UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

When saving seems like the right choice : the role of utility and space in hoarding disorder Kellman-McFarlane, Kirstie 2017

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2018_february_kellman-mcfarlane_kirstie.pdf [ 1.18MB ]
Metadata
JSON: 24-1.0362867.json
JSON-LD: 24-1.0362867-ld.json
RDF/XML (Pretty): 24-1.0362867-rdf.xml
RDF/JSON: 24-1.0362867-rdf.json
Turtle: 24-1.0362867-turtle.txt
N-Triples: 24-1.0362867-rdf-ntriples.txt
Original Record: 24-1.0362867-source.json
Full Text
24-1.0362867-fulltext.txt
Citation
24-1.0362867.ris

Full Text

	 WHEN SAVING SEEMS LIKE THE RIGHT CHOICE: THE ROLE OF UTILITY AND SPACE IN HOARDING DISORDER  by KIRSTIE KELLMAN-MCFARLANE M.A., The University of British Columbia, 2013   A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIRMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Psychology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2017  © Kirstie Kellman-McFarlane, 2017  	 ii	Abstract Despite substantial advances in research on hoarding disorder over the past decade, the mechanisms driving hoarding behaviour remain poorly understood. It is well established that people who hoard feel justified in keeping objects that most individuals would consider worthless. However, there has been little empirical investigation of the decision-making process that provides people who hoard with justification to save objects that most people would discard. The current research proposed that people who hoard are routinely biased towards saving because they engage in a decision-making process that excessively prioritizes factors that favour keeping possessions (i.e., potential future utility) and neglects factors that would normatively encourage discarding (i.e., infrequent past usage, lack of household space, object dysfunction). This model builds on previous research suggesting that people who hoard are abnormally concerned about needing possessions in the future, and Frost and Hartl’s (1996) theoretical proposition that people who hoard have a higher threshold for discarding useless objects.  To test this hypothesized model, the current research developed three novel decision-making tasks about household possessions: the Information Seeking Task, the Vignette-Based Task, and the Dysfunction Tolerance Scale. Once validated, these tasks were completed by a community sample (N = 174) of individuals with varying levels of hoarding symptoms. Analyses examined the relationship between task performance and hoarding as both a categorical (i.e., hoarding disorder (n = 53) vs. healthy control (n = 76)) and dimensional construct (i.e., severity of self-reported hoarding symptoms).  As predicted, when making discarding decisions, hoarding participants were less responsive to information that would decrease the perceived value of objects, such as 	 iii	infrequent use, poor condition, and lack of household space. The hoarding group also displayed a greater preference than the healthy control group for considering information that would promote saving (i.e., potential future uses of objects) when making discarding decisions. Although the results supported a consistent view of individual differences in discarding decision-making in hoarding, several of the effect sizes were small. The current findings provide several promising directions for future research. Moreover, the current work offers practical suggestions for enhancing current hoarding interventions that target maladaptive decision-making about possessions.              	 iv	Lay Summary Despite substantial advances in hoarding research over the past decade, little is known about the decision-making process that leads people who hoard to save objects others would typically discard (e.g., broken electronics, old newspapers, ill-fitting clothing). The current research proposed that people who hoard are more likely to save such possessions because they engage in a decision-making process that excessively prioritizes information that favours saving (i.e., potential future utility) and neglects information that encourages discarding (i.e., low past usage, low household space, object dysfunction). To test this model, the current research developed and administered three novel decision-making tasks to a large community sample (N = 174) of individuals with varying levels of hoarding symptoms. The results strongly supported the hypothesized model, but effects were small. The findings suggest promising directions for future research and enhancing current hoarding interventions that target maladaptive decision-making about possessions.          	 v	Preface The work for this dissertation was conducted under the supervision of Dr. Sheila Woody in the Centre for Collaborative Research on Hoarding. I devised the theory behind this research, developed the experimental tasks, trained undergraduate research assistants to assist with data collection, analysed the data, and wrote this dissertation. Grace Truong programmed one of the experimental tasks within this dissertation, the Information Seeking Task. Funds provided by the Canadian Institutes of Health Research (CIHR) to Sheila Woody, and the University of British Columbia were used to conduct this research. This research was approved by the UBC Behavioural Research Ethics Board, UBC BREB #H15-00612.                 	 vi	Table of Contents Abstract ..................................................................................................................................... ii Lay Summary ........................................................................................................................... iv Preface ....................................................................................................................................... v Table of Contents ..................................................................................................................... vi List of Tables ......................................................................................................................... viii List of Figures .......................................................................................................................... ix Acknowledgements ................................................................................................................... x Chapter 1: Introduction and Literature Review ........................................................................ 1 Background Information About Compulsive Hoarding ....................................................... 5 Decision-Making in Compulsive Hoarding .......................................................................... 9 Reasons for Saving and Discarding .................................................................................... 16 Normative Reasons ............................................................................................................. 16 Reasons for Saving in Hoarding ......................................................................................... 21 Current Research ................................................................................................................. 27 Chapter 2: Methods ................................................................................................................. 31 Design Overview ................................................................................................................ 31 Procedure ............................................................................................................................ 31 Measures ............................................................................................................................. 33 Participants .......................................................................................................................... 39 Chapter 3: Vignette-Based Task ............................................................................................. 46 Task Construction and Refinement ..................................................................................... 46 	 vii	Results ................................................................................................................................. 56 Selection of Final Items for the VBT .............................................................................. 58 Analysis of Utility Vignettes .......................................................................................... 61 Analysis of Space Vignettes ........................................................................................... 65 Discussion of Vignette-Based Task .................................................................................... 69 Chapter 4: Information Seeking Task ..................................................................................... 75 Task Construction and Refinement ..................................................................................... 75 Results ................................................................................................................................. 79 Task Validation ............................................................................................................... 81 Future-Oriented Information Seeking ............................................................................. 82 Discussion of Information Seeking Task ............................................................................ 84 Chapter 5: Dysfunction Tolerance Scale ................................................................................ 91 Task Construction and Refinement ..................................................................................... 91 Results ................................................................................................................................. 95 Discussion of Dysfunction Tolerance Task ........................................................................ 96 Chapter 6: Discussion ........................................................................................................... 100 Review of Study Rationale and Hypotheses ..................................................................... 100 Discussion and Illustration of Integrated Model ............................................................... 105 References ............................................................................................................................. 113 Appendix A: Vignette-Based Task ....................................................................................... 131 Appendix B: Information Seeking Task ............................................................................... 132 Appendix C: Dysfunction Tolerance Scale ........................................................................... 134 	 viii	List of Tables  Table 1: Demographics of the Sample and Each Group ......................................................... 43 Table 2: Means and Standard Deviations of Self-Report Measures ....................................... 45 Table 3: Discarding Scores in Utility and Space Conditions .................................................. 49 Table 4: Discarding Scores in Utility and Space Conditions of Final Vignettes .................... 59 Table 5: Means for Utility Conditions by Diagnostic Group ................................................. 63 Table 6: Means for Space Conditions by Diagnostic Group .................................................. 68                	 ix	List of Figures Figure 1: Graph of Means by Utility Conditions and Diagnostic Group ................................ 64 Figure 2: Graph of Means by Space Conditions and Diagnostic Group ................................. 66                    	 x	Acknowledgements First and foremost, I would like to thank my advisor and mentor Sheila Woody for her dependable guidance, insightful suggestions, kindness, patience, and evident commitment to furthering my development as a researcher and clinician since my arrival at UBC. I would like to thank Todd Handy and David Klonsky for agreeing to be on my committee and for their thoughtful feedback. I would also like to thank Peter Lenkic and Kyle Kotowick for their guidance and assistance when I was learning the software used to analyse the majority of the data collected in this dissertation. The graduate students and undergraduate research assistants who helped to recruit participants and conduct phone screens are also deserving of thanks, as this project would not have been possible without them. Thank you Patricia, Brent, May, and Becca. I would like to thank my family and friends for their support and encouragement throughout my graduate education. Lastly, I would like to thank the University of British Columbia for providing funding for this work.     	 1	Chapter 1: Introduction and Literature Review Compulsive hoarding is a debilitating mental disorder associated with a range of serious risks to the health and safety of those who are affected, their families, and their surrounding communities. Hoarding is characterized by the excessive accumulation of possessions that amass in disorganized piles around the home. The stacks of clutter found in hoarded homes often pose significant risks to residents’ physical welfare by creating the potential for heavy objects to fall and injure residents, facilitating the growth and spread of fires, blocking exits during emergencies, and producing health hazards when unsanitary (e.g., mold, pests). In addition to physical dangers, severe clutter adversely impacts mental well-being through its association with stigma, feelings of shame, social difficulties, family conflict, occupational impairment, and housing insecurity (Frost, Steketee, & Williams, 2000; Tolin, Frost, Steketee, & Fitch, 2008; Tolin, Frost, Steketee, Gray, & Fitch, 2008). One of the central mysteries about hoarding behaviour is that despite the intense hardships conferred by clutter, people who hoard are extremely resistant to letting go of objects most people would consider useless or trivial. When faced with the decision to keep or discard an object most people would quickly and easily let go of, people who hoard typically experience indecisiveness, emotional distress, and ultimately avoidance (Frost & Hartl, 1996). Clearly, there are important differences between people who hoard and those who do not hoard with regard to the process of making decisions about whether to discard their possessions. The current research seeks characterize some aspects of the decision-making process that renders discarding an unfeasible option for people who hoard and thereby inform aspects of treatment attempting to ease difficulty discarding and reduce clutter. 	 2	People who hoard report greater levels of indecisiveness (Fitch, 2011; Frost & Gross, 1993; Grisham et al., 2010; Steketee, Frost, & Kyrios, 2003; Tolin & Villavicencio, 2011; Wincze, Steketee, & Frost, 2007), as well as anxiety and distress when those decisions pertain to objects (Grisham et al., 2010; Hayward, 2011; Luchian, McNally, & Hooley, 2007; Wincze et al., 2007). Moreover, behavioural studies have shown evidence of decision-making differences wherein people who hoard took significantly longer to make decisions about possessions (Grisham et al., 2010; Luchian et al., 2007; Wincze et al., 2007) and made a greater number of ‘keeping’ decisions than non-hoarding individuals when asked to discard possessions (Tolin et al., 2009; Tolin et al., 2012). Despite evidence of the struggle that ensues when people who hoard make decisions about possessions, research has not yet pinpointed where in the decision-making process people who hoard diverge from people who do not hoard such that discarding a trivial object becomes a distressing and ultimately unfeasible option. Overall, studies employing gambling tasks developed to measure decision-making abilities suggest that people who hoard do not suffer from a general deficit in the ability to detect patterns and make advantageous decisions in the long term (Woody, Kellman-McFarlane, & Welsted, 2014). Given that individuals who hoard do not appear to exhibit generally poor decision-making skills, it is possible that their decisions are guided by a different set of considerations that disproportionately favour object retention. The current research will investigate several factors that may unduly and regularly bias decision-making in favour of keeping possessions. The potential reasons for wanting to keep a possession are almost infinite even in normative populations, ranging from simple immediate utility to preserving identity in times of change (Kleine & Baker, 2004).  However, among the many rationales and dysfunctional 	 3	beliefs thought to contribute to difficulty discarding, several studies suggest that people may be particularly motivated to save items for reasons related to instrumental value (i.e., perceived usefulness) and the need to be prepared for the future (Coulter & Ligas, 2003; Frost & Gross, 1993; Nordsletten, de la Cruz, Billoti, & Mataix-Cols, 2013).  The importance placed on the utility of an object by people who hoard may initially seem counterintuitive because clutter often renders many possessions inaccessible in hoarded homes and therefore functionally useless the majority of the time. However, as most items have the potential to be used in some capacity (however trivial), it is always possible for an item to possess some future utility. For example, one is unlikely to have worn a torn shirt in the recent past but one could potentially use it to wipe up a mess or cut up the fabric for re-use in the future. As such, emphasizing possible future utility relative to present or past utility is likely to skew the decision-making process in favour of keeping a possession that would otherwise seem to be disposable. Therefore, one potential explanation for chronic saving is that people who hoard overly focus on potential future uses of possessions and neglect decision criteria reflecting current and past usage that would lead to negative appraisals of utility, making it difficult to justify discarding.  In comparison, people who do not hoard may be more likely to base their discarding decisions on past and current usage of possessions, thereby more easily generating negative appraisals of utility. Although many previous studies of hoarding have noted the preoccupation with possibility of discarding something that may be needed in the future, this decision-making bias has not been explored experimentally.  A second factor that may carry less weight in the decision-making processes of people who hoard is household space. The fact that hoarding is not the norm even in heavily 	 4	consumerist societies such as North America suggests that most individuals work to maintain a balance between their space and the volume of their possessions. By definition, individuals who hoard do not maintain this balance between space and the desire to keep possessions, instead allowing rooms to become so full of possessions that they are no longer functional. Therefore, while individuals who hoard suffer severe ramifications from uncontrolled clutter, they may neglect decision-criteria concerning available space when making discarding decisions about individual items. Currently, only one study, by Preston, Muroff, and Wengrovitz (2009), has experimentally explored individual differences in the impact of space on decisions to keep or discard items. Their results suggest space may have less of an impact on the retention habits of people who tend to acquire more than others. However, aspects of Preston et al.’s experimental design limit its generalizability to individuals with hoarding disorder. The current study will build on their work by investigating whether people with clinical hoarding symptoms exhibit a lesser responsiveness to household space constraints. In sum, the current research seeks to determine whether people who hoard differ from those who do not in how they integrate utility and household space into their decision-making process. I propose that people who hoard overly focus on the future consequences of discarding decisions and neglect decision criteria reflecting past use of possessions. Furthermore, in addition to a greater emphasis on future use, I propose that people who hoard require a much lower level of potential usefulness to justify keeping possessions. Consequently, people who hoard would have a higher threshold for the level of dysfunction needed to justify discarding a possession. The idea of a higher threshold for discarding has been proposed theoretically as a means of explaining the tendency of people who hoard to overvalue seemingly trivial possessions (Frost & Hartl, 1996) but has received limited 	 5	empirical attention. Lastly, I propose that people who hoard are less likely to integrate the relative availability of household space in their decision-making. As the preservation of household space appears to serve as a significant motivator to discard items in normative decisional processes, the relative neglect of such a factor would help to explain the reduced frequency of hoarding decisions that favour discarding.  The current program of research is designed to provide insight into decision-making processes in compulsive hoarding and to inform cognitive behavioural treatment approaches that seek to facilitate discarding attempts and enhance decision-making skills. While current research has shown decision-making to be more difficult and distressing for people with compulsive hoarding, little is known about where in the decision-making process individuals with hoarding disorder diverge from non-hoarding individuals such that discarding is difficult even when intent is present. This research will begin to characterize this dysfunctional decision-making process and guide future research in this area. Furthermore, specifying how decision-making differs in compulsive hoarding will inform the development of treatment approaches that directly target decision-making difficulties specific to hoarding rather than relying on general problem solving skills. Overall, this research has the potential to advance our understanding of compulsive hoarding and to improve current treatment techniques. Background Information About Compulsive Hoarding Hoarding and Other Obsessive-Compulsive Related Disorders  Hoarding disorder has recently been codified as a mental disorder in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM–5; American Psychiatric Association, 2013). The core diagnostic criteria for hoarding disorder in the DSM-5 require 	 6	that individuals experience persistent difficulty discarding or parting with possessions, regardless of objective value, and that this difficulty be due to a perceived need to save these items or distress associated with discarding them. Furthermore, these symptoms must result in the accumulation of possessions that congest and clutter living areas (e.g., bedroom, living room, kitchen) and substantially compromise their intended use (e.g., difficulty moving through home, being unable to sleep comfortably in the bed because it is cluttered with possessions). Living areas may be uncluttered if third parties, such as family members, often intervene to reduce clutter. These symptoms must result in clinically significant distress or impairment in social, occupational, or other important areas of functioning. As low insight and excessive acquiring are common issues in hoarding, these characteristics are included as specifiers to the diagnosis. These criteria are consistent with the formulations of hoarding used in many psychopathological, epidemiological, and treatment studies conducted in the decade prior to release of the DSM-5 (Mataix-Cols et al., 2010).  The addition of hoarding disorder to the DSM-5 is due to the growing empirical and clinical recognition of hoarding as a distinct mental disorder (Pertusa et al., 2008; Mataix-Cols et al., 2010; Steketee & Frost, 2003). Prior to this recognition, hoarding behaviour was conceptualized as a subtype of obsessive-compulsive disorder (OCD) or a symptom of obsessive compulsive personality disorder (OCPD). While hoarding behaviours can manifest as a form of OCD, hoarding disorder as described in the DSM-5 occurs without other OCD symptoms and has demonstrated many distinct features in terms of symptom presentation, prevalence, insight, demographics, neuroimaging patterns, and comorbidity. (See Mataix-Cols et al., 2010 for a review.) Unlike the intrusive and disturbing nature of obsessions that characterize OCD (Foa et al., 1995), individuals with hoarding disorder typically experience 	 7	their desires to save possessions as ego-syntonic and display poorer insight into their dysfunctional thinking than individuals with OCD. Moreover, unlike compulsions, saving and acquiring behaviours in hoarding are more likely to be associated with positive emotions such as delight in cherished items (Grisham & Barlow, 2005; Steketee & Frost, 2003). Given these notable differences, it is not surprising that individuals with primary hoarding symptoms have tended to benefit less from cognitive behavioural treatments designed for OCD (Abramowitz, Franklin, Schwartz, & Furr, 2004; Black et al., 1998; Mataix-Cols, Marks, Greist, Kobak & Baer, 2002; Rufer, Fricke, Kloss, & Hand, 2006), spurring the need for a better understanding of hoarding. Nonetheless, OCD and compulsive hoarding display similarities (i.e., avoidance, indecisiveness, responsibility, and doubt; Steketee & Frost, 2003) and an elevated rate of comorbidity (18%; Frost, Steketee, & Tolin, 2011). Accordingly, hoarding disorder is currently grouped with OCD in the DSM-5 under the category of Obsessive Compulsive Related Disorders.  Hoarding as a disorder has also been separated from OCPD. Within the DSM-IV, OCPD diagnostic criteria included hoarding behaviour, defined as the inability to discard worn out worthless objects even when they are not sentimentally attached to them (APA, 2000).  This criterion was found to be too restrictive to capture the diversity of hoarding behaviour. As discussed by Frost and Steketee (2014), individuals with hoarding experience difficulty parting with objects through many means (e.g., donating, recycling, lending), not only discarding. Moreover, the types of possessions that individuals with hoarding disorder save in excess are not restricted to worn out worthless objects (Frost & Gross, 1993) and often include objects saved for sentimental reasons (Steketee, Frost, & Kyrios, 2003). However, OCPD has been found to share several features with hoarding such as 	 8	perfectionism and indecisiveness (Frost & Gross, 1993) and is the most common personality disorder found in hoarding samples (Frost, Steketee, & Tolin, 2011). Past research has produced mixed findings concerning whether OCPD is more common in hoarding than non-hoarding OCD when the hoarding criterion is removed from the OCPD diagnosis (Frost et al., 2011; Frost, Steketee, Williams, & Warren, 2000; Pertusa et al., 2008; Samuels et al., 2008). Prevalence  In terms of prevalence, based on the occurrence of hoarding as a prominent subtype of obsessive-compulsive disorder, hoarding has been estimated to have a lifetime prevalence of 0.04% (Steketee & Frost, 2003). However, when hoarding is considered independent of OCD, its prevalence has been estimated to be between 2 to 6% of the general population in Western countries (Bulli et al., 2014; Iervolino et al., 2009; Mueller et al., 2009; Samuels et al., 2008; Timpano et al., 2011). However, research on the prevalence of hoarding considered in isolation from OCD is a more recent development and has varied in how hoarding status is defined, the assessment measures used, and the populations assessed. Among these prevalence studies, two conducted respectively by Timpano et al. (2011) and Iervolino et al. (2009) stand out for their use of hoarding criteria that approximates those used to define hoarding disorder in the DSM-5. Both of these studies examined the point prevalence of hoarding in large European samples (Italy and Germany) using validated cut-off scores on a self-report version of the Hoarding Rating Scale (HRS-SR; Tolin et al., 2010). The HRS-SR assesses several core features of hoarding (e.g., difficulty discarding, clutter, distress and impairment) and maps onto DSM-5 hoarding disorder criteria A, C, and D. However, the HRS does not directly examine the hoarding disorder criterion concerning motivation for 	 9	saving and does not rule out other potentially explanatory disorders and conditions (e.g., depression).   Decision-Making in Compulsive Hoarding In 1996, Frost and Hartl (1996) proposed a cognitive behavioural model of hoarding based on their clinical observations and the limited amount of research available at the time. Within this model, decision-making difficulties were proposed to be a core problem in hoarding. Frost and Hartl (1996) suggested fear of making mistakes, combined with uncertainty about the probability of needing an object in the future, resulted in indecisiveness about whether or not to discard objects. They proposed perfectionism would interfere with decision-making as perfectionistic individuals strive to find a solution that will satisfy all possible relevant factors, resulting in a prolonged weighing of the pros and cons of each option. In addition, they expected people with hoarding disorder to have a higher threshold about what to discard. This decision threshold could involve distorted perceptions of the probability of future need, anticipated consequences of making an incorrect decision (e.g., discarding an object for which a need later arises), and self-efficacy for handling such consequences. Although most people experience some uncertainty deciding whether a particular item of clothing should be discarded or donated, Frost and Hartl implied that people with hoarding disorder would require clothing to be more worn out, more stained, or more ill-fitting before making a decision to discard the clothing.  Frost and Hartl’s (1996) model is a landmark paper within the hoarding literature and has stimulated several lines of research. Most of the research testing Frost and Hartl’s (1996) propositions about decision-making in hoarding has investigated self-reported indecisiveness, most frequently with the Frost Indecisiveness Scale, which assesses how respondents 	 10	approach decisions, including the tendency to postpone decisions, and post-decisional regret. Items include, “I try to put off making decisions” and “Once I make a decision, I stop worrying about it” (reverse scored). This scale shows moderately high correlations with hoarding symptom severity (Frost & Gross, 1993; Frost & Shows, 1993; Steketee, Frost, & Kyrios, 2003; Wincze, Steketee, & Frost, 2007). Overall, hoarding participants complain more of indecisiveness than do healthy controls, with an average effect size of d = 1.60 across four studies (Grisham, Norberg, Williams, Certoma, & Kadib, 2010; Steketee et al., 2003; Tolin & Villavicencio, 2011; Wincze et al., 2007). However, the difference between hoarding participants and clinical controls such as non-hoarding OCD and mood or anxiety disorders is much smaller (about d = 0.50; Grisham et al., 2010; Steketee et al., 2003; Tolin & Villavicencio, 2011; Winze et al., 2007). Turning to performance-based tests of decision-making, latency to complete tasks that involve decisions has been taken as an index of decision-making difficulty. To operationalize difficulty discarding – a central feature of the disorder – Tolin's group has used a behavioural paradigm involving discarding decisions in two imaging studies (Tolin, Kiehl, Worhunsky, Book, & Maltby, 2009; Tolin et al., 2012). Participants brought their own paper items (e.g., mail, flyers) to the lab and, while in the scanner, made decisions about whether to immediately shred each item as the experimenter presented it. Even after controlling for depression and non-hoarding OCD, those with hoarding took significantly longer to make the decision to shred versus keep their own papers and also reported more anxiety during the decision-making. Not surprisingly, they also chose to discard fewer papers. Participants with hoarding, but not healthy controls, had more difficulty (i.e., longer 	 11	decision latency and higher anxiety) when deciding about their own papers than when deciding about the experimenter's papers.  Going beyond indecisiveness during discarding (part of the definition of the disorder), other studies have examined broader decision-making performance. Three studies have examined decision latency during categorization tasks wherein participants are asked to sort (not discard) a variety of objects using a self-generated scheme (Grisham et al., 2010; Luchian, McNally, & Hooley, 2007; Wincze et al., 2007). These studies found that hoarding participants take longer than healthy or clinical controls to categorize or sort objects, suggesting indecisiveness is not limited to discarding tasks.  Decision-making in hoarding has also been examined using gambling tasks, such as the Cambridge or Iowa Gambling Tasks, which are often used in research on cognitive functioning as indicators of ability to make advantageous decisions. The Iowa Gambling Task (IGT; Bechara, Damasio, Damasio, & Anderson, 1994) is a widely used test of decision-making ability that involves learning the reward and penalty schedules of four decks of cards in order to make choices between them that will result in greater long term, rather than short term, gains. Scoring well on the IGT involves responding flexibly to changing contingencies, monitoring prior responses and their outcomes, generating solutions to a novel problem, and inhibiting a dominant response (da Rocha et al., 2008). Better performance also indicates an ability to balance immediate rewards against longer-term negative consequences (Cavallaro et al., 2003; Lawrence et al., 2006; Starcke, Tuschen-Caffier, Markowitsch, & Brand, 2009). As such, the IGT is not a simple measure of cost-benefit decision-making, but rather involves making decisions while integrating emotional and cognitive information (Bechara, 2001; Bechara, Damasio, Damasio, & Lee, 1999; Grisham et al., 2007; Malloy-	 12	Diniz, Fuentes, Leite, Correa, & Bechara, 2007) under ambiguous and uncertain conditions (Starcke et al., 2009). Using the IGT, researchers have compared participants with hoarding problems with various relevant comparison groups. Nakaaki et al. (2007) reported a 48-year old man with frontotemporal lobar degeneration with initial symptoms of hoarding and pathological gambling consistently selected more cards from disadvantageous decks across 5 blocks of trials, whereas 10 age-matched participants consistently selected more cards from advantageous decks and showed greater improvement in performance across the blocks. In another study, hoarding and washing, but not other types of OCD symptoms, predicted worse performance on the IGT, controlling for OCD severity, depressive symptoms, anxious symptoms, age, education, and verbal IQ (Lawrence et al., 2006). In contrast, three other studies have found no group differences between hoarding and healthy controls on IGT performance or learning trajectories (Blom et al., 2011; Grisham et al., 2007; Tolin & Villavicencio, 2011). Similarly, Grisham et al. (2010) found no group differences on the Cambridge Gambling Task (CGT). On each trial of the CGT, participants are presented with an array of red and blue cards and are asked to bet on whether a yellow token is hidden behind a red card or a blue card. Participants choose how many points to bet on each trial, and the proportion of red and blue cards changes across trials, shifting the probability of a correct guess. Once the bet has been placed, the cards are removed to reveal the actual location of the yellow token, and the amount of the bet is added to or taken away from the participant's accumulated points. Grisham et al. found no significant between-group differences on either quality of decision-making (choosing the color with more cards in the array on a given trial) or risk adjustment (e.g., betting higher amounts on trials in which a 	 13	large majority of the cards are one color), although the effect size for the difference between the hoarding and healthy control groups on percentage bet (risk adjustment) was moderate in size  (d = 0.65). Because both the IGT and CGT are sensitive to risk tolerance, they are probably not the best tools for examining decision-making in the context of hoarding. Making risky choices results in poor scores, a response profile that has been observed among participants with problems such as substance abuse that include a propensity for risk-taking (Bechara, 2001; Petry, Bickel, & Arnett, 1998). Although hoarding patients often demonstrate problems with response inhibition and impulse control (Woody, Kellman-McFarlane, & Welsted, 2014), suggesting they might score poorly on these gambling tasks, hoarding also involves risk aversion. Hoarding patients can be excessively cautious, for example, taking steps to save or acquire items in preparation for low probability future events. Such risk aversion would enhance IGT or CGT performance by leading participants to avoid disadvantageous decks once they learn the reward schedule (Blom et al., 2011; Grisham et al., 2007). Broadly speaking, these studies do not present a convincing case that indecisiveness is an issue that is specific to hoarding. Even in those studies that reported significant group differences in indecisiveness, depression is an important confound given the high comorbidity of depression and hoarding (Frost, Steketee, & Tolin, 2011). Depressed patients report more decisional conflict (the aversive experience that accompanies indecisiveness) than do healthy controls (van Randenborgh, de Jong-Meyer, & Hu ̈ffmeier, 2010). Other more general trait-like factors may also play a role in the indecisiveness expressed by individuals with hoarding problems. For example, indecisiveness has been found to be related to low 	 14	self-esteem (Effert & Ferrari, 1989), neuroticism (Jackson, Furnham, & Lawty-Jones, 1999), procrastination (Beswick, Rothblum, & Mann, 1988), and perfectionism (Frost & Shows, 1993), characteristics that commonly occur in hoarding as well as many other forms of psychopathology. Similarly, although decision-making appears to be slowed in comparison to healthy controls, the degree to which this is specific to hoarding is not yet clear. For example, depression is associated with psychomotor retardation and is often comorbid with hoarding, although not always controlled for in research.  Given the results of studies employing the IGT and the CGT, the extreme bias toward saving decisions in hoarding cannot be convincingly explained by a generally poor ability to detect patterns over time and engage in rational or advantageous decision-making. Likewise, general indecisiveness and avoidance cannot fully explain why excessive saving is not a common problem in other forms of psychopathology that are accompanied by indecision and avoidance. Therefore, research that seeks to explain why individuals with hoarding problems make so many saving decisions would benefit from moving away from designs that seek to establish general deficits in rational decision-making abilities. Instead, research should more closely examine the process through which people who hoard arrive at saving decisions and where in this process people who hoard diverge from people who do not hoard. One possibility is that the process of determining whether one should save or discard an object differs in hoarding because it is guided by different priorities or considerations from people who do not hoard, such that saving items that most people would consider useless becomes justifiable. The identification of such factors would be both empirically and clinically useful, as it would help to characterize decision-making difficulties evident in hoarding and provide targets for treatment. 	 15	The Preston, Muroff, and Wengrovitz (2009) examination of how space constraints may differentially impact saving decisions is a good example of this kind of research but is unfortunately somewhat unique in the literature. Preston et al. designed a computerized task wherein unselected undergraduate participants (N = 89) made discarding and acquiring decisions about everyday objects under increasing space limitations (i.e., unlimited space vs. the space within a shopping cart vs. the space in a paper bag). Hierarchical cluster analyses were used to identify three subgroups of participants: Acquiring cluster (i.e., those who selected and retained the most items), the Spartan cluster (i.e., those who selected and retained the fewest items), and the Intermediate cluster (i.e., those who selected an intermediate number of items and also discarded many). The Acquiring cluster scored significantly higher than other groups on self-report measures of hoarding and indecisiveness, suggesting this pattern may be most reflective of hoarding difficulties.  One of the strengths of Preston et al.’s (2009) study compared to the previously mentioned gambling tasks is that it employed a task better suited to capturing the types of decision-making difficulties people who hoard experience in real life (i.e., patterns of acquiring and saving under pressure to discard). Additionally, it investigated a decision-making factor likely to function differently in hoarding (i.e., responsiveness to space constraints) in contrast to more common decision-making difficulties such as indecisiveness. However, their study also has several characteristics that limit its generalizability to individuals with hoarding disorder. Specifically, they utilized a non-clinical sample and the cluster they identified as most representative of hoarding was defined by over-acquiring, which unlike difficulty discarding, is not a core feature of hoarding disorder. Overall, Preston et al.’s findings suggest that responsiveness to space constraints is a lucrative area for 	 16	understanding individual variation in the propensity to discard possessions, but requires replication and is currently limited in its generalizability to hoarding populations.  Extant studies have not addressed the specific types of indecision about objects that Frost and Hartl (1996) hypothesized to play a role in hoarding, such as over-cautiousness and concern about needing possessions in the future. The current research seeks to more directly examine problematic decisional processes that are both specific to hoarding (i.e., decisions about whether to save or discard) and related to Frost and Hartl’s propositions about the inhibiting role of potential future need. To do this, I use novel decision-making tasks that evaluate the propensity to consider certain types of information about an object’s future usefulness and neglect other information that would lead to more negative appraisals of an object’s value when making decisions about whether to keep or discard possessions. Moreover, the current research will build on Preston et al.’s work by further investigating how space may play a role in the decision-making processes of people who hoard compared to those who do not hoard.  Reasons for Saving and Discarding Normative Reasons  Research within psychology and consumer behaviour has produced a rich and extensive literature on the diverse reasons ordinary people save possessions. Many theorists suggest that the tendency to have meaningful connections with owned objects is universal and has biological and sociocultural roots (Dittmar, 1992; Ellis, 1985; McCracken, 1986; Wallendorf & Arnould, 1988). According to this line of thinking, possessions play an important role in self-concept through their contribution to development of self (i.e., learning what is ‘me’ involves learning what is ‘mine’) and differentiation of self from others; their 	 17	ability to symbolize group membership, economic status, personal values and interests, and relationships with others; and their function in preserving self-continuity during major life transitions. (For a review, see Kleine and Baker, 2004 and Pierce, Kostova and Dirks, 2002.) Belk (1998) proposed that possessions are so intertwined with identity that they can be seen as extensions of the self, and went as far as to conclude, “we are what we have” and this “may be the most basic and powerful fact of consumer behavior” (p. 160).  Interestingly, exceptional circumstances are not needed to create these connections between the self and objects. Possessions do not need to be expensive, rare, or extraordinary to accrue deep personal meaning, although this may happen more easily for unique possessions such as heirlooms; people derive meaning and attachment from many different mundane possessions (Dittmar, 1992; Kleine & Baker, 2004). Moreover, creating associations between self-concept and objects can be accomplished quite easily. Gawronski, Bodenhausen, and Becker (2007) found that merely asking individuals to choose which of two objects they would like to take home was sufficient to increase implicit self-associations with that chosen object and positive implicit evaluations. Ownership has been found to increase valuing and preference for objects, even quite soon after the moment of perceived acquisition (Kahneman, Knetsch, & Thaler, 1990; Turk et al., 2011). This tendency for people to overvalue things they own is called the endowment effect and is typically displayed in tasks involving money-for-goods trades (Thaler, 1980). In these tasks, people demand higher prices to relinquish goods they own than they would be willing to pay to acquire them. The mechanism behind this effect has traditionally thought to be loss aversion (i.e., the inclination for losses to be weighted more heavily than potential gains; Kahneman & Tversky 1979). However, more recent researchers have presented evidence that the endowment effect 	 18	may be explained by the connection between possessions and their relevance to the self (Morewedge, Shu, Gilbert, & Wilson, 2009).   In addition to their role in identity, possessions have been noted to provide people with many other important functions that encourage saving. One of the landmark papers on the meaning of possessions was a review by Furby (1978). Based on her review of the extant research, Furby established two prominent rationales for saving items: instrumental saving (i.e., objects with a functional or practical purpose) and sentimental saving (i.e., objects to which we are emotionally attached). Furby concluded that one of the central reasons possessions are meaningful for most people is because they provide their owners with an increased sense of control over their environment, especially when there is a possibility a possession will be needed (i.e., useful) in the future. Many potential motivations for saving possessions can be subsumed under the categories of instrumental and sentimental saving. For example, instrumental reasons for saving would include use of a possession to accomplish any functional goal, from turning on a light to transportation. Sentimental saving would include much of the previously discussed relevance of identity to possession, as well as many other emotional motivators such as the association between possessions and loved ones, meaningful events, and accomplishments. Research that has tried to differentiate patterns of saving and associated rationales in hoarding have found that people with hoarding problems tend to save the same kinds of objects for some of the same reasons as people who do not hoard, but in greater volume (Frost et al., 1998; Frost & Gross, 1993).  In sum, everyday possessions are capable of acquiring deep personal meaning for most people, and there are an almost infinite number of compelling reasons to save possessions. This raises the question of how anyone, let alone people who hoard, push past 	 19	these reasons to ultimately discard possessions. Research on normative consumer behaviour has generally devoted a much greater amount of attention to acquiring and saving than to disposal practices (e.g., Jacoby, Berning, & Dietvorst, 1977; Shim, 1995). The literature that is available on disposal practices has identified multiple motives for why people discard, sell, or recycle possessions (e.g., lack of space, financial incentives, waste avoidance, philanthropy, advances in technology or style, decreases in item functionality, poor reflection of social status, unattractive form, changes in life circumstances, situational pressures such as moving (e.g., Cooper, 2004; Domina & Koch, 1998; Jacoby et al., 1977; Laitala, 2014; Paden & Shell, 2005; Roster, 2001; Savas, 2002)). More generally, research on psychological ownership suggests that items are easier to part with when they no longer serve the underlying motive that once compelled keeping them (e.g., a change in identity that no longer fits with the possession, the item is no longer useful for controlling one’s environment; Pierce, Kostova & Dirks, 2002).  While the literature suggests many potential motivations on both sides of the discarding/saving decision, there is less information about how these factors compete with each other to result in normative disposal practices and the maintenance of uncluttered homes. A better understanding of what factors ultimately enable discarding in normative populations, despite temptations to save, would provide insight into what decision-making factors are likely to operate differently in hoarding. One method of identifying these factors is to compare the decision-making rationales of people who are highly effective discarders to those who are avid keepers. Unfortunately, studies that broadly explore differences in decision-making rationales between individuals who differ in their discarding practices are relatively rare.  	 20	Coulter and Ligas (2003) interviewed ‘purgers’ (n = 14; i.e., individuals who discard easily and continually take stock of their needs) and ‘packrats’ (n = 14; i.e., individuals who tend to keep things and do not discard easily) about their reasoning for discarding and retaining possessions and their perceptions of each other. One of the key differences that emerged between these two groups is that purgers typically did not consider the future use of old and used items. One purger stated, “There is no need to keep something once I am done using it. I probably won’t use it again, or I will just go buy it again the next time. It is useless after it gets old.” (Coulter & Ligas, 2003, p. 40). In contrast, packrats reported potential future use as a key rationale for keeping possessions; they often reported keeping things because they or someone else might need them later on. Similarly, packrats reported that they kept items because they could be repurposed it in the future and therefore still have functional value. When packrats were asked about their perceptions of purgers, they expressed concern that purgers will not have what they need in the future. Purgers felt that packrats were cheap and not resourceful. Another difference between the two groups was that purgers were more likely to relinquish items due to space constraints. Furthermore, packrats were more sentimentally attached to possessions and more likely to want to find them a good home, whereas purgers preferred methods of discarding for their ease. These differences in rationales suggest that the degree to which individuals think about the future use of possessions (e.g., needing it, re-purposing it) may be one of the key differences in the decision-making processes of successful and unsuccessful discarders.  Specifically, Coulter and Ligas’s (2003) study suggests that how one evaluates utility may be an important determinant of discarding practices, with more frequent keeping resulting from a greater focus on potential future uses and possibilities. As will be discussed 	 21	in more detail later, this focus on potential future need is commonly observed in hoarding populations. These findings raise the question of whether discarding is generally more likely to occur when an object’s utility is evaluated more heavily on the basis of actual past usage. One explanation for why considerations related to past usage would be more likely to result in discarding is that past usage is grounded in concrete behaviour. As actual use of possessions is limited by factors such as time, interests, and availability, evaluating an object’s utility through past behaviour is likely generate a more negative conclusion about its personal utility if the item is not frequently used. In contrast, potential future uses are often hypothetical and can therefore be generated to provide a positive basis for keeping virtually any possession. The current research will expand on Coulter and Ligas’s work by using an experimental paradigm to test whether individuals with extreme saving habits (i.e., people with hoarding disorder) are more likely to consider information about future utility to inform their decision-making about possessions and how this preference ultimately relates to saving.  Reasons for Saving in Hoarding Frost and Hartl’s (1996) cognitive behavioural model of hoarding also provided propositions about reasons for saving in hoarding. According to Frost and Hartl, decisions about discarding objects in hoarding are largely based on the same instrumental valuing (i.e., judgments about the future use or need for a possession) and sentimental valuing (i.e., emotional attachment) processes that are believed to underlie normative saving. However, people who hoard are proposed to define a greater number of objects as sentimentally or instrumentally valuable. The following is a review of studies on motivations for saving in hoarding. This review will focus primarily on instrumental reasons for saving, as it is the focus of the current research.  	 22	 Instrumental Saving. Frost and Hartl’s (1996) description of instrumental value as an important underlying rationale for hoarding behaviour is consistent with earlier theory and anecdotal observations about excessive saving behaviour in obsessionals by Salzman (1973) and hoarding by Warren and Ostrom (1988). Frost and Hartl proposed several potential explanations for why people who hoard are more likely to retain items that could be used in the future. One explanation is that they erroneously overestimate the probability of needing possessions in the future. Alternatively, people who hoard may judge the probability of needing a possession correctly, but are more fearful of the consequences of not having a possession when they need it and underestimate their ability to cope in that situation. Furthermore, Frost and Hartl suggested that people who hoard may be especially motivated to avoid mistakenly discarding an object they may need in the future due to perfectionism and an exaggerated sense of responsibility for being prepared.  Several studies support the importance of instrumental saving in hoarding. Frost, Hartl, Christian, and Williams (1995) found that hoarding behaviour was significantly correlated with a sense of responsibility for being prepared for future events (r = .42 - .48) in a mixed undergraduate and community sample. The concept of heightened responsibility for keeping possessions for the future is also represented in the Saving Cognitions Inventory (SCI; Steketee, Frost, & Kyrios, 2003). The SCI is a measure of hoarding related beliefs that contains six subscales: memory, value of possessions, emotional comfort, loss, control over possessions, and responsibility. The desire to be prepared for future events is represented in the responsibility subscale (i.e., “I’m ashamed when I don’t have something like this when I need it” and “If this possession may be of use to someone else, I am responsible for saving it for them”) which has been found to be a significant predictor of hoarding severity controlling 	 23	for age, mood state, OCD severity, indecisiveness, and the other hoarding belief subscales (Steketee et al., 2003). A study by Frost and Gross (1993) on saving patterns and rationales in hoarding suggests that concern about potential future need may be one of the more important barriers to discarding in hoarding. At the time, this study was unique in its goal to experimentally explore the relative importance of different reasons for saving in compulsive hoarding. In their study, individuals with probable hoarding problems (defined by self-identification as a chronic saver or packrat, not being a collector, and having a large amount of typically unused possessions) rated the frequency on a 5-point scale of different types of thoughts while deciding whether to save or discard items. The four types of thoughts were: “I might need this some day” (instrumental value), “This means too much to me to throw away” (sentimental value), “This may be worth something someday” (monetary value), and “This is too good to throw away” (intrinsic value). The most frequently reported thought was related to future utility (M = 4.8), followed by intrinsic value (M = 4.3), sentimental value (M = 3.8), and monetary value (M = 3.6) respectively. While these scores suggest concern for future need may be a factor of primary importance for saving in hoarding, Frost and Gross did not statistically compare these scores to each other or to scores of a healthy control group. However, the importance of being prepared for future needs was further supported in a follow up experiment, wherein individuals with probable hoarding problems were significantly more likely than non-clinical controls to carry “just-in-case” items and to buy extras of things to prepare for the event they run out and need another. These findings support the relative importance of instrumental saving in hoarding, but are mostly descriptive and limited in their interpretation due to the initial lack of control group.  	 24	 Sentimental Saving. Another factor Frost and Hartl (1996) proposed to motivate excessive saving is an intensified sentimental valuing and attachment to possessions, termed hypersentimentality. According to this conceptualization, people who hoard define a much larger number of household possessions as sentimentally valuable, even items that are not associated with particularly meaningful events. Possessions provide people who hoard with a source of comfort when distressed (Frost & Hartl, 1996; Frost et al., 1995; Nedelisky & Steele, 2009), a sense of safety and security (Cherrier & Ponnor, 2010; Frost & Hartl, 1996), and a means of remembering important life events (Cherrier & Ponnor, 2010). Hypersentimentality is also associated with overidentification with objects (Frost et al., 1995; Kyrios, Steketee, Frost, & Oh, 2002) and the tendency to imbue objects with sentient qualities (Timpano & Shaw, 2013).  Multiple case studies and anecdotal observations of hoarding have noted this heightened emotional attachment to objects (Frankenburg, 1984; Greenberg, 1987; Frost & Hartl, 1995; Warren & Ostrom, 1988), as have empirical studies (Hartl, Duffany, Allen, Steketee & Frost, 2005; Steketee et al., 2003). Despite the strong evidence of an abnormal emotional valuing of objects in hoarding, this tendency is unlikely to fully or even primarily explain why objects accumulate in hoarding. In studies designed to compare the importance of different reasons for saving among individuals who hoard, sentimentality has typically emerged as equal to or less important than future utility (Frost & Gross, 1993; Frost, Steketee, Tolin, Sinopoli, & Ruby, 2015; Nordsletten, de la Cruz, Billoti, & Mataix-Cols, 2013). However, as will be discussed below in more detail, studies comparing different reasons for saving are in their nascency and require further development before conclusions can be drawn about the relative importance of different reasons for saving. Overall, while 	 25	excessive sentimental valuing is a vibrant example of dysfunction and has been the subject of much empirical research, current evidence suggests it may not be the most fruitful area of exploration for research that seeks to understand the accumulation of clutter in hoarded homes.  Comparative Reasons for Saving. While past research has suggested many potential reasons for saving in hoarding, far less investigation has been conducted on the relative importance of these motivations. Since the Frost and Gross initial study on saving patterns in 1993, three studies have compared reasons for saving in hoarding (Dozier & Ayers, 2014; Frost, Steketee, Tolin, Sinopoli, & Ruby, 2015; Nordsletten, de la Cruz, Billoti, & Mataix-Cols, 2013). Generally, these studies have involved presenting participants with hoarding disorder with different reasons for saving and asking them to rate or otherwise indicate the degree to which each motive influences their discarding practices. However, these three studies vary significantly in important aspects of their methodology, including which motives they investigated, the methods of assessing these motives (i.e., self-report questionnaires vs. coded interviews), and the presence and type of comparison group (i.e., self-identified collectors, healthy controls and OCD, no control group).  These three studies produced mixed findings, with some preliminary evidence of the relative importance of potential future usefulness. No motives emerged as more significant than others within Dozier and Ayer’s hoarding sample, while Nordsletten et al. (2013) found that individuals with hoarding disorder rated potential future usefulness as their most frequently endorsed reason for saving. Frost et al. (2015) found that reasons related to waste avoidance and retaining information were rated more highly than other motives in their hoarding group. The interview questions used in Frost et al.’s (2015) study to assess the 	 26	waste avoidance and informational motives suggests that both of these motives may partially overlap with future need. The question assessing waste avoidance was, “Are you afraid of wasting a potentially useful object when you try to discard something? That is, you are concerned about being wasteful because the object could eventually be put to good use” (emphasis added) and the question assessing information was, “Are you afraid of losing important information? That is, you are afraid you will mistakenly throw out information that you will need someday.” Furthermore, sentimentality was often also a prominently featured rationale within these studies but was typically somewhat lower in relative importance.  Overall, results from the Nordsletten et al. (2013) and Frost et al. (2015) studies provide preliminary support for the relative importance of potential future usefulness, but further investigation with more consistent methodologies is required. Furthermore, this area of research would generally benefit from a shift to methods of assessing saving motivations that go beyond self-report. The current use of self-report to study motivations for saving in hoarding is questionable for two reasons. Firstly, hoarding is highly stigmatized disorder and likely to result in feelings of shame when participants are asked to report on their saving practices. Therefore, hoarding participants may be motivated to provide responses they consider to be more rational and socially desirable over more stigmatized responses such as excessive sentimentality. Secondly, efforts at honest self-assessments may not be accurate due to limitations in introspective ability. Many studies in social psychology have shown that people are surprisingly limited in their ability to accurately report or predict their own behaviour. (For a review, see Pronin, 2009.) The current research improves upon these limitations by using performance tasks to measure differences in saving strategies. In the current research, participants were exposed to different types of information and asked to 	 27	make choices that reflect different aspects of their decision-making process about possessions. Current Research  People who hoard feel compelled to save possessions that most individuals would consider trivial or useless despite the detriments caused by the resulting clutter. Currently, the mechanisms underlying this clearly maladaptive style of decision-making about possessions remains unclear. People who hoard do not appear to suffer from a general deficit in decision-making abilities, yet they often report saving possessions due to their potential usefulness despite the lower actual usage of possessions in hoarded homes (Frost, Hartl, Christian, & Williams, 1995). The limited number of studies investigating saving rationales in packrats and clinical hoarding suggests that greater focus on future utility may be an important aspect of maladaptive saving practices. However, research has not empirically tested the relationship between saving behaviour and excessive focus on future utility. Furthermore, research has also not explored normative decision-making about potentially useful items. The current research aims to address these gaps.     Specifically, the current research aims to determine whether focusing on potential future uses when evaluating an object’s utility contributes to the excessive saving characteristic of hoarding. Hoarding symptoms are expected to be associated with a greater focus on the potential future uses of possessions when making discarding decisions and neglect for decision-criteria reflecting past usage of possessions. As a result of these biases, people with greater hoarding symptoms would often neglect factors that would decrease the value of a possession (e.g., low past usage) and instead would focus on information that would support keeping possessions (e.g., potential future uses, consequences of not having 	 28	the possession when needed). The current research will also begin to address Frost and Hartl’s (1996) proposition that hoarding is associated with a higher threshold for what to discard. Specifically, I hypothesize that people who hoard will require objects to be in worse condition (e.g., more dysfunctional, damaged, currently unusable) than people without hoarding problems before discarding them. Such a tendency would also positively bias appraisals of an object’s usefulness, such that people who hoard would be more likely to see value in objects that non-hoarding individuals would consider useless.  In addition to these hypotheses about utility, I hypothesize that people who hoard are more likely to neglect negative information about the amount of space they have available to store objects. Although inability to maintain a balance between space and the volume of possessions in one’s home is a core feature of hoarding, very little research has been conducted on how space constraints differentially influence discarding decisions for individuals who hoard. To the best of my knowledge, only one study, by Preston, Muroff, and Wengrovitz (2009), has experimentally explored individual differences in the impact of space on decisions to keep or discard items. Their results suggest space may have less of an impact on the retention habits of people who tend to acquire more than others, but these results require replication and extension to a clinical hoarding sample. The current research will help to extend and expand on their findings to better understand how space is integrated into the decision-making process of people who hoard.  I investigate these hypotheses using three methods. Firstly, I develop a novel computer-based information seeking task to examine specific patterns of information seeking during discarding decisions. This task sequentially presents participants with hypothetical possessions and asks them to decide whether to keep or discard each object. Before making 	 29	this decision, participants have the opportunity to access information about each object’s past utility or its potential future utility to inform their decision. I predict that individuals with more severe hoarding problems are more likely to choose to access information about future utility than about past utility, and individuals without hoarding problems choose to access more information about past utility. As a means of validating the task, I also examine between-group differences in rates of discarding and task-related distress. If individuals who hoard make decisions on the task in the same manner they would at home, I expect that hoarding problems would be associated with greater task-related distress and a greater number of decisions in favour of keeping objects.  Secondly, I develop a novel vignette-based task to examine hypotheses pertaining to neglect of negative information about past utility and space constraints. In this task participants are presented with vignettes containing either positive or negative information about an object’s past utility and the amount of household space available to store it. After each vignette, participants indicate how likely they would be to discard the described objects. I predict that the decision-making processes of individuals who hoard are less sensitive to changes in information about past utility and space constraints than that of healthy controls. In contrast, healthy controls will be inclined to discard objects when presented with negative information about past utility and space constraints. Lastly, I construct a Guttman-style scale to examine whether people who hoard require objects to be in worse condition before discarding them compared to healthy discarders. The items on this scale describe different objects with diminishing levels of function, condition, or utility. Participants indicate the lowest level of functioning at which they would be willing to retain an object. I predict that hoarding problems will be associated 	 30	with higher scores indicating hoarding is associated with greater willingness to keep dysfunctional objects.      	 31	Chapter 2: Methods Design Overview The current research was an online study that began with an extensive preliminary phone screen. At the end of the phone screen, eligible participants were sent a link to complete experimental tasks and self-report questionnaires via online platforms. Validity questions were embedded within two of the tasks to ensure the quality of the data obtained through these online methods. These validity questions assessed attention to study materials, and participants who did not pass these questions were excluded from the study.  The data collected through these procedures were then used to test hypotheses about the relationship between hoarding psychopathology and task performance through a combination of linear mixed modeling, regression, and analysis of covariance. Within these analyses, hoarding was examined as both a categorical (i.e., hoarding group vs. healthy control group) and dimensional construct (i.e., severity of self-reported hoarding symptoms). Chapters 3 through 5 describe further details about the development of the experimental tasks and data analysis for each task.  Procedure Advertisements were placed in areas where they are likely to be encountered by the general public and in places where they are especially likely to be seen by individuals with clinical hoarding problems. Specifically, poster advertisements were placed in community centers, libraries, and stores around Vancouver (e.g., coffee shops, grocery stores). Online advertisements were posted on our laboratory’s website, the International Obsessive Compulsive Disorder Foundation’s website, and in the sections of community websites (e.g., 	 32	Craigslist, Kijiji) advertising volunteer positions and opportunities to acquire free possessions.  Participants were also recruited through an established registry held by the Centre for Collaborative Research on Hoarding. This registry contained the contact information of individuals who had either participated in past research or otherwise contacted the laboratory and indicated they were interested in participating in future research. The participants in this registry had been previously interviewed by someone with knowledge and training in hoarding and accordingly categorized as having or not having hoarding problems.  Interested individuals who contacted the laboratory or were contacted through the research registry were scheduled for a phone screen with a trained research assistant. The phone screen was used to assess inclusion and exclusion criteria and to determine group assignment. Participants who met inclusion and exclusion criteria were sent a link to the Clutter Image Rating Scale (CIR; Frost, Steketee, Tolin, & Renaud, 2008) and asked to provide a rating of the most cluttered room in their home. The CIR was used to increase the efficiency of the phone screen by screening individuals who had no problems with household clutter from a more time intensive clinical interview. Participants who provided very low ratings on the CIR (i.e., 1-2) were immediately sent a link (via email) to the study tasks and questionnaires. Participants who scored higher than 2 on the CIR completed a structured clinical interview assessing hoarding disorder (i.e., the Hoarding Rating Scale or another structured hoarding interview used in our lab) before being sent a link to the study materials.  Furthermore, the phone screen was also used to inform participants that several of the study measures contained validity questions that assessed attention to questionnaire content. Participants were provided with an example of a validity question and notified that 	 33	compensation would be withheld if any of the validity questions were answered incorrectly. Before ending the phone screen, participants were coached to complete the study in a quiet environment and encouraged to contact the laboratory for technical assistance if needed. Once participants clicked on the provided link, they were directed to a page containing the information necessary for informed consent and encouraged to contact the laboratory if they had any questions before proceeding. Once participants consented, they were directed to complete the experimental tasks and self-report measures of hoarding and depression. Age, gender, education level and income were assessed via self-report. Options for income brackets were based on those used by Statistics Canada for British Columbia (Statistics Canada, 2015). After completing the study materials, participants were directed to a debriefing page and given the option of receiving compensation in either the form of cash, an Amazon gift certificate, or a donation to a mental health charity. Participants who completed the entire study and passed the validity questions were compensated $10/hour according to their preferred method. Measures Hoarding Rating Scale-Interview (HRS-I) The HRS-I (Tolin, Frost, & Steketee, 2010) is a semi-structured interview that assesses DSM-5 symptoms of hoarding: difficulty using rooms due to clutter, excessive acquisition of objects, difficulty discarding possessions, distress due to hoarding behaviours, and functional impairment due to hoarding. The scale has good internal consistency (α = 0.96), good test-retest reliability varying across time (1-12 weeks) and contexts (home visit and clinic; r = 0.96 for total scale score), good convergent validity with the CIR (r = 0.72) and the Saving Inventory - Revised, which is a self-report measure of hoarding behaviour  	 34	(SI-R; r = 0.91), and clearly distinguishes individuals with compulsive hoarding from community controls and OCD patients without hoarding (Tolin et al., 2010).  MINI-Hoarding Module  Our lab has been using the Mini-International Neuropsychiatric Interview (MINI; Sheehan et al., 1998) for diagnostic screening for other studies, and we have developed a Hoarding module based on the DSM-5 criteria and the interview conventions of the MINI. For example, the question on clutter accumulation is, “Does the number of possessions or clutter make it difficult for you to use at least one room in your home?”. To meet criteria for hoarding disorder, an individual must endorse all DSM-5 diagnostic criteria (i.e., difficulty discarding, intentional saving, distress or impairment, and verification that symptoms are not due to another disorder or medical condition). Failure to endorse any diagnostic criterion results in termination of the interview and the conclusion that hoarding disorder is absent. This MINI-Hoarding module interview was administered during the phone screen. Phone screeners were trained to follow the interview and also to ask additional questions when they did not have sufficient information to code a diagnostic criterion as present or absent.   Saving Inventory-Revised (SI-R) The SI-R (Frost, Steketee, & Grisham, 2004) is a 23-item self-report questionnaire with three subscales; each scale measures the severity of one of the three core components of compulsive hoarding: difficulty discarding, clutter and excessive acquisition. Items are rated according to a 4-point rating scale from 0 (none/not at all/never) to 4 (almost all/extreme/very often). The SI-R subscales have test-retest reliabilities exceeding r = 0.78, internal consistencies exceeding r = 0.87 and have displayed good convergent validity with both self-report scales and observer-rated scales of hoarding (Frost et al., 2004). Moreover, 	 35	the SI-R also displayed discriminant validity with measures unrelated to hoarding such as those measuring OCD symptoms. The mean scores obtained from people who hoard and non-hoarding individuals by Frost et al. (2004) clearly differentiated these two groups (M = 53, SD = 14 vs. M = 24, SD = 12). The SI-R is therefore a valid, reliable and suitable measure for assessing symptom severity of compulsive hoarding in both clinical and non-clinical groups. The SI-R was used as a secondary means of validating group assignment; mean scores of the clinical hoarding group were expected to meet a commonly used cut-off score (M = 41) and to be significantly higher than those of the healthy control group. SI-R total scores were also used as continuous measures of hoarding psychopathology. Clutter Image Rating Scale (CIR) The CIR (Frost, Steketee, Tolin, & Renaud, 2008) is a visual rating tool for measuring the severity of clutter in a home and can be completed as a self-report measure or an observer-rated measure. The CIR has three separate cards for rating kitchens, living rooms, and bedrooms, each containing 9 standardized colour photographs depicting increasingly severe levels of clutter. Research participants or observers make ratings ranging from 1 (no clutter) to 9 (severe clutter) for each type of room by selecting the picture that best corresponds to the amount of clutter in the home. Clinically significant clutter is defined by a rating of 4 or more for any of the rooms (Steketee & Frost, 2006). The CIR demonstrated good convergent validity with the clutter subscale of the SI-R (r = 0.72), discriminant validity through weaker correlations with other subscales of the SI-R not assessing clutter (r = 0.37 – 0.56) and good internal consistency for the composite score (i.e., the mean score of the three principal rooms α = 0.84; Frost et al., 2008).  	 36	One of the inclusion criteria for the hoarding group in the present study was a score of 4 or higher on any room in the home not intentionally used for storage. This criterion for clinically significant clutter was used instead of a composite rating of all rooms because the composite score may underrepresent the extent of hoarding problems in some situations. For example, participants may restrict their clutter to particular rooms in the home at the behest of roommates or spouses. A score of 7 in one’s bedroom and 1 in common spaces such as the kitchen and living room, would result in a composite score of 3 which is below the threshold for clinically significant clutter. Moreover, a high score in two commonly used spaces may be enough to create distress and impairment. For example, a score of 5 in one’s living room, 2 in one’s kitchen, and 4 in one’s bedroom, would result in a composite score below 4 even though this person is coping with an impairing amount of clutter in major areas of their home.  The Center for Epidemiologic Studies Depression – Revised Scale (CESD-R; Eaton, Smith, Ybarra, Muntaner, & Tien, 2004): The CESD-R is a revised version of a widely used measure of depression created for research and epidemiological studies, the CES-D (Radloff, 1977). The CESD-R has demonstrated high internal consistency, divergent and convergent validity with positive and negative affect, anxiety, and schizotypy (Van Dam & Earleywine, 2011). It measures symptoms of depression outlined in the Diagnostic and Statistical Manual of Mental Disorder, Fourth Edition, Text Revision, (DSM-IV-TR) such as sadness, poor appetite, and sleep disturbances, experienced in the past week on a four-point scale from 0 (rarely or none of the time) to 3 (most or all of the time). The CESD-R will be used in the current study to measure depression symptom severity. Depression severity was 	 37	included in analyses to investigate whether depression accounted for differences between the healthy control and hoarding group in task performance. Subjective Units of Distress Scale (SUDS) The SUDS (Wolpe, 1990) is a commonly used single question anxiety barometer used to determine how anxious or distressed a subject feels from 0 (no distress) to 100 (most distress possible). It has shown convergent validity with measures of autonomic arousal (Thyer, Papsdorf, Davis, & Vallecorsa, 1984). Subjects rated their SUDS level at the beginning of the IST and their peak SUDS at the end of the task. In the current study, task related distress was calculated by subtracting peak SUDS from the SUDS recorded at the beginning of the task.  Experimental Tasks Vignette-Based Task (VBT)  The VBT (see Appendix A) is a novel vignette-based task that assessed the impact of information about past utility and household space on discarding decisions. In each vignette on the VBT, participants read a brief description of a hypothetical household possession and indicated their likelihood of discarding that possession from 0 (absolutely keep) to 100 (absolutely discard) by moving a slider scale that is initially set at 50 (equal probability of discarding or keeping). The VBT contained 33 experimental vignettes and 10 foil vignettes. Of the experimental vignettes, 17 presented information on utility of the hypothetical possession, and 16 presented information on available household space. Utility-based vignettes contained information about how often a possession has been used in the past, and space-based vignettes contained information about the amount of household space available to store a possession. Vignettes were presented with either a “low” or “high” condition, 	 38	reflecting whether the information about past utility or household space indicated little past utility or available space (“low” condition) or ample previous use or available space (“high” condition). The impact of negative information about past utility or household space on discarding decisions was determined by comparing discarding scores across low and high conditions. For more detailed information about the construction and validation of the VBT, see chapter 3. Information Seeking Task (IST)  The IST (see Appendix B) is a novel task measuring information-seeking preferences when making discarding decisions about possessions. This task is described fully in Chapter 4. Briefly, participants were presented with hypothetical possessions and given the opportunity to access information about either the past or future utility of those items before making a discarding decision (i.e., to keep or discard the item). The two main outcome measures of the IST are the number of future utility questions that participants choose to unlock and the number of items discarded. Dysfunction Tolerance Scale (DTS)  The DTS (see Appendix C) is a Guttman-style scale developed for this study that measures differences in utility thresholds for discarding. Each item presents a household object and four descriptions of that object in states of increasing disrepair. For each item, participants select the highest (i.e., most dysfunctional) option they would still be willing to keep. For details regarding the development and validation of the DTS, see Chapter 5.  Validity Questions To ensure participants were providing meaningful answers to measures administered online, nine validity questions were interspersed among items on the DTS and VBT. These 	 39	validity questions were constructed to blend into the surrounding material. Validity questions placed before or after DTS items were presented in a multiple choice format similar to the DTS and required participants to choose the response option that was true or false (depending on the question). For example, one validity question presented participants with three true statements about cars (i.e., Most cars have four wheels; A van is a type of car; Cars have steering wheels) and one obviously false statement (i.e., Cars can fly), and asked participants “Which one is false?”. Validity questions within the VBT asked participants to move the slider toward true or false about a given statement (e.g., The sky is blue). Participants who provided incorrect answers to any one of these questions were assumed to be responding to study materials without thoroughly reading them. Therefore, these participants were excluded from data analysis and denied compensation. Twelve participants were disqualified on this basis.  Participants Participants were eligible for the study if they were 18-75 years old, fluent in English, and had not suffered a traumatic head injury. This upper age limit was chosen to help exclude individuals who exhibited hoarding behaviours associated with dementia. Traumatic brain injuries were screened to minimize the influence of neurocognitive deficits on decision-making processes.  To facilitate group-based analyses, participants were divided into a clinical hoarding group and a healthy control group. Participants in the healthy control group were screened for the presence of current mood or anxiety problems severe enough to cause impairment in order to limit the presence of information processing biases associated with mood and anxiety disorders. 	 40	Group assignment was primarily based on the results of a hoarding clinical interview administered during the phone screen and the Clutter Image Rating (CIR) scale. During the beginning of the study, the clinical interview was the Hoarding Rating Scale Interview (HRS; Tolin, Frost, & Steketee, 2010). In that case, participants were required to meet cut-off scores on three of the four items of the HRS to be included in the hoarding group. These four items assessed the core components of hoarding disorder according to the diagnostic criteria in the DSM-5. These items were: 1) difficulty discarding ordinary objects, 2) difficulty using one’s home due to clutter, 3) emotional distress due to hoarding problems and 4) impairment in daily life due to hoarding problems. The fifth item on the HRS addressed acquiring habits and therefore was not be used as an inclusion criterion because excessive acquisition is a specifier, not a core feature, of hoarding disorder in the DSM-5. Participants in the hoarding group were required to have scores in the high range (4-8) on items concerning difficulty discarding and clutter, plus a score in the high range on either impairment or distress. Participants in the healthy control group were required to have scores in the low range (0-2) on these HRS items, indicating that they experienced little to no difficulty with discarding or clutter. Several weeks into the study, the HRS was replaced with the MINI-Hoarding interview. Like the HRS, the MINI-Hoarding is a structured clinical interview that assesses the core DSM-5 criteria for hoarding disorder. A substantial benefit of the MINI-Hoarding within the context of a phone screen is that participants without clinical hoarding problems are progressively screened out when they fail to meet any core criterion for hoarding disorder. Conversely, the HRS requires interviewers to continue asking questions even after it becomes clear that the interviewee does not struggle with hoarding in order to make the 	 41	required ratings. Therefore, the MINI-Hoarding provided a more efficient alternative to the HRS for group assignment. As further means of validating group assignment, individuals in the hoarding group were required to report a score of 4 or higher on at least one room of their home not set aside for storage on the CIR. Individuals in the healthy control group were required to score 1-2 on all CIR items, suggesting they have little to no clutter in their home. The only exception to these cut-offs was when household clutter was being restrained by the ongoing intervention of another person. In these cases, participants were not required to have a CIR score of 4 or above if they met criteria for hoarding disorder according to the HRS or MINI-Hoarding. Participants who attained a combination of scores that did not meet group assignment criteria were not used in group-based analyses. Examples of participants who fell into this category include those who obtained scores in the middle range (3) on HRS items concerning difficulty discarding or clutter, and participants who appeared healthy or hoarding according to clinical interviews but provided scores on the CIR that were too low or high to meet criteria for those respective groups. Instead, the data collected from these participants were used in continuous analyses that examined hoarding as a dimensional construct.   Furthermore, group assignment was determined for 48 participants on the basis of a concurrent study on cognitive functioning in hoarding disorder. This concurrent study entailed a more thorough in-person assessment of hoarding symptoms and household clutter, and was therefore used for group assignment when participants overlapped between studies. This concurrent study utilized the same phone screen and group assignment criteria as the current study, with the added advantage of two procedural features that are likely to have enhanced diagnostic decision-making. Firstly, the concurrent study on cognition provided a 	 42	more stringent verification of CIR ratings. Specifically, CIR ratings were based on the interviewers’ clinical judgment of actual photos of participants’ homes. Within the current study, participants were provided with instructions on how to complete the CIR during the phone screen but these ratings ultimately depended on each participant’s judgement. Secondly, the cognition study employed the full Mini-International Neuropsychiatric Interview (Sheehan et al., 1998), including the Hoarding module developed by our lab. This full interview was not under the same time constraints as the phone screen. Therefore, interviewers in the cognition study had a greater amount of information to make their diagnostic decisions and were therefore more likely to provide accurate differential diagnoses. Due to these advantages, CIR scores and interviewer assessments from the cognition study were used to enhance the accuracy of group assignment when available. Sample Characteristics The same recruitment procedures, exclusion criteria, and methods of group assignment were used to collect data for the finalized versions of the VBT, IST, and the DTS. However, the amount of data collected for each task differed due to different rates of pilot testing and funding limitations. The following is a description of demographic characteristics of participants collected across all three tasks. The characteristics of the subsamples of participants who completed each task were consistent with the overall sample and will therefore not be presented.  The entire sample had a mean age of 42.60 years (SD = 15.33), and 72.4% of participants were female. No significant differences between the hoarding and healthy control group were found for gender, income, education, or age (see Table 1).  	 43	Table 1 Demographics of the Sample and Each Group   Group     Variable Full Sample (N = 174) % Healthy Control (n = 76) % Hoarding (n = 53) % X2   p Female 72.4 71.1 75.0 .39 .83 Highest Degree     6.09 .11 High school or less 23.0 17.1 30.2   College 19.0 14.5 22.6   Bachelor’s  36.8 43.4 30.2   Graduate degree 21.3 25.0 17.0   Income    4.21 .38 < $10, 000 14.9 14.5 13.2   $10, 000 – 19, 999 21.8 19.7 20.8   $20, 000 – 49, 999 31.0 28.9 35.8   $50, 000 – 99, 999 23.0 30.3 17.0      ≥ $100, 000  9.2 6.6 13.2   Age (M, SD) 42.60 (15.33) 43.39 (15.25) 43.21 (15.24)    Validation of Group Assignment As a means of validating group assignment across tasks, independent samples t-tests were conducted to examine differences between the healthy control and hoarding groups on self-report measures of hoarding severity (i.e., SI-R and CIR) and depression symptoms (i.e., 	 44	CESD-R; see Table 2 for means, t-tests, and effect sizes). As expected, the hoarding group reported substantially higher scores on the SI-R and the CIR than did the healthy control group. Six healthy controls and five hoarding participants did not provide complete data for CIR ratings during the online survey and therefore were not included in the analysis of CIR scores. SI-R and CIR scores reported by hoarding participants in this study were similar to those demonstrated in other studies of hoarding psychopathology (Frost, Steketee, & Grisham, 2004; Grisham et al., 2010; Tolin, Frost, & Steketee, 2010), suggesting that this sample was representative of other research samples. Also, as would be expected for a typical hoarding sample due to the comorbidity of hoarding and depression, the hoarding group scored significantly higher on the CESD-R than did the healthy control group.	 45	Table 2 Means and Standard Deviations of Self-Report Measures by Diagnostic Group   Measure Healthy Control (n = 76) M (SD) Hoarding (n = 53) M (SD)   t   p   d Saving Inventory –Revised      Clutter 7.96 (5.97) 21.16 (6.15) 12.20 < .001 2.18 Difficulty discarding 10.68 (4.70) 17.49 (3.44) 9.42 < .001 1.61 Acquisition 8.02 (4.78) 15.41 (5.42) 8.17 < .001 1.46 Total 26.67 (13.24) 54.07 (12.2) 11.91 < .001 2.13 Clutter Image Rating  1.54 (0.52) 3.37 (1.18) 10.05 < .001 2.15 Center for Epidemiologic Studies Depression Scale Revised 8.76 (10.69) 20.84 (18.95) 4.20 < .001 0.82          	 46	Chapter 3: Vignette-Based Task Task Construction and Refinement  The Vignette-Based Task (VBT) is a novel vignette task designed to measure the impact of information about past utility and household space on how likely a participant is to discard a possession. The development of the vignette task involved the generation of 32 experimental vignettes containing information about past utility or household space, and 10 un-scored foil vignettes designed to obscure the hypotheses being tested. Participants are asked to read each vignette and then decide how likely they would be to discard each item on a 100-point scale from 0 (absolutely keep) to 100 (absolutely discard). To make a rating on this scale, participants were required to drag a cursor from a midpoint towards either absolutely keep or absolutely discard. The midpoint of this scale can be conceptualized as a neutral point; dragging the cursor to the left (towards absolutely keep) indicated a decision to keep the item and dragging the cursor to the right (towards absolutely discard) indicated a decision to discard the item. The proximity of where participants placed the cursor to either end of this scale was used to indicate their level of certainty about that decision. For example, a rating of 95 on this scale would indicate that a participant would like to discard the hypothetical possession and is very certain about that decision. Conversely, a rating of 45 on this scale would indicate that a participant would like to keep the hypothetical possession but is not very confident about that decision.   Vignettes designed to assess neglect of information about past utility contained a sentence that indicated hypothetical possessions had either a high frequency or low frequency of past use. Information about the items’ future utility was held constant (i.e., one potential future use). For example, one of the utility vignettes stated: “You see a tent. You got the tent 	 47	some time ago, with the intention of going on camping trips but have not done so in the past 8 years. You could use it if you go camping with friends in the future.” Vignettes designed to assess neglect of information about household space contained information indicating that there was either a high or low amount of space available in participants’ homes to store hypothetical possessions while all other information about the possessions remained positive or neutral (e.g., item quality, item availability, item colour). For example, one space vignette stated: “You see a mug. It is white and can be bought at most stores carrying dishes. There is enough space in your kitchen cupboard to easily place it next to your other mugs.”  The foil vignettes presented descriptive information (e.g., colour, length, condition) about target items and required participants to choose a discarding method (i.e., keep, donate, throw away, give away to friend or family member). These three types of vignettes (i.e., utility, space, foils) were interspersed such that no more than two to three vignettes of the same type were presented consecutively.  Similarly to the IST, the items used in vignettes were selected to be diverse in terms of size, cost, frequency of use, and relevance to hoarding. Moreover, a reading level analyzer (https://readability-score.com/) was used to ensure that the final vignettes did not exceed a Grade 8 reading. Participants read the following context for their decision-making during the final version of the task: Imagine that you need to empty a storage area that contains a number of household possessions. You have been managing without these items, which means they are either not critical or not the only items you have. Some of these items are large and some are small things you’ve found in a junk drawer. As you go through the possessions there, you will need to decide which items to relocate to others areas in 	 48	your home and which items to discard (e.g., throw away, donate, recycle). To do this, you will need to go through many of your possessions, one at a time, and decide whether they are worth keeping. Please read the information in each item and indicate how likely you would be to discard each item, or what you would do with each item (e.g., keep, throw, donate, give away), based on the information provided. Pilot Studies Pilot Study 1. A preliminary pilot study was conducted to validate and refine the 42 vignettes initially developed for the vignette task. Fifteen personal contacts of laboratory members, who had no prior knowledge of the research, were recruited for this initial pilot study. Participants read each vignette and rated how likely they would be to discard each item on a 100-point scale as described earlier. Participants in this pilot study also completed a knowledge-based question about each vignette immediately after completing it. These knowledge questions were designed to act as a manipulation check by assessing the degree to which each vignette conveyed the intended information about space and utility (i.e., low past utility, high past utility, low household space, high household space).  For vignettes about past utility, the knowledge question asked participants how often they had used the object in the past based on the information in the vignette. For example, one utility-based vignette about a television was followed by the question “How much have you previously used the TV described in the last vignette?” from 0 (never) to 100 (frequently). For vignettes about household space, participants were asked to how much space was available to store the item. For example, one space based vignette about a mug was followed by the question, “How much space is there in your kitchen cupboard according to the last vignette?” from 0 (Completely Full) to 100 (Lots of Space). To be judged as successfully communicating low 	 49	ratings of space or past utility, items were required to have a mean rating of 30 or lower on their corresponding knowledge questions. Similarly, questions successfully communicating high space and past utility were required to have a mean rating of 70 or higher on their corresponding knowledge questions. Two pilot participants were excluded from the analysis as they were considered outliers due to extreme responding styles (i.e., providing either scores of 0 or 100 for most responses) and a total score of more than two standard deviations above the mean. Paired t-tests were used to compare the low and high conditions of the utility and space vignettes as a further manipulation check. As expected, participants were significantly more likely to discard items in the low utility condition compared to the high utility condition (see Table 3 for t-tests, mean scores, standard deviations, and effect sizes).  Similarly, participants were significantly more likely to discard items in the low space condition compared to the high space condition. The most divergent item from each high and low condition was removed to increase internal consistency.  Table 3 Mean Discarding Scores in Utility and Space Conditions from Pilot Study One Condition Discarding Score M (SD) Independent t-test  d High utility 20. 73 (16.41) t(12) = 7.53, p < .001 2.01 Low utility 53.73 (16.4)   High space Low Space 24.74 (17.50) 61.78 (15.71) t(12) = 7.72, p < .001 2.23 Note: N = 13 	 50	Internal consistency was acceptable for the space vignettes but below the acceptable range for the utility vignettes. Cronbach’s alpha was .81 for the high space vignettes and .76 for the low space vignettes. Cronbach’s alpha was .69 for the high utility vignettes and .64 for the low utility vignettes. Two items intended to communicate high utility failed to meet the criterion of a mean score of 70 on the knowledge question. These items were edited to increase their perceived utility. Five additional items were edited due to several participants providing incorrect responses on the knowledge questions. In addition to these changes, three items were added to the utility subscales in an effort to increase internal consistency.  Pilot Study 2. A second pilot study was conducted to evaluate the internal consistency of the scale after the modifications following the first pilot analysis had been made. This second pilot included data from the first fourteen healthy controls recruited as per the study procedures outlined in chapter 2. As with the first pilot analysis, participants read each vignette and rated how likely they would be to discard each item on a 100-point scale from 0 (absolutely keep) to 100 (absolutely discard). One participant was excluded from the analysis due to an extreme responding style (i.e., providing either scores of 0 or 100 for most responses) and a total score of more than three standard deviations below the mean.  The internal consistency of the subscales varied from low to moderate. Cronbach’s alpha for the vignettes in the high and low utility conditions was respectively .36 and .77. Cronbach’s alpha for the vignettes in the high and low space conditions was respectively .67 and .57. Examination of the raw data suggested that the low internal consistency of these subscales was due to extreme responding to many of the subscale questions, causing a lack of variation in participant responses. Extreme responding refers to participant responses of either 0 or 100, conveying that they would absolutely keep or discard an item respectively, 	 51	suggesting that the item evoked a strong negative or positive reaction. Extreme responses comprised 37% to 47% of all responses across conditions.  One potential explanation for this pattern of extreme responding was that the content of the vignettes carried overly positive or negative valences that provoked strong reactions from participants. As explained above, the space vignettes were originally written to include positive or negative information about household space flanked by neutral information. Utility vignettes were originally written to include positive or negative information about past utility flanked by a positively valenced distractor (i.e., one potential future use). Due to this structure, the high space and high utility vignettes contained only positive information about the items. Similarly, the low space vignettes contained only negatively valenced information about the items. By comparison, the low utility vignettes, which contained one piece of negative information (i.e., low past usage) and one piece of positive information (i.e., one potential future use), yielded more variation in participant responses and better internal consistency.  The presence of only positive or negative information was therefore hypothesized to unduly influence participant decisions about discarding by making the decision to keep or discard too easy. For example, one of the high utility vignettes described a flashlight that had frequently been useful in the past and was likely to be useful in the future: “You turn your attention to a flashlight. You can recall using the flashlight many times in the past five years during power outages and to look for things around the house. You could use it for light if the power goes out in the future.” Ten of the fourteen participants provided a score of 0 to this vignette, indicating that they would keep the flashlight with absolute certainty. This pattern of responding was hypothesized to be due to a lack of negative or conflicting 	 52	information that would provide a rationale to discard the flashlight, thereby making the decision-making process overly simplistic and less likely to vary across participants.  Pilot Study 3. To create vignettes that were more balanced and therefore less likely to provoke extreme responses, all vignettes were modified to contain one positively valenced and one negatively valenced distractor statement in addition to the experimental statement that indicated whether space or past utility was high or low. To preserve independence of the experimental conditions, the distractor statements within the utility vignettes did not contain information about household space, and the distractor statements in the space vignettes did not contain information about utility. The distractor statements commented negatively on features of the object unrelated to household space or past utility such as stylishness, condition, or quality. At this point in the VBT’s development, each vignette contained information that varied around a different hypothetical possession in order to maintain the participants’ interest in the task and to obscure the study’s hypotheses. One potential confound of this design was that differences between the high and low conditions of utility and space could be due to differences in distractor information and the hypothetical items featured in the vignettes. Therefore, two versions of the VBT were created to balance the non-experimental information (i.e., items, distractor information) across conditions. In this second version of this task (i.e., Version B), the valence of the critical experimental information (i.e., the information indicating whether past utility and household space were low or high) was flipped but the flanking distractor information and hypothetical possession were kept the same. This meant that the vignettes in the low conditions of Version B of the task contained the same distractor and item information as the vignettes in the high conditions of Version A. 	 53	Similarly the vignettes in the high conditions of Version B contained the same distractor and item information as the vignettes in the low conditions of Version A. The two vignettes below illustrate how one of the low utility vignettes in Version B was created from one of the high utility vignettes in Version A: valence of the critical statement (bolded) was flipped while all other aspects of the vignette remained identical.  Version A (High Utility): You pick up a space heater. It’s an older model that consumes quite a bit of electricity. You have used it many times in the past 5 years to keep warm during cold weather. You could use it to warm your living room next time there is a cold snap.  Version B (Low Utility): You pick up a space heater. It’s an older model that consumes quite a bit of electricity. You can’t think of a time in the past 5 years when your home has been cold enough to require a space heater. You could use it to warm your living room next time there is a cold snap. This design offered two key advantages. Firstly, non-experimental information was balanced across different vignette conditions by randomly assigning these two task versions to participants. Secondly, this design permits an additional type of manipulation check, wherein vignettes that vary with regard to experimental condition but are identically constructed in terms of distractor information and hypothetical possessions can be directly compared. This comparison enables verification that significant differences in discarding between the low and high conditions of each vignette type are due to differences in critical information between conditions and not extraneous information. I performed this manipulation check by comparing each vignette condition in Version A with its flipped counterpart in Version B in the next pilot.  	 54	The third pilot study was conducted to examine whether these changes improved the internal validity of the task. To facilitate rapid data collection from non-hoarding participants for this pilot, we recruited 54 participants using Amazon’s Mechanical Turk (mTurk) crowdsourcing tool. mTurk enables individuals and businesses to create and post jobs known as Human Intelligence Tasks (HITs; e.g., surveys, audio-visual tasks) on their server. Workers (i.e., people who sign up with Amazon to participate in these tasks) can then browse among existing jobs and complete them in exchange for a monetary payment. The two versions (A and B) of the vignette task were posted as separate tasks on mTurk.  Several steps were taken to enhance the quality of the data collected through mTurk and to verify that participants were relatively free of psychological problems that would impact task performance and fit within the study’s broader inclusion criteria. Firstly, potential participants had to pass a pre-survey screening measure. This screening measure included a series of questions that verified that participants were between the ages of 19-75, had never suffered a major head injury that caused loss of consciousness for more than a few minutes, and were not currently struggling with any serious emotional or mental health issues. To ensure the quality of the data, participants were also required to pass a series of validity questions that were embedded in the survey in order to be compensated and for their data to be included. These validity questions are described in chapter 2. As a further means of ensuring the quality of the data, participants were required to have achieved mTurk Masters status. mTurk Masters are “users who have consistently completed HITs of a certain type with a high degree of accuracy across a variety of Requesters” (Worker Web Site, 2017). These users were conceptualized as being the most likely to complete the tasks thoughtfully.  	 55	As expected, the revised versions of the vignette task yielded higher internal consistencies than those obtained for the previous version of the task. Cronbach’s alphas obtained for vignette conditions in Version A (n = 25) of the VBT were in the acceptable to good range. The high and low utility conditions for Version A were respectively .82 and .84. Cronbach’s alpha for the high and low space subscales of Version A was respectively .75 and .78. The Cronbach’s alphas for Version B (n = 29) of the VBT were also in the good to adequate ranges, with the exception of the low space condition that was slightly below adequate. Cronbach’s alpha for the high and low utility conditions of Version B were respectively .84 and .79. Cronbach’s alpha for the high and low space conditions of Version B were respectively .75 and .69. These improvements in internal consistency suggested that the changes made to the vignette task successfully corrected many of the problems in the previous version. With these new versions achieving internal consistency ratings in an adequate to good range, the study proceeded to collect data from a community sample. It was expected that these estimates would continue to improve with the collection of more data.   The data collected through mTurk was also used to confirm that the significant differences observed between the low and high vignette conditions were due to the intended manipulation rather than differences in peripheral information. This validity check involved the comparison of space and utility vignettes from Version A of the task with their Version B counterparts which only differed with regard to experimental condition (i.e., whether utility/space was high or low). As expected, discarding was significantly higher in the low utility (t(52) = -7.47, p < .001 , d = 2.05 ) and space (t(52) = -6.18, p < .001, d = 1.70) conditions of Version A than in the corresponding high utility and space conditions of Versions B. Similarly, discarding was significantly lower in the high utility (t(52) = 4.24, p < 	 56	.001, d = 1.16) and space (t(52) = 6.75, p < .001, d = 1.84) conditions of Version A than in the corresponding low utility and space conditions of Version B. These results suggest that the differences in discarding scores between the high and low conditions of the VBT are due to the intended experimental manipulation and not other differences in vignette content.  Results  Overview of Data Analytic Approach After pilot testing was complete, data on the VBT continued to be collected through the study procedures described in chapter 2. Participants who met inclusions criteria were directed to complete the VBT through an online platform alongside the other tasks and questionnaires. At the end of data collection, I evaluated the utility and space vignettes to determine which ones functioned adequately as part of their respective low and high conditions and therefore would be retained for the final analyses. Once poorly functioning items (defined below), were dropped from the VBT, I tested that the condition manipulation was preserved in the remaining vignettes. Subsequently, I modeled task version as a predictor of discarding score to verify that the two versions of the VBT did not account for meaningful variation in task performance.   Once the final set of vignettes was selected, the study’s theoretical predictions were tested using Linear Mixed-Effects Modeling (LMM) in R (R Core Team, 2017) using the lme4 package (Bates, Maechler, Bolker, & Walker, 2015). Like linear regression, LMM models the relation between a response variable and explanatory predictor variables. An advantage of LMM over simpler approaches for repeated measurement data is its ability to model individual differences between participants and test items that influence the outcome variable but are unrelated to variables of interest (Baayen, Davidson, & Bates, 2008). Within 	 57	LMM, individual differences can be modeled by assuming different random intercepts and/or slopes for each subject (Winter, 2013). These intercepts and slopes are referred to as random effects within LMM. The current study modeled by-subject and by-item variability by including random intercepts for both subject and item. Random slopes for space and utility conditions were also specified to model individual differences in extent to which discarding scores changed between conditions of these variables.   The space and utility vignettes were analysed separately.  In each set of analyses, experimental condition (i.e., high or low), hoarding psychopathology, and the interaction were incrementally entered as predictors of likelihood of discarding. Due to the potential influence of depression on decision-making tasks, CESD-R score was entered last to assess whether the effects could be accounted for by depression. Overall model improvement was assessed at each step of predictor entry by comparing the log-likelihood values of former and current models through chi-squared analyses. In each set of analyses, hoarding psychopathology was first analyzed as a categorical variable based on group assignment (i.e., hoarding or healthy control) and then as a continuous variable based on SI-R total scores. Categorical analyses only included participants who met inclusion criteria for the hoarding group (n = 52) or the healthy control group (n = 52). Continuous analyses utilized the entire sample, including participants who did not meet criteria for either diagnostic group (N = 144).  	 58	Selection of Final Items for the VBT. Once data collection was complete, Cronbach’s alpha and corrected item-total correlations were calculated for each VBT condition (i.e., low utility, high utility, low space, high space) to determine which items functioned adequately and would therefore be retained for analysis. When items were removed from Version A of the VBT, their Version B counterparts were also removed in order to maintain the integrity of the manipulation. This strategy aimed to balance the psychometric properties of the task with the internal validity of the experimental design by only retaining items that functioned adequate across both versions of the task and ensuring that all participants were exposed to the same information regardless of which version they received.  While there is no strict minimum for acceptable item-total correlations, Field (2009) recommends 0.30 as a minimum value for retaining items during scale development and notes that items below that limit are likely problematic. Therefore, a lower limit of .30 for corrected item-total correlations was used to identify items for removal. After the removal of items below this threshold, Cronbach’s alpha for the low and high subscales of utility and space ranged from .74 to .86. These values indicated that the internal consistency of all final VBT subscales fell within an acceptable range (Tavakol & Dennick, 2011).  To verify that the condition manipulation was preserved in this final set of vignettes, paired t-tests were used to compare average discarding scores between the high and low conditions of the utility and space. As expected, participants were significantly more likely to discard items in the low utility condition compared to the high utility condition (see Table 4 for t-tests, mean scores, standard deviations, and effect sizes). Similarly, participants were 	 59	significantly more likely to discard items in the low space condition compared to the high space condition. Table 4 Means Discarding Scores in Utility and Space Conditions of Final Vignettes Condition Discarding Score M (SD) Independent t-test  d High utility 34.60 (21.44) t(143) = -10.84, p < .001 0.84 Low utility 53.79 (24.07)   High space Low Space 29.93 (21.74) 56.20 (26.28) t(143) = -14.87, p < .001 1.08 Note: N = 144 I then verified that task version did not account for meaningful variation in VBT responses. LMM models were constructed with vignette version as a predictor of discarding score, and participant-intercept and item-intercept as random effects. As expected, vignette version was not a significant predictor of discarding score for utility vignettes, b = 2.71, SE = 6.01, 95% CI [-9.19, 14.62], p = .65, or space vignettes, b = 4.09, SE = 6.48, 95% CI [-8.84, 17.03], p = .53. Moreover, the addition of vignette version as a predictor did not significantly improve the goodness of fit of the intercept-only model for utility, χ2(1) = 0.20, p = .65, or space vignettes, χ2(1) = 0.39, p = .52. These results demonstrated that Version A and B of the VBT did not result in significant differences in how participants responded to the task, and participant responses could therefore be compared across task versions.  Coding, Statistical Software, and Assumptions. Continuous predictors were mean centered to reduce the impact of multicollinearity and to improve interpretability. Categorical 	 60	variables such as vignette condition (i.e., low and high) and diagnostic group (i.e., hoarding an healthy control) were dummy coded before being entered into the analysis. The high condition in analyses of utility and space vignettes were coded as 0 and the low conditions were coded as 1. Within the diagnostic group variable, the healthy control group was coded as 0 and the hoarding group was coded as 1. Additionally, participant responses were recoded to center around the neutral midpoint on the response scale (i.e., 50) to ease interpretation. Accordingly, scores below 50, which indicated a decision to keep an item, became negative. Scores above 50, which indicated a decision to discard an item, became positive. Increasing numbers in either direction indicated participants’ levels of certainty in those decisions. For example, both scores of -20 and -40 would indicate that a participant decided to keep an item but -40 indicates a greater level of certainty in that decision.  Additional R packages were used in combination with the lme4 package to supplement the statistics produced. The lmerTest package (Kuznetsova, Brockhoff Christensen, 2013) was used to produce p-values for fixed effects based on Satterthwaite approximations for denominator degrees of freedom, which has been recommended to control for Type 1 error in LMM (Luke, 2016). The MuMIn package (Barton, 2015) was used to produce marginal and conditional R2 values based on Nakagawa and Schielzeth (2013) and Johnson (2014). Marginal R2 values reflect the variance explained by fixed effects, whereas conditional R2 values reflect both mixed effects plus random effects. The current study will report the marginal R2 as the fixed effects are of primary interest for the hypotheses.   I conducted several tests to verify whether the assumptions underlying LMM were met for this data set before running the main analyses. Multivariate normality and 	 61	homoscedasticity were assessed through examination of residual distributions. To assess the impact of multivariate outliers, I compared models that excluded cases having outlier standardized residuals (Z ≥ 3.0) to models utilizing all data. When the interpretation between these reduced and full models was the same, the models utilizing all data were reported. Multicollinearity between predictor variables was assessed through variance inflation factors and was found to be within acceptable ranges.  Analysis of Utility Vignettes Hoarding as a Categorical Predictor of Discarding. I hypothesized that the interaction between diagnostic group and utility condition would be a significant predictor of discarding score, indicating that the impact of utility on discarding scores differs depending on hoarding status. To investigate this hypothesis, I first specified a null model that contained discarding score as the dependent variable and participant-intercept and item-intercept as random factors. I then incrementally added predictor variables (utility condition [high, low], diagnostic group [healthy control, hoarding], and their interaction [utility condition x diagnostic group]) to the null model to evaluate which terms improved the model’s predictive ability. Model improvement was evaluated using chi-square tests on the log-likelihood values (i.e., a measure of model fit in LMM) between models.  The addition of utility condition and diagnostic group as predictors significantly improved the null model, χ2(2) = 44.09, p < .001, suggesting that both utility and diagnostic group accounted for significant variance in discarding scores. The interaction between utility condition and diagnostic group was then added to this model, also producing a significant improvement in model fit: χ2(1) = 11.04, p < .01. Lastly, CESD-R was added to the model as a predictor to evaluate whether depression symptoms accounted for the effects observed thus 	 62	far. The addition of CESD-R did not significantly improve model fit, χ2(1) = 0.04, p = .85, and did not alter the significance of the other predictors. Therefore, CESD-R was not included in the final model.  Following guidelines by Barr (2013) regarding the specification of random effects structures in LMMs containing interactions, a random slope for utility condition was added to the model. The addition of a random slope for utility condition resulted in a significant improvement in fit for the final model, suggesting that participants differed in the extent to which utility condition influenced their discarding score: χ2(2) = 26.08, p < .001.  The final model of discarding scores included participant-intercept, item-intercept, and slope of utility condition as random effects, as well as utility condition, diagnostic group and their interaction as fixed effects. Within this final LMM (Marginal R2 = 0.13), low utility was a significant predictor of probability of discarding, b = 24.62, SE = 4.67, 95% CI [15.46, 33.76], p < .001. Similarly, hoarding group membership was a significant negative predictor of probability of discarding, b = -13.56, SE = - 4.12, 95% CI [-22.81, -6.69], p < .001. The interaction between utility condition and diagnostic group was also a significant predictor of probability of discarding, b = -9.81, SE = 4.00, 95% CI [-15.06, -0.39], p = .02.  The interaction effect was of primary interest, and as hypothesized, indicated that the effect of group depended on utility condition. Although the hoarding group was significantly more likely to keep objects than the healthy control group in both conditions, the difference between groups was greater at low levels of utility (d = 0.86) than at high levels of utility (d = 0.52; see Table 5 for cell means and t-tests).   	 63	Table 5 Means and Standard Deviations for Discarding Score by Condition and Diagnostic Group   Condition Healthy control M (SEM) n = 52 Hoarding M (SEM) n = 52  Independent t-test High utility -7.67 (3.92) -22.43 (3.90) t(101.97) = 3.59, p < .001 Low utility 16.94 (3.99) -7.62 (3.96) t(102.73) = 5.82, p < .001 Note: Values below zero indicate greater probabilities of keeping items and values above zero indicate greater probabilities of discarding items.   The disparity in the size of between-group differences was due to the healthy control group’s greater responsiveness to negative information about past utility. As displayed in Figure 1, both groups kept objects in the high utility condition, but only the healthy control group became willing to discard objects (indicated by crossing the dotted red line) in the low utility condition. Although participants in the hoarding group were responsive to the condition manipulation, their shift in discarding scores was not large enough to move them over the threshold of discarding (i.e., from negative scores to positive scores).  Overall, this pattern suggests that people in the hoarding group are responsive to changes between utility conditions (i.e., they differ in their certainty about keeping objects), but they are not as responsive as the healthy control group, which became willing to discard objects on the basis of low utility. This pattern of findings supports the hypothesis that information indicating an object has not been used is less likely to persuade people who hoard to discard that object, likely because negative information about past utility is not viewed as unfavorably as it is by healthy controls.  	 64	Figure 1 Mean Discarding Score and 95% CI by Utility Conditions and Diagnostic Group    Note: The red dotted line represents threshold for discarding a hypothetical object. Values below the dotted red line indicate keep decisions, and values above the dotted red line indicate discard decisions.  Hoarding Symptom Severity as a Continuous Predictor of Discarding. The analyses Propensity to DiscardGroupControlHoarding-40 -20 0 20 40 = utility highControlHoarding = utility low	 65	described above were performed again with SI-R total score substituted for diagnostic group as a continuous measure of hoarding psychopathology. Similarly, I hypothesized that the interaction between SI-R total score and utility condition would be a significant predictor of discarding score, indicating that the impact of utility on discarding score differs depending on severity of hoarding symptoms.  Consistent with the prior analysis, utility condition, b = 21.20, SE = 4.07, 95% CI [13.20, 29.19], p < .001, SI-R total score, b = -0.27, SE = 0.09, 95% CI [-0.45, -0.09], p = .003, and the interaction between utility condition and SI-R total score, b = -0.37, SE = 0.09, 95% CI [-0.54, -0.19], p < .001, were all significant predictors of the probability of discarding hypothetical possessions on the VBT (Marginal R2 = 0.13). As hypothesized, the significance and direction of the interaction suggests that the relationship between hoarding severity and discarding differs depending on an item’s history of use. Specifically, greater hoarding severity decreases the probability of discarding when an item has low past utility. This finding reinforced the earlier interpretation that hoarding severity is associated with a different way of responding to low utility items when deciding whether items should be kept or thrown away.  Analysis of Space Vignettes Hoarding as a Categorical Predictor of Discarding. The analyses performed above were conducted again with the data obtained from the space vignettes to investigate the impact of information about household space on discarding decisions. As with the utility vignettes, the addition of space condition and diagnostic group as predictors significantly improved the null model, χ2(2) = 49.47, p < .001, suggesting that space condition and diagnostic group accounted for significant variance in discarding scores. The addition of the 	 66	interaction between space condition and diagnostic group to this model did not produce a significant improvement in model fit, χ2(1) = 2.08, p < .15. Consistent with the analyses of the utility vignettes, the addition of depression (i.e., CESD-R) as a predictor did not significantly improve model fit, χ2(1) = 0.22, p = .55, and was therefore not retained in the final model. Following guidelines by Barr (2013) regarding the specification of random effects structures in LMMs containing interactions, a random slope for space condition was added to the model. The addition of a random slope for space condition resulted in a significant improvement in fit for the final model, suggesting that participants differed in the extent to which space condition influenced their discarding score, χ2(2) = 17.55, p < .001.  The final LMM of discarding scores included participant-intercept, item-intercept, and a random slope of space as random effects, as well as fixed effects of space condition, group, and the interaction between space condition and group. Within this final model (Marginal R2 = 0.15), low space, b = 25.83, SE = 4.61, 95% CI [16.80, 34.86], p < .001 and hoarding group membership, b = -17.39, SE = 3.9, 95% CI [-25.06, -9.72], p < .001 were significant predictors of probability of discarding.  Unlike the results of the utility vignettes, the interaction between space condition and group did not reach statistical significance, b = -5.44, SE = 3.77, 95% CI [-12.86, 1.96], p = .16, and therefore did not support the study’s hypothesis regarding the reduced impact of low space on intent to discard among individuals with hoarding problems. However, the pattern of results observed in the utility analyses wherein healthy controls became willing to discard items in the low utility condition, and individuals in the hoarding group became significantly more willing but did not ultimately shift above the discarding threshold (i.e., 0), was also 	 67	apparent in the space analyses (see Figure 2).  Figure 2 Mean Discarding Score and 95% CI by Space Condition and Diagnostic Group   Note: The red dotted line represents threshold for discarding a hypothetical object. Values below the red dotted line indicate keep decisions and values above the red dotted line indicate discard decisions.  Also consistent with the results of the utility vignettes, there were significant Propensity to DiscardGroupControlHoarding-40 -20 0 20 40 = space highControlHoarding = space low	 68	difference between groups in both conditions, but these differences were greater at low levels of space (d = 0.78) than at high levels of space (d = 0.62; see Table 6 for cell means and t-tests).  Table 6 Means and Standard Deviations for Discarding Score by Space Condition and Diagnostic Group  Condition Healthy control M (SEM) n = 52 Hoarding M (SEM) n = 52 Independent t-test High space -10.15 (3.84) -27.55 (3.82) t(101.99) = 4.45, p < .001 Low space 15.68 (4.10) -7.16 (4.09) t(102.58) = 5.17, p < .001 Note: Values below zero indicate greater probabilities of keeping items and values above zero indicate greater probabilities of discarding items.   Hoarding Symptom Severity as a Continuous Predictor of Discarding. Mirroring the analyses for utility vignettes, the space condition analyses performed again with SI-R total score as a continuous measure of hoarding psychopathology. In this model (Marginal R2 = 0.15), the addition of the interaction term between SI-R total score and space condition did result in a significant improvement in model fit, χ2(1) = 22.01, p < .001, and this interaction was a significant negative predictor, b = -0.31, SE = 0.08, 95% CI [-0.47, -0.15], p < .001, of probability of discarding in the final model. Given the relatively small difference in effect sizes between space conditions in the group-analyses, it is likely that the continuous analyses yielded significant results due to the larger sample size and therefore greater power. Taken together, this pattern of results provides partial support for the study’s hypothesis that 	 69	individuals in the hoarding group are less responsive to negative information about household space when making discarding decisions. However, the size of this effect is likely to be smaller than the equivalent effect for utility.  Discussion of Vignette-Based Task   Findings   The response scale for the VBT was constructed as deviations from a central point of perfect indecision about keeping versus discarding. Participants dragged a visual indicator to the right to indicate the strength of a ‘more likely to discard’ decision and to the left to indicate a ‘more likely to keep’ decision.  With this conceptualization, continuous responses could be categorized as “keep” or “discard” according to which side of the response continuum participants used.   As predicted, information suggesting an object had rarely been used in the past led participants to make decisions on the “discard” side of the continuum. Healthy control participants changed their responses to the “keep” side of the continuum when vignettes indicated that objects had frequently been used in the past. In contrast, hoarding participants were significantly less sensitive to this information; their decisions remained on the “keep” side of the decision continuum even for objects that had been rarely used. Similarly to controls, hoarding participants were more likely to discard objects that had rarely been used compared to objects that had frequently been used. However, this shift in willingness to discard was not strong enough to prompt hoarding participants to move the response indicator into the “discard zone”, as it did for healthy control participants. The finding that people with hoarding problems are less likely to alter their discarding decisions on the basis of information about past utility was true both when considering diagnosed hoarding (i.e., 	 70	categorical) and dimensional hoarding symptoms.  These findings support the study’s hypothesis that people who hoard are more likely to neglect negative information about past utility in their thinking about whether to keep or discard possessions.  The vignettes manipulating information about amount of household space provided weaker support for the study’s hypotheses. Although the pattern of participants’ responses appeared very similar to the utility vignettes, with hoarding participants displaying a smaller shift in responsiveness to low space than the healthy control participants, this interaction between space condition and diagnostic group was not statistically significant. The extent to which hoarding participants and healthy controls differed in their willingness to discard possessions changed more with the utility information manipulation (d = 0.86 vs. 0.52, for low and high utility, respectively) than it did with the manipulation of information on household space (d = 0.78 vs. 0.62 for low and high space, respectively). When hoarding symptoms were measured dimensionally, the interaction with space condition was significant in the predicted direction and consistent with the results observed for utility. Overall, these findings provide weak support for the hypothesis that people with hoarding problems are more likely to neglect negative information about available household space when making discarding decisions. Further testing is needed to clarify whether this interpretation also applies to categorical conceptualizations of hoarding.   Implications  Overall, findings on the VBT suggest that hoarding symptoms are associated with neglect of negative information about past utility and household space when making decisions about possessions. This offers a partial explanation for why people with hoarding 	 71	problems make different decisions than non-hoarding individuals about possessions that have not been used in a long time and in situations of limited household space.  If intention to keep a possession can be taken as an indicator of valuing, then these results suggest that negative information about past usage and household space devalues items more for healthy individuals than for hoarding individuals. For the healthy control group, information indicating items were not frequently used in the past or could not be stored comfortably in the home devalued items so much that participants moved their decisions into the “discard zone” (on average). This suggested that low past utility and low space acted as significant motivators for discarding in the control group. The hoarding group’s valuation of items changed so their “keep” decisions were less strong, but they did not go so far as to move toward the “discard” side of the decision. That is, the hoarding group’s decisions were still toward “keeping”, but the shift in strength of this decision shows that these participants did attend and respond to the information. This shallower decline in value may make it more difficult for people who hoard to justify discarding items that have positive qualities but have become less useful or are taking up too much space. Conversely, healthy people may find it easier or may be more motivated to let go of possessions in these circumstances, thereby curtailing the accumulation of clutter.  The finding that hoarding symptoms are associated with less sensitivity to negative information about past utility and household space has clinical implications. Clinicians could strive to help hoarding clients restructure their decision-making processes to more carefully consider negative information about utility and space. For example, clinicians could develop structured approaches to decision-making about possessions similar to those available for problem-solving. Similarly to how problem-solving strategies have discrete steps for patients 	 72	to follow, possession based decision-making strategies could include steps that explicitly encourage patients with hoarding to consider and weigh evidence that they might otherwise dismiss (e.g., low frequency of past use, low amount of household space).  Moreover, clinicians could also help clients to craft discarding rules that set more realistic standards for the degree of past usage and household space to justify keeping an item. The results of the VBT suggest that individuals who hoard may implicitly know that items that have not been used frequently in the past are of somewhat lower value. However, they may still be unsure of where to ‘draw the line’ for when an object should be discarded. Similarly, individuals who hoard may be willing to discard possessions when space is restricted, but remain unsure of an appropriate threshold for ultimately discarding possessions. CBT for hoarding currently involves decision trees that help clients make concrete plans about where they will store the items that they choose to keep during sorting exercises. The current research provides support for this practice, and suggests the utility of developing additional strategies for helping people who hoard to think more critically about space constraints. Limitations and Future Directions The finding that people with hoarding problems were less likely to discard items on the basis of lack of household space raises the question of why space is not as compelling a motivator for discarding in hoarding. One theory is that people who hoard have a greater baseline tolerance for cluttered living environments than people who do not hoard. Hypothetically, there may be individual differences in the extent to which clutter impedes enjoyment and comfort in living spaces. Some people may be at ease living and working in cluttered and disordered environments, whereas others may require clear environments and 	 73	become distressed when their homes become disorderly and cluttered. As demonstrated on the space vignettes, healthy controls were less likely to decide to keep items that they did not have space for, possibly because they knew that keeping things that they do not space for will result in household clutter. The healthy control participants may find living in cluttered spaces to be more aversive than do the hoarding participants, with this preference being reflected in their discarding choices. Alternatively, even if both groups find uncluttered living environments to be equally appealing, perhaps people who hoard judged the items to be more valuable than the living space they would occupy. Future research could explore this question of individual differences in aversion to clutter in living spaces, and whether such differences serve as a vulnerability or maintenance factor for hoarding.   One limitation of the space vignettes that may help to explain their weaker results is that the information presented in the vignettes does not map neatly onto the challenges people face when making decisions about objects and space at home.  The space vignettes presented clear scenarios where one specific target area was either full or had plenty of room to store each item. These scenarios presented more straightforward options than occur in cluttered homes, which likely involve a much higher level of uncertainty about whether there is “enough” space for each item in the target space. Furthermore, more areas may be considered as potential target storage locations for a given item in hoarded homes. For example, if one’s closet is full, finding space for a cherished sweater may involve looking beyond the closet to other target areas of the home such as the hallway or living room. In turn, choosing to “find space” for clothing items in those locations may come with additional consequences to consider (i.e., reduced use of those areas). Therefore, the space vignettes 	 74	may not have optimally elicited the kind of distorted thinking about space that occurs in hoarding (if it does occur).  A more general limitation of the VBT is that it artificially provided information that participants may not have otherwise generated themselves when making decisions about possessions. Each vignette explicitly provided information about utility and space to guide decision-making. It is possible that many people who hoard would not spontaneously consider these types of information when thinking about their possessions at home. However, as previously discussed, past research suggests that utility is a commonly cited reason for keeping possessions in hoarding and is therefore likely to be considered frequently outside of the laboratory.                	 75	Chapter 4: Information Seeking Task Task Construction and Refinement The following describes the development of the Information Seeking Task (IST), which is a forced choice decision-making task used to measure information-seeking preferences when making discarding decisions about hypothetical possessions. The IST was constructed using Inquisit software (Inquisit 4, 2015). During the IST, participants were presented with the names and images of objects and asked to decide whether to keep or discard them based on information about the item’s past or future utility. The task contained 30 experimental trails and 15 foil trials. In each experimental trial, participants were given the choice to unlock the answer to one of two questions about hypothetical possessions, one question assessing future utility or one question assessing past utility. Questions assessing past utility focused on frequency (e.g., How often have I used this in the past year?), timing (e.g., When was the last time I used this?), or purpose of past usage (e.g., What did I last use this for?). Questions assessing future utility focus on anticipated frequency of future use (e.g., How often do I think I will use this in the next year?), future events in which the object could be used (e.g., What occasions could I use this for?), or purpose of future usage (e.g., How could I use this in the future?). Variants of these questions were counterbalanced across trials.  After a participant selected the question they wished to unlock, they were provided with an answer to their selected question. These answers are not of experimental interest, as they were primarily constructed to maintain participant engagement in the task. To this end, answers to selected questions were written in first person perspective and referenced various scenarios that someone may generate when reflecting on the past or potential future use of a 	 76	possession. For example, on the experimental trial where participants are presented with a pair of hiking boots, participants who unlocked the question about future utility (i.e., “What occasions could I use this for?”) would then be told: “I could wear these the next time I go hiking in North Vancouver.” In contrast, a participant who selected the question about past utility (“How often have I used this in the past 2 years?”) would be told: “I used this on a group hiking trip in Capilano last year… nothing other than that in the past few years”. These answered were constructed such that items assessing past and future utility had similar average word counts and were balanced in providing positive (i.e., that objects had been used in the past or would be used in the future) and negative information (i.e., that objects had not been used in the recent past or were unlikely to be used in the future). Furthermore, questions and answers did not exceed a grade 8 reading level to increase their likelihood of being understood by a wide range of participants. After participants read the answers to their selected questions, they are asked to make a decision to keep or discard the hypothetical possession. To make the decision-making process more realistic and challenging in these experimental trials, 20 of the 30 hypothetical possessions were selected to reflect low frequency of use (e.g., tennis racket, fondue pot, rain poncho, snow globe). Items that are likely to be used on a daily basis were considered less likely to provoke indecision about whether an item is “useful enough” to keep. The remaining ten experimental trial possessions were chosen to reflect typically hoarded items that are likely to be especially relevant to hoarding group (e.g., paper items, toiletries, containers).  Foil trials were designed to obscure the hypothesis being testing by the IST. Foil trials were functionally similarly to experimental trials, but the questions participants could unlock 	 77	were related to expense (e.g., What is the price of this item?), sentimentality (e.g., Does this item have special meaning to me?), brand (e.g., What brand is this item?), or the age of the possession (e.g., What version or generation is this item?). The fictional items used for foil trials were chosen to be appropriate to question content. For example, trials where participants could ask questions about the age of possessions included electronics. Foils appeared in every third trial, such that only two experimental trials were presented consecutively at a time.  Before participants began the IST, they were given a cover story for their decisions and incentive for providing a realistic representation of their discarding practices at home. In terms of incentive, participants were reminded of the larger implications of their participation in the study. Specifically, they were informed that the task they were about to complete was new and required their help to develop it into a realistic measure of how people make decisions. Moreover, they were told that the successful development of this task would have important implications for consumer behaviour, social psychology, and mental health. Following this information, participants read the following context for their decision-making during the task: Imagine that you no longer have enough space for all of your possessions and have set aside time to conduct a ‘spring cleaning’ of your home. You will be going though many of your possessions, one at a time, and deciding whether they are worth keeping. In order to figure out what possessions are worth keeping and what possessions are simply taking up space, you will be able to access different kinds of information about each possession. However, you will only be able to access one piece of information before making each decision, so choose carefully. 	 78	Lastly, participants were prompted to provide a Subjective Units of Distress (SUDS) rating at the beginning of the IST and a second SUDS rating at the end of the IST to indicate their peak level of distress during the task. These ratings were collected as an additional means of validating whether the IST is reflective of real-life decision-making, as people in the hoarding group would be expected to find it more stressful to make discarding decisions than individuals without hoarding problems. These differences in distress were expected due to indecisiveness and distress experienced during discarding in hoarding (Fitch, 2011; Frost & Gross, 1993; Grisham et al., 2010; Steketee, Frost, & Kyrios, 2003; Tolin & Villavicencio, 2011; Wincze et al., 2007). In previous research, individuals with hoarding disorder, as would be expected, report greater distress than healthy controls in tasks that involve discarding possessions (Tolin et al., 2012) and sorting possessions into categories (Grisham et al., 2010; Hayward, 2011; Kellman-McFarlane, 2013; Wincze et al., 2007). Therefore, individuals with hoarding symptoms were expected to experience significantly greater distress during the IST compared to the healthy control group. The initial pilot analysis took place after data were collected from the first 21 healthy controls recruited as per the procedures outlined in chapter 2. The purpose of this initial analysis was to evaluate the internal consistency of participants’ question selections across trials, as this was the primary outcome variable and central to the hypothesis testing. Specifically, I wished to determine whether participants were consistent in their choices, rather than selecting items at random or responding in a manner that produced limited variability. Future question selections were coded as 1 and past question selections were coded as 0 for these analyses. Cronbach’s alpha for items in the experimental trials was .86, suggesting good internal consistency. At the time of this pilot analysis, the average number 	 79	of future based questions chosen across trials was 15.09 (out of 30) with a standard deviation of 6.03. Overall, these figures suggested that the task displayed internal consistency and that participants were not engaging in extreme responding styles. Results  Overview of Data Analysis   The IST data were analyzed using Generalized Linear Mixed-Effects Modeling (GLMM) in R (R Core Team, 2017) using the lme4 package (Bates, Maechler, Bolker, & Walker, 2015). GLMM was selected as an analytic strategy because it enables logistic regression modeling with random and fixed effects. The current study modeled by-subject and by-item variability by including both subject and item as random intercepts. Random slopes for SI-R total score or diagnostic group (i.e., hoarding or healthy control), depending on the model, were also specified to model individual differences in extent to which response variables scores were influenced by these variables.   As an additional means of task validation, I examined the relation between hoarding and discarding and distress during the IST. Specifically, participants’ discarding decisions (i.e., whether they chose to keep or discard hypothetical possessions) on each trial were predicted from SI-R scores or diagnostic group while controlling for depression. With regard to distress, I tested for between-group differences in the change in SUDS collected at the beginning and end of the IST. If the task adequately reflects true discarding behavior, hoarding symptoms are expected to be a significant predictor of decisions to keep possessions and task related distress.   The hypothesis that hoarding is associated with a bias toward thinking about the future utility of possessions when making discarding decisions was examined by fitting a 	 80	GLMM with information choice (i.e., whether participants chose to unlock information about future or past utility) on each trial as the response variable, and hoarding psychopathology and depression (i.e., CESD-R) as predictor variables. These analyses were performed twice; first with hoarding psychopathology measured as a categorical variable based on diagnostic group (i.e., hoarding or healthy control) and then with dimensional hoarding symptoms represented as a continuous variable (SI-R total score).  Categorical analyses only included participants who met inclusion criteria for the hoarding group (n = 42) or the healthy control group (n = 56). Continuous analyses utilized the entire sample, including participants who did not meet criteria for either diagnostic group (N = 138).   Coding, Statistical Software, and Assumptions. Continuous predictors were mean centered to reduce the impact of multicollinearity and to improve interpretability. Categorical variables such as information choice, discarding decision, and diagnostic group were dummy coded before being entered into the analysis. Within the discarding decision variable, choices to discard possessions were coded as 0 and choices to keep possessions were coded as 1. For the diagnostic group variable, the healthy control group was coded as 0 and the hoarding group was coded as 1. In the information choice variable, choices to access information about past utility were coded as 0 and choices to access information about future utility were coded as 1.  Additional R packages were used in combination with the lme4 package to supplement the statistics produced. The lmerTest package (Kuznetsova, Brockhoff, Christensen, 2013) was used to produce p-values for fixed effects based on asymptotic Wald tests. The rcompanion package (Mangiafico, 2017) was used to produce Nagelkerke’s pseudo R2.  	 81	 To assess the impact of multivariate outliers, I compared models that excluded cases with standardized residuals of Z ≥ 3 at a distance greater than 3 standard deviations from 0 with models utilizing all data. When the interpretation for these reduced and full models was the same, the models utilizing all data were reported. In the event that an assumption of parametric testing was not met, non-parametric testing was used to examine study hypotheses. When the conclusions of non-parametric testing were consistent with parametric testing, parametric test results were reported for ease of interpretation. Task Validation Discarding Choices. If the IST reflects real life discarding behaviour, hoarding should be associated with more frequent decisions to keep hypothetical possessions. As a means of testing the validity of the task, a GLMM was specified with discarding decision as the response variable, diagnostic group [healthy control, hoarding] as a predictor, and subject-intercept and item-intercept as random effects. This model was a significant improvement in fit from the intercept-only model, χ2(3) = 13.57, p < .01. The further addition of CESD-R (i.e., depression symptoms) as a predictor significantly improved model fit, χ2(1) = 9.49, p = .003, and was therefore retained. As expected, diagnostic group was a significant predictor of decisions to keep possessions, b = 0.80, SE = 0.23, 95% CI [0.35, 1.27], p < .001, OR = 2.25. These results indicate that individuals in the hoarding group were 2.25 times more likely to keep objects than individuals in the healthy control group. CESD-R score was also a significant predictor of keep decisions, b = 0.02, SE = 0.01, 95% CI [0.02, 0.04], p = .003, OR = 1.02. The pseudo-R2 estimate for this overall model was small (Nagelkerke = .011). 	 82	Task-Related Distress. I hypothesized that participants in the hoarding group would experience significantly greater distress during the IST compared to the healthy control group. To examine this hypothesis, a SUDS change score was created for each participant by subtracting the amount of distress reported at the beginning of the task from the peak amount of distress reported at the end of the task. A one-way ANCOVA was conducted to examine whether there was a significant difference between the hoarding and healthy control groups in these SUDS change scores, while controlling for depression symptoms (i.e., CESD-R).  As expected, the hoarding group (M = 5. 77, 95% CI [1.01, 10.52], SEM = 2.39) displayed a significantly greater increase in distress than the healthy control group (M = -1.05, 95% CI [-5.14, 3.03], SEM = 2.05) which was not accounted for by depression symptoms, F(1, 95) = 4.43, p = .04. The effect size for this difference was small to medium in size (ηp2 = .04). This result supports the validity of the IST as a discarding task because it indicates that decision-making during the IST was engaging enough to provoke heightened distress in the hoarding group.  Future-Oriented Information Seeking Hoarding as a Categorical Predictor of Information Seeking. I hypothesized that hoarding would be a significant predictor of future-oriented information seeking about utility. To test this hypothesis, a GLMM was specified with information seeking as the response variable, diagnostic group [healthy control, hoarding] as a fixed effect, and subject-intercept, item-intercept, and slope of diagnostic group as random effects. The addition of depression as a predictor was modeled but did not significantly improve model fit, χ2(1) = 1.69, p = .19 or alter the significance of the other predictors. Therefore, depression was not included in the final model. 	 83	Diagnostic group was a significant predictor of the choice to access information about future utility, b = 0.44, SE = 0.20, 95% CI [0.05, 0.82], p = .03, OR = 1.55. Furthermore, this model, which included diagnostic group (and its corresponding random slope term) as a predictor, was a significant improvement from the intercept only model, χ2(3) = 8.34, p = .04. These results indicate that individuals in the hoarding group were 1.55 times more likely to select future-oriented questions. However, the pseudo-R2 estimate for the overall model was very small (Nagelkerke = .005). Hoarding Symptom Severity as a Continuous Predictor of Information Seeking. The analyses described above were performed again with SI-R total score as a continuous measure of hoarding psychopathology. Similarly, I hypothesized that hoarding would be a significant predictor of future-oriented information seeking about utility. To test this hypothesis, a GLMM was specified with information seeking as the response variable, SI-R total score as a fixed effect, and subject-intercept, item-intercept, and SI-R slope as random effects. The addition of depression as a predictor was modeled but did not significantly improve model fit, χ2(1) = .45, p = .50, or alter the significance of the other predictors. Therefore, depression was not included in the final model.  Consistent with the group-based analysis, SI-R total score was a significant predictor of future question choice, b = 0.018, SE = 0.005, 95% CI [0.008, 0.028], p < .001, OR = 1.02. Again, the pseudo-R2 estimate was very small (Nagelkerke = .005).  Overall, these findings are consistent with the hypothesis that individuals in the hoarding group would be more likely to seek out future-oriented information when making discarding decisions.  	 84	Discussion of Information Seeking Task  Findings   As expected, when given the choice between basing discarding decisions on information about future or past utility, people in the hoarding group demonstrated a greater preference for future utility than did the healthy control group. This finding is consistent with the hypothesis that people with hoarding disorder excessively prioritize information about potential future uses of possessions over actual past usage when making keep or discard decisions.   The IST was primarily designed to examine the relationship between information seeking preferences about utility and hoarding. Nevertheless, every trial on the IST asked participants to follow-up their information seeking choices with a decision to keep or discard the hypothetical possession. Therefore, it was also possible to examine the broader relationship between information seeking preferences and saving behaviour. However, this relationship was not presented because the IST was not designed to optimize the internal validity of this specific analysis. After selecting a question about future or past utility and before making a discarding decision, participants were presented with a response to their selected question. These answers were primarily constructed to be engaging and entertaining. Efforts were made to balance the valence of these answers and thereby prevent one type of question from becoming more associated with negative or positive answers. However, psychometric properties of these answers were not pilot-tested. Therefore, participants may have varied in their reactions to the content of these answers and made discarding decisions accordingly. Moreover, the valence of an answer’s content was not randomly assigned – each question had a fixed answer that was either positive or negative. Therefore, the total number 	 85	of positively and negatively valenced answers each participant received was not constant regardless of how many future or past questions they chose. Therefore, the relationship between information seeking preferences and discarding choices is an important theoretical question to investigate but should be viewed with caution in the current research design.   When the relationship between information seeking and discarding choices was examined across participants, there was a large sized correlation (r = .49) between future utility and subsequent decisions to keep objects. This finding supports the study’s hypothesis that a preference for information about future utility may bias people who hoard against discarding. Moreover, this relationship suggests a general link between information seeking about future utility and keeping decisions that is universally present, but may be overrepresented in hoarding due to a disproportionate focus on future utility during discarding decisions.   Implications  Overall, findings from the IST provide support for the proposition that a bias towards assessing objects on the basis of potential future uses is part of the decision-making process that results in pathological saving. Furthermore, the current study provided preliminary evidence that there is a general relation between seeking information about future utility and decisions to keep objects. This result raises the question of the specific mechanism through which a tendency to speculate about potential future uses of objects leads to greater saving. One explanation for this link is that the potential future uses of a possession are theoretically unlimited; someone who contemplates potential future uses for a possession is likely to find one, if not many, reasons to justify keeping that possession. Therefore, a decision-making process that relies more on this means of evaluating objects is more likely to retrieve 	 86	information that favours saving. Conversely, past usage is less flexible because it is grounded in real behaviour. As such, seeking out information about past usage is more likely to yield negative information in favour of discarding when objects have objectively outlived their usefulness.  Theoretically, a style of decision-making that routinely favours fantasizing about future uses over recalling history of past usage could help to explain the development and maintenance of hoarding symptoms. While the current study did not investigate acquiring behaviour, the finding that people who hoard excessively focus on future utility during discarding decisions may also have implications for understanding decision-making about prospective possessions. Hypothetically, focusing more on how possessions could be used in the future than on current habits or needs could also yield more favourable assessments of prospective possessions and thereby increase the number of objects that seem like reasonable acquisitions. In contrast, non-hoarding individuals may be more likely to consider negative information about current need when making acquiring decisions and therefore see a more limited range of objects as reasonable acquisitions.  The distinction between need and want is often a consideration when most people are deciding whether to acquire something. Helping clients who hoard to learn to make this distinction, and thereby acquire fewer items that are wanted rather than needed, is a focus of cognitive behavioural treatment designed to reduce excessive acquisition in the context of hoarding disorder. Theoretically, this greater emphasis on want over current need, and potential future uses of objects over objective past usage, are both related to basing decisions on an idealized future self rather than the actual self. A tendency to match decisions with a pleasant vision of one’s future self (e.g., buying skis in the hopes of becoming a skier) rather 	 87	than one’s current circumstances (e.g., not having the space to store skis or the time, skills, or funds to go skiing) would help to explain why both acquiring and discarding decisions are often out of sync with current needs in hoarding.  Pending replication, the finding that prioritizing future utility was positively associated with subsequent keeping decisions could be used to guide clinical interventions. Clinicians could provide psychoeducation about the relationship between how one evaluates utility (i.e., potential future uses versus actual past usage) and saving behaviour, and encourage clients with hoarding problems to use that knowledge to their advantage by deliberately focusing more on past usage when trying to de-clutter (and reduce acquiring). Clients who initially find it difficult to base discarding decisions on past behaviour may find it more helpful to engage in cognitive rehabilitation exercises that generally train them to use past behaviour as a reliable cue for predicting future behaviour. For example, people who hoard could practice observing the amount of time it takes them to complete tasks and using that knowledge to increase their accuracy in allotting time for subsequent tasks. Once this skill is mastered, they could then move on to observe past utility and use that information to make discarding decisions with a greater degree of confidence.  Moreover, current cognitive behavioural strategies for treating hoarding involve directing clients to develop rules and decision-trees for when to discard possessions as a means of raising awareness and retraining the decision-making process. The results of this study suggest that developing rules pertaining to past utility may be an especially important practice because people who hoard may be less likely to spontaneously generate such rules on their own due to an excessive focus on future utility. Fortunately, rules focusing on past usage are already commonly used in hoarding treatment (e.g., item must have been used in 	 88	the past two years to justify keeping). This study provides empirical support for this practice and encourages that it become a standardized guideline for rule generation.  Limitations and Future Directions  One limitation of the current research design is that it was not ideally constructed to examine the relationship between information seeking strategies and saving decisions. The current research does suggest that evaluating possessions on the basis of potential future use instead of past usage is more likely to lead to saving decisions. However, due to the previously discussed limitations of the IST, this finding requires testing in an experimental design that can provide a cleaner evaluation of the relationship between decision-making strategies that favour future utility and subsequent saving decisions. A possible next step would be to directly manipulate whether participants think about past or future utility and then observe discarding outcomes. For example, participants could be presented with a series of possessions to sort into keep and discard piles, and specifically directed to evaluate those objects on the basis of how they might be of use in the future or how often they’ve been useful in the past. A finding that people who think about potential future uses are more likely to keep possessions than those who think about past utility would reinforce the current study’s findings.  Another limitation of the current research is that it could not provide insight into possible differences in the quality of hoarding and non-hoarding individuals’ estimates of future use. One fruitful line of inquiry for future research could be whether non-hoarding individuals generate more realistic estimates of future use than do hoarding individuals and how both groups arrive at those estimates. Presumably, all decision-making strategies about keeping or discarding possessions are ultimately attempting to estimate future use. If one 	 89	could be 100% certain that an item would never be used in the future, then no decision-making process would be necessary as it would be obvious that the item should be discarded. Since this level of certainty is impossible, people employ strategies for estimating future use. Differences in these strategies may produce more or less realistic predictions of future use. For example, there are several ways one could evaluate the probability of using a pair of skis in the future. More realistic assessments may take into account factors such as past usage of the specific item or others like it, self-knowledge about one’s degree of athleticism and monetary restrictions, whereas less realistic assessments may focus more on the possibility of enjoying skiing. A better understanding of how these assessments are made in hoarding could help direct interventions that shape these evaluations to yield more realistic discarding decisions. A related line of inquiry could address whether people who hoard are less willing to tolerate uncertainty when evaluating whether an item will be needed in the future. While all decisions about the future contain uncertainty, it is possible that people who do not hoard are willing to accept a greater degree of uncertainty about future need before discarding possessions. Frost and Hartl (1996) posited that people who hoard are abnormally concerned about erroneously discarding possessions they may need in the future. Currently, the presence of a lower threshold for uncertainty regarding future use of possessions has not been empirically tested.  Although the focus of this research is on cognitive factors involved in decision-making, emotional factors are undoubtedly relevant as well for these decisions. For example, hoarding symptoms are correlated with intolerance of uncertainty in undergraduates (Oglesby et al., 2013; Wheaton et al., 2015). Intolerance of uncertainty is defined as a “dispositional 	 90	characteristic that results from a set of negative beliefs about uncertainty and its implications and involves the tendency to react negatively on an emotional, cognitive, and behavioural level to uncertain situations and events” (Buhr & Dugas, 2009, p. 216). Intolerance of uncertainty may reduce or eliminate willingness to discard items in the face of even a very small degree of uncertainty about whether the items may be useful in the future. This line of thinking is consistent with Frost and Hartl’s (1996) other proposition that people who hoard keep possessions to avoid negative emotions associated with decision-making about possessions.               	 91	Chapter 5: Dysfunction Tolerance Scale Task Construction and Refinement  The Dysfunction Tolerance Scale (DTS) is a novel measure that was designed to test the hypothesis that people who hoard have a higher threshold for how dysfunctional objects must be to warrant being discarded. The following describes the construction and initial pilot analysis conducted to develop the DTS.   The DTS uses a Guttman scale to measure differences in utility thresholds for discarding. Each item presents a household object and four descriptions of that object in states of increasing disrepair. To standardize the levels of disrepair across questions, the following guidelines were used to construct the four response options for each item of the DTS: Level 1) Item cannot fulfill primary function perfectly, Level 2) Primary function is severely limited, Level 3) Item is completely unable to fulfill its primary function but could be repaired easily, Level 4) Item requires intense repair to make it usable. For example, one item from the DTS described a suitcase with the following four response options: 1) The zipper on one of the outer pockets is stuck., 2) One of the four wheels is broken. The suitcase still rolls but not smoothly., 3)The zipper on the main compartment is ripped. It won't close completely., 4)There is a large tear in the back of the suitcase. Anything inside would be at risk of falling out. For each item, participants selected the highest response option where they would still be willing to keep the hypothetical possession.  The DTS was introduced to participants with the following instructions:  Imagine that you are going through items you have placed in storage and are evaluating what is worth keeping. Each item is described in four different states of disrepair. All of these descriptions suggest the items are imperfect in some way (e.g., 	 92	damaged, missing pieces, not fully functioning) but may still have some potential to be useful. Please read all of the descriptions and select the highest option where you would still keep the item. For example, if you would keep the item as described in options 1, 2, and 3, but would discard it as described in option 4, you would select option 3. Please keep potential secondary uses in mind. For example, most people would no longer drink out of a chipped cup (i.e., primary use). However, some people would re-purpose it as a pencil holder or a plant pot (i.e., secondary uses) and therefore still consider it useful enough to keep. Others would consider it completely useless and throw it away.  Pilot Studies Pilot Study 1. Thirty items, each with four response options, were initially generated for the DTS, for a total of 120 descriptive statements. The objective of the initial pilot work was to ensure that response options were presented in the correct order and ranged appropriately in their perceived defectiveness (i.e., that the Guttman scale was functioning properly). Items were evaluated using an online reading level analyzer (https://readability-score.com/); items were edited to avoid exceeding a grade eight reading level. Fourteen personal contacts of members of the laboratory rated how dysfunctional each item was from 1 (Item is as functional as a new item) to 10 (I can't see myself using this in any way). None of these pilot participants were aware of the purposes of the research. Descriptive statements (i.e., response options) were presented in random order within each item.  One participant’s responses were removed due to failure to complete the survey. Cronbach’s alpha was .96 for the total scale, suggesting excellent internal consistency. 	 93	Response options were placed in rank order according to average ratings from the pilot participants. Eight items were removed due to multiple response options being rated too similarly. Twenty-two items required minor to moderate changes to increase the difference between response options or reduce variability. To be considered sufficiently different, ratings of response options were required to be at least one point apart between levels. In total, forty-eight response options were edited to clarify content or adjust perceived defectiveness.  Pilot Study 2. A second pilot study was conducted on the remaining 22 items to further develop the Guttman scale. Fourteen additional experiment-naïve personal contacts of laboratory members were recruited and asked to rank response options from 1 (most functional/useful) to 4 (least functional/useful). Descriptive statements were randomly arranged to prevent order of presentation from influencing rankings. Items were then evaluated in terms of how well average rankings fit the expected order, the distance between average rankings, and the variability of assigned ranks. Five items were deleted because two or more response options were ranked too similarly (i.e., less than 0.5 units apart), rankings were not in the correct order, or rankings were too variable across participants (e.g., three ranks were equally represented within one or more descriptors). Five items performed very well according to these standards and were not altered. Twelve items were edited to increase the distance between their rankings. For example, the first and second descriptors for the item Running Shoes were 0.42 ranks apart due to disagreement about which statement portrayed a greater level of dysfunction. Therefore, the first statement (“The laces are frayed and missing their caps”) was edited to decrease the perceived level of dysfunction (“The laces are 	 94	frayed”). Four additional items were constructed by modeling the properties of the successful items and their response options.  Pilot Study 3. A third pilot study was conducted to further evaluate the Guttman scale properties of the 21 items remaining or generated from the second pilot. In Guttman-style scales, response options for each item are arranged in an order such that when an individual selects one response option as true, they would consider the lower order response options to be true as well. Within the context of the DTS, this would mean that if a participant indicates they would keep an item based on the third response option, they would also keep the item as described in the earlier two response options. To evaluate this property within the DTS, 14 experiment-naïve personal contacts of laboratory members read each response option (presented in random order) for the remaining items and indicated whether they would keep or discard/recycle/donate each item according to that descriptor.  The pattern of keeping and discarding responses was then examined for each item for evidence of failure of the Guttman scale (i.e., a participant indicated they would discard a lower ordered item but keep higher ordered items, or vice versa) as well as basement (discarding all items) and ceiling effects (keeping all items). Twelve items had reasonable variability as well as two or fewer ordering errors across all statements and all participants. Two additional items had two or fewer ordering errors, but more than half of participants indicated that they would discard the least dysfunctional descriptor. These two items were modified with a new (less dysfunctional) descriptor for the first response option. For example, the first descriptor statement for one of these items was, “Mug: The handle has two or three small chips, revealing a lighter colour underneath”.  After modification, this descriptor statement was changed to be less dysfunctional by describing the colour as faded 	 95	instead of chipped (i.e.,  “Mug: The outer colour is faded”). The final version of the DTS contained 14 items.  Results Overview of Data Analyses  Continuous analyses for DTS performance were examined using multivariate regression to predict DTS total scores from hoarding psychopathology while controlling for depression. Continuous analyses utilized the entire sample of participants who completed the DTS (N = 171). Categorical analyses were conducted using ANCOVA and only included participants who met inclusion criteria for the hoarding group (n = 53) or the healthy control group (n = 74).  Data Cleaning and Assumption Testing I conducted several tests to verify whether the parametric assumptions underlying independent samples t-tests and regression were met for this dataset before running the main analyses. Normality was assessed through visual examination of data and residual distributions, Shapiro-Wilk tests, and Normal QQ plots. Potential outliers were detected using boxplots, Normal QQ plots, and Cook’s values. Outliers more than 3.5 SD from the mean were winsorized to the next closest data point. Moreover, heterogeneity of variances was evaluated through Levene’s test of equal variances. To reduce multicollinearity in regression analyses, continuous variables were mean centered. Tests of multicollinearity were also examined and found to be in acceptable ranges (i.e., VIF).   Categorical Analyses of Hoarding and Dysfunction Tolerance   An ANCOVA was conducted to compare the hoarding and healthy control group on total DTS score while controlling for depression (i.e., CESD-R score). As hypothesized, the 	 96	results showed a significant effect of group, F(1,124) = 15.30, p < .001, η2 = .11. Similarly to the analyses for the VBT and IST, depression was not a significant predictor of DTS scores, F(1,124) = 0.58, p = .45, η2 = .005. LSD post-hoc tests indicated that the hoarding group (M = 31.51, SE = 1.20) scored significantly higher on the DTS than did the healthy control group (M = 25.19, SE = 1.00). The difference between groups was medium-large in size (d = 0.73) according to Cohen’s standards of effect size. These findings are consistent with the hypothesis that people who hoard have a higher threshold for when dysfunctional objects should be discarded.  Dimensional Analyses of Hoarding and Dysfunction Tolerance Multivariate regression was used to investigate the relationship between hoarding as a dimensional construct and performance on the DTS. To test the hypothesis that hoarding symptoms predicted DTS scores above and beyond depression, depression was also included as a predictor in the analysis. A single multivariate regression was conducted with DTS total score as a response variable, and SI-R total score and CESD-R score as predictors.  The overall model accounted for 13% of the variance in DTS total score. Consistent with the group-based analyses, the SI-R total score was a significant predictor of DTS total score, β = 0.32, t = 3.67, p = < .001, but depression was not, β = 0.06, t = 0.75, p = .45. These findings reinforce the conclusion of the group-based analyses that hoarding is associated with a significantly higher threshold for when to discard broken and dysfunctional possessions. Discussion of Dysfunction Tolerance Task   Findings   Hoarding participants chose a higher level of dysfunction as the tipping point at which an object should be discarded. Control participants’ judgments of whether items were 	 97	worth keeping occurred at earlier stages of dysfunction, suggesting that people who hoard saw value in retaining broken items that controls did not see. This was consistent with the hypothesis that people who hoard require objects to be more broken or dysfunctional before they can justify discarding them. The difference between groups on the DTS controlling for depression was medium to large (d = 0.73), suggesting that this difference is likely to have practical significance and to be observable in real world discarding practices. Furthermore, this result was observed over and above the potentially confounding effects of depression, suggesting that higher dysfunction tolerance among hoarding participants was not the result of apathy or lethargy caused by greater symptoms of depression in the hoarding group.  Implications  The current research is the first to provide empirical support for Frost and Hartl’s (1996) proposition that people with hoarding problems have a generally higher threshold for deciding to discard objects that may be useful in the future. Frost and Hartl proposed this higher threshold as a means of explaining the “decision-making deficit” in which individuals who hoard save a much larger volume of possessions than do non-hoarding individuals despite keeping the same kinds of possessions and expressing the same reasons for saving. The current research is the first to empirically examine this proposition. The medium-large effect size found suggests that Frost and Hartl’s threshold model is a meaningful way of understanding decision-making about possessions in hoarding and warrants further investigation.  The finding that people who hoard require objects to be more dysfunctional or broken down before they can justify discarding could explain one potential pathway by which clutter accumulates. Within consumerist cultures, many objects are constructed with planned 	 98	obsolescence in mind (Slade, 2007) and therefore are intended to require replacement or otherwise become obsolete as they wear out or better products are developed. Most people cope with this planned obsolescence by discarding and replacing objects regularly. People with hoarding disorder may continue to see value in declining objects past the point non-hoarding individuals would. Discarding such objects may involve negative emotions, such as guilt or sadness, that people who hoard would rather avoid. As a result, individuals with hoarding problems may be more inclined to replace objects without discarding their predecessors, thereby incrementally growing clutter. As with the other tasks, the finding that people who hoard set unreasonably high standards for how useless objects should be before discarding has implications for treatment. Specifically, an additional goal of treatment for hoarding could be to assist clients in developing more normative standards for when to discard items that are no longer functioning adequately. For example, clinicians could encourage hoarding clients to poll non-hoarding friends and family about their standards for discarding common household items such as clothing, electronics, paper items, and old food containers that are no longer in prime condition. Once hoarding clients become more aware of the kinds of standards for item functionality that enable others to keep their homes relatively free of clutter, they can integrate these standards into discarding rules and decision trees.  Limitations and Future Directions   A limitation of the current research design is that it could not assess the rationale behind the hoarding group’s reluctance to part with dysfunctional possessions. Future research should endeavour to understand the thinking behind these decisions in order to provide an empirical foundation for interventions that focus on restructuring beliefs, 	 99	decision-making strategies, and behaviours that lead to excessive saving. Frost and Hartl (1996) suggested several potential mechanisms for a higher discarding threshold including erroneously inflated judgements about the probability that an object will be needed in the future, greater perceived risk in discarding potentially needed objects, decreased confidence in one’s ability to re-acquire objects later on, and the belief that possessions may become more valuable over time regardless of current circumstances. To add to this list, I posit that people who hoard may have more unrealistically optimistic expectations about their ability to repurpose or fix broken objects, therefore requiring a much lower level of functionality before deeming objects as useful and therefore worth keeping. Further research is needed to evaluate these potential mechanisms.    	 100	Chapter 6: Discussion In this chapter I first present a review of the rationale and hypotheses of the current research. I will then discuss how the combined findings from this research can be used to inform our understanding of maladaptive decision-making in hoarding. I will conclude with a review of the contributions of this work and several directions for future research in this area.  Review of Study Rationale and Hypotheses  Hoarding disorder is characterized by the excessive accumulation of possessions that amass in disorganized piles around the home. These possessions accumulate to such an extent that they impair the use of the home for intended functions (e.g., eating in one’s kitchen, sleeping in one’s bed, moving around the home) and cause significant distress or impairment in other important life domains. In severe cases, the clutter in hoarded homes becomes a threat to the safety of the individual and the surrounding community by acting as a fire hazard, blocking egress during emergencies, and creating unsanitary living conditions (i.e., pests, mold growth). Hoarding has also been shown to cause emotional hardship through its association with stigma, feelings of shame, social dysfunction, familial conflict, occupational impairment, and housing insecurity (Frost, Steketee, & Williams, 2000; Tolin, Frost, Steketee, & Fitch, 2008; Tolin, Frost, Steketee, Gray, & Fitch, 2008).  Despite the intense hardships conferred by clutter, people who hoard are extremely resistant to letting go of objects that most people would consider to be useless or trivial (e.g., out-of-date newspapers/magazines/books, broken appliances, duplicates of toiletries). One potential explanation for the seemingly irrational pattern of decision-making observed in hoarding (i.e., repeatedly choosing to keep conventionally useless possessions despite little to no household space) is that people who hoard exhibit a general deficit in decision-making 	 101	ability. However, studies employing neuropsychological measures to assess decision-making deficits (i.e., gambling tasks) suggest that people who hoard do not suffer from a general inability to detect patterns and make advantageous decisions (Woody, Kellman-McFarlane, & Welsted, 2014). As a next step to understanding decision-making in hoarding, the current research proposed that decision-making about possessions in hoarding is guided by a different set of considerations that disproportionately favour object retention.  Previous research and theory suggests that people who hoard keep many of the same kinds of objects as non-hoarding individuals and share many of the same underlying motivations for saving (e.g., utility, sentimentality, intrinsic value), but ultimately define a much greater number of objects as worth keeping (Frost & Gross, 1993; Frost & Hartl, 1996; Frost et al., 1998). This pattern of findings suggests important differences in the decision-making process through which hoarding and non-hoarding individuals make choices about whether to keep or discard possessions. The current research sought to provide insight into several differences in this process that could routinely bias people who hoard against discarding objects that other people would consider worthless due to low past usage or low functionality, or less valuable than the household space they occupy. Firstly, the current research hypothesized that excessive saving partially results from differences in how people who hoard use information about utility to guide their discarding decisions. Among the many rationales and dysfunctional beliefs thought to contribute to difficulty discarding, practical utility is regarded as an especially motivating reason to save items (Frost & Gross, 1993; Frost & Hartl, 1996; Nordsletten, de la Cruz, Billoti, & Mataix-Cols, 2013; Warren & Ostrom, 1988). However, extreme object accumulation renders many possessions inaccessible in hoarded homes and therefore functionally useless. If utility 	 102	remains an important motivator for keeping in hoarding despite these issues, then it is probable that people who hoard evaluate the utility of their possessions differently than non-hoarding individuals. The current research hypothesized that people who hoard focus excessively on the potential future uses of possessions and neglect actual past usage when making decisions about possessions. Future utility is conceptually a very flexible rationale for saving possessions because it is uncertain and limited only by judgment and imagination. Conversely, assessments of utility based on past usage (e.g., “How often have I used this?”) are grounded in actual behaviour and therefore more likely to result in negative appraisals of an object’s utility, especially in hoarded homes where access to possessions can be limited. Therefore, a bias to focus on the ways objects could be used or needed in the future, paired with a neglect of actual past usage, may help to explain why people who hoard save so many conventionally useless objects but continue to cite utility as an important rationale for saving.  In support of this hypothesis, previous research and clinical observation have found that people who hoard often report saving objects due to a perceived future need for them (Frost & Gross, 1993; Frost & Hartl, 1996; Nordsletten, de la Cruz, Billoti, & Mataix-Cols, 2013). Currently, only one qualitative study has provided insight into the focus on future utility in hoarding. In this study, Coulter and Ligas (2003) interviewed “purgers” and “packrats” about their saving and discarding rationales. One of the key differences that emerged between these groups was that packrats reported potential future use as a key rationale for saving, whereas purgers typically did not consider the future use of old and used items. This finding is in line with my hypothesis that a focus on future utility is related to saving behaviour. However, Coulter and Ligas’s findings are limited because they did not 	 103	use a clinical sample or an experimental paradigm. The current research extended their findings by using an experimental paradigm to test whether individuals with hoarding disorder are more likely to base their discarding decisions on information about future utility.    Secondly, the current research hypothesized that excessive saving partially results from differences in how people who hoard use information about household space to guide discarding decisions.  The fact that most people in consumerist societies such as Canada and the United States do not live in hoarded homes indicates that most individuals maintain a balance between available household space and the volume of their possessions. Presumably, part of what allows people in consumerist societies, who are to some extent perpetually acquiring throughout their lifetime, to prevent their homes from becoming excessively cluttered is that they discard items when household space decreases. Conversely, hoarding is practically defined by an imbalance between household space and the volume of one’s possessions; a diagnosis of hoarding requires that objects have been allowed to accumulate to the point that living spaces become impaired. The current research hypothesized that part of what enables hoarded homes to become pathologically cluttered is that people with hoarding disorder are less responsive to the increasing space constraints that ordinarily trigger discarding behaviour. Currently, only one study, by Preston, Muroff, and Wengrovitz (2009), has experimentally investigated individual differences in the impact of space constraints on decisions to keep or discard items. Their results suggested that space has less of an impact on the saving decisions of people who tend to acquire more than others. However, their sample was composed of undergraduate students, and therefore requires replication and extension to a clinical hoarding sample. The current research will build on their work by examining 	 104	differences in how information about household space is integrated into the decision-making process of people who meet criteria for hoarding disorder.  In summary, the current research proposed that the excessive saving characteristic of hoarding partially results from a biased decision-making process that habitually prioritizes factors that favour keeping possessions (i.e., potential future utility) and neglects factors that would typically encourage discarding (i.e., low past usage and low household space). The model implies that non-hoarding individuals discard items when utility or household space become low, an idea that has not yet been empirically tested. Furthermore, the current research hypothesized that people who hoard have a higher threshold for when an object becomes ‘useless’ or too dysfunctional to justify keeping. This would mean that people who hoard would require objects to be more worn out, more broken, or in worse condition than people who do not hoard before they see objects as worthy of discarding. This idea of a higher threshold for discarding was proposed by Frost and Hartl (1996) as a means of explaining excessive saving in hoarding but has not yet received empirical attention. Specifically, a higher threshold for dysfunction would lead to judging a larger number of objects as potentially useful and therefore worth keeping. Together, these decision-making factors would bias people who hoard towards keeping possessions that other people would consider useless (due to low past usage or low functionality) or less valuable than the household space required to store them (in situations of restricted space). These hypotheses were tested using three novel decision-making tasks developed for the current research: the Vignette-Based Task (VBT), the Information Seeking Task (IST), and the Dysfunction Tolerance Scale (DTS). Relative to the control group, people in the hoarding group were expected to prefer information about future rather than past utility 	 105	during the IST, to be less responsive to negative information about past utility and household space in deciding whether to keep or discard on the VBT, and to display significantly greater tolerance for dysfunction on the DTS.  Discussion and Illustration of Integrated Model   The pattern of findings in this research supports the biased model of decision-making proposed in this dissertation. In the proposed model, people with hoarding disorder fail to fully integrate factors into their decision making process that would decrease the perceived value of possessions (e.g., low past usage, poor condition, reduced functionality, restricted living space) and instead focus on information that is more likely to lead to keeping possessions (i.e., potential future uses). Due to this disproportionate focus on motivations for saving and neglect of motivations for discarding, individuals who hoard are more likely to keep objects that non-hoarding individuals would feel compelled to discard on the basis of low utility or household space.    The increased emphasis on potential future utility observed in hoarding participants may be especially problematic given that people in the hoarding group also displayed a higher threshold for discarding – seeing potential utility in objects that were in poor condition. This threshold shift suggests that people who hoard see value in a wider range of objects; therefore, when speculating about future uses for possessions, people who hoard are more likely to make positive assessments of the potential utility of damaged objects. Having such a clear vision of the potential utility of damaged objects reduces the justification to discard them.   To illustrate this process, I will explain how these proposed differences would theoretically influence decision-making about a t-shirt for a healthy individual and an 	 106	individual with hoarding disorder. In this scenario, the t-shirt has been used once in the past 10 years and is in poor condition (i.e., old and ill fitting). There is little to no space to store it in an appropriate location such as a closet or wardrobe. Disregarding for the moment all other potential valuing factors (e.g., aesthetic preference, monetary value), the value of the t-shirt would strongly decline for the healthy individual in view of how rarely they had worn the shirt, its poor condition, and the lack of space to store it. Although the healthy individual may also consider potential future uses for the shirt (e.g., using it as a cleaning rag, wearing it for messy chores), these substandard uses would have a relatively smaller influence on the value of the shirt than it would in hoarding due to a lower tolerance for dysfunction. In the end, the healthy person’s decision-making process would favour discarding the t-shirt due to the strong devaluing influence of low past utility and space.  In contrast, this model would predict that an individual with hoarding disorder would disproportionately focus on the potential future utility of the t-shirt, viewing those potential future uses more favourably due to a higher tolerance for dysfunction. Negative information about past utility and space may also be considered but would not be sufficiently powerful to devalue the item. The item would therefore be deemed too valuable to justify discarding. In the absence of sufficient motivation to discard the shirt, and the potential gains from keeping the shirt, there is little reason to discard. The mechanism behind this pattern of attention to information that would support keeping objects and relative neglect of information that promotes discarding could be conceptualized in two ways. Firstly, this model could reflect a cognitive bias that renders certain individuals more vulnerable to hoarding behaviour through a greater proclivity to fantasize about future uses of objects (even if they are not in good working order), an 	 107	inherent tendency to use less realistic strategies for estimating future need (e.g., history of use), and a lesser aversion to cluttered living environments. Alternatively, this pattern may represented a form of emotion driven cognition in which people who hoard selectively attend to information that supports the outcome that would cause them the least distress or the most pleasure, not parting with their possessions. In this second case, the pattern of results observed in this study would be related to emotion-based factors that interfere with discarding in hoarding such as object attachment, intolerance of uncertainty, perfectionism, and inflated responsibility for possessions (Frost & Hartl, 1996; Wheaton et al., 2015). The current research design did not enable inference into which of these mechanisms is more likely, but provides a foundation for future research that seeks to understand the origins of biased decision-making in hoarding disorder. The presence of significant findings for each component of the predicted model provides substantial support for the overall theory of biased decision making proposed in this dissertation. However, it is worth noting that the effect sizes for the main findings on the VBT and the IST were small.  Specifically, participants in the hoarding group were only 1.55 times more likely than the healthy control group to select future-oriented questions about utility on the IST. Similarly, the statistical models used to analyze the relationship between hoarding and different conditions of utility and space on the VBT accounted for a relatively small amount of variance in discarding behaviour (R2 = .13 - .15). These small effect sizes may reflect the artificial nature of the tasks and the associated costs to ecological validity.  Asking participants to make decisions about hypothetical possessions instead of real ones allowed me to standardize items and conditions across participants and thereby optimize internal validity. However, not using owned possessions is likely to have reduced 	 108	participants’ investment and interest in the consequences of their decisions. On the IST, this may have reduced hoarding participants’ inclination to fantasize about future uses, and therefore, their interest in information about future utility. Specifically, people who hoard may have been less interested in fantasizing about potential future uses for objects that they could not ultimately carry out due to the artificial nature of the task (e.g., it may be more rewarding to fantasize about using an old tent to go camping if there is a potential to act on that fantasy).  Moreover, the small effect sizes of VBT models may have resulted in part from the artificially simple decision-making environment created within the vignettes. The VBT was designed to examine one decision-making factor in isolation (i.e., space or utility) as a means of optimizing my ability to draw accurate inferences about how each specific factor influenced discarding choices. However, real decision-making about possessions is likely to employ multiple factors concurrently (e.g., utility, space, sentimentality, aesthetic appeal, monetary value, condition). The influence of each factor in isolation may be relatively small, but they may add or multiply to exert a large impact on discarding behaviour. For example, one might expect a healthy person to be much more likely to discard an inexpensive possession that they have rarely used, have no space to store, and for which they have only a weak sentimental association (e.g., an old Christmas card from their bank) than a possession that is lacking on only one of these factors. In contrast, someone who hoards would be less responsive to this accumulation of negative information, and consequently would move little in their willingness to discard the possession. Therefore, one explanation for the small effects of utility and space found in this study is that these two factors are small but relevant pieces 	 109	of a decision-making process that cumulatively creates a large bias in judgment about possessions.  Despite the artificial decision-making environments of the VBT and IST, several indicators suggest that the tasks captured realistic hoarding behaviour and therefore reflect some degree of external validity. On the VBT, the hoarding group’s pattern of discarding responses mapped on to what would be expected based on a hoarding diagnosis, and for what would be expected given the manipulation of space and utility. Specifically, hoarding participants were significantly less likely to discard than controls in every condition of the VBT while still demonstrating the expected responsiveness to changes in conditions (i.e., being more likely to discard in low conditions of space and utility than in high conditions). With regard to the IST, the hoarding group was more than twice as likely to keep hypothetical possessions than was the control group. Secondly, participants in the hoarding group reported greater subjective distress while engaging in decision-making during the IST than did the healthy participants. This finding is also consistent with distress and indecision surrounding real discarding in hoarding disorder. Overall, while the IST and VBT may not have been as engaging as decision-making about real possessions, they were apparently engaging enough to prompt realistic group-differences in decision-making. Conclusions  Although our understanding of compulsive hoarding has evolved substantially over the past decade, little research has addressed the mechanisms responsible for the development and maintenance of hoarding symptoms. The literature on decision-making in hoarding is particularly in need of a new approach to studying these problems, as previous lines of research on indecisiveness and gambling tasks have not convincingly characterized 	 110	hoarding-specific difficulties in this area. The current research took a novel experimental approach to studying decision-making by focusing on differences in the types of information people prioritize in their discarding-decisions and their relative impact on saving behaviour. The results of this work suggest that this approach is a promising avenue for better understanding decision-making about possessions in hoarding. The aforementioned findings provide preliminary support for the biased model of decision-making explored in this dissertation, but as with any new theory, require replication for validation. Furthermore, this model also requires expansion to include other important decision-making factors that may play a role in hoarding. The current research established that people who hoard integrate information about utility and space differently into their decision-making process about possessions. However, utility and space are unlikely to be the only valuing dimensions that are weighted differently in the decision-making process of people who hoard. As noted earlier, previous research and theory has identified several other values (e.g., sentimentality, aesthetics, intrinsic value, wastefulness, memory) that may be of particular interest in better understanding why people who hoard experience such great difficulty in parting with possessions of little or no worth to others. However, researchers currently have few experimental paradigms for exploring how these differences may exert influence on discarding behaviour. The current work has helped to fill this gap and advance research in this area with three novel decision-making tasks that can be adapted to examine how the prioritization of different types of information about possessions impacts discarding decisions. For example, the VBT could be modified to include low and high sentimentality conditions as a means of exploring whether people who hoard are less responsive to differences in sentimental value (e.g., a card from a beloved relative versus a stranger). 	 111	Similarly, the IST could be of adapted to measure preferences for information about other decisions-making factors, such as sentimentality, by changing the content of the questions participants choose between.  In addition to contributing to the research literature on hoarding, the tasks developed as a part of this dissertation have applications for the broader field of consumer behaviour. Although the literature on ordinary saving suggests many potential reasons for why people save possessions (e.g., Cooper, 2004; Domina & Koch, 1998; Jacoby et al., 1977; Laitala, 2014; Paden & Shell, 2005; Roster, 2001; Savas, 2002), the differential importance of such factors in discarding decisions is not understood. In the current research, the IST was used to evaluate the relative importance of future and past utility through information seeking preferences. In research on consumer behaviour, it could easily be adapted to evaluate the relative importance of competing factors in typical discarding practices. For example, participants could be asked to choose between obtaining information about the sentimental or monetary value of possessions to inform their discarding decisions. The VBT could be similarly adapted to research that focuses on individual differences in normative behaviour toward possessions. Instead of examining differences between hoarding and healthy controls, the VBT could be used to explore differences in other groupings where responsiveness to utility and space may differ such as life stage, culture, and socio-economic status.    Overall, the current research has helped to establish a foundation for a new approach to understanding decision-making and excessive saving in hoarding disorder. The findings from this research provided convergent evidence in support of the proposed theory that people who hoard engage in a biased decision-making process that is more likely to result in saving. According to the findings of the current research, people with hoarding disorder are 	 112	not as sensitive to information that would decrease the perceived value of possessions (e.g., low past usage, poor condition, reduced functionality, restricted living space) and were more likely entertain information that was more likely to lead to keeping possessions (i.e., potential future uses).  These findings have contributed to the literature by providing a better understanding of the process by which people who hoard justify saving possessions that others would find useless or less valuable than the space they occupy. Overall, the current research represents a promising step towards a line of research that can elucidate the mechanisms behind hoarding behaviour and thereby improve interventions for people who struggle with this debilitating disorder.    	 113	References Abramowitz, J., Franklin, M., Schwartz, S., & Furr, J. (2003). Symptom presentation and outcome of cognitive–behavior therapy for obsessive–compulsive disorder. Journal of Consulting and Clinical Psychology, 71(6), 1049–1057. doi:10.1037/0022-006X.71.6.1049 Albinsson, P. A., & Perera, B. Y. (2009). From trash to treasure and beyond: the meaning of voluntary disposition. Journal of Consumer Behaviour, 8(6), 340-353. doi:10.1002/cb.301 American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing. Baayen, R. H., Davidson, D. J., & Bates, D. M. (2008). Mixed-effects modeling with crossed random effects for subjects and items. Journal of memory and language, 59(4), 390-412. https://doi.org/10.1016/j.jml.2007.12.005 Bar-Haim, Y. (2010). Research review: Attention bias modification (ABM): A novel treatment for anxiety disorders. Journal of Child Psychology and Psychiatry, 51(8), 859-870. doi:10.1111/j.1469-7610.2010.02251.x Barr, D. J. (2013). Random effects structure for testing interactions in linear mixed-effects models. Frontiers in Psychology, 4, 328. http://doi.org/10.3389/fpsyg.2013.00328 Barton (2015). MuMIn: Multi-Model Inference. R package (1.15.6) [Computer Software]. Retrieved on July 1, 2017 from https://CRAN.R-project.org/package=MuMIn Bates, Maechler, Bolker, & Walker. (2015). Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software, 67(1), 1-48. doi:10.18637/jss.v067.i01 	 114	Bechara, A. (2001). Neurobiology of decision-making: Risk and reward. Seminars in Clinical Neuropsychiatry, 6(3), 205. doi:10.1053/scnp.2001.22927 Bechara, A., Damasio, H., Damasio, A.R., & Lee, G. P. (1999). Different contributions of the human amygdala and ventromedial prefrontal cortex to decision-making. The Journal of Neuroscience, 19(13), 5473–5481. Retreived on August 30, 2017 from http://www.jneurosci.org/content/19/13/5473.short Belk, R. W. (1988), Possessions and the extended self. Journal of Consumer Research, 15(2), 139-68. https://doi.org/10.1086/209154 Beswick, G., Rothblum, E. D., & Mann, L. (1988). Psychological antecedents of student procrastination. Australian Psychologist, 23(2), 207–217. doi:10.1080/00050068808255605 Black, D. W., Monahan, P., Gable, J., Blum, N., Clancy, G., & Baker, P. (1998). Hoarding and treatment response in 38 nondepressed subjects with obsessive-compulsive disorder. Journal of Clinical Psychiatry, 59(8), 420–425. doi: 10.4088/JCP.v59n0804 Blom, R.M., Samuels, J.F., Grados, M.A., Chen, Y., Bienvenu, J., Riddle, M.A., Liang, K.Y., Brandt, J., & Nestadt, G. (2011). Cognitive functioning in compulsive hoarding. Journal of Anxiety Disorders, 25(8), 1139-1144. doi:10.1016/j.janxdis.2011.08.005 Brosch, T., Pourtois, G., & Sander, D. (2010): The perception and categorization of emotional stimuli: A review. Cognition and Emotion, 24(3), 377-400. doi:10.1080/02699930902975754 Brown, T.A., Chorpita, B.F., Korotitsch, W., & Barlow, D.H. (1997). Psychometric properties of the depression anxiety stress scales (DASS) in clinical samples. Behaviour Research and Therapy, 35(1), 79-89. doi:10.1348/014466506X158996 	 115	Buhr, K., & Dugas, M. J. (2009). The role of fear of anxiety and intolerance of uncertainty in worry: An experimental manipulation. Behaviour Research and Therapy, 47(3), 215-223. doi: https://doi.org/10.1016/j.brat.2008.12.004 Bulli, F., Melli, G., Carraresi, C., Stopani, E., Pertusa, A., & Frost, R. (2014). Hoarding behaviour in an Italian non-clinical sample. Behavioural and cognitive psychotherapy, 42(3), 297-311. doi: https://doi.org/10.1017/S1352465812001105 Carleton, R. N., Norton, M. P. J., & Asmundson, G. J. (2007). Fearing the unknown: A short version of the Intolerance of Uncertainty Scale. Journal of Anxiety Disorders, 21(1), 105-117. doi: https://doi.org/10.1016/j.janxdis.2006.03.014 Cavallaro, R., Cavedini, P., Mistretta, P., Bassi, T., Angelone, S., Ubbiali, A., et al. (2003). Basal-corticofrontal circuits in schizophrenia and obsessive-compulsive disorder: A controlled, double dissociation study. Biological Psychiatry, 54(4), 437–443. doi: 10.1016/S0006-3223(02)01814-0 Cherrier, H, & Ponnor, T. (2010). A study of hoarding behavior and attachment to material possessions. Journal of Qualitative Market Research: An International Journal, 13(1), 8-23. doi:10.1108/13522751011013945 Clara, I., Cox, B., & Enns, M. (2001). Confirmatory factor analysis of the Depression Anxiety Stress Scales in depressed and anxious patients. Journal of Psychopathology and Behavioral Assessment, 23(1), 61-67. doi:10.1136/oem.60. suppl_1.i77 Clark, D. A. (2004). Cognitive-behavioral therapy for OCD. New York, NY: Guilford Press. Cooper, T. (2004). Inadequate life? Evidence of consumer attitudes to product obsolescence. Journal of Consumer Policy, 27(4), 421-449. doi: 10.1007/s10603-004-2284-6 	 116	Coulter, R. A., & Ligas, M. (2003). To retain or to relinquish: Exploring the disposition practices of packrats and purgers. Advances in Consumer Research, 30, 38-43. Retrieved on March 10, 2015 from http://www.acrwebsite.org/volumes/8732/volumes/v30/NA-30 Crawford, J., & Henry, J. (2003). The Depression Anxiety Stress Scales (DASS): Normative data and latent structure in a large non-clinical sample. British Journal of Clinical Psychology, 42(2), 111-131. doi:10.1348/014466503321903544 da Rocha, F. F., Malloy-Diniz, L., Lage, N. V., Romano-Silva, M.A., de Marco, L. A., & Correa, H. (2008). Decision-making impairment is related to serotonin transporter promoter polymorphism in a sample of patients with obsessive–compulsive disorder. Behavioural Brain Research, 195(1), 159–163. doi: 10.1016/j.bbr.2008.05.015 de la Cruz, L.F., Landau, D., Iervolino, A.C., Santo, S., Pertusa., Singh, S., & Mataix-Cols, D. (2013). Experiential avoidance and emotion regulation difficulties in hoarding disorder. Journal of Anxiety Disorders, 27(2), 204-209. doi:10.1016/j.janxdis.2013.01.004 Dittmar, H. (1992). The social psychology of material possessions: To have is to be. Hemel Hempstead: Harvester Wheatsheaf. Domina, T., & Koch, K. (1998). Environmental profiles of female apparel shoppers in the Midwest, USA. International Journal of Consumer Studies, 22(3), 147-161. doi: 10.1111/j.1470-6431.1998.tb00726.x Dozier, M. E., & Ayers, C. R. (2014). The predictive value of different reasons for saving and acquiring on hoarding disorder symptoms. Journal of Obsessive-Compulsive and Related Disorders, 3(3), 220-227. doi: https://doi.org/10.1016/j.jocrd.2014.05.002 	 117	Eaton, W. W., Smith, C., Ybarra, M., Muntaner, C., Tien, A. (2004). Center for Epidemiologic Studies Depression Scale: Review and revision (CESD and CESD-R). In ME Maruish (Ed.). The Use of Psychological Testing for Treatment Planning and Outcomes Assessment (3rd Ed.), Volume 3: Instruments for Adults, pp. 363-377. Mahwah, NJ: Lawrence Erlbaum. Effert, B. R., & Ferrari, J. R. (1989). Decisional procrastination: Examining personality correlates. Journal of Social Behavior & Personality, 4, 151–161. Retrieved from https://www.researchgate.net/publication/232562755_Decisional_procrastination_Examining_personality_correlates Ellis, L. (1985). On the rudiments of possessions and property. Social Science Information, 24(1), 113-143. doi:10.1177/053901885024001006 Field, A. P. (2009). Discovering statistics using SPSS: (and sex and drugs and rock 'n' roll). Los Angeles: SAGE Publications. Fitch, K. E. (2011). Information Processing Deficits in Nonclinical Compulsive Hoarding (Master’s Thesis, Florida State University). Retrieved from http://diginole.lib.fsu.edu/etd/4448 Foa, E. B., Kozak, M. J., Goodman, W. K., Hollander, E., Jenike, M. A., & Rasmussen, S. A. (1995). DSM-IV field trial: Obsessive-compulsive disorder. American Journal of Psychiatry, 154(7), 90–96. doi: 10.1176/ajp.152.1.90 Frankenburg, F. R. (1984). Hoarding in anorexia nervosa. British Journal of Medical Psychology, 57(1), 57-60. doi: 10.1111/j.2044-8341.1984.tb01581.x Freeston, M. H., Rhéaume, J., Letarte, H., Dugas, M. J., & Ladouceur, R. (1994). Why do people worry?. Personality and Individual Differences, 17(6), 791-802. 	 118	Frost, R., & Gross, R. (1993). The hoarding of possessions. Behaviour Research and Therapy, 31(4), 367–382. doi:10.1016/000579679390094B Frost, R., & Hartl, T. (1996). A cognitive–behavioral model of compulsive hoarding. Behaviour Research and Therapy, 34(4), 341–350. doi:10.1016/0005-7967(95)000712 Frost, R., Hartl, T., Christian, R., & Williams, N. (1995). The value of possessions in compulsive hoarding: Patterns of use and attachment. Behaviour Research and Therapy, 33(8), 897 902. doi:10.1016/0005-7967(95)00043. Frost, R., & Shows, D. L. (1993). The nature and measurement of compulsive indecisive- ness. Behaviour Research and Therapy, 31(7), 683–692. doi:10.1016/0005- 7967(93)90121-A. Frost, R., & Steketee, G. (Eds.). (2014). The Oxford handbook of hoarding and acquiring. Oxford University Press. Frost, R., Kim, H. J., Morris, C., Bloss, C., Murray-Close, M., & Steketee, G. (1998). Hoarding, compulsive buying and reasons for saving. Behaviour Research and Therapy, 36(7), 657-664. doi: https://doi.org/10.1016/S0005-7967(98)00056-4 Frost, R., Steketee, G., & Grisham, J. (2004). Measurement of compulsive hoarding: Saving Inventory-Revised. Behaviour Research and Therapy, 42(10), 1163–1182. doi:10.1016/j.brat.2003.07.006 Frost, R.O., Steketee, G., & Tolin, D.F. (2011). Comorbidity in hoarding disorder. Depression and Anxiety, 28(10), 876-84. doi:10.1002/da.20861 Frost, R.O., Steketee, G., Tolin, D.F., & Renaud, S. (2008). Development and validation of the Clutter Image Rating. Journal of Psychopathology and Behavioral Assessment, 	 119	30(3), 180-192. doi:10.1007/s10862-007-9068-7 Frost, R., Steketee, G., Tolin, D. F., Sinopoli, N., & Ruby, D. (2015). Motives for acquiring and saving in hoarding disorder, OCD, and community controls. Journal of Obsessive-Compulsive and Related Disorders, 4, 54-59. doi: https://doi.org/10.1016/j.jocrd.2014.12.006 Frost, R., Steketee, G., & Williams, L. (2000). Hoarding: A community health problem. Health and Social Care in the Community, 8(4), 229–234. doi:10.1046/j.1365- 2524.2000.00245.57 Frost, R.O., Steketee, G., Williams, L.F., & Warren, R. (2000). Mood, personality disorder symptoms and disability in obsessive-compulsive hoarders: A comparison with clinical and nonclinical controls. Behaviour Research and Therapy, 38(11), 1071–1081. doi:10.1016/S0005-7967(99)00137-0 Frost, R., Tolin, D. F., Steketee, G., Fitch, K. E., & Selbo-Bruns, A. (2009). Excessive acquisition in hoarding. Journal of Anxiety Disorders, 23(5), 632-639. doi:10.1016/j.janxdis.2009.01.013 Furby, L. (1978). Possessions: Toward a theory of their meaning and function throughout the life cycle. Life-span development and behavior. 1, 297-336. Gawronski, B., Bodenhausen, G. V., & Becker, A. P. (2007). I like it, because I like myself: Associative self-anchoring and post-decisional change of implicit evaluations. Journal of Experimental Social Psychology, 43(2), 221-232. doi: https://doi.org/10.1016/j.jesp.2006.04.001 Greenberg, D. (1987). Compulsive hoarding. American Journal of Psychotherapy, 41(3), 409– 416. Retrieved from http://web.ebscohost.com/ehost/detail?sid=ac07a5aa-7685-	 120	4d33- 9756-71a54e9f60ac%40sessionmgr115&vid=1&hid= 107&bdata=JnNpd GU9ZWhvc3QtbGl2ZQ%3d%3d#db=mnh&AN=3631351 Grisham, J. R., & Barlow, D. H. (2005). Compulsive hoarding: Current research and theory. Journal of Psychopathology and Behavioral Assessment, 27(1), 45-52. doi: 10.1007/s10862-005-3265-z Grisham, J. R., Brown, T. A., Savage, C. R., Steketee, G., & Barlow, D. H. (2007). Neuropsychological impairment associated with compulsive hoarding. Behaviour research and therapy, 45(7), 1471-1483. doi: https://doi.org/10.1016/j.brat.2006.12.008 Grisham, J.R., Kim, H-J, Frost, R.O., Steketee, G., Tarkoff, A., & Hood, S. (2009). Formation of attachment to possessions in compulsive hoarding. Journal of Anxiety Disorders, 23(3), 357-361. doi:10.1016/j.janxdis.2008.12.006 Grisham, J.R., Norberg, M.M., Williams, A.D., Certoma, S.P., & Kadib, R. (2010). Categorization and cognitive deficits in compulsive hoarding. Behaviour Research and Therapy, 48(9), 866–872. doi:10.1016/j.brat.2010.05.011 Hartl, T. L., Duffany, S. R., Allen, G. J., Steketee, G., & Frost, R. (2005). Relationships among compulsive hoarding, trauma, and attention-deficit/hyperactivity disorder. Behaviour Research and Therapy, 43(2), 269-276. doi:10.1016/j.brat.2004.02.002 Hayward, L. C. (2011). Emotion and categorization: Do hoarders have difficulties with categorization, emotion, and categorizing when emotional?. (Doctoral Dissertation) State University of New York at Binghamton. 	 121	Hayward, L.C., & Coles, M. (2009). Elucidating the relation of hoarding to obsessive- compulsive disorder and impulse control disorders. Journal of Psychopathology Behavioral Assessment, 31(3), 220–227. doi:10.1007/s10862-008-9106-0 Iervolino, A. C., Perroud, N., Fullana, M. A., Guipponi, M., Cherkas, L., Collier, D. A., et al. (2009). Prevalence and heritability of compulsive hoarding: A twin study. American Journal of Psychiatry, 166(10), 1156-1161. doi:10.1176/appi.ajp.2009.08121789 Inquisit 4 [Computer software]. (2015). Retrieved from http://www.millisecond.com. Jackson, C. J., Furnham, A., & Lawty-Jones, M. (1999). Relationship between indecisiveness and neuroticism: The moderating effect of a tough-minded culture. Personality and Individual Differences, 27(4), 789–800. doi:10.1016/S0191-8869(99)00027-6 Jacoby, J., Berning, C. K., & Dietvorst, T. F. (1977). What about disposition?. The Journal of Marketing, 41(2), 22-28. doi: 10.2307/1250630 Johnson, P. (2014). Extension of Nakagawa & Schielzeth’s R2GLMM to random slopes models. Methods in Ecology and Evolution, 5(9), 944-946. doi: 10.1111/2041-210X.12225. Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1990). Experimental tests of the endowment effect and the Coase theorem. Journal of Political Economy, 98(6), 1325-1348. doi: 10.1086/261737 Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica: Journal of the Econometric Society, 47(2), 263-291. doi:10.2307/1914185 Kellman-McFarlane, K. (2013). Clarifying the relationship between compulsive hoarding and categorization deficits (Doctoral dissertation, University of British Columbia). 	 122	Khawaja, N.G. & Yu, L. (2010) A comparison of the 27-item and 12-item Intolerance of Uncertainty Scales. Clinical Psychologist, 14(3), 97­106. doi: 10.1080/13284207.2010.502542 Kleine, S. S., & Baker, S. M. (2004). An integrative review of material possession attachment. Academy or Marketing Science Review, 2004(1), 1-29. Kuznetsova, Brockhoff & Christensen (2016). LmerTest: Tests in linear mixed effects models. R package (2.0-33) [Computer Software]. Retrieved on August 1, 2017, from https://CRAN.R-project.org/package=lmerTest Laitala, K. (2014). Consumers' clothing disposal behaviour – A synthesis of research results. International Journal of Consumer Studies, 38(5), 444-457. doi:10.1111/ijcs.12088 Lawrence, N.S., Wooderson, S., Mataix-Cols, D., David, R., Speckens, A., & Phillips, M.L. (2006). Decision making and set shifting impairments are associated with distinct symptom dimensions in obsessive-compulsive disorder. Neuropsychology, 20(4), 409- 419. doi:10.1037/0894-4105.20.4.409 Luchian, S.A., McNally, R.J., & Hooley, J.M. (2007). Cognitive aspects of nonclinical obsessive-compulsive hoarding. Behaviour Research and Therapy, 45(7), 1657-1662. doi:10.1016/j.brat.2006.08.014 Luke, S. G. (2016). Evaluating significance in linear mixed-effects models in R. Behavior research methods, 49(4), 1494–1502. doi:10.3758/s13428-016-0809-y Mackin, R., A.P., Delucchi, K., & Mathews, C.A. (2010). Cognitive functioning in individuals with severe compulsive hoarding behaviors and late life depression. International Journal of Geriatric Psychiatry, 26(3), 314–321. doi:10.1002/gps.2531 	 123	Malloy-Diniz, L., Fuentes, D., Leite, W. B., Correa, H., & Bechara, A. (2007). Impulsive behavior in adults with attention deficit/hyperactivity disorder: Characterization of attentional, motor and cognitive impulsiveness. Journal of the International Neuropsychological Society, 13(4), 693–698. doi:10.1017/S1355617707070889 Mangiafico S. (2017). rcompanion: Functions to Support Extension Education Program Evaluation. Available at: https://cran.r-project.org/package=rcompanion Mataix-Cols, D., Frost, R., Pertusa, A., Clark, L. A., Saxena, S., Leckman, J. F…,Wilhelm, S. (2010). Hoarding disorder: A new diagnosis for DSM-V? Depression and Anxiety, 27(6), 556-572. doi:10.1002/da.20693 Mataix-Cols, D., Marks, I. M., Greist, J. H., Kobak, K. A., & Baer, L. (2002). Obsessive-compulsive symptom dimensions as predictors of compliance with and response to behaviour therapy: results from a controlled trial. Psychotherapy and psychosomatics, 71(5), 255-262. doi: https://doi.org/10.1159/000064812 McCracken, G. (1986). Culture and consumption: A theoretical account of the structure and movement of the cultural meaning of consumer goods. Journal of Consumer Research, 13(1), 71-84. doi: http://dx.doi.org/10.1086/209048 Morewedge, C. K., Shu, L. L., Gilbert, D. T., & Wilson, T. D. (2009). Bad riddance or good rubbish? Ownership and not loss aversion causes the endowment effect. Journal of Experimental Social Psychology, 45(4), 947-951. https://doi.org/10.1016/j.jesp.2009.05.014 mTurk Worker Website FAQs. (n.d.). In Amazon Mechanical Turk. Retrieved August 10, 2017, from https://www.mturk.com/mturk/help?helpPage=worker Mueller, A., Mitchell, J. E., Crosby, R. D., Glaesmer, H., & de Zwaan, M. (2009). The 	 124	prevalence of compulsive hoarding and its association with compulsive buying in a German population-based sample. Behaviour Research and Therapy, 47(8), 705-709. doi:10.1016/j.brat.2009.04.005 Murray, N., Sujan, H., Hirt, E., & Sujan, M. (1990). The influence of mood on categorization: A cognitive flexibility interpretation. Journal of Personality and Social Psychology, 59(3), 411-425. doi:10.1037/0022-3514.59.3.411 Nakaaki, S., Murata, Y., Sato, J., Shinagawa, Y., Hongo, J., Tatsumi, H., et al. (2007). Impairment of decision-making cognition in a case of frontotemporal lobar degeneration (FTLD) presenting with pathologic gambling and hoarding as the initial symptoms. Cognitive and Behavioral Neurology, 20(2), 121–125. doi:10.1097/WNN. 0b013e31804c6ff7 Nakagawa, S. & Schielzeth, H. (2013). A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods in Ecology and Evolution, 4(2), 133–142. doi:10.1111/j.2041-210x.2012.00261.x Nedelisky, A., & Steele, M. (2009). Attachment to people and to objects in obsessive-compulsive disorder: an exploratory comparison of hoarders and non-hoarders. Attachment & human development, 11(4), 365-383. doi: https://doi.org/10.1080/14616730903016987 Nordsletten, A. E., de la Cruz, L. F., Billotti, D., & Mataix-Cols, D. (2013). Finders keepers: The features differentiating hoarding disorder from normative collecting. Comprehensive Psychiatry, 54(3), 229-237. doi: 10.1016/j.comppsych.2012.07.063 	 125	Oglesby, M. E., Medley, A. N., Norr, A. M., Capron, D. W., Korte, K. J., & Schmidt, N. B. (2013). Intolerance of uncertainty as a vulnerability factor for hoarding behaviors. Journal of Affective Disorders, 145(2), 227-231. doi: https://doi.org/10.1016/j.jad.2012.08.003 Paden, N., & Stell, R. (2005). Consumer product redistribution: Disposition decisions and channel options. Journal of Marketing Channels, 12(3), 105-123. doi: http://dx.doi.org/10.1300/J049v12n03_06 Persons, J. B., & Foa, E. B. (1984). Processing of fearful and neutral information by obsessive-compulsives. Behaviour Research and Therapy, 22(3), 259–265. doi:10.1016/0005- 7967(84)90006-8 Pertusa, A., Fullana, M. A., Singh, S., Alonso, P., Menchón, J. M., & Mataix-Cols, D. (2008). Compulsive hoarding: OCD symptom, distinct clinical syndrome, or both?. American Journal of Psychiatry, 165(10), 1289-1298. doi: 10.1176/appi.ajp.2008.07111730 Petry, N. M., Bickel, W. K., & Arnett, M. (1998). Shortened time horizons and insensitivity to future consequences in heroin addicts. Addiction, 93(5), 729–738. Pierce, J. L., Kostova, T., & Dirks, K. T. (2003). The state of psychological ownership: Integrating and extending a century of research. Review of General Psychology, 7(1), 84. Preston, S. D., Muroff, J. R., & Wengrovitz, S. M. (2009). Investigating the mechanisms of hoarding from an experimental perspective. Depression and Anxiety, 26(5), 425-437. doi:10.1002/da.20417 	 126	Pronin, E. (2009). The introspection illusion. Advances in experimental social psychology, 41, 1-67. doi: https://doi.org/10.1016/S0065-2601(08)00401-2 R Core Team (2017). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Retrieved from https://www.R-project.org/. Rasmussen, J. L., Steketee, G., Frost, R., Tolin, D. F., & Brown, T. A. (2014). Assessing squalor in hoarding: The Home Environment Index. Community Mental Health Journal, 50(5), 591-596. doi:10.1007/s10597-013-9665-8 Roster, C. A. (2001). Letting go: The process and meaning of dispossession in the lives of consumers. Advances in Consumer Research, 28, 425-430. Retrieved on March 12 2016, from http://acrwebsite.org/volumes/8537/volumes/v28/NA-28 Rufer, M., Fricke, S., Moritz, S., Kloss, M., & Hand, I. (2006). Symptom dimensions in obsessive-compulsive disorder: Prediction of cognitive-behavior therapy outcome. Acta psychiatrica Scandinavica, 113(5), 440–6. doi:10.1111/j.16000447.2005.00682.x Salzman L. (1973). The obsessive personality. New York, NY: Jason Aronson. Samuels, J. F., Bienvenu, O. J., Grados, M. A., Cullen, B., Riddle, M. A., Liang, K. Y., Eaton, W.W., & Nestadt, G. (2008). Prevalence and correlates of hoarding behavior in a community-based sample. Behaviour Research and Therapy, 46(7), 836–844. doi:10.1016/j.brat.2008.04.004 Savaş, Ö. (2002). A perspective on the person-product relationship: attachment and detachment (Doctoral dissertation). Middle East Technical University, Turkey. 	 127	Saxena S., Brody, A.L., Maidment, K.M., Smith, E.C., Zohrabi, N., Katz, E., Baker, S.K., & Baxter, L.R. (2004). Cerebral glucose metabolism in obsessive-compulsive hoarding. American Journal of Psychiatry, 161(6), 1038–1048. doi:10.1176/appi.ajp.161.6.1038 Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Janavs, J., Weiller, E., Keskiner, A., ... & Dunbar, G. C. (1997). The validity of the Mini International Neuropsychiatric Interview (MINI) according to the SCID-P and its reliability. European Psychiatry, 12(5), 232-241. doi: https://doi.org/10.1016/S0924-9338(97)83297-X Shim, S. (1995). Environmentalism and consumers' clothing disposal patterns: An exploratory study. Clothing and Textiles Research Journal, 13(1), 38-48. doi:10.1177/0887302X9501300105  Slade, G. (2009). Made to break: Technology and obsolescence in America. Cambridge, MA: Harvard University Press. Starcke, K., Tuschen-Caffier, B., Markowitsch, H. -J., & Brand, M. (2009). Skin conductance responses during decisions in ambiguous and risky situations in obsessive-compulsive disorder. Cognitive Neuropsychiatry, 14(3), 199–216. doi:10.1080/13546800902996831 Steketee, G., & Frost, R. (2006). Compulsive hoarding and acquiring: Therapist guide. New York, NY: Oxford University Press.   Steketee, G., & Frost, R. (2003). Compulsive hoarding: Current status of the research. Clinical Psychology Review, 23(7), 905–927. doi:10.1016/j.cpr.2003.08.002 Steketee, G., Frost, R., & Kyrios, M. (2003). Cognitive aspects of compulsive hoarding. Cognitive Therapy and Research, 27(4), 463-479. doi:10.1023/A:1025428631552 	 128	Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach’s alpha. International Journal of Medical Education, 2, 53–55. doi: http://doi.org/10.5116/ijme.4dfb.8dfd Thaler, R. (1980). Toward a positive theory of consumer choice. Journal of Economic Behavior & Organization, 1(1), 39-60. https://doi.org/10.1016/0167-2681(80)90051-7 Thyer, B. A., Papsdorf, J. D., Davis, R., & Vallecorsa, S. (1984). Autonomic correlates of the subjective anxiety scale. Journal of Behaviour Therapy and Experimental Psychiatry, 15(1), 3–7. doi:10.1016/0005-7916(84)90115-0 Timpano, K., Exner, C., Rief, W., Keshaviah, A., Brahler, E., & Wilhelm, S. (2011). The epidemiology of the proposed DSM-5 hoarding disorder: exploration of the acquisition specifier, associated features, and distress. Journal of Clinical Psychiatry, 72(6), 780-6.	doi: 10.4088/JCP.10m06380 Timpano, K., & Shaw, A. M. (2013). Conferring humanness: The role of anthropomorphism in hoarding. Personality and Individual Differences, 54(3), 383-388. doi: https://doi.org/10.1016/j.paid.2012.10.007 Tolin, D.F., Frost, R.O., & Steketee, G. (2010). A brief interview for assessing compulsive hoarding: The Hoarding Rating Scale-Interview. Psychiatry Research, 178(1), 147-152. doi:10.1016/j.psychres.2009.05.001 Tolin, D. F., Frost, R., Steketee, G., & Fitch, K. E. (2008). Family burden of compulsive hoarding: Results of an internet survey. Behaviour Research and Therapy, 46(3), 334-344. doi: https://doi.org/10.1016/j.brat.2007.12.008 Tolin, D. F., Frost, R., Steketee, G., Gray, K. D., & Fitch, K. E. (2008). The economic and social burden of compulsive hoarding. Psychiatry Research, 160(2), 200–211. doi:10.1016/j.psychres.2007.08.008 	 129	Tolin, D. F., Kiehl, K. A., Worhunsky, P., Book, G. A., & Maltby, N. (2009). An exploratory study of the neural mechanisms of decision-making in compulsive hoarding. Psychological Medicine, 39(2), 325–336. doi:10.1017/S0033291708003371 Tolin, D. F., Meunier, S. A., Frost, R., & Steketee, G. (2010). Course of compulsive hoarding and its relationship to life events. Depression and Anxiety, 27(9), 829-838. doi: 10.1002/da.20684. Tolin, D.F., Stevens, M.C., Villavicencio, A.L., Calhoun, V., Frost, R.O., Steketee, G., Rauch, S., & Pearlson, G. (2012). Neural mechanisms of decision making in hoarding disorder. Archives of General Psychiatry. 69(8), 832-841. doi:10.1001/archgenpsychiatry.2011.1980 Tolin, D. & Villavicencio, A. (2011). Inattention, but not OCD, predicts the core features of hoarding disorder. Behaviour Research and Therapy, 49(2), 120-125. doi:10.1016/j.brat.2010.12.002 Turk, D. J., Van Bussel, K., Brebner, J. L., Toma, A. S., Krigolson, O., & Handy, T. C. (2011). When “it” becomes “mine”: Attentional biases triggered by object ownership. Journal of Cognitive Neuroscience, 23(12), 3725-3733. doi: 10.1162/jocn_a_00101  Van Dam, N. T., & Earleywine, M. (2011). Validation of the Center for Epidemiologic Studies Depression Scale—Revised (CESD-R): Pragmatic depression assessment in the general population. Psychiatry Research, 186(1), 128-132. doi:10.1016/j.psychres.2010.08.018 van Randenborgh, A., de Jong-Meyer, R., & Hu ̈ffmeier, J. (2010). Decision making in depression: Differences in decisional conflict between healthy and depressed 	 130	individuals. Clinical Psychology and Psychotherapy, 17(4), 285–298. doi:10.1002/cpp.651 Wallendorf, M., & Arnould, E. J. (1988). “My favorite things”: A cross-cultural inquiry into object attachment, possessiveness, and social linkage. Journal of Consumer Research, 14(4), 531-547. doi: 10.1086/209134 Warren, L. W., & Ostrom, J. C. (1988). Pack rats: World-class savers. Psychology Today, 22, 58-62. Retrieved on September 10 2016, from http://ezproxy.library.ubc.ca/login?url=https://search-proquest-com.ezproxy.library.ubc.ca/docview/214478562?accountid=14656 Wheaton, M. G., Abramowitz, J. S., Jacoby, R. J., Zwerling, J., & Rodriguez, C. I. (2016). An investigation of the role of intolerance of uncertainty in hoarding symptoms. Journal of Affective Disorders. 193, 208-214. doi: https://doi.org/10.1016/j.jad.2015.12.047 Wincze, J.P., Steketee, G., & Frost, R.O. (2007). Categorization in compulsive hoarding. Behaviour Research and Therapy, 45(1), 63–72. doi:10.1016/j.brat.2006.01.012 Winter, B. (2013). A very basic tutorial for performing linear mixed effects analyses. arXiv preprint arXiv:1308.5499. Retrieved from July 12 2017, from http://arxiv.org/pdf/1308.5499.pdf Wolpe, J. (1990). The practice of behavior therapy (4th ed.) New York: Pergamon Press. Woody, S. R., Kellman-McFarlane, K., & Welsted, A. (2014). Review of cognitive performance in hoarding disorder. Clinical Psychology Review, 34(4), 324-336. doi:10.1016/j.cpr.2014.04.002  	 131	Appendix A Vignette-Based Task  Instructions Imagine that you need to empty a storage area that contains a number of household possessions. You have been managing without these items, which means they are either not critical or not the only items you have. Some of these items are large and some are small things you’ve found in a junk drawer. As you go through the possessions there, you will need to decide which items to relocate to others areas in your home and which items to discard (e.g., throw away, donate, recycle). To do this, you will need to go through many of your possessions, one at a time, and decide whether they are worth keeping. Please read the information in each item and indicate how likely you would be to discard each item, or what you would do with each item (e.g., keep, throw, donate, give away), based on the information provided.  Screenshot: The following is a screenshot of a high space vignette from Version A of the VBT.                  	 132	Appendix B Information Seeking Task Introduction. You are about to be a participant in the first study to use this decision-making task. That means we need your help to develop this task into an accurate measure of how people make decisions about their possessions. All you have to do to help us with this goal is try to make decisions in this task like you would at home. Knowledge of how people make these decisions has important implications for consumer behaviour, social psychology, and mental health. Thank you for your help!  Task Instructions. Imagine that you no longer have enough space for all of your possessions and have set aside time to conduct a ‘spring cleaning’ of your home. You will be going though many of your possessions, one at a time, and deciding whether they are worth keeping. In order to figure out what possessions are worth keeping and what possessions are simply taking up space, you will be able to access different kinds of information about each possession. However, you will only be able to access one piece of information before making each decision, so chose carefully.   Screenshots The following are three screenshots from a trial of the IST. In the first phase of IST trials, participants are presented with a hypothetical possession (e.g., a throw pillow) and asked to indicate if they would like to access information related to future utility (e.g., How could I use this in the future?) or past utility (e.g., When was the last time I used this?). In the second phase of the trial, participants are presented with the answer to their selected question, in this case referring to future utility. In the third phase of the trial, participants are asked to choose whether they wish to keep or discard the hypothetical possession.  Phase I of IST Trial    	 133	  Phase II of IST Trial     Phase III of IST Trial     	 134	Appendix C Dysfunctional Tolerance Scale  Instructions Imagine that you are going through items you have placed in storage and are evaluating what is worth keeping. Each item is described in four different states of disrepair. All of these descriptions suggest the items are imperfect in some way (e.g., damaged, missing pieces, not fully functioning) but may still have some potential to be useful. Please read all of the descriptions and select the highest option where you would still keep the item. For example, if you would keep the item as described in options 1, 2, and 3, but would discard it as described in option 4, you would select option 3. Please keep potential secondary uses in mind. For example, most people would no longer drink out a chipped cup (i.e., primary use). However, some people would re-purpose it as a pencil holder or a plant pot (i.e., secondary uses), and therefore still consider it useful enough to keep. Others would consider it completely useless and throw it away.  Screenshot The following is an example item from the DTS.   

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0362867/manifest

Comment

Related Items