummy variable and rejection sensitivity.  Finally, the third analysis was conducted to determine if rejection sensitivity moderated the relationship between the conditions and negative affect.  To examine this, a third multiple linear regression was conducted.  The dependent variable was the difference score for T2 and T3 positive affect.  The independent variables in this regression model was post-stress task levels of positive affect, the two dummy coded variables for the criticism conditions, rejection sensitivity, and the interaction between each dummy variable and rejection sensitivity. To address hypotheses 2c and 2d, three moderated regression analyses were run following the exact structure outlined for hypotheses 2a and 2b but with rumination replacing rejection sensitivity as the moderator variable.  All analyses were run using SPSS version 20 with p values less than .05 being considered statistically significant.    32 Chapter 7: Results Data Screening/Cleaning  Initially, the key study variables (positive and negative affect, self-reported stress, rumination, and rejection sensitivity) were examined for missing data, univariate and multivariate outliers, normality, and multicollinearity as recommended by Field (2009) and Tabachnick and Fidell (2007).   Missing Data.  Of the key study variables, less than 2% of data were missing.  Little’s MCAR test revealed that the data were missing completely at random, 2 = 3828.08, p = .32.  Given these results, mean scores for positive and negative affect, rumination, and rejection sensitivity were computed using data from participants who completed at least 2/3 of responses on each scale.  This approach leaves the data in its most natural form therefore avoiding the known issues associated with other approaches to deal with missing data (see Tabachnick & Fidell, 2007 pp.  66-70).  Based on this criteria, no participants were excluded. Univariate Outliers.  Examination of transformed z-scores revealed several univariate outliers on a number of scales (z-scores greater than 3.29 are considered to be extreme outliers; Tabachnick & Fidell, 2007).  For baseline scores on the anxiety and hostility subscales of the DABS, one individual outlier was found for each.  At the third time point (post-criticism), two outliers were identified on the hostility subscale.  Inspection of these four univariate outliers led to the assumption that these participants were indeed part of the intended population and therefore their scores were retained and unaltered.   Multivariate Outliers.  Multivariate outliers are cases that have extreme scores on more than one variable and can have undue influence over results.  As recommended by Tabachnick and Fidell (2007), Mahalanobis distances were used to determine if there were any multivariate outliers on the independent variables of interest (T1 and T2 self-reported stress scores, T1 and T2 joy, contentment, anxiety, and hostility subscale scores, rumination, and rejection sensitivity).   33 Mahalanobis scores greater than 32.91 (p < .001) were considered to be extreme.  Using this cut-off value, three multivariate outliers were identified.  Using Cook’s D (Tabachnick & Fidell, 2007), these cases were analysed to see if they had influence over the data.  None of the cases were found to have influence using the cut-off score of 1.0 (i.e., all scores were below 1.0).  However, upon visual inspection of a scatterplot with Cook’s D scores, one case was identified as having a dramatically higher score than the other cases and was therefore removed from the dataset.   Normality.  Initially, histograms, normal and detrended Q-Q plots, stem & leaf plots, and box plots were produced to help aid in the assessment of normality.  As recommended by Field (2009), Kolmogorov-Smirnov (K-S) tests were conducted.  Baseline scores on positive, D(168) = 0.09, p <.05, and negative affect, D(168) = 0.171, p <.001 and the visual analogue scale, D(168) = 0.14, p <.001, were significantly non-normal.  Similarly, at times two, D(168) = 0.15, p <.001, D(168) = 0.13, p <.001, D(168) = 0.09, p <.001 (respectively) and three, D(168) = 0.11, p <.001, D(168) = 0.17, p <.001, D(168) = 0.07, p <.001 (respectively) all scores were significantly non-normal.  Non-normality was also significant for scores on rumination, D(168) = 0.09, p <.001, but not for scores on rejection sensitivity.  Visual inspection of the graphs and plots confirmed these findings for the most part.  Due to violations of the assumption of normality and some univariate outliers, bootstrapping was conducted for the applicable tests. Multicollinearity.  Multicollinearity means that multiple predictor variables are highly correlated resulting in a number of problems such as inability to generalize outside of the dataset and difficulty determining the individual importance of each predictor (Field, 2009).  Initially, a correlation matrix including all predictor variables was assessed for highly correlated variables.  All correlations among the predictor variables were below Field’s (2009) recommended cut-off of .80.  Additionally, variance inflation factors (VIF) and tolerance levels were assessed.  According to Field (2009), VIF values above 10 and tolerance levels below 0.1 indicate  34 problems.  All statistics fell within the acceptable range leading to the conclusion that multicollineraity was not an issue for this dataset. Preliminary Descriptive Analysis  The final sample consisted of 170 participants.  The majority of participants were female (57.8%) with a mean age of 20 (ranging from 18-25).  The sample was predominantly Caucasian (50.3%), followed by Asian (31.2%), and other (18.5%) with 66.5% being born in Canada.  Means, standard deviations, and ranges of the scores on positive affect at each time point are presented in Table 1.  Scores are presented for each condition separately as well as aggregated scores across conditions.  Similar descriptive statistics are presented for negative affect and self-reported stress in tables 2, and 3, respectively.  The aggregated mean scores for rejection sensitivity and rumination were 2.88 (SD = 0.67; range = 1.56-4.13) and 3.21 (SD = 0.55, range = 1.58-4.36), respectively. 35  Table 1 Means, Standard Deviations, and Ranges for Positive Affect        Positive affect total score  M (SD) range  M (SD) range  M (SD) range Control   Text Criticism   Face-to-Face Criticism      Baseline (T1) 1.86 (0.70) 0.40-3.50  1.65 (0.67) 0.30-2.90  1.62 (0.68) 0.30-3.00    Post Stress (T2) 0.96 (0.75) 0.00-2.70  0.73 (0.64) 0.00-2.60  0.80 (0.66) 0.00-2.70    Post Criticism (T3) 1.31 (0.77) 0.10-2.90  0.87 (0.69) 0.00-2.90  0.91 (0.64) 0.00-2.70 Aggregated            Baseline (T1) 1.71 (0.69) 0.30-3.50          Post Stress (T2) 0.83 (0.69) 0.00-2.70          Post Criticism (T3) 1.04 (0.73) 0.00-2.90       Note: Possible scores for positive affect ranged from 0 (Very slightly or not at all) to 4 (Extremely).  M (SD) = mean (standard deviation).    36  Table 2 Means, Standard Deviations, and Ranges for Negative Affect        Negative affect total score  M (SD) range  M (SD) range  M (SD) range Control   Text Criticism   Face-to-Face Criticism      Baseline (T1) 0.42 (0.44) 0.00-2.00  0.48 (0.37) 0.00-1.80  0.56 (0.48) 0.00-2.30    Post Stress (T2) 1.16 (0.80) 0.00-3.68  1.21 (0.84) 0.10-3.70  1.27 (0.80) 0.00-2.90    Post Criticism (T3) 0.49 (0.52) 0.00-2.20  0.78 (0.78) 0.00-3.20  0.79 (0.71) 0.00-3.00 Aggregated            Baseline (T1) 0.48 (0.44) 0.00-2.30          Post Stress (T2) 1.21 (0.81) 0.00-3.70          Post Criticism (T3) 0.68 (0.69) 0.00-3.20       Note: Possible scores for negative affect ranged from 0 (Very slightly or not at all) to 4 (Extremely).  M (SD) = mean (standard deviation).    37 Table 3 Means, Standard Deviations, and Ranges for Stress   Self-reported stress score  M (SD) range  M (SD) range Control   Text Criticism      Baseline (T1) 27.56 (21.89) 0.00-78.00  32.91 (24.95) 0.00-81.00    Post Stress (T2) 55.71 (25.80) 5.00-100.00  62.04 (28.93) 0.00-100.00    Post Criticism (T3) 31.64 (23.60) 0.00-92.00  44.28 (29.06) 0.00-100.00 Face-to-Face Criticism   Aggregated      Baseline (T1) 31.98 (23.88) 0.00-92.00     30.74 (23.53) 0.00-92.00    Post Stress (T2) 58.39 (26.64) 6.00-100.00     58.58 (27.04) 0.00-100.00    Post Criticism (T3) 44.14 (24.22) 3.00-93.00     39.85 (26.17) 0.00-100.00 Note: Possible scores for self-reported stress ranged from 0 (Not at all stressed) to 100 (Extremely stressed).   M (SD) = mean (standard deviation).  38 Randomization Check.  In order to ensure that the randomization process was effective, one-way ANOVAs were run to determine if there were significant differences in demographic characteristics, rumination, rejection sensitivity, and baseline scores of subjective stress and positive and negative affect.  With respect to demographics, there was no significant difference between the three study conditions based on sex, F(2, 144) = 0.96, p = .38; age, F(2, 163) = 2.21, p = .11; or ethnicity, F(2, 163) = 0.24, p = .79.   Analysis of key variables using one-way ANOVAs revealed that there were no between group differences in baseline positive affect, F(2, 165) = 2.18, p = .12; negative affect, F(2, 165) = 1.44, p = .24; or self-reported stress, F(2, 165) = 0.87, p = .42.  Similarly, there were no differences immediately following the stress task on positive affect, F(2, 165) = 1.94, p = .15; negative affect, F(2, 165) = 0.37, p = .69; or self-reported stress, F(2, 165) = 0.76, p = .47.  Finally, no differences were found based on group assignment for the moderator variables of rejection sensitivity, F(2, 165) = 0.53, p = .59; or rumination, F(2, 165) = 1.05, p = .35. Manipulation Check.  In order to assess the effectiveness of the TSST at inducing a psychological stress response, a manipulation check was performed using paired samples t-tests.  On average, participants experienced significantly lower levels of positive affect post stress task (M = .83, SE = 0.05) than at baseline (M = 1.71, SE = 0.05), t(167) = 16.35, p = .001, r = .62.  The opposite was found for participants related to negative affect, with scores being higher post stress task (M = 1.22, SE = 0.06) than at baseline (M = 0.48, SE = 0.04), t(167) = -12.78, p = .001, r = .49.  Similarly, participants felt significantly more stressed post stress task (M = 58.58, SE = 2.00) than at baseline (M = 30.61, SE = 1.82), t(167) = -14.93, p = .001, r = .57.  These findings confirm that the stress manipulation (TSST) was effective in creating changes in the expected direction in positive and negative affect, as well as subjective feelings of stress.  39 Main Analysis Research Question 1: Effect of Text Message and Face-to-Face Criticism on Mood and Stress  In order to determine if receiving criticism (face-to-face or via text message) had a significant impact on positive affect, negative affect, and self-reported stress after the stress task, three one-way analyses of covariance (ANCOVAs) were run.  First, the effect of group assignment on positive affect was assessed (with post-stress task levels of positive affect included as a covariate).  T3 positive affect adjusted means for the control, text message criticism, and face-to-face criticism groups were 1.20, 0.97, and 0.95, respectively.  Levene’s test of equality of error variances was not violated F(2, 167) = 2.71, p = .07.  Similarly, the interaction between condition and post-stress-task (T2) positive affect was not significant, F(2, 164) = 0.36, p = .70, indicating that the assumption of homogeneity of regression slopes was met.  The covariate of positive affect at T2 was significantly related to positive affect post-criticism (T3) F(1, 164) = 378.46, p < .001, r = .48, 95% bootstrap CI [0.71, 0.96].  There was also a significant effect of condition on levels of T3 positive affect after controlling for the effect of T2 positive affect, F(2, 164) = 3.86, p < .05, partial n2 = .05.  Planned contrasts revealed that being in the control group led to significantly higher levels of positive affect after the stress task, compared to those in the face-to-face criticism group, t(164) = -2.20, p < .05, r = .03, 95% bootstrapped CI [-0.49, -0.02] and the text message criticism group, t(164) = -2.58, p < .05, r = .04, 95% bootstrapped CI [-0.52, -0.09].  When comparing the text message criticism group and the face-to-face criticism group, there was no significant difference in levels of positive affect, t(164) = -0.27, p = .78, r = .00, 95% bootstrapped CI [-0.15, 0.11].  Next, the effect of group on negative affect was assessed.  T3 negative affect adjusted means for the control, text message criticism, and face-to-face criticism groups were 0.52, 0.78, and 0.76, respectively.  Levene’s test of equality of error variances was not violated F(2, 167) =  40 2.54, p = .08.  However, the interaction between condition and T2 negative affect was significant, F(2, 164) = 4.72, p < .05, indicating that the assumption of homogeneity of regression slopes was violated.  The covariate of negative affect at T2 was significantly related to negative affect post-criticism (T3) F(1, 164) = 245.65, p < .001, r = .24, 95% bootstrapped CI [0.35, 0.63].  However, there was a non-significant effect of condition on levels of T3 negative affect after controlling for the effect of T2 negative affect, F(2, 164) = 1.16, p = .32, partial n2 = .01.  Planned contrasts revealed that being in the control group did not lead to significantly different levels of T3 negative affect compared to those in the face-to-face criticism group, t(164) = 0.91, p = .29, r = .01, 95% bootstrapped CI [-0.12, 0.35] or the text message criticism group, t(164) = -0.65, p = .33, r = .00, 95% bootstrapped CI [-0.28, 0.09].  When comparing the text message criticism group and the face-to-face criticism group, there was no significant difference in levels of negative affect, t(164) = -0.33, p = .75, r = .00, 95% bootstrapped CI [-0.19, 0.14].  However, due to the violation of homogeneity of regression slopes, the interpretability of these results is questionable.  Finally, the effect of condition on self-reported stress was assessed.  T3 self-reported stress adjusted means for the control, text message criticism, and face-to-face criticism groups were 33.44, 42.14, and 43.53, respectively.  Levene’s test of equality of error variances was not violated F(2, 165) = 0.16, p = .85.  Similarly, the interaction between condition and post-stress-task (T2) self-reported stress was not significant, F(2, 162) = 1.19, p = .31, indicating that the assumption of homogeneity of regression slopes was met.  The covariate of self-reported stress at T2 was significantly related to self-reported stress at T3, F(1, 162) = 198.47, p < .001, r = .24, 95% bootstrapped CI [0.43, 0.79].  However, there was no significant effect of condition on levels of T3 self-reported stress after controlling for the effect of T2 self-reported stress, F(2, 162) = 0.90, p = .41, partial n2 = .01.   Planned contrasts revealed that T3 levels of self-reported stress in the control group did not significantly differ from those in the face-to-face criticism  41 group, t(162) = 1.09, p = .16, r = .01, 95% bootstrapped CI [-4.01, 19.85] or the text message criticism group, t(162) = -0.18, p = .83, r = .00, 95% bootstrapped CI [-14.48, 11.77].  When comparing the text message criticism group and the face-to-face criticism group, there was no significant difference in levels of self-reported stress, t(164) = -0.32, p = .74, r = .00, 95% bootstrapped CI [-5.71, 7.45]. Research Question 2: Moderating Role of Rejection Sensitivity and Rumination. In order to answer research question two, six moderated regression analyses were conducted to examine the moderating effect of rejection sensitivity and rumination on the relationship between study condition and the outcome variables (positive affect, negative affect, and self-reported stress).  Difference scores between T2 and T3 positive affect, negative affect, and self-reported stress were used in the analysis.  For the following analyses, rejection sensitivity and rumination were each centered in order to compute their respective interaction terms with the condition variable. Rejection Sensitivity  Results for the moderation analysis of rejection sensitivity can be seen in Table 4.  Related to positive affect, when comparing the text message group to the other conditions the interaction between rejection sensitivity and the experimental condition was non-significant (B = 0.01, 95% bootstrapped CI [-0.19, 0.25]).  When comparing the face-to-face group to the other conditions the interaction between rejection sensitivity and the experimental condition was also non-significant (B = 0.05, 95% bootstrapped CI [-0.15, 0.28]).   The main effect of rejection sensitivity on positive affect was also non-significant (B = -0.05, 95% bootstrapped CI [-0.14, 0.04]).  Related to negative affect, when comparing the text message group to the other conditions the interaction term was non-significant (B = 0.15, 95% bootstrapped CI [-0.11, 0.42]).  When comparing the face-to-face group to the other conditions the interaction between rejection sensitivity and the experimental condition was also non-significant (B = 0.13, 95%  42 bootstrapped CI [-0.25, 0.45]).  However, the main effect of rejection sensitivity significantly predicted changes in negative affect (B = -0.15, 95% bootstrapped CI [-0.27, -0.03]).  Related to self-reported stress, when comparing the text message group to the other conditions the interaction between rejection sensitivity and the experimental condition was non-significant (B = 4.72, 95% bootstrapped CI [-4.69, 16.09]).  When comparing the face-to-face group to the other conditions the interaction between rejection sensitivity and the experimental condition was also non-significant (B = 7.36, 95% bootstrapped CI [-2.63, 18.22]).  The main effect of rejection sensitivity on self-reported stress was non-significant (B = -3.49, 95% bootstrapped CI [-7.34, 0.52]).   Rumination  Results for the moderation analysis for rumination can be seen in Table 5.  Related to positive affect, when comparing the text message group to the other conditions the interaction between rumination and the experimental condition was non-significant (B = 0.02, 95% bootstrapped CI [-0.27, 0.34]).  When comparing the face-to-face group to the other conditions the interaction between rejection sensitivity and the experimental condition was also non-significant (B = 0.10, 95% bootstrapped CI [-0.23, 0.45]).  The main effect of rumination did not significantly predicted changes in positive affect (B = -0.11, 95% bootstrapped CI [-0.23, 0.00]).  Related to negative affect, when comparing the text message group to the other conditions the interaction term also failed to reach significance (B = 0.24, 95% bootstrapped CI [-0.12, 0.59]).  When comparing the face-to-face group to the other conditions the interaction between rejection sensitivity and the experimental condition was also non-significant (B = 0.05, 95% bootstrapped CI [-0.37, 0.46]).  Similarly, the main effect of rumination did not significantly predict negative affect (B = -0.08, 95% bootstrapped CI [-0.24, 0.06]).  Related to self-reported stress, when comparing the text message group to the other conditions the interaction between rumination and the experimental condition was non-significant (B = 5.82, 95% bootstrapped CI [-7.67, 19.70]).   43 When comparing the face-to-face group to the other conditions the interaction between rejection sensitivity and the experimental condition was also non-significant (B = 4.23, 95% bootstrapped CI [-9.74, 18.16]).  The main effect of rumination on self-reported stress was also non-significant (B = 0.94, 95% bootstrapped CI [-4.71, 6.65]).      44 Table 4 Moderating role of Rejection Sensitivity on the Relationship Between Experimental Condition and Positive Affect, Negative Affect, and Stress     Positive affect   Negative affect Stress  B SEb 95% CI β R2 ΔR2 B SEb 95% CI β R2 ΔR2 B SEb 95% CI β R2 ΔR2 Step 1     .05 .07     .03 .04     .03 .04    Text  -.20 .08 [-.36, -.05] -.23   .24 .09 [.07, .41] .21   6.93 3.71 [-0.56, 13.83] .16      F-to-F -.23 .07 [-.37, -.09] -.27   .19 .10 [-.02, .39] .17   9.09 3.53 [2.15, 16.13] .23   Step 2     .06 .01     .06 .04     .04 .01    Text  -.20 .08 [-.35, -.05] -.22   .25 .09 [.09, .43] .22   7.31 3.70 [-0.28, 14.22] .18      F-to-F -.23 .07 [-.38, -.09] -.27   .19 .10 [-.02, .38] .17   9.05 3.52 [2.17, 16.04] .22      RS -.05 .04 [-.14, .04] -.08   -.15 .06 [-.27, -.03] -.19   -3.49 2.09 [-7.34, 0.52] -.12   Step 3     .04 .00     .05 .01     .04 .01    Text  -.20 .08 [-.35, -.05] -.22   .25 .09 [.08, .42] .22   7.30 3.66 [-0.22, 14.27] .18      F-to-F -.23 .07 [-.37, -.08] -.27   .19 .10 [-.02, .38] .17   9.19 3.53 [2.17, 16.04] .23      RS -.07 .09 [-.25, .09] -.11   -.25 .12 [-.48, -.01] -.31   -7.45 4.17 [-16.05, 0.42] -.25      Text x RS .01 .11 [-.19, .25] .01   .15 .14 [-.11, .42] .12   4.72 5.41 [-4.69, 16.09] .10      F-to-F x RS .05 .11 [-.15, .28] .05   .13 .18 [-.25, .45] .09   7.36 5.40 [-2.63, 18.22] .14   Note: 95% CI = 95% bootstrapped confidence intervals [lower limit, upper limit].  SEb = bootstrap standard error of B.  Difference scores (T3 – T2) used for outcome variables.  RS = rejection sensitivity.  Text = dummy variable with all members of the text message group coded as 1; F-to-F = dummy variable with all members of the face-to-face group coded as 1.     45 Table 5 Moderating Role of Rumination on the Relationship Between Experimental Condition and Positive Affect, Negative Affect, and Stress    Positive affect   Negative affect Stress  B SEb 95% CI β R2 ΔR2 B SEb 95% CI β R2 ΔR2 B SEb 95% CI β R2 ΔR2 Step 1     .06 .07     .03 .04     .03 .04    Text  -.20 .08 [-.36, -.06] -.23   .24 .06 [.08, .41] .21   6.93 3.61 [-0.24, 13.68] .17      F-to-F -.23 .08 [-.38, -.09] -.27   .19 .10 [-.07, .39] .17   9.01 3.51 [2.10, 15.76] .22   Step 2     .07 .02     .03 .01     .03 .00    Text  -.19 .08 [-.34, -.04] -.21   .25 .09 [.09, .42] .22   6.80 3.59 [-0.45, 13.57] .16      F-to-F -.22 .08 [-.37, -.07] -.25   .20 .10 [-.01, .40] .18   8.98 3.54 [1.61, 15.80] .22      Rum -.11 .06 [-.23, .00] -.15   -.08 .07 [-.24, .06] -.09   0.94 2.82 [-4.71, 6.65] .03   Step 3     .06 .00     .03 .01     .02 .01    Text  -.18 .08 [-.01, -.04] -.20   .25 .09 [.08, .42] .22   6.92 3.60 [-0.31, 13.69] .16      F-to-F -.22 .08 [-.38, -.07] -.25   .21 .10 [-.00, .41] .19   9.21 3.58 [2.21, 15.89] .23      Rum -.16 .14 [-.44, .11] -.21   -.18 .16 [-.49, .12] -.18   -2.38 5.12 [-12.54, 7.22] -.07      Text x Rum .02 .16 [-.27, .34] .02   .24 .18 [-.12, .59] .14   5.82 7.02 [-7.67, 19.70] .09      F-to-F x Rum .10 .17 [-.23, .45] .08   .05 .20 [-.37, .46] .03   4.23 6.98 [-9.74, 18.16] .07   Note: 95% CI = 95% bootstrapped confidence intervals [lower limit, upper limit].  SEb = bootstrap standard error of B.  Difference scores (T3 – T2) used for outcome variables.  Rum = rumination. Text = dummy variable with all members of the text message group coded as 1; F-to-F = dummy variable with all members of the face-to-face group coded as 1.  46 Chapter 8: Discussion Social relationships can have a significant impact on wellbeing, with those being more satisfied with their relationships having better physical and mental health outcomes (Gilbert et al., 2013).  However, a number of negative aspects of relationships have been associated with poor health outcomes (Bolger et al., 1989; Cohen et al., 1998; Rook, 2001).  With the recent and dramatic increases in the use of text messaging to develop and maintain relationships (Cupples & Thompson, 2010; Skierkowski & Wood, 2012; Smith, 2011), it has become increasingly important to understand the potential negative outcomes that text messaging may have on those who use it.  A number of theories have been developed to try to explain and predict differences in communication styles and relational outcomes based on computer-mediated communication (for a review see Walther, 2011).  The goals of the present study were to determine if the impact of receiving criticism via text message or face-to-face communication following a stressful event would differ, and to determine who might be more or less impacted by criticism occurring via text message.  Based on previous theories and research it was hypothesized that rejection sensitivity and trait rumination would play a moderating role on the relationship between criticism and feelings of positive affect, negative affect, and self-reported stress.  To the author’s knowledge, the current study is the first to compare the effect of criticism received via face-to-face communication versus text message using a lab-based experimental design.  Additionally, it is believed to be the first to consider rejection sensitivity and rumination as potential moderator variables. Summary of Results  The current study provides evidence that receiving critical feedback via text message and face-to-face communication both have a statistically significant negative impact on positive affect compared to those who do not receive any feedback following an acute laboratory stressor.  There were no significant differences in levels of negative affect or self-reported stress across the  47 three conditions.  However, due to the violation of the assumption of homogeneity of regression slopes, the findings related to negative affect must be interpreted with caution (a violation of this assumption indicates that the regression equation may not represent each group accurately; Field, 2009).  The moderation analyses for rejection sensitivity did not reveal significant interactions for positive affect, negative affect, or self-reported stress.  However, across conditions, those higher in rejection sensitivity reported significantly higher levels of negative affect.  Related to rumination, there were similarly no significant interactions for positive affect, negative affect, or self-reported stress.  Similarly, although there was a trend for those higher in rumination to report lower levels of positive affect, this main effect did not reach significance.   Research Question 1: Effect of Criticism on Mood and Stress  The first study aim was to examine the impact of criticism via text message and face-to-face communication on levels of positive affect, negative affect, and stress following an acute laboratory stressor, compared to a control group (who received no criticism).  The first hypothesis was partially supported with reports of positive affect being significantly lower in the text message and face-to-face conditions, compared to the control condition.  There was no significant difference between the two criticism groups on positive affect.  Although the criticism conditions did not lead to significantly greater levels of negative affect and stress compared to the control group, the results were in the expected directions.  Specifically, participants in both criticism conditions reported higher levels of negative affect and self-reported stress than the control group.    Taken together, these findings suggest that criticism, whether it be face-to-face or via text message, may lead to similar changes in mood, and that the impact of this criticism appears to be greater on positive affect.  These findings are consistent with previous research indicating that text messaging can lead to lower mood (Choi & Toma, 2014) and relationship conflict (Ogletree  48 et al., 2014).  Given that people are often more willing to express critical remarks via text than face-to-face (Allen, 2012; Patchin & Hinduja, 2006; Reid & Reid, 2010), this mode of communication not only has the potential to lower mood, but also to increase relational harm.  What’s more, there is evidence that positive interactions are less impactful via text message (Seltzer et al., 2012; Sherman et al., 2013) indicating that although a negative interaction is likely to result in lowered positive affect (and potentially increased negative affect and stress), efforts to make amends using this technology may not be successful.  While it appears that there is some support for the aforementioned CFO theories related to positive interactions, the findings from the current study indicate that negative interactions via text message or face-to-face can have a similar effect on mood despite differences in the number of available cues.  These findings are more consistent with SIP theory and EPT (as participants were only permitted one mode of communication in the study).  As mentioned earlier, Walther and Bazarova (2008) found that when there was only one available communication modality, differences in satisfaction were not found between different types.  The findings in this study add to EPT indicating that when there is only one available option for communication, users will be equally affected by the interaction regardless of the mode used.       Given that the criticism received in the current study was fairly mild (and given by a confederate) it is worth considering that more extreme cases of criticism (or cases of criticism from close others) might have a much more dramatic impact.  For obvious ethical reasons it was not possible to provide overly critical feedback in this study.  It is likely that the mild nature of the feedback was not enough to contribute to statistically significant differences in negative affect.  Had the criticism been more dramatic, it may have produced more marked changes.    There have been a number of findings related to the harmful impact that cyberbullying has on its victims such as feelings of anger, sadness, embarrassment, and even depression (for a review see Slonje, Smith, & Frisén 2013).  While these studies often include all types of CMC interactions,  49 a study done by Raskauskas (2010) found that 43% of students had experienced bullying via text message.  Given the findings of the current study it can be inferred that bullying done via text message may be as impactful as traditional bullying.  In fact, some studies have even reported that cyberbullying can lead to worse outcomes related to self-harm or suicide when compared to traditional bullying (Hay, Meldrum, & Mann, 2010).  Considering that cyberbullying extends beyond the schoolyard and into the victim’s home, and often takes place over longer periods, the impact of these interactions is likely very large.   Despite the finding that there was a significant difference in levels of positive affect between the control and criticism groups, the effect size for this finding was very small.  There are a number of reasons that may contribute to the low effect sizes seen here and to the non-significant findings related to negative affect and self-reported stress.  First, it was apparent that the stress task was capable of producing differences in the variables of interest as indicated by the manipulation check.  These changes in affect and levels of stress associated with the post-stress-task measures were quite substantial as indicated by their large to very-large effect sizes.  It could be that these changes were so significant that they carried over to the post-criticism measures.  In other words, the effects of the stress task may have overpowered or dominated that of the critical remarks following the stress task.  Also, although there were precautions taken in order to ensure that the protocol was followed and that the stress task and subsequent criticism session were consistent between participants, a number of issues related to this were unavoidable.  For example, some participants inevitably performed better at the TSST and felt relatively confident following the task.  Although the research assistant’s role was to point out negative aspects of each participant’s performance, those who felt good about their performance may have been much less impacted by the research assistant’s critical feedback and this may have lessened the magnitude of the results.  Similarly, although the research assistant made every attempt to give equally critical feedback to every participant, the conversations in the feedback  50 period likely went in different directions based on participant responses.  It is also likely that, due to the lack of available cues in text messaging, some participants may not have perceived the critical remarks in a negative manner.  For example, some may have taken the RA’s comments to be helpful, constructive feedback.  Further protocol and design issues will be discussed in the limitations section to follow. Research Question 2: Moderation role of Rejection Sensitivity and Rumination  Another major goal of the current study was to determine if there were differences in affect or stress that resulted from individual differences variables – specifically rejection sensitivity and trait rumination.  Due to some of the inherent differences between text messaging and face-to-face communication (e.g., text messaging conversations often occur over a more extended period of time and lack tone, which can lead to potential miscommunications, etc.), it was hypothesized that those higher in rejection sensitivity and rumination in the text message criticism condition would have reported significantly lower positive affect and significantly higher negative affect and stress, compared to those lower in these traits.  Moderation analyses did not reveal any significant interaction effects.  However, there was a main effect of rejection sensitivity on levels of negative affect (but not positive affect or stress).  This finding is not surprising given that past research on rejection sensitivity has shown a relationship to negative interpersonal outcomes and a tendency to respond more negatively to social stressors (Downey & Feldman, 1996; Downey et al., 1998).  There was also a near significant main effect of rumination on positive affect (but not negative affect or stress).  Although non-significant, the trend found here is in line with previous research that has shown that those higher in trait rumination experience increased distress following a difficult event (Nolen-Hoeksema & Davis, 1999).    51 Limitations and Future Directions One of the main limitations of the current study may also have been one of its strengths.  Although the use of an experimental laboratory-based paradigm is ideal for controlling extraneous variables, it may have resulted in the criticism feeling less severe than it might in a real world setting.  The fact that the participants had no prior relationship with the RA may have resulted in a less impactful criticism.  Although an attempt was made to establish rapport between the RA and participant at the beginning of the study protocol, it is likely that there is a tendency for people to take criticism from a friend, relative, or romantic partner much more seriously than from a stranger.  Additionally, participants may have felt that the feedback (criticism) provided by the RA was contrived, especially for those who did better on the stress task.  To further this problem, the participants may have viewed the task as unimportant and therefore been less impacted by the criticism.  These issues likely contributed to the small and sometimes non-significant effects of the criticism on participants’ mood and stress.  One reason why significant effects were found for positive, but not negative affect, could have been that the negative affect items assessed here (anxiety and hostility) may not have accurately captured feelings associated with criticism.  Terms such as “ashamed”, “embarrassed”, or “guilty” in the context of criticism might have more accurately captured the participants’ experiences and led to larger effects.  As mentioned above, issues related to homogeneity of regression slopes could also have impacted the present findings.   Every effort was made to ensure that the criticism in the face-to-face and text message conditions were as similar as possible in both length and content.  This consistency was meant to allow for a direct comparison that likely wouldn’t be possible in a “real world” setting.  The RA was given specific things to say in the criticism, but was also to attempt to tailor the feedback to each participant based on the notes from the committee so that it was not obvious that it was scripted.  While personalization of the criticism likely made the feedback more realistic, this  52 resulted in slightly different content for each participant.  As outlined in SIP theory, difficulties with consistency between the criticism conditions arose from the different time periods required to hold conversations face-to-face versus through text message.  Due to the different pace that conversations run face-to-face and via text, many of the sessions for the text messaging group were longer in time (11-12 minutes versus 7-8 minutes), yet likely contained less content.  This may have led to an underestimate of the effects of text message criticism on emotional wellbeing for a couple of reasons.  First, having the criticism period extended may have allowed for more time for the participant to regress towards baseline levels of stress and affect (i.e., they had a longer period between the stress task and the final measures).  This may have eliminated any differences that may have arose between the face-to-face and text message groups. Second, the limited content in the text message groups likely reduced the effect of the criticism on levels of affect and stress.  However, it would be difficult to increase the amount of content without extending the time making these two issues very difficult to address in a laboratory based experiment (i.e., attempts to adjust for one contribute to more problems with the other).   Given how people interact in a naturalistic setting it is important to consider the differences between these types of communication before drawing any conclusions related to the moderation analyses.  An argument that takes place face-to-face is likely to resolve (or at least end for a period) much quicker than one via text message.  An argument that lasts for an hour in person may last throughout the entire day if the only available option is text messaging.  There is likely to be more opportunities for miscommunication during this time.  This creates a drawn out period that is potentially harmful for people higher in rejection sensitivity or rumination.  Furthermore, the exclusion of any participants who had been diagnosed with a psychological disorder may have excluded people who were higher on these traits and may have been more dramatically impacted by the criticism.  Although this was done for the safety of the participants, it is unlikely that these people avoid the use of text messaging.   53 Another limitation of the current study, and a possible reason for lack of significant findings, could be lack of statistical power as a consequence of a somewhat small sample size.  This was likely most problematic for the moderation analyses.  Although the current study may have benefitted from the inclusion of additional participants, it should be noted that in comparison with other studies using similar protocols (e.g., the TSST) or looking at similar questions (i.e., CMC vs face-to-face) the sample size was fairly large.  Previous studies have typically included well under 100 participants (for example, Kirschbaum et al., 1993; Kirschbaum et al., 1999; Seltzer et al., 2012; and Sherman et al., 2013).  Despite the relatively large sample size of the current study, it still may not have been large enough to detect differences in the overall impact of criticism provided by a confederate or the moderating effects of rumination or rejection sensitivity.   Future Directions Finding the right balance between creating a realistic setting in the lab and keeping stringent control over extraneous variables can be difficult.  While the current study focused on one type of computer-mediated-communication (text messaging), it is likely that many young adults do not rely solely on this type of communication.  The use of multiple modes of communication makes it challenging to isolate the effects of a single mode of communication and remain realistic.  Isolation of single communication modalities is likely not reflective of real life and therefore may not lead to results that are generalizable to real world situations.  This is compounded by information and communication technologies (ICT) succession theory, which posits that use of multiple modes of communication results in the most effective forms of communication (Walther, 2011).  This suggests that, although text messaging is likely to have some weaknesses (such as the lack of tone/potential for miscommunication, etc.) individuals can use other types of communication to compensate for this.  For example, a conversation that starts via text message may be resolved with a phone call if it becomes negative.  Future studies would  54 benefit from the inclusion of multiple modes of communication as this is likely more reflective of how people interact.  However, as mentioned above this makes isolation of effects difficult, and it may make it difficult to extract meaningful results.   In future research, a more ecologically-valid approach to capturing the effects of negative interactions via text-messaging would be to use intensive longitudinal designs, which would allow for the naturalistic assessment of text message use in day-to-day life.  Another consideration to increase ecological validity in a similar TSST paradigm would be to include a close friend, romantic partner, or parent to provide the criticism.  However, this raises additional issues related to control and consistency with different types of relationships, different communications styles, and differing levels of connectedness, closeness, or superiority in the relationship.   The findings in this study run counter to some past findings that support the CFO theories.  Specifically, Seltzer et al. (2012) and Sherman et al. (2013) found that for positive, supportive interactions, digitally-mediated communication was less effective at eliciting user satisfaction and creating feelings of bonding and support.  On the other hand, the current study indicates that the mode of communication may not matter when engaging in negative exchanges.   This is consistent with past research and theory suggesting that humans have evolved to be highly sensitive to interpersonal threat and rejection.  Since interpersonal threat may be processed at a more unconscious, automatic level than positive social interactions, the specific method of delivery of that threat may be less important (Smith & Williams, 2004; Wesselmann et al., 2012).   Future studies are needed to explore how the impact of text messaging may differ, depending on whether interactions are positive or negative.   Future studies may also benefit from ensuring recruitment of a more diverse sample and consideration of other potential moderating variables.  For example, the current sample included only those between the ages of 18 and 25, and the majority of participants were university  55 students.  Greater variance in the scores on mood, stress, and moderator variables could have been obtained by including young adults who are not currently attending university, as well as other age groups, such as adolescents and older adults.  This would also have the benefit of increasing the generalizability of the current findings, as well as allow for an examination of whether there are age differences in the way in which criticism via text messaging may impact social and emotional wellbeing.  The moderating role of other relevant variables, such as gender and the perceived importance of computer-mediation communication, also warrant attention in future research.   Conclusions  Text messaging is becoming increasingly ubiquitous for young adults.  However, the impact that this technology has on its users is currently not entirely understood.  Although there is evidence of its ability to create positive outcomes, the current study highlights its potential to lead to negative emotional outcomes.  Although the results indicate that criticism may lead to similar outcomes regardless of the mode of communication (face-to-face or via text message), it is important to consider that past studies have highlighted a number of issues associated with text messaging that may exacerbate the effects of negative interactions.  First, several studies have found that people are willing use text messaging to say negative things that they would not say if the interaction took place face-to-face (Allen, 2012; Patchin & Hinduja, 2006; Reid & Reid, 2010).  Considering that these statements may be just as impactful as if they were exchanged face-to-face, text messaging may be leading to conflict, stress, or reduced positive affect that may have otherwise been avoided.  These findings may be understood in the context of the Social Identity Model of Deindividuation Effects (SIDE; Reicher, Spears, & Postmes, 1995) This model purports that in online interactions, visual anonymity leads to a dehumanizing of the people receiving communications.  Individuals therefore may not be acutely aware of the effect  56 of their hurtful comments or may be less considerate of the feeling of the other person while texting.   The main objective of this study was to examine the impact of criticism via text message and face-to-face communication on levels of positive affect, negative affect, and stress following an acute laboratory stressor and to determine if there were differences in affect or stress that resulted from individual differences in rumination and rejection sensitivity.  To the author’s knowledge, this was the first study to examine the role of critical feedback of this nature in a controlled laboratory setting with the inclusion of the aforementioned variables.  The findings outlined here warrant further investigation in order to better understand the complex nature of young adults’ communication patterns.  Further research in this area is likely to help identify individuals who are more adversely affected by text messaging.  Psychoeducational interventions related to the appropriate use of this technology may help to avoid unnecessary distress and relational strain among young adults, who are at a critical point in their emotional and social development.       57 References Afifi, T., Walid Afifi, Anne F.  Merrill, Amanda Denes, & Sharde Davis.  (2013).  ‘You need to stop talking about this!’: Verbal rumination and the costs of social support.  Human Communication Research, 39,(4): 395–421.  doi:10.1111/hcre.12012. Allen, K.  P.  (2012).  Off the radar and ubiquitous: Text messaging and its relationship to ‘drama’ and cyberbullying in an affluent, academically rigorous US high school.  Journal of Youth Studies, 15,(1), 99–117. Angster, A., Frank, M., & Lester, D.  (2010).  An exploratory study of students’ use of cell phones, texting, and social networking sites.  Psychological Reports, 107(2), 402–404.  doi:10.2466/17.PR0.107.5.402-404. Arnett, J.  J.  (2000).  Emerging adulthood: A theory of development from the late teens through the twenties.  American Psychologist 55(5), 469–80.  doi:10.1037//0003-066X.55.5.469. Baumeister, R.  F., & Leary, M.  R.  (1995).  The need to belong: Desire for interpersonal attachments as a fundamental human motivation.  Psychological Bulletin, 117(3), 497–529.  http://www.ncbi.nlm.nih.gov/pubmed/7777651. Bolger, N., DeLongis, A., Kessler, R.  C., & Schilling, E.  (1989).  Effects of daily stress on negative mood.  Journal of Personality and Social Psychology, 57(5), 808–818.  doi:10.1037//0022-3514.57.5.808. Butler, J.  C., Doherty, M.  S., & Potter, R.  M.  (2007).  Social antecedents and consequences of interpersonal rejection sensitivity.  Personality and Individual Differences, 43(6), 1376–85.  doi:10.1016/j.paid.2007.04.006. Cacioppo, J.  T., Hawkley, L.  C., Crawford, L.  E., Ernst, J.  M., Burleson, M.  H., Kowalewski, R., … Berntson, G.  G.  (2002).  Loneliness and health: Potential mechanisms.  Psychosomatic Medicine, 64, 407-417. Canadian Wireless Telecommunications Association.  (2013).  Facts and figures.  Retrieved  58 from  http://cwta.ca/facts-figures/ Carlson, J.  R., & Zmud, R.  W.  (1999).  Channel expansion theory and the experiential nature of media richness perceptions.  Academy of Management Journal, 42, 153-170.   Choi, M., & Toma, C.  L.  (2014).  Social sharing through interpersonal media: Patterns and effects on emotional well-being.  Computers in Human Behavior, 36, 530–541.  doi:10.1016/j.chb.2014.04.026. Cohen, S., Frank, E., Doyle, W.  J., Skoner, D.  P., Rabin, B.  S., & Gwaltney, J.  M.  (1998).  Types of stressors that increase susceptibility to the common cold in healthy adults.  Health Psychology : Official Journal of the Division of Health Psychology, American Psychological Association, 17(3), 214–223.  http://www.ncbi.nlm.nih.gov/pubmed/9619470. Cohen, S., & Wills, T.  A.  (1985).  Stress, social support, and the buffering hypothesis.  Psychological Bulletin, 98(2), 310–57. Cohen, S.  (2004).  Social relationships and health.  American Psychologist, 59(8), 676-684. Coviello, L., Sohn, Y., Kramer, A.  D.  I., Marlow, C., Franceschetti, M., Christakis, N.  A., & Fowler, J.  H.  (2014).  Detecting emotional contagion in massive social networks.  PloS One, 9(3), e90315.  doi:10.1371/journal.pone.0090315. Coyne, S.  M., Stockdale, L., Busby, D., Iverson, B., & Grant, D.  M.  (2011).  ‘I Luv U :)!’: A descriptive study of the media use of individuals in romantic relationships.  Family Relations 60(2), 150–162.  doi:10.1111/j.1741-3729.2010.00639.x. Culnan, M., & Markus, M.  L.  (1987).  Information technologies.  In F.  M.  Jablin, L.  L.  Putnam, K.  H.  Roberts, & L.  W.  Porter, (Eds.), Handbook of organizational communication: An interdisciplinary perspective (pp.  420-444).  Newbury Park, CA: Sage. Cupples, J., & Thompson, L..  (2010).  Heterotextuality and digital foreplay.  Feminist Media Studies 10(1), 1–17.  doi:10.1080/14680770903457063.  59 Davie, R., Panting, C., & Charlton, T.  (2004).  Mobile phone ownership and usage among pre-adolescents.  Telematics and Informatics 21(4): 359–73.  doi:10.1016/j.tele.2004.04.001. DeClerck, D., Meisner, K., & Holtzman, S.  (2014, June).  Young adults' views of text messaging and relationships: A qualitative study.  Poster session at the Canadian Psychological Association Convention, Vancouver, B.C. Derogatis, L.  R.  (1996) Derogatis Affects Balance Scale (DABS): Preliminary scoring, procedures & administration manual.  Baltimore, MD, Clinical Psychometric Research. Desanctis, G., & Scott, M.  (1994).  Capturing the complexity in advanced technology use : Adaptive Structuration Theory.  Organization Science, 5(2), 121–148.  doi = 10.1287/orsc.5.2.121 Downey, G, & Feldman, S.  I.  (1996).  Implications of rejection sensitivity for intimate relationships.  Journal of Personality and Social Psychology 70(6), 1327–1343.  doi:10.1037/0022-3514.70.6.1327 Downey, G., Freitas, A.  L., Michaelis, B., & Khouri, H.  (1998).  The self-fulfilling prophecy in close relationships : Rejection sensitivity and rejection by romantic partners.  Journal of Personality and Social Psychology, 75(2), 545–560.  doi:10.1037/0022-3514.75.2.545  D'Urso, S.  C., & Rains, S.  A.  (2008).  Examining the scope of channel expansion: A test fo channel expansion theory with new and traditional communication media.  Management Quarterly, 21, 486-507.  doi:10.1177/0893318907313712 Ewart, C.  K., Taylor, C.  B., Kraemer, H.  C., & Agras, W.  S.  (1991).  High blood pressure and marital discord: Not being nasty matters more than being nice.  Health Psychology, 10(3), 155–163.  doi:10.1037//0278-6133.10.3.155. Faulkner, X., & Culwin, F.  (2005).  When fingers do the talking: A study of text messaging.  Interacting with Computers, 17(2), 167–185.  doi:10.1016/j.intcom.2004.11.002.  60 Feldner, M.  T., Leen-Feldner, E.  W., Zvolensky, M.  J., & Lejuez, C.  W.  (2006).  Examining the association between rumination, negative affectivity, and negative affect induced by a paced auditory serial addition task.  Journal of Behavior Therapy and Experimental Psychiatry, 37(3), 171–187.  doi:10.1016/j.jbtep.2005.06.002. Field, A.  (2009).  Discovering statistics using SPSS (3rd Ed.) London, UK: Sage Publications Limited. Gerber, J., & Wheeler, L.  (2009).  A meta-analysis of experimental research on rejection.  Perspectives on Psychological Science, 4(5), 468–489.  doi:10.1111/j.1745-6924.2009.01158.x Gilbert, K.  L., Quinn, S.  C., Goodman, R.  M., Butler, J., & Wallace, J.  (2013).  A meta-analysis of social capital and health: A case for needed research.  Journal of Health Psychology, 18(11), 1385–1399.  doi:10.1177/1359105311435983. Hall, J.  A., & Baym, N.  K.  (2012).  Calling and texting (too much): Mobile maintenance expectations, (over)dependence, entrapment, and friendship satisfaction.  New Media & Society, 14(2), 316–331.  doi:10.1177/1461444811415047. Hay, C., Meldrum, R., & Mann, K.  (2010).  Traditional bullying, cyber bullying, and deviance: A general strain theory approach.  Journal of Contemporary Criminal Justice, 26(2), 130–147.  doi:10.1177/1043986209359557  Hofferth, S.  L., & Moon, U.  J.  (2012).  Cell phone use and child and adolescent reading proficiency.  Psychology of Popular Media Culture, 1(2), 108–122.  doi:10.1037/a0027880. Holtgraves, T.  (2011).  Text messaging, personality, and the social context.  Journal of Research in Personality, 45(1), 92–99.  doi:10.1016/j.jrp.2010.11.015. Igarashi, T., & Motoyoshi, T.  (2008).  No Mobile , No Life : Self-perception and text-message dependency among Japanese high school students.  Computers in Human Behavior, 24, 2311–2324.  doi:10.1016/j.chb.2007.12.001.  61 Joormann, J., Dkane, M., & Gotlib, I.  H.  (2006).  Adaptive and maladaptive components of rumination? Diagnostic specificity and relation to depressive biases.  Behavior Therapy, 37(3), 269–280.  doi:10.1016/j.beth.2006.01.002. Kamibeppu, K., & Sugiura, H.  (2005).  Impact of the mobile phone on junior high-school students’ friendships in the Tokyo metropolitan area.  Cyberpsychology & Behavior : The Impact of the Internet, Multimedia and Virtual Reality on Behavior and Society, 8(2), 121–130.  doi:10.1089/cpb.2005.8.121. Kiecolt-glaser, J.  K., Malarkey, W.  B., & Chee, M.  (1993).  Negative behavior during marital conflict is associated with immunological down-regulation.  Psychosomatic Medicine, 55, 395–409.  doi:10.1097/00006842-199309000-00001 Kirschbaum, C., Pirke, K.  M., & Hellhammer, D.  H.  (1993).  The ‘Trier Social Stress Test’ - A tool for investigating psychobiological stress responses in a laboratory setting.  Neuropsychobiology, 28, 76-81.  doi:10.1159/000119004 Korzenny, F.  (1978).  A theory of electronic propinquity: Mediated communication in organizations.  Communication Research, 5, 3-24.  doi:10.1177/009365027800500101 Kothgassner, O.  D., Kafka, J.  X., & Rudyk, J.  (2014).  Does social exclusion hurt virtually like it hurts in real-life ? The role of agency and social presence in the perception and experience of social exclusion.  Proceedings of the International Sociaety fo Presence Research.  45–56. Kross, E., Verduyn, P., Demiralp, E., Park, J., Lee, D.  S., Lin, N., … Ybarra, O.  (2013).  Facebook use predicts declines in subjective well-being in young adults.  PloS One, 8(8), e69841.  doi:10.1371/journal.pone.0069841. Kudielka, B.  M., Hellhammer, D.  H., & Kirschbaum, C.  Ten years of research with the Trier Social Stress Task - Revisited.  In E.  Harmon-Jones & P.  Winkielman (Eds.), Social  62 neuroscience: Integrating biological and psychological explanations of social behavior, (pp.  56–83).  New York: The Guilford Press. Leatherdale, S.  T.  (2010).  Factors associated with communication-based sedentary behaviors among youth: Are talking on the phone, texting, and instant messaging new sedentary behaviors to be concerned about? The Journal of Adolescent Health : Official Publication of the Society for Adolescent Medicine 47(3), 315–18.  doi:10.1016/j.jadohealth.2010.02.012. Lee, Y.  K., Chang, C.  T., Lin, Y.  & Cheng, Z.  H.  (2014).  The dark side of smartphone usage: Psychological traits, compulsive behavior and technostress.  Computers in Human Behavior, 31, 373–83.  doi:10.1016/j.chb.2013.10.047. Lenhart, A.  (2010).  Cell phones and American adults.  1–42.  Retrieved from the Pew Research Internet Project Website: http://www.pewinternet.org/2010/09/02/cell-phones-and-american-adults/  Lenhart, A., Ling, R., Campbell, S., & Purcell, K.  (2010).  Teens and mobile phones: text messaging explodes as teens embrace it as the centerpiece of their communication strategies with friends.  Retrieved from the Pew Research Internet Project Website: http://www.pewinternet.org/2010/04/20/teens-and-mobile-phones/ Lepp, A., Barkley, J.  E., & Karpinski, A.  C.  (2014).  The relationship between cell phone use, academic performance, anxiety, and satisfaction with life in college students.  Computers in Human Behavior, 31, 343–350.  doi:10.1016/j.chb.2013.10.049. Lyubomirsky, S., Kasri, F., & Zehm, K.  (2003).  Dysphoric rumination impairs concentration on academic tasks.  Cognitive Therapy and Research, 27(3), 309–30.  doi: 0147-5916/03/0600-0309/0 Manago, A.  M., Taylor, T., & Greenfield, P.  M.  (2012).  Me and My 400 Friends: The anatomy of college students’ facebook networks, their communication patterns, and well-being.  Developmental Psychology, 48(2), 369–380.  doi:10.1037/a0026338.  63 Masten, C.  L., Telzer, E.  H., Fuligni, A.  J., Lieberman, M.  D., & Eisenberger, N.  I.  (2012).  Time spent with friends in adolescence relates to less neural sensitivity to later peer rejection.  Social Cognitive and Affective Neuroscience, 7(1), 106–114.  doi:10.1093/scan/nsq098. Murdock, K.  K.  (2013).  Texting while stressed: Implications for students’ burnout, sleep, and well-being.  Psychology of Popular Media Culture, 2(4), 207–221.  doi:10.1037/ppm0000012. Nolen-Hoeksema, S., & Davis, C.  G.  (1999).  "Thanks for sharing that": Ruminators and their social support networks.  Journal of Personality and Social Psychology, 77(4), 801-814.  doi:10.1037/0022-3514.77.4.801 Ogara, S.  O., Koh, C.  E., & Prybutok, V.  R.  (2014).  Investigating factors affecting social presence and user satisfaction with mobile instant messaging.  Computers in Human Behavior, 36, 53–59.  doi:10.1016/j.chb.2014.03.064. Ogletree, S.  M., Fancher, J., & Gill, S.  (2014).  Gender and texting: Masculinity, femininity, and gender role ideology.  Computers in Human Behavior, 37, 49–55.  doi:10.1016/j.chb.2014.04.021. Patchin, J.  W.  (2006).  Bullies move beyond the schoolyard: A preliminary look at cyberbullying.  Youth Violence and Juvenile Justice, 4(2), 148–169.  doi:10.1177/1541204006286288. Pettigrew, J.  (2009).  Text messaging and connectedness within close interpersonal relationships.  Marriage & Family Review, 45(6-8): 697–716.  doi:10.1080/01494920903224269. Pierce, T.  (2009).  Social anxiety and technology: Face-to-face communication versus technological communication among teens.  Computers in Human Behavior, 25(6), 1367–1372.  doi:10.1016/j.chb.2009.06.003.  64 Pollet, T.  V., Roberts, S.  G., & Dunbar, R.  I.  (2011).  Use of social network sites and instant messaging does not lead to increased offline social network size, or to emotionally closer relationships with offline network members.  Cyberpsychology, Behavior and Social Networking, 14(4), 253–258.  doi:10.1089/cyber.2010.0161. Reicher, S.  D., Spears, R., & Postmes, T.  (1995).  A social identity model of deindividuation phenomena.  European Review of Social Psychology, 6, 161–198.  doi:10.1080/14792779443000049 Reid, F.  J., & Reid, D.  J.  (2010).  The expressive and conversational affordances of mobile messaging.  Behaviour & Information Technology, 29(1), 3–22.  doi:10.1080/01449290701497079. Robles, T.  F., & Kiecolt-Glaser, J.  K.  (2003).  The physiology of marriage: Pathways to health.  Physiology & Behavior, 79(3), 409–416.  doi:10.1016/S0031-9384(03)00160-4. Rook, K.  S.  (2001).  Emotional health and positive versus negative social exchanges: A daily diary analysis.  Applied Developmental Science, 5(2): 86–97.  doi:10.1207/s1532480xads0502_4 Rosengren, A., Orth-Gomér, K., Wedel, H., & Wilhelmsen, L.  (1993).  Stressful life events, social support, and mortality in men born in 1933.  BMJ (Clinical Research Ed.), 307(6912), 1102–1105.  doi:10.1136/bmj.307.6912.1102 Salmela-Aro, K., Aunola, K., & Nurmi, J.  (2008).  Trajectories of depressive symptoms during emerging adulthood: Antecedents and consequences.  European Journal of Developmental Psychology, 5(4), 439–65.  doi:10.1080/17405620600867014. Seltzer, L.  J., Prososki, A.  R., Ziegler, T.  E., & Pollak, S.  D.  (2012).  Instant messages vs.  speech: Hormones and why we still need to hear each other.  Evolution and Human Behavior, 33(1), 42–45.  doi:10.1016/j.evolhumbehav.2011.05.004.  65 Shallow, M., DeClerck, D., Hudson, S., Rempel, A.  & Holtzman, S.  (2014, June).  Text messaging: Costs and benefits for socially anxious young adults.  Poster session at the Canadian Psychological Association Convention, Vancouver, B.C. Sherman, L.  E., Michikyan, M., & Greenfield, P.  M.  (2013).  The effects of text, audio, video, and in-person communication on bonding between friends.  Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 7(2).  doi:10.5817/CP2013-2-3. Skierkowski, D., & Wood, R.  M.  (2012).  To text or not to text? The importance of text messaging among college-aged youth.  Computers in Human Behavior, 28(2), 744–756.  doi:10.1016/j.chb.2011.11.023. Smith, A.  (2014).  Older Adults and Technology Use: Adoption Is Increasing, but Many Seniors Remain Isolated from Digital Life.  Retrieved from the Pew Research Internet Project Website:  http://www.pewinternet.org/files/2014/04/PIP_Seniors-and-Tech-Use_040314.pdf Smith, A.  (2011).  Americans and Text Messaging.  Retrieved from the Pew Research Internet Project Website: http://www.pewinternet.org/2011/09/19/americans-and-text-messaging/ Smith, A., & Williams, K.  D.  (2004).  R U there? Ostracism by cell phone text messages.  Group Dynamics: Theory, Research, and Practice, 8(4): 291–301.  doi:10.1037/1089-2699.8.4.291. Smith, J.  M., & Alloy, L.  B.  (2009).  A roadmap to rumination: A review of the definition, assessment, and conceptualization of this multifaceted construct.  Clinical Psychology Review, 29(2), 116–128.  doi:10.1016/j.cpr.2008.10.003. Song, H., Zmyslinski-Seelig, A., Kim, J., Drent, A., Victor, A., Omori, K., & Allen, M.  (2014).  Does facebook make you lonely?: A meta analysis.  Computers in Human Behavior, 36, 446–452.  doi:10.1016/j.chb.2014.04.011.  66 Sultan, A.  J.  (2014).  Addiction to mobile text messaging applications is nothing to ‘lol’ about.  The Social Science Journal, 51(1), 57–69.  doi:10.1016/j.soscij.2013.09.003. Tabachnick, B.  G., & Fidell, L.  S.  (2007).  Using multivariate statistics, Fifth Edition (pp.  60-116).  Boson, MS: Pearson. Takao, M., Takahashi, S., & Kitamura, M.  (2009).  Addictive personality and problematic mobile phone use.  Cyberpsychology & Behavior : The Impact of the Internet, Multimedia and Virtual Reality on Behavior and Society, 12(5), 501–507.  doi:10.1089/cpb.2009.0022. Thomée, S., Härenstam, A., & Hagberg, M.  (2011).  Mobile phone use and stress, sleep disturbances, and symptoms of depression among young adults--a prospective cohort study.  BMC Public Health, 11(1), 66.  doi:10.1186/1471-2458-11-66. Thompson, L., & Cupples, J.  (2008).  Seen and not heard? Text messaging and digital sociality.  Social & Cultural Geography, 9(1): 95–108.  doi:10.1080/14649360701789634. Trapnell, P.  D., & Campbell, J.  D.  (1999).  Private self-consciousness and the five-factor model of personality: Distinguishing rumination from reflection.  Journal of Personality and Social Psychology, 76(2), 284–304.  doi:10.1037//0022-3514.76.2.284. Twenge, J.  M., Catanese, K.  R., & Baumeister, R.  F.  (2002).  Social exclusion causes self-defeating behavior.  Journal of Personality and Social Psychology, 83(3): 606–615.  doi:10.1037//0022-3514.83.3.606. Uchino, B.  N.  (2009).  Understanding the Links Between Social Support and Physical Health: A Life-Span Perspective With Emphasis on the Separability of Perceived and Received Support.  Perspectives on Psychological Science, 4(3), 236–255.  doi:10.1111/j.1745-6924.2009.01122.x  Underwood, Marion K., Lisa H.  Rosen, David More, Samuel E.  Ehrenreich, & Joanna K.  Gentsch.  2012.  The BlackBerry Project: Capturing the Content of Adolescents’ Text Messaging.  Developmental Psychology 48 (2): 295–302.  doi:10.1037/a0025914.  67 Walsh, S., White, K.  M., & Young, R.  (2010).  Needing to connect: The effect of self and others on young people’s involvement with their mobile phones.  Australian Journal of Psychology, 62(4), 194–203.  doi:10.1080/00049530903567229. Walther, J.  B.  (1992).  Interpersonal effects in computer-mediated interaction: A relational perspective.  Communication Research, 19(1), 52–90.  doi:10.1177/009365092019001003. Walther, J.  B.  (2011).  Theories of computer-mediated communication and interpersonal relations.  In M.  L.  Knapp & J.  A.  Daly (Eds.), The Handbook of Interpersonal Communication (4th ed., pp.  443-479).  Thousand Oaks, CA: Sage. Walther, J.  B., Anderson, J.  F., & Park, D.  W.  (1994).  Interpersonal effects in computer-mediated interaction: A meta-analysis of social and antisocial communication.  Communication Research, 21(4), 460–487.  doi:10.1177/009365094021004002. Walther, J.  B., & Bazarova, N.  (2008).  Validation and application of electronic propinquity theory to computer-mediated communication in groups.  Communication Research, 35, 622-645.  doi:10.1177/0093650208321783 Wesselmann, E.  D., Nairne, J.  S., & Williams, K.  D.  (2012).  An evolutionary social psychological approach to studying the effects of ostracism.  Journal of Social, Evolutionary, and Cultural Psychology, 6(3), 309–328.  doi:10.1037/h0099249  White, B.  A., & Kistner, J.  K.  (2011).  Biased self-perceptions, peer rejection, and aggression in children.  Journal of Abnormal Child Psychology, 39(5), 645–656.  doi:10.1007/s10802-011-9506-6. Williams, K.  D., Cheung, C.  K., & Choi, W.  (2000).  Cyberostracism: Effects of being ignored over the internet.  Journal of Personality and Social Psychology, 79(5), 748–62.  http://www.ncbi.nlm.nih.gov/pubmed/11079239. Williams, K.  D.  (2007).  Ostracism.  Annual Review of Psychology, 58, 425–452.  doi:10.1146/annurev.psych.58.110405.085641.  68 Williams, K.  D., Govan, C.  L., Croker, V., Tynan, D., Cruickshank, M., & Lam, A.  (2002).  Investigations into differences between social- and cyberostracism.  Group Dynamics: Theory, Research, and Practice, 6(1), 65–77.  doi:10.1037//1089-2699.6.1.65. Williams, K., & Zadro, L.  (2001).  Ostracism: On being ignored.  In Mark R Leary (Eds.), Interpersonal Rejection, (pp.  21-53).  New York: Oxford University Press. Wirth, J.  H., & Williams, K.  D.  (2009).  `They don’t like our kind': Consequences of being ostracized while possessing a group membership.  Group Processes & Intergroup Relations, 12(1), 111–127.  doi:10.1177/1368430208098780. Zadro, L., Williams, K.  D., & Richardson, R.  (2004).  How low can you go? Ostracism by a computer is sufficient to lower self-reported levels of belonging, control, self-esteem, and meaningful existence.  Journal of Experimental Social Psychology, 40(4), 560–567.  doi:10.1016/j.jesp.2003.11.006. Zimmer-Gembeck, M.  J., Nesdale, D., McGregor, L., Mastro, S., Goodwin, B., & Downey, G.  (2013).  Comparing reports of peer rejection: Associations with rejection sensitivity, victimization, aggression, and friendship.  Journal of Adolescence, 36(6), 1237–1246.  doi:10.1016/j.adolescence.2013.10.002.      69 Appendix A - Timeline of the 75-minute Protocol 0 Minutes:  Participant arrives, is introduced to the project and completes the informed consent procedure. 5 Minutes:  Acclimation (rest) period (15 minutes) 20 Minutes: Self-report measures (Baseline) 25 Minutes: Participant is introduced to stress task 30 Minutes: Stress task 40 Minutes: Self-report measures (Post-stress) 45 Minutes: Support/control condition 55 Minutes: Self-report measures (Post-feedback) 75 Minutes: Compensation/debriefing  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0223353/manifest

Comment

Related Items