A N E X P E R I M E N T A L ANALYSIS O F C O M P U L S I V E O R D E R I N G A N D A R R A N G I N G by A D A M S C O T T R A D O M S K Y B . S c , The University of Toronto, 1994 M . A . , The University of Brit ish Columbia, 1998 A T H E S I S S U B M I T T E D IN PARTIAL F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F PHILOSOPHY in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Department of Psychology) We accept this thesis as conforming^t0'tn^> required standard T H E UNIVERSITY O F BRITISH C O L U M B I A January , 2001 © A d a m Scott Radomsky, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shal make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shal not be alowed without my writen permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract Compulsive ordering and arranging, and a preoccupation with symmetry have been documented within the context of obsessive-compulsive disorder (OCD) symptomatology, but have not been examined experimentally. Three connected studies were conducted to examine this phenomenon, including the development of a self-report measure of this behaviour, the validation of this scale, and a test of anxiety provocation and memory in association with compulsive ordering and arranging. Results indicated that the self-report measure had good psychometric properties, and external validity. Also, participants who had a strong preference for order were made more anxious by having to complete a difficult task in a disorganized environment. Hypotheses about memory bias in association with this behaviour were not confirmed. Like other types of abnormal behaviour, these symptoms are likely extensions of normal and adaptive ordering and arranging behaviour. They are consistent with cognitive-behavioural conceptualizations of the disorder, however some aspects of this behaviour may not necessarily be reflected by current theory. Results are discussed in terms of the phenomenology of compulsive ordering and arranging, and its relationship to both OCD and normal human behaviour. T A B L E O F C O N T E N T S Abstract ii List of Tables .< iv List of Figures v Acknowledgements vi Introduction 1 Study 1 13 Study 2 37 Study 3 46 Overall Discussion 64 References 72 Appendix I 78 Appendix II 81 iv List of Tables Table 1 - Participant characteristics, Study 1, Phase 1 16 Table 2 - Principal components analysis, Study 1, Phase 1 18 Table 3 - Factor loadings, item means and standard deviations, Study 1, Phase 1 20 Table 4 - Participant characteristics, Study 1, Phase 2 26 Table 5 - Principal components analysis of the SOAQ, Study 1, Phase 2 28 Table 6 - Factor loadings, item means and standard deviations for the SOAQ, Study 1, Phase 2 29 Table 7 - Hotelling's t-tests for differences between dependant correlations, Study 1, Phase 2 32 Table 8 - Participant characteristics, Study 2 39 Table 9 - Participant characteristics, Study 3 51 Table 10 - Baseline SUDS ratings, Study 3 52 Table 11 - Post-provocation SUDS ratings, Study 3 53 List of Figures Figure 1 - Scatterplot of degree of preference of ordered target items over disordered target items and S O A Q scores, Study 2 ... 42 Figure 2 - Post provocation S U D S ratings, Study 3 54 Figure 3 - Target item recall scores by group and condition, Study 3 57 vi Acknowledgements I would like to thank Doreen Radomsky, Nikolas Dixon, members of the Fear and Anxiety Disorders Laboratory at U B C (including D a n a Thordarson, David Hammond, and Michael Morris), my dissertation committee and especially J a c k Rachman for their support and assistance during this seemingly unending process. Portions of this research were presented at the 2000 Association for the Advancement of Behavior Therapy Conference in New Orleans, and at invited addresses at the University of Miami and Concordia University. This research was supported in part by an N S E R C P G S - B scholarship #207769-1998. 1 Introduction Obsessive compulsive disorder (OCD) is characterized by the presence of either obsessions or compulsions, which produce either interference with daily functioning or significant distress (American Psychiatric Association, 1994). While certain examples are provided, the content of these obsessions and compulsions is not specified in the Diagnostic and Statistical Manual of Mental Disorders (4 t h Edition). However, for both obsessions and compulsions, the content of these thoughts and behaviours is not random -it appears to commonly occur in one of several different categories (Pigott, 1998; Rachman & Hodgson, 1980). More recently, this observation has been supported by empirical data from the factor analyses of OCD symptoms (e.g., Baer, 1994; Foa et al., 1998; Leckman et al., 1997; Rasmussen & Eisen, 1992; Summerfeldt et al., 1999; Thordarson et al. , 1993). Within obsessions, thoughts often tend to encompass one or more of aggressive, blasphemous, or inappropriate sexual thoughts. Similarly, compulsions tend to include washing, checking, ordering/arranging, ritualistic tapping/touching, and mental compulsions such as counting or praying (de Silva 85 Rachman, 1998, Rachman & Hodgson, 1980). There are of course, other types of obsessions and compulsions that occasionally occur, but these have rarely been documented. 2 Initially, the content of obsessions a n d / o r compuls ions was not essential to the unders tanding of O C D theory or treatment (Wolpe, 1958). Repetitive behaviour was treated through the appl icat ion of reciprocal inh ib i t ion treatment strategies (Wolpe, 1958) and later by exposure and response prevention (Meyer, 1966). (Incidentally, this treatment modal i ty remains the most supported and applied psychotherapeutic intervention for O C D to this day.) However, as our unders tanding of O C D theory and treatment has increased, there has also been an increasing emphasis on different theoretical and applied conceptualizations of O C D subtypes, i n addi t ion to conceptual iz ing O C D problems as occurr ing w i th in a "spectrum" (Hollander, 1993). Ea r ly cognit ive-behavioural theories of O C D (e.g., Sa lkovskis , 1985) tended to explain the complete phenomenology of O C D wi th one set of psychological factors. In the case of Pau l Salkovskis ' s theory of O C D (1985), the major psychological factor involved is proposed to be inflated responsibil i ty. That is,, most O C D symptomatology can be accounted for by an increased perception of one's responsibi l i ty for the safety and well-being of oneself a n d / o r of one's friends and family members. For example, a person wi th O C D checks household appliances repeatedly because they feel responsible for the safety and security of the household and its contents; a person who washes repeatedly does so because they feel responsible for the heal th of themselves or those a round them; a person who prays excessively 3 does so because they feel responsible for neutra l iz ing or undo ing the effects of their own intrusive b lasphemous thoughts, etc. As this theory has developed and been increasingly applied (through the appl icat ion of responsibi l i ty reduct ion strategies), it has become debatable the extent to w h i c h responsibi l i ty, as an overarching psychological factor, plays the same role i n each O C D subtype. It appears that responsibi l i ty is certainly a core factor i n the nature of compulsive checking; however the importance of responsibi l i ty might well be d imin i shed i n someone who washes compuls ively - and possibly be even further d imin i shed i n someone wi th unwanted , intrusive thoughts - when compared wi th the perceived responsibi l i ty of those who check compulsively. Because of the uncer ta inty about the degree to w h i c h responsibi l i ty plays a role i n a l l cases and subtypes of O C D , other cognitive behavioural theories of the disorder have been proposed - par t icular ly for those subtypes w h i c h do not lend themselves wel l to being completely explained by inflated responsibi l i ty. A prime example of this is the cognit ive-behavioural theory of obsessions by R a c h m a n (1997, 1998), i n w h i c h personal significance is hypothesized to be the psychological factor of foremost importance. Whi le the concept of inflated responsibi l i ty is not incompatible w i th this newer formulat ion and conceptual izat ion of obsessions, it doesn't appear to explain as m u c h of the phenomenology as a formulat ion that centralizes the role of personal significance. S imi lar ly , one would expect that other 4 subtypes might benefit from revised and more specific cognitive-behavioural explanations. One subtype that has received very little attention in the literature is a compulsion to order and arrange one's surroundings, to avoid disorderly surroundings, and to ensure that objects are arranged in "exactly the right way". This may often include an increased preference for symmetry, either generally, or with respect to a few select possessions. Patients within this subtype of OCD will often report that they need to ensure that their belongings are "just right" before they can proceed with their day, often spending several hours in front of a shelf or table, repositioning objects unti l their anxiety decreases to a tolerable level. It is surprising that this subtype of OCD has been so neglected in the literature, particularly given the results of epidemiological data, which indicate that ordering and arranging is one of the more common presentations of OCD generally (e.g., Rasmussen & Eisen, 1992; Sasson et al., 1997), in adults (e.g., Leckman et al., 1997; Summerfeldt et al. , 1999), and in children (e.g., Flament et al., 1988; Valleni-Bassile et al., 1994). While studies like these document the presence of this subtype of OCD, there is a notable absence of both critical experimental analyses of ordering and arranging, and of theoretical explanations of this behaviour. As such, little is known about the specific phenomenology of ordering and arranging apart from the experiences of those who treat the problem clinically. This 5 investigation will hopefully provide the basis for future investigations of the prevalence of ordering and arranging within populations of people who suffer from OCD (something which has not been well documented in the literature). We would expect that this compulsion serves the same functional purpose as other compulsions - to reduce anxiety. One might hypothesize that people with ordering and arranging compulsions feel that by obtaining control over a small part of their external environment (which may also include an internal sense of being "just right"), they can feel safe and/or more in control of the rest of their environment. This is similar to some conceptualizations of eating disorders, in which it is hypothesized that one of the reasons these patients restrict their diets is because they feel that weight is one of the few things they can control - and this limited control provides some relief from a perceived chaotic world (Fairburn, Shafran 85 Cooper, 1999; Rodin, 1977). This similar formulation may account for the strong overlap between eating disorders and the specific subtype of ordering and arranging within OCD found by Matsunaga and colleagues (1999). One obvious overlap with ordering and arranging behaviour occurs within obsessive-compulsive personality disorder (OCPD). This personality disorder is characterized by a "preoccupation with orderliness, perfectionism, and mental and interpersonal control" (American Psychiatric 6 Associa t ion , 1994, p.669). There is no reason to assume that ordering and arranging behaviour i n O C D should be any different from the same behaviour i n O C P D . In fact, the conceptual izat ion of this behaviour as an attempt at gaining more control is consistent across both disorders. We would expect however, that O C D ordering and arranging shou ld serve a different function - as anxiety reducing behaviour; whereas O C P D ordering and arranging is l ikely carried out because of a simple preference for orderly environments . The better psychometric properties of Ax i s I diagnoses when compared wi th personali ty or Ax i s II disorders w i l l lead us to examine and conceptualize ordering and arranging behaviour wi th in the context of O C D rather than O C P D ; however it is not argued that the two types of behaviour are fundamental ly different. Another important psychological factor i n compulsive ordering and arranging is l ikely to be inflated responsibil i ty. If these patients d id not feel responsible for the stabili ty or control of their environment, they might not be so consumed by this behaviour - or by the desire to achieve the in ternal and external state of things being "just right". This formulat ion is consistent wi th Salkovskis ' s (1985) theory of O C D but also enables more detailed cognitive analyses of concepts l ike perceived control , and feeling "just right". The concept of inflated responsibi l i ty alone wou ld not necessari ly implicate these ideas i n the development or maintenance of this disordered behaviour. 7 Much like other symptoms of OCD, we would expect that ordering and arranging behaviour lies on a continuum from normal, adaptive, everyday ordering and arranging behaviour to more significant, clinically impairing compulsive ordering and arranging. Certainly, we all need some degree of order in our lives. We tend to sort things naturally - both for convenience (e.g., plates, glasses, pots etc., are all kept in groups/piles in the kitchen) and for aesthetics (e.g., pictures tend to hang on our walls level with the floor/ceiling, organized rooms are more appealing, etc.). Without any ordering and arranging, it would be more difficult to find the things we need, to manoeuvre safely through our homes, and to live in our complicated physical and social environments. Some of us tend to prefer more order than others and many people feel as if some of the objects in their home (e.g., trophies, prized posessions, etc.) have an exact and proper place. Obsessions and compulsions do occur in normal people (Gibbs, 1996). We know that normal individuals experience the same types of thoughts as obsessionals (de Silva & Rachman, 1978; Salkovskis & Harrison, 1984). We also know that behaviour almost identical to compulsions is commonly found in normal individuals (Muris, Merkelbach & Clavan, 1997). It is reasonable to expect then, that a preference for order and some amount of arranging behaviour is prevalent in a majority of people. This hypothesis needs to be tested however, and evidence from other studies of normal undergraduates and 8 community adults suggests that, if it is present, it can be easily detected (Gibbs, 1996). In order to assess the degree to which this behaviour exists in normal populations, there is a need for a tool that measures ordering and arranging behaviour. One of the most popular pencil and paper measures of OCD symptomatology has been the Maudsley Obsessional Compulsive Scale (MOCI - Rachman 85 Hodgson, 1980). Of the thirty true/false items in the scale, none of them assesses ordering or arranging behaviour or a desire for symmetry. The revision of the MOCI, the Vancouver Obsessive Compulsive Scale (VOCI - Thordarson, Radomsky, Rachman, Sawchuk 8B Shafran, 1993), contains one item within its Routine/Counting/Slowness subscale, "I feel upset if my furniture or other possessions are not always in exactly the same position". This item does not adequately assess the full range of ordering and arranging behaviour hypothesized to occur among compulsive orderers and arrangers. There is currently no well-validated scale available which assesses this behaviour. Another good argument for the development of a psychometric assessment tool is that it would provide an excellent opportunity to determine if ordering and arranging behaviour is a singular construct or if it comprises more than one set of beliefs and/or behaviours. Unlike claustrophobia, which is comprised of both a fear of restriction and a fear of suffocation (Rachman 8B Taylor, 1993; Radomsky et al., in press), there is no reason to hypothesize that ordering and arranging behaviour is made up of separate fears. The assumption that we would not expect people to be arranging their objects according to multiple sets of rules or with separate goals in mind, primarily results from the phenomenology of ordering and arranging behaviour seen in the clinic. Our experience indicates that some anxiety disorders appear to fall into several classes (e.g., claustrophobic patients seem to be primarily concerned with either an inability to breathe in or an inability to escape from enclosed places). Compulsive ordering and arranging concerns however, appear to present as a simple distress in the presence of disorder and a calmness in the presence of order. While there are the occasional cases that present with concerns about only symmetry, for example, the overwhelming majority appear to have general preferences for order over disorder. There will certainly be idiosyncratic cognitions underlying this behaviour (e.g., "if my belongings are not properly arranged, my husband will be in an accident" or "if my belongings are not just right, I will have bad luck") - as there appears to be in association with other compulsive behaviour. However, clinicians have reported that ordering and arranging compulsions often occur in the absence of traditional fear-related cognitions (Wilhelm, 2000). The demonstration of a unitary factor structure within a self-report measure of ordering and arranging behaviour would lend support to this hypothesis. It may be revealed that this type of compulsive 10 behaviour is motivated by a basic prepared calming response associated with orderly surroundings. In addition to a psychometric test of the structure of the behaviour, and its prevalence in normal populations, an experiment in which participants are asked to indicate their preference for both orderly and disorderly environments would provide a slightly more ecologically valid indication of the degree to which this behaviour is prevalent. Also, provoking individuals who do have a preference for orderly surroundings, by having them complete a difficult task in a disorderly and disorganized environment, should provide an excellent opportunity to examine the beliefs and cognitions operating in these individuals in vivo, as well as to test the hypothesis that these people are more anxious when their surroundings are disorganized. This will obviously need to be contrasted with participants without this preference, as well as with participants placed in an orderly and organized environment. Should we find elevated anxiety among participants who have a preference for order, who are then placed in disorderly surroundings, it would provide some indication of how these preferences lead to distress and interference in daily functioning. A set of well developed psychometric, experimental, and ecologically valid studies will enable us to test if the clinically observed phenomenology of ordering and arranging is consistent with current cognitive behavioural theories of OCD. 11 Memory bias in compulsive ordering and arranging This research grew out of some observations made during a study of memory bias in compulsive washing (Radomsky & Rachman, 1999). This study required participants who washed compulsively to watch as each of a set of 50 objects was either contaminated, or touched but not contaminated. After watching this procedure and completing a neuropsychological memory assessment, participants were given several memory tests. Results indicated that participants with OCD and a fear of contamination had a biased recall of contaminated items, whereas a clinically anxious control group, and an undergraduate student group did not. One of the clinical experimental group participants who was excluded from the Radomsky & Rachman (1999) study was someone who did not have a fear of contamination but instead ordered and arranged many of her belongings at home. This participant was run through the experimental protocol accidentally (only compulsive washers were to be included in the experimental group - and a diagnosis of OCD was an exclusionary criteria for the clinical control group), resulting from a miscommunication between the assessor and the experimenter. During the recall task, this participant recalled all of the objects that were "not put right" on the table "because 12 they stood out - like they needed to be moved a bit", and had trouble recalling the objects that had been 'properly placed'. Just as memory biases in favour of threatening information have been demonstrated in compulsive washers (Radomsky & Rachman, 1999) and among compulsive checkers (Radomsky, Rachman 8& Hammond, in press), we would expect that compulsive orderers and arrangers will have a biased recall for objects which are out of place or disorderly. An experimental analysis of compulsive ordering and arranging therefore provides an excellent opportunity to test for the presence of this bias. Summary Increasingly, researchers are paying more and more attention to OCD subtypes. While some of these (especially washing, checking, and more recently, obsessions without overt compulsions) have been receiving particular attention in the literature, there is a great need for an experimental analysis of ordering and arranging. This behaviour can be explained somewhat by current cognitive behavioural theories of OCD but might well benefit from a more precise conceptualization. There is therefore a need to develop a scale which measures this phenomenon, to assess whether it (like other OCD symptoms) occurs on a continuum from normal everyday behaviour to more significant clinically impairing behaviour, and' 13 to test if people who do compulsively order and arrange their surroundings are indeed upset by disorder and disarray. Finally, these studies provide an excellent opportunity to test for the presence of memory bias among compulsive orderers and arrangers - which would be consistent with memory biases that have been found in association with other subtypes of OCD. Study 1 Aim: The aim of this study is to develop a psychometric scale (The Symmetry, Ordering and Arranging Questionnaire - SOAQ) to assess beliefs and behaviour associated with compulsive ordering and arranging. Predictions: 1. The SOAQ will have sufficient psychometric properties to be used as a self-report measure for ordering and arranging behaviour within a sample of undergraduate students. 2. The SOAQ will measure a single factor. 3. The SOAQ will have a stable factor structure. 14 Method - Phase I Thirty-four items pertaining to compulsive ordering, arranging and a desire for symmetry were formulated by the U B C Fear and Anxiety Disorders Laboratory from both theoretical conceptualizations and clinical presentations of ordering and arranging in O C D . The working group in the lab included a professor with international authority on the research and treatment of obsessive compulsive disorder, a post-doctoral fellow with several years of senior experience in the research and treatment of O C D , several doctoral level graduate students, all of whom had conducted research and treatment with O C D patients, and one of whom does a great deal of ordering and arranging in her private life. Items were based on clinical reports, on logical extensions of items developed for the Vancouver Obsessional Compulsive Inventory (VOCI - Thordarson et al., 1993), and on extensions of current cognitive behavioural conceptions of O C D . This scale (see Appendix I) was distributed to 250 undergraduate psychology students, who were asked to rate each item on a 0 to 4 scale to indicate how strongly they agree with each statement. Participants were also asked to complete scales that assess related phenomena - e.g., the Maudsley Obsessional-Compulsive Inventory - M O C I (Rachman & Hodgson, 1980), the Vancouver Obsessive Compulsive Inventory - V O C I (Thordarson, Radomsky, Rachman, Sawchuk 8B Shafran, 1993), the Beck Anxiety 15 Inventory - BAI (Beck et al., 1990), and the Beck Depression Inventory II -BDI (Beck, Steer, 85 Garbin, 1996). Participants responded to announcements in undergraduate psychology lectures and to signs in the psychology department building indicating where they could obtain questionnaire packages. Participants were offered course credit for completing the questionnaires. Participants We received 211 completed questionnaire packages (a return rate of 84.4%). 74.3% of participants who completed the questionnaires were female. Other participant characteristics are displayed in Table 1. 16 Table 1 - Participant characteristics, Study 1, Phase I Mean SD Minimum Maximum Maximum possible Age 19.34 1.48 16.00 30.00 BDI 1 12.74 9.12 0.00 52.00 63.00 B A I 2 12.11 9.39 0.00 43.00 63.00 M O C I 3 8.64 5.85 0.00 25.00 30.00 Total V O C I 4 0.83 0.56 0.02 2.69 4.00 Total 1 - Beck Depression Inventory 2 - Beck Anxiety Inventory 3 - Maudsley Obsessional Compulsive Inventory 4 - Vancouver Obsessional Compulsive Inventory 17 Results - Phase 1 Inter-item reliability The inter-item reliability of the 34 items was extremely high, Cronbach's oc=0.98. Factor Structure Items were subjected to a principal components analysis and resulting Eigenvalues are displayed in Table 2. The best solution (both statistically and conceptually) was a one-factor solution that accounted for 56.7% of the variance. Factor loadings, item means and standard deviations are displayed in Table 3. Table 2 -Principal components analysis, Study 1, Phase 1 18 Initial Eigenvalues Component Total % of variance Cumulative % 1 19.287 56.73 56.73 2 1.566 4.60 61.33 3 1.243 3.66 64.99 4 1.042 3.06 68.05 5 0.876 2.58 70.63 6 0.796 2.34 72.97 7 0.674 1.98 74.95 8 0.625 1.84 76.79 9 0.606 1.78 78.57 10 0.576 1.69 80.27 11 0.539 1.58 81.85 12 0.532 1.57 83.41 13 0.472 1.39 84.80 14 0.451 1.33 86.13 15 0.427 1.26 87.39 16 • 0.414 1.22 88.60 17 0.354 1.04 89.65 18 0.346 1.02 90.66 19 0.318 0.94 91.60 20 0.300 0.88 92.48 21 0.289 0.85 93.33 22 0.265 0.78 94.11 23 0.247 0.73 94.84 24 0.238 0.70 95.54 25 0.211 0.62 96.16 26 0.205 0.60 96.76 27 0.175 0.51 97.28 28 0.168 0.50 97.77 29 0.156 0.46 98.23 30 0.149 0.44 98.67 3 1 0 . 1 4 3 0 . 4 2 9 9 . 0 9 3 2 0 . 1 1 5 0 . 3 4 9 9 . 4 3 3 3 0 . 1 0 0 0 . 2 9 9 9 . 7 2 3 4 0 . 0 9 6 0 . 2 8 1 0 0 . 0 0 20 Table 3 -Factor loadings, item means and standard deviations, Study 1, Phase 1 Item Factor Loading Item Mean Standard deviation 1. I feel upset if my furniture is not 0.760 0.670 0.931 always in exactly the same position. 2. Other people think I spend too much 0.735 0.609 0.955 time ordering and arranging my belongings. 3. It is essential that I arrange my 0.727 0.868 1.102 clothing in a particular and specific way. 4. I am more at ease when my belongings 0.720 1.308 1.058 are "just right". 5. I cannot go to sleep unless my 0.748 0.528* 0.931 belongings have been arranged properly. 6. I must keep my papers, receipts, 0.751 0.821 1.000 documents, etc. organized according to a specific set of rules. 7. Ordering and arranging take up so 0.740 0.420* 0.747 much time that I am prevented from doing important activities in my daily life. 8. It is important that my belongings are 0.770 0.623 0.892 placed in a symmetrical and evenly distributed way. 9. When I am anxious, I find myself 0.676* 0.835 1.019 arranging and ordering my surroundings much more than usual. 10. The wall hangings (pictures, posters, 0.624* 1.283 1.150 etc.) in my home must be exactly even or straight. 21 11. If someone accidentally disturbs my 0.779 0.844 1.028 belongings - however slightly, I become bothered or upset. 12.1 feel compelled to arrange my 0.846 0.829 1.037 possessions until it feels "just right". 13. When I think that my belongings are 0.843 0.733 0.995 out of place, I am uncomfortable or anxious. 14. I cannot leave my desk or workspace 0.847 0.564* 0.883 until it has been arranged so that it is "just right". 15. When I put things away, I feel 0.757 0.844 0.925 compelled to do it carefully and precisely. 16. My main problem is that I spend too 0.801 0.415* 0.765 much time arranging my belongings 17. Putting objects in order helps to calm 0.685* 0.835 1.024 me down. 18. The furniture in my home must be in 0.855 0.706 0.941 exactly the "right" spot. 19.1 feel calm and relaxed only when 0.779 0.807 0.991 objects around me are organized and placed correctly. 20.1 feel compelled to arrange cans or 0.759 0.611 0.957 boxes of food on my kitchen shelves in a specific way. 21. When I see that my belongings are out 0.867 0.632 0.912 of place, I become anxious until I can arrange them properly. 22.1 feel compelled to arrange objects so 0.800 0.687 0.950 that they are balanced and evenly spaced. 23. Ordering and arranging take up so 0.671* 0.332* 0.752 much of my time that I am often late for appointments, events, etc. 24.1 am much more concerned about 0.812 0.514* 0.857 22 ordering and arranging than is necessary. 25.1 insist that people not touch,my belongings because they might move them out of place. 26.1 feel calm/at ease only when my surroundings are neat and tidy. 27. Even when my home is messy, I keep things organized according to a specific set of rules. 28. Things in my home have a proper and exact place. 29. If I cannot have all of my belongings in a properly arranged manner, I feel that something awful might happen. 30.1 cannot concentrate unless things are in the right place. 31. It is sometimes difficult to tell whether or not I have arranged things properly. 32.1 don't like to disturb objects once they are properly arranged. 33. Messiness in my home always makes me uncomfortable or anxious. 34. Cooking items in my kitchen (e.g. pots, pans, measuring cups) must be arranged in a particular way. Items from the original pool with factor loadings of less than 0.7 (items 9, 10, 17, 23, 29, 31, 33 and 34) were removed. Additionally, items with means at or less than 0.6 (items 5, 7, 14, 16, 24 and 25) were deleted 0.784 0.704 0.735 0.744 0.686* 0.771 0.607* 0.775 0.672* 0.687* 0.597* 1.000 0.925 1.142 0.311* 0.678 0.613 0.807 0.920 0.811 0.943 0.993 1.116 1.073 0.707 0.838 0.898 0.991 0.943 1.041 23 in order to produce a scale that will better discriminate between normal and excessive orderers/arrangers. Convergent and divergent validity A preliminary analysis of convergent validity was conducted by computing correlations between this new 20-item scale, the MOCI, the VOCI, the "just right" subscale of the VOCI, as well as with the BAI and BDI. Hotelling's t-tests for testing differences between dependant correlations (Hotelling, 1940) indicated that the correlation between the ordering and arranging scale and the MOCI total score (r=0.63, p_<0.001) was greater than the correlation between the ordering scale and the BDI (r=0.40, p<0.001), t(208)=6.24, p_<0.001, and than the correlation between the ordering scale and the BAI (r=0.46, p<0.001), t(208)=3.07, p<0.005. Similarly, the correlation between the ordering scale and the VOCI total score (r=0.74, P<0.001) was significantly greater than the correlation between the ordering scale and the BDI, t(208)=7.40, p<0.001, and than the correlation between the ordering scale and the BAI, t(208)=6.19, p<0.001 . Additionally, the correlation between the ordering scale and the VOCI "just right" subscale score (r=0.77, p<0.001) was significantly greater than the correlation between the ordering scale and the BDI, t(208)=7.95, p<0.001, and than the correlation between the ordering scale and the BAI, t(208)=6.81, p<0.001. 24 Method - Phase 2 The SOAQ (Appendix 2) was distributed to approximately 250 undergraduate psychology students, along with other related questionnaires (MOCI, VOCI). Questionnaires which assess problems that are NOT hypothesized to be related to compulsive ordering and arranging, such as the BDI, the BAI, the Claustrophobia Questionnaire - CLQ (Radomsky et al., in press), the Agoraphobic Cognitions Questionnaire - ACQ (Chambless et al., 1984), and the Body Sensations Questionnaire - BSQ (Chambless et al., 1984) were included in the package. Participants responded to announcements in undergraduate psychology lectures and to signs in the psychology department building indicating where they could obtain questionnaire packages. Participants received course credit for completing the questionnaires. Participants There were 190 questionnaire packages returned by psychology undergraduate students (a return rate of 76.0%). 72.3% of the participants 2 5 who completed the questionnaire packages were female. Other participant characteristics are displayed in Table 4. 26 Table 4 -Participant characteristics, Study 1, Phase 2 Mean SD Minimum Maximum Maximum possible Age 19.1 1.81 17.0 31.0 BDI 1 11.4 9.00 0.0 49.0 63.0 BAI 2 14.7 10.42 0.0 49.0 63.0 MOCI 3 9.8 10.13 0.0 26.0 30.0 Total VOCI 4 0.78 0.56 0.02 2.84 4.00 Total ACQ 5 26.3 9.16 14.0 57.0 64.0 BSQ 6 37.5 11.10 16.0 64.0 64.0 CLQ 7 32.8 16.85 1.0 88.0 104.0 1 - Beck Depression Inventory 2 - Beck Anxiety Inventory 3 - Maudsley Obsessional Compulsive Inventory 4 - Vancouver Obsessional Compulsive Inventory 5 - Agoraphobic Cognitions Questionnaire 6 - Body Sensations Questionnaire 7 - Claustrophobia Questionnaire 27 Results - Phase 2 Inter-item reliability The inter-item reliability of the SOAQ was extremely high, Cronbach's a=0.96. Factor structure The SOAQ data were subjected to a principal components analysis to confirm the factor structure used in Phase 1. Again, there was strong evidence for a one-factor solution, accounting for 58.5% of the variance. Eigenvalues are displayed in Table 5. Factor loadings, item means and item standard deviations are displayed in Table 6. 28 Table 5 -Principal components analysis of the SOAQ, Study 1, Phase 2 Initial Eigenvalues Component Total % of variance Cumulative % 1 11.691 58.46 58.46 2 0.992 4.96 63.42 3 0.891 4.45 67.87 4 0.819 4.10 71.96 5 0.654 3.27 75.24 6 0.614 3.07 78.31 7 0.593 2.97 81.27 8 0.495 2.48 83.75 9 0.449 2.24 85.99 10 0.400 2.00 87.99 11 0.356 1.78 89.78 12 0.323 1.61 91.38 13 0.312 1.56 92.94 14 0.284 1.42 94.36 15 0.239 1.19 95.56 16 0.222 1.11 96.67 17 0.203 1.02 97.68 18 0.183 0.92 98.60 19 0.155 0.78 99.38 20 0.125 0.62 100.00 29 Table 6 - Factor loadings, item means and standard deviations for the SOAQ, Study 1, Phase 2 Item Factor Item Standard Loading Mean deviation 1. I feel upset if my furniture is not 0.758 0.687 0.901 always in exactly the same position. 2. Other people think I spend too much 0.639 0.615 0.849 time ordering and arranging my belongings. 3. It is essential that I arrange my 0.724 0.844 0.999 clothing in a particular and specific way. 4. I am more at ease when my belongings 0.712 1.447 1.132 are "just right". 5. I must keep my papers, receipts, 0.739 1.111 1.121 documents, etc. organized according to a specific set of rules. 6. It is important that my belongings are 0.783 0.838 1.097 placed in a symmetrical and evenly distributed way. 7. If someone accidentally disturbs my 0.751 0.888 1.022 belongings - however slightly, I become bothered or upset. 8. I feel compelled to arrange my 0.844 0.860 1.004 possessions unti l it feels "just right". 9. When I think that my belongings are 0.818 0.721 1.017 out of place, I am uncomfortable or anxious. 10. When I put things away, I feel 0.789 0.972 1.046 compelled to do it carefully and precisely. 11. The furniture in my home must be in 0.810 0.620 0.931 exactly the "right" spot. 30 12.1 feel c a l m and relaxed only when objects a round me are organized and placed correctly. 13.1 feel compelled to arrange cans or boxes of food on my k i tchen shelves i n a specific way. 14. W h e n I see that my belongings are out of place, I become anxious u n t i l I can arrange them properly. 15.1 feel compelled to arrange objects so that they are balanced and evenly spaced. 16.1 feel c a l m / a t ease only when my sur roundings are neat and tidy. 17. Even when my home is messy, I keep things organized according to a specific set of rules. 18. Things i n my home have a proper and exact place. 19. I cannot concentrate unless things are i n the right place. 20.1 don't l ike to d is turb objects once they are properly arranged. Convergent and divergent validity Convergent validity was examined through correlations between the S O A Q and the V O C I total score (r=0.64, p<0.001), and between the S O A Q and the V O C I "just right" subscale score (r=0.72, p<0.001). In order to assess divergent validity, these correlations were contrasted wi th 0.756 0.692 0.873 0.794 0.764 0.770 0.754 0.793 0.689 0.972 0.905 0.777 0.821 1.212 1.123 1.168 0.888 1.084 1.109 0.993 1.009 1.045 1.176 1.160 1.109 1.091 1.101 31 correlations between the S O A Q and other scales (e.g., A C Q , B S Q , C L Q , B A I , B D I , etc.). Correlat ions between the S O A Q and other self report measures, the B D I (r=0.34, p<0.001), BAI (r=0.44, p<0.001), C L Q (r=0.41, p<0.001), B S Q (r=0.37, p<0.001) and A C Q (r=0.37, p<0.001) were significantly lower than those between the S O A Q and O C D measures reported above (all Hotelling's t-tests for testing differences between dependant correlations were significant, p_<0.001, and are displayed i n Table 7). 32 Table 7 - Hotelling's t-tests for differences between dependant correlations V O C I 1 Total V O C I "just right" B D I 2 5.90 7.70 B A I 3 4.31 6.01 C L Q 4 4.11 5.70 A C Q 5 5.65 7.03 B S Q 6 5.16 6.50 1 - Vancouver Obsess ional Compuls ive Inventory 2 - Beck Depress ion Inventory 3 - Beck Anxie ty Inventory 4 - C laus t rophob ia Quest ionnaire 5 - Agoraphobic Cognit ions Quest ionnaire 6 - Body Sensat ions Quest ionnaire Table shows Hotelling's t values for comparisons of dependant correlations between the S O A Q and co lumn scales, wi th correlations between the S O A Q and row scales. A l l p's<0.001. 33 Normative Da ta This r andom sample of 190 undergraduate students obtained a mean of 18.6 wi th a s tandard deviation of 16.0 on the S O A Q . The range of scores obtained by this sample was from 0 to 70 wi th a possible m a x i m u m range from 0 to 80. D i scus s ion Th i s s tudy resulted i n the development of the Symmetry, Order ing and Ar rang ing Quest ionnaire (SOAQ). It is a 20-i tem self-report scale that was designed to assess ordering and arranging beliefs and behaviour. Phase 1 of this s tudy reduced a pool of 34 items to the final 20 items and provided a pre l iminary look at a one-factor structure, very good inter- i tem reliabil i ty, and good convergent and divergent validity. Phase 2 of this s tudy confirmed that the S O A Q has a strong and stable one-factor solut ion account ing for a large port ion of the variance i n the scale. It also confirmed that the S O A Q has very good inter-i tem reliabil i ty as wel l as very good convergent and divergent validity. Test-retest reliabil i ty w i l l be investigated i n S tudy 2. 34 Of course, there are some l imitat ions to this investigation of the psychometr ic properties of the S O A Q . A l l of the par t ic ipants used i n the development of the scale were undergraduate students, and while some of them certainly cou ld have been compulsive orderers and arrangers, this was not a focus of the val idat ion. We sought to base the development of this scale on an undergraduate sample for several reasons. The most important one is that ordering and arranging behaviour is hypothesized to occur on a con t inuum. Therefore, we expected to find a wide range of this behaviour i n an undergraduate student sample. In fact, the range obtained by this sample (0 to 70) reflects a wide dis t r ibut ion of this behaviour i n undergraduate students. A c l in ica l val idat ion was not inc luded i n the proposal for this project because of the natura l ly occurr ing preferences for order and symmetry i n the general populat ion (also see S tudy 2), and because of time constraints . Va l ida t ion of the S O A Q u s i n g a c l in i ca l sample of patients wi th O C D is now underway at Massachuset ts Genera l H o s p i t a l / H a r v a r d Medica l School , i n Bos ton . The one-factor solut ion of the S O A Q provides valuable information about the nature and structure of this behaviour. It is possible to assume that a preference for order and symmetry is either present i n some degree or absent, and is not l ikely the result of mult iple beliefs, fears, concerns, etc. We chose in i t i a l i tems for the scale ant ic ipat ing the possibi l i ty that there may be several domains of ordering and arranging behaviour (e.g., only at 35 home, only i n the k i tchen, motivated by anxiety, motivated by pleasure, etc.) but, consistent w i th our hypotheses, there appears to be only one. The un i ta ry factor structure of the S O A Q does however warrant further investigation. It may be that the nature of the popula t ion used i n this val idat ion of the scale is such that more complicated var iants or types of ordering and arranging behaviour were not adequately represented and this leaves the possibi l i ty that a c l in ica l val idat ion would resul t i n a different factor structure. However, even if only a smal l number of "different-type" orderers (e.g., people wi th only a need for symmetry, people who only arrange things at home, etc.) were i n the sample, it wou ld be un l ike ly that the single factor solut ion would have accounted for as m u c h variance as it d id . The S O A Q is strongly correlated wi th other measures of O C D (e.g., the VOCI) and wi th symptoms of O C D that are consistent w i th ordering and arranging behaviour (e.g., the V O C I "just right" subscale). These correlations are significantly larger than those between the S O A Q and other scales measur ing constructs less relevant to ordering and arranging (i.e., c laus t rophobia (CLQ), symptoms of depression (BDI), symptoms of anxiety (BAI), and symptoms of panic disorder and agoraphobia (ACQ, BSQ)). Th is convergent and divergent val idi ty lends support to the specificity of the scale. 36 However, it is of note that the S O A Q was significantly correlated wi th a l l of the scales completed i n this investigation. This result cou ld be interpreted as indicat ing that ordering and arranging behaviour is a non-specific coping strategy for emotional distress; however another interpretation is l ikely more accurate. It wou ld not be surpr i s ing to find that people who engage i n a great deal of ordering and arranging behaviour are par t icu lar ly depressed. In fact, it has long been established that depression is strongly associated wi th vi r tual ly every anxiety disorder, w i th the possible exception of specific phobia (see Bar low, 1988 or Craske , 1999). It wou ld also not be surpr i s ing to find that individuals who spend a great deal of time ordering and arranging their sur roundings are l ikely quite anxious , par t icular ly since we are conceptual izing ordering and arranging behaviour as a component of an anxiety disorder (OCD). The l i nks between the S O A Q and measures of c laust rophobic- and panic-related anxiety however are a bit less straightforward u n t i l one considers that these measures have consistently correlated wi th each other i n other investigations (see Radomsky et a l . , i n press). The best l ikely explanat ion for this is that indiv iduals who are chronical ly depressed a n d / o r anxious (i.e., neurotic) exhibit mult iple and diverse symptoms and use mul t ip le coping strategies consistent wi th depression and anxiety. Th is wou ld inc lude not only physiological symptoms of anxiety (as measured by the BAI) but also symptoms of depression (as measured by the BDI) , of O C D 37 (as measured by the M O C I , V O C I and SOAQ), of panic disorder and agoraphobia (as measured by the A C Q and BSQ) and of c laus t rophobia (as measured by the CLQ) . A replicat ion of this psychometric analysis (particularly the convergent and divergent validity analyses) u s i n g a measure of neuro t ic i sm as a covariate by might be useful to consider as a possible means of addressing this issue. We anticipate that the S O A Q w i l l be useful i n a variety of research and c l in ica l applicat ions inc lud ing measur ing treatment progress, examinat ions of ordering and arranging behaviour i n no rma l adul ts , patients w i th O C D , patients wi th O C P D , and patients w i th related problems (e.g., eating disorders). While more psychometric data on the scale is current ly being collected, there is already good evidence to support its use i n the field. It is provided here for publ ic use. S tudy 2 A i m : The a i m of this experiment is to test the hypothesis that people generally have a preference for orderly and arranged objects and environments over disorderly objects and environments. Predictions: 1. Par t ic ipants w i l l indicate a preference for orderly environments (as depicted i n photographs). 2. Par t ic ipants w i l l indicate more uneasiness about being i n disorderly environments (as depicted i n photographs) than they w i l l about being orderly and arranged environments. 3. The degree/ amount of uneasiness i n Predict ion 2 w i l l be positively correlated wi th S O A Q scores. Method Part ic ipants Seventy-four psychology undergraduate students volunteered to participate i n this study. Part icipants had a mean age of 21.3 (SD=4.27) years. 81 .1% of the part icipants were female. Other par t ic ipant characterist ics are displayed i n Table 8. 39 Table 8 - Participant characteristics, Study 2 Mean Standard Minimum . Maximum Deviation Post-secondary education 2.47 1.70 BDI 1 11.40 '8.81 BAP 10.90 9.03 1 - Beck Depression Inventory 2 - Beck Anxiety Inventory Procedure Participants were asked to sort 60 photographs into one of six piles, indicating how comfortable/uneasy they would feel if they were in the environment depicted in each of the pictures. The piles were labeled "very uncomfortable", "uncomfortable", "slightly uncomfortable", "slightly comfortable", "comfortable", and "very comfortable". Participants were asked to visualize themselves in the scene depicted in each picture before placing it in one of the six piles. 0.0 9.0 0.0 47.0 0.0 44.0 40 Photographs were randomly ordered before being given to the part ic ipants for sort ing and consisted of 15 pairs of target i tems and 15 pairs of distractor items. Target i tems were 30 photographs of 15 scenes. E a c h scene was photographed i n both an order ly/arranged condi t ion and a disorderly condi t ion. Distractor items were 30 photographs of different objects or scenes, and consisted most ly of photographs of scenery or rooms inside houses or apartments. After sort ing the target and distractor items, par t ic ipants completed the S O A Q , and were asked to re turn i n two week's time to complete it again (test-retest rel iabil i ty - see Study 1). Resul ts Overal l rankings of organized target photographs (M=3.5, SD=0.73) was significantly higher than those of disorganized target photographs (M=2.0, SD=0.79), paired-samples t(73)=14.0, p<0.001. Th is preference, measured by the difference between the organized target mean r ank and the disorganized target mean rank, was highly correlated wi th S O A Q scores, r=0.57, p<0.001, indicat ing that part icipants who scored higher on the S O A Q h a d a stronger difference between their rankings of organized a n d disorganized target photographs. A more conservative analysis of this relat ionship was conducted us ing a part correlation between S O A Q scores and r ank differences, control l ing for the influence of r anks made of organized targets. This part correlation demonstrated a s imi la r significant relat ionship, r=0.55, p<0.001. A scatterplot of rank difference scores and S O A Q scores is displayed i n Figure 1. 42 Figure 1 - Scatterplot of degree of preference of ordered target items over disordered target items and SOAQ scores, Study 2 -2 -1 0 1 2 3 4 5 Difference in rank between ordered and disordered targets 43 Test-retest rel iabil i ty of the S O A Q Of the 74 part icipants who completed the S O A Q at the time of sort ing the photographs, 33 returned to complete the S O A Q a mean of 16.2 (SD=9.0) days later. Test-retest rel iabil i ty was very h igh , Pearson's r=0.92, rj<0.001. There were no significant differences between the S O A Q scores of par t ic ipants who returned to complete the scale a second time (M=20.52, SD=17.40) and those who d id not (M=18.58, SD=21.25). D i scuss ion Resul ts indicated that part icipants ranked photographs of organized scenes higher than photographs of disorganized scenes. That is , par t ic ipants indicated that they wou ld feel more comfortable / relaxed i n organized environments than they would i n disorganized environments . Moreover, this preference was present i n a l l but two of the 74 par t ic ipants who completed Study 2 (See Figure 1). Addi t ional ly , the degree of this preference was strongly correlated wi th S O A Q scores, indica t ing that increases i n ordering and arranging beliefs and behaviour were associated wi th increases i n the degree to w h i c h part icipants preferred the photographs of ordered scenes over disordered scenes. This relat ionship was 44 demonstrated th rough a s tandard correlat ion coefficient between S O A Q scores a n d mean rank difference scores, and also through a more conservative part correlation between S O A Q scores and difference scores, control l ing for the variance at tr ibuted to ranks of organized photographs. These results support the hypothesis that there is a general preference for order over disorder i n no rma l adults . Th i s preference is adaptive (imagine a wor ld wi thout order) and it is not surpr i s ing that we find an associat ion between disorder and discomfort. Of course, it wou ld be more ecologically val id to have used ac tua l rooms i n ordered a n d disordered states, a l though m u c h less pract ical . Addi t ional ly , it wou ld have been interest ing to have a th i rd type of photograph, one i n w h i c h the scene was only sl ightly disordered. Target photographs used i n the s tudy were pairs of photographs of the same scene i n drast ical ly different states. For example, one of the organized target items was a photograph of approximately 1 0 pairs of shoes, a l l neatly l ined u p i n a row. The disorganized target match ing this i tem showed the same shoes bu t pi led randomly. Since few of u s live a n d / o r function i n surroundings w h i c h are completely disordered (with the possible exception of some of the offices i n the Kenny building), it might be useful i n future investigations of this type to include a set of photographs w h i c h show scenes that are only part ial ly disordered (e.g., 1 0 pairs of shoes, 9 of w h i c h are neatly l ined u p i n a row and one of w h i c h has been set askew or upside down). This wou ld provide an indica t ion of how 45 sensitive people are to smal l amounts of disorder instead of to almost complete disorder. Th is s tudy also demonstrates addi t ional convergent val idi ty of the S O A Q . The fact that S O A Q scores were strongly and significantly correlated wi th the degree of preference of ordered scenes over disordered scenes, whether measured by a s tandard correlation coefficient or through a more conservative part correlation, indicates that the S O A Q score can predict the degree to w h i c h part ic ipants feel relaxed / comfortable whe n imag in ing themselves i n scenes of varying states of order. F ina l ly , the S O A Q had an extremely high test-retest rel iabil i ty coefficient. Th is is very promis ing for the potential use of the S O A Q i n longer term applicat ions but also might provide an indica t ion of the stabili ty of these types of preferences. C l in ic i ans often report that the treatment of compulsive orderers and arrangers is very difficult and slow going compared wi th the treatment of patients wi th other subtypes of O C D . Whi le it has been established that the course of O C D fluctuates and cycles over time (Rachman 8B Hodgson, 1980), perhaps the course of compuls ive ordering and arranging concerns is m u c h more stable than that of other subtypes. Alternatively, O C P D , an Axi s 2 or personality disorder is thought to be incredibly stable over time (APA, 1994), a n d perhaps the h igh test-retest rel iabil i ty of the scale is s imply a reflection of this. One way to resolve these 4 6 issues would be to examine the test-retest reliability of the scale in a larger sample and over longer periods of time. Study 3 Aim: The aim of this experiment is to test the hypothesis that people who excessively order and arrange their surroundings are more likely to be disturbed (e.g., feel anxious) when forced to complete a difficult task in disorderly surroundings than are people who do not excessively order and arrange their surroundings. A supplementary aim of this experiment is to test for the presence of a memory bias for threat-relevant information (disorderliness) among people who feel compelled to keep their surroundings orderly and arranged. Predictions: 1. Participants who score highly on the SOAQ will be more anxious after preparing a difficult task in a disorderly environment than will high SOAQ scoring participants who prepare the task in an orderly environment. 2. This difference in anxiety will not be significant among participants with low SOAQ scores. 47 Supplementary predictions: 3. H i g h S O A Q scoring part icipants w i l l remember more threat-relevant information (groups of objects that were disordered) than threat-irrelevant information (groups of objects that were wel l ordered and arranged). 4. Par t ic ipants who score low on the S O A Q w i l l not demonstrate this bias and w i l l recal l s imi lar amounts of threat-relevant and threat-irrelevant information. Method Par t ic ipants were selected for this experiment based on their S O A Q scores. Twenty-four part icipants wi th S O A Q scores greater than 0.5 s tandard deviations above the normative mean were assigned to the High S O A Q group, and twenty-four part icipants wi th S O A Q scores lower than 0.5 s tandard deviations below the normative mean were assigned to the Low S O A Q group. A l l par t ic ipants completed a smal l questionnaire package that inc luded the V O C I , B D I , and the B A I . After complet ion of the quest ionnaire package, par t ic ipants were asked to provide a baseline measure of anxiety on a subjective un i t s of distress scale (SUDS - reported from 0 - 100). 48 Part ic ipants were then told to prepare a smal l 5 -minute speech about a topic of their choice, wh ich w i l l be presented to a panel of three Univers i ty of B r i t i s h C o l u m b i a faculty members. Part icipants were informed that their speech was to be graded on both its content and del ivery/style . They were told that because this task often produces some anxiety, they wou ld be given a few minutes and a space i n w h i c h to prepare their speech. Par t ic ipants were then be taken to a desk i n a room w h i c h has either been ordered and arranged, or w h i c h has been disordered, p roduc ing two condit ions - organized workspace and disorganized workspace. Two experimenters verified that the desk was i n an ordered or disordered state. Regardless of w h i c h workspace condi t ion the part icipant was assigned to, there were four groups of objects on the windowsi l l i n front of the desk. Two were organized (e.g., books arranged by height, crayons sorted i n box by colour) and two were disorganized (e.g., books pi led randomly, crayons i n r andom pile out of box). After 3 minutes , the experimenter asked part ic ipants to provide another (post-provocation) S U D S rat ing to indicate how anxious they were after preparing their speech. Par t ic ipants were not given ins t ruct ions about whether or not to alter their workspace, a l though they were free to do so. The experimenter then took the part icipant away from the workspace and asked each part icipant to write a l is t of a l l objects that were on the 49 windowsill in front of the desk in the room in which the speech was prepared. Participants were given 2 minutes to complete this task. Participants in the High S O A Q group were then given a very brief interview to assess their cognitions and beliefs about disorderly objects, environments, etc., and about why it is important to them to order and arrange their surroundings. The structured part of the interview consisted of four questions: 1. What kinds of thoughts go through your mind while you are ordering and arranging your surroundings or when you are unable to order or arrange your surroundings? 2. Why do you think you engage in ordering and arranging behaviour? 3. What do you think would happen if you were unable to order or arrange your surroundings? 4. Do you think that this behaviour is excessive? How m u c h of it do you think is acceptable? Following each question, the interviewer was free to probe for more details or to ask for more information. Finally, all participants were told that there was no speech to be given and were debriefed from the study. Participants 50 Part ic ipants i n this study had a mean age of 21.7 (SD=5.5) years. 79 .2% of par t ic ipants were female. Other part icipant characterist ics are displayed i n Table 9. 51 Table 9 - Par t ic ipant characterist ics, S tudy 3 Group S O A Q Score V O C I 1 Total V O C I B D I 2 B A I 3 Score "just right" Subscale Low S O A Q 4.75 0.26 0.33 6.22 6.25 (3.00) (0.33) (0.42) (5.13) (6.58) High S O A Q 44.42 1.09 1.30 16.71 15.13 (12.96) (0.58) (0.72) (10.38) (11.06) 1 - Vancouver Obsess ional Compuls ive Inventory 2 - Beck Depress ion Inventory 3 - Beck Anxie ty Inventory Resul ts Basel ine S U D S ratings for the two groups and two condit ions are displayed i n Table 10, and an A N O V A on these data confirm that there were significant differences i n baseline S U D S ratings between the two condit ions, F(l ,44)=6.26, p_<0.02. Post-provocation S U D S ratings are displayed i n Table 11 and i n Figure 2. Table 10 - Basel ine S U D S ratings, S tudy 3 Group Low S O A Q H i g h S O A Q Organized workspace 4.25 8.33 Condi t ion (7.28) (7.22) Disorganized workspace 12.66 16.92 (12.16) (17.33) 53 Table 11 - Post-provocation S U D S ratings, Study 3 Group Low S O A Q H i g h S O A Q Organized workspace 22.92 45.00 Condi t ion (30.01) (27.80) Disorganized workspace 21.83 70.00 (13.56) (14.92) Figure 2 - Post provocation S U D S ratings. S tudy 3 54 • Low SOAQ group • High SOAQ group Organized Room Disorganized Room 55 Because of baseline differences, a 2 (group) by 2 (room condition) between, between A N C O V A was conducted on post-provocation S U D S means wi th baseline S U D S ratings as covariates. Levine's equali ty of variance test indicated that the assumpt ion of homogenaity of var iance was tenable, F(3,44)=2.67, n.s. Addi t ional ly , there were no violat ions of assumpt ions of co-linearity between the covariate and the dependant variable, or of paral lel regression lines wi th in each group. Resul ts indicated that while part icipants i n the High S O A Q group were significantly more anxious at post-provocation than were par t ic ipants i n the Low S O A Q group, F(l ,43)=25.74, p<0.001, there was a significant interact ion between condi t ion a n d S O A Q group, F(l ,43)=5.00, p_<0.05. P lanned univariate one-way A N C O V A ' s confirmed that High S O A Q part ic ipants were significantly more anxious after having to prepare their speech i n a disorganized environment than i n an organized environment, F(l ,21)=5.89, p_<0.02, while Low S O A Q part icipants were not, F( l ,21)=2.71, n.s. Memory resul ts Recal l scores for organized items and disorganized items, from Low S O A Q and H i g h S O A Q part icipants i n the organized workspace and 56 disorganized workspace condit ions are displayed i n Figure 3. Of a possible total score of 4 for each part icipant, no single par t ic ipant achieved a score greater than 1. A 2 (group) by 2 (room condition) by 2 (type of object), between, between, wi th in M A N O V A was conducted on these data. There were no significant effects found (all F's <1). Figure 3 - Target item recall scores by group and condition, Study 3 57 2-i 1.8-1.6-1.4-1.2-1-0.8-0.6-0.4-0.2-0- J Z Z L • Organized room -Organized items • Organized room -Disorganized items • Disorganized room • Organized items • Disorganized room • Disorganized items Low SOAQ High SOAQ 58 Interview Resul ts The 24 part ic ipants i n the H i g h S O A Q group indicated a narrow range of explanat ions and beliefs associated wi th their ordering and arranging behaviour. Two part icipants indicated that they engaged i n ordering and arranging behaviour because it reduced feelings of impending doom, a l though neither was able to indicate a specific threat that they were t rying to prevent. The other 22 part icipants i n this group reported a variety of s imi lar explanat ions for their ordering behaviour. These inc luded "not feeling right" u n t i l things were arranged properly, wai t ing for an in ternal "cl ick" that wou ld enable them to stop ordering and arranging, and feelings that this type of behaviour wou ld be the only way to reduce general or unspecified anxiety. A l l 24 part icipants i n this group reported that they felt that this behaviour was somewhat i r rat ional , a l though a l l of them also reported that some degree of it was necessary for anyone to funct ion wel l i n their lives. D i scuss ion Resul ts of this study confirmed our hypothesis that (after account ing for baseline variances i n anxiety) people wi thout strong ordering concerns 59 (Low S O A Q participants) do not report more anxiety about completing a difficult task in a disordered environment than they do about completing a difficult task in an ordered environment. However, people with strong ordering concerns (High S O A Q participants) are significantly more anxious about completing a difficult task (speech preparation) in a disordered environment than they are after completing the same task in an ordered environment. These participants also appear to be significantly more anxious generally than are participants without strong concerns about ordering and arranging. Results from the memory portion of the study did not confirm our hypotheses and indicated that participants in both groups were unable to remember m u c h at all (including both ordered and disordered groups of objects) from the windowsill in front of them, regardless of the condition of the room. Incidentally, when told to write a list of all of the groups of objects that were on the windowsill in front of the desk, 4 participants asked, "was there a windowsill there or is this a trick question?" Post-provocation anxiety results are completely consistent with those found in other subtypes of O C D . It has been well established that compulsive washers become anxious when exposed to contamination, and that compulsive checkers become anxious when exposed to doubt or uncertainty about the state of the compulsively checked object. It is therefore not surprising that compulsive orderers and arrangers become 60 anxious w h e n asked to complete a difficult task i n a disordered environment. However, since the s tudy of O C D has been almost exclusively l imited to compulsive wash ing and checking (with a few recent contr ibut ions i n the areas of compulsive hoarding and repugnant obsessions), it was important to es tabl ish that the same basic phenomenology applies to compulsive orderers and arrangers as well . It is of some concern that the baseline differences i n anxiety were not between groups but between condit ions that had yet to be assigned. It had been p lanned that baseline anxiety levels wou ld be used as a covariate i n the analyses, but wi th the assumpt ion that any differences w o u l d be between the Low S O A Q group and the High S O A Q group and not between part ic ipants who were about to prepare a speech at a tidy desk and those who were about to prepare a speech at a messy desk. Th is may s imply be the result of r andom variance i n the populat ion and a relatively smal l sample size. However, it is also quite possible that on the way to the desk where part ic ipants were asked to give their in i t ia l consent to participate i n the s tudy and to fill out questionnaires, they were able to see the (organized or disorganized) workspace, w h i c h had been set up i n advance. One wou ld have ant icipated though, that even if this were the case, it wou ld have differentially effected the two groups but this does not seem to be the case. A repl icat ion of this s tudy would almost certainly resolve this issue. 61 Resul ts from the brief interview about cognition and beliefs indicate that there may be something about compulsive ordering and arranging behaviour that is quite different from other compulsive behaviour. In most manifestations of O C D , the person engages i n compulsive activity to prevent some dreaded event from occurr ing (Rachman 8& Hodgson, 1980). Compuls ive washers and cleaners are often attempting to reduce the l ike l ihood of disease or i l lness. Compuls ive checkers are often at tempting to reduce the l ike l ihood that their stove wi l l cause a fire, that they accidental ly ran someone over and didn' t stop, that they have left the water r u n n i n g and w i l l r u n u p a large b i l l , etc. Even what we know about other subtypes of O C D is consistent w i th this conceptual izat ion (e.g., compuls ive hoarders are often t rying to ensure that they wi l l always be prepared by having appropriate suppl ies or documentat ion, patients wi th b lasphemous obsessions are often trying to avoid going to hel l , etc.). However, this pre l iminary investigation of ordering and arranging has shown that this may not be the case i n this subtype. It appears as if compulsive orderers and arrangers engage i n this behaviour as an anxiety reducing technique but not as a threat reducing technique. None of the part icipants i n the H i g h S O A Q group were able to articulate a specific threat that wou ld present itself shou ld they be unable to order and arrange their sur roundings . It i s possible that this resul ts from the use of undergraduate s tudents i n this study; however it is highly un l ike ly that c l in ica l manifestations of 6 2 compulsive ordering and arranging behaviour wou ld be drast ical ly different from that seen i n the High S O A Q group used i n this investigation. It cou ld be a concern to cognitive theorists that there do not appear to be any threat related cognitions involved i n compulsive ordering and arranging, a l though this pre l iminary investigation is by no means definitive. One addi t ional explanat ion is that instead of direct threat-related cognit ion main ta in ing compulsive ordering and arranging (e.g., "I m u s t keep my belongings orderly and arranged or I w i l l never be able to f ind anyth ing i n my home"), perhaps there is some k i n d of broader based meta-cognit ion at work (e.g., "Ordering and arranging makes me feel more i n control of my life" or "When I order my belongings, I can achieve a sense that things are 'just r ight ' "). Meta-cognit ion has already been implicated i n generalized anxiety disorder (Wells, 1995) as a way to justify the excessive worry associated wi th that disorder (e.g., "When I worry, it makes me feel l ike I a m helping out i n some way"). Meta-cognit ion is not t h ink ing about the direct result of the thoughts or behaviour; it is about the indirect function of those thoughts or behaviour. These types of thoughts and beliefs were not assessed i n the current study but cou ld shed some light on our unders tanding of compulsive ordering and arranging concerns a n d on the apparent lack of direct, threat-relevant cognition i n associat ion wi th this type of compuls ive behaviour. 63 The memory results of the current s tudy d id not confirm our hypotheses. Whi le this may s imply be the result floor effects (e.g., the task was too difficult), it is more l ikely because of the design of this par t icular experiment. A t least two studies have confirmed that if ecologically va l id s t imul i are used that are perceived as personally threatening, a memory bias i n O C D is strong and detectable (Radomsky 8B R a c h m a n , 1999; Radomsky , R a c h m a n 8B H a m m o n d , i n press). The current s tudy was l ikely able to achieve the basic a ims as far as the s t imul i were concerned but the provocation used i n the experiment l ikely had a greater effect on at tention than it d id on memory. It has been proposed that i n social anxiety disorder (or social phobia), anxiety provocations result i n self-focused attention (Rapee & Heimberg, 1997; Woody, 1996). A s k i n g part icipants to prepare a speech to be given to three faculty members is a common provocation associated wi th investigations of social anxiety. Even though the current s tudy had s t imu l i and even a workspace that would be appropriate to test for phenomena associated wi th compulsive ordering and arranging, it is quite l ikely that the provocation used directed par t ic ipants ' at tention inward and away from their external environment, leading to very poor memory about their sur roundings at the later recall test. U s i n g an alternate provocation, (e.g., a sk ing part icipants to tidy up the desk) might enable better encoding and subsequent testing of both recall and recognit ion memory associated wi th this subtype of O C D . It might also be fruitful to 64 consider a provocation i n w h i c h a room is only slightly disordered (e.g., a very t idy desk wi th a plant t ipped over i n one corner) as a means of examining the sensitivity of part icipants to the degree of disorder i n their sur roundings and its subsequent effects on information processing and emotional arousal . Of course, it is also possible that the recal l task was s imply too difficult, i n w h i c h case, a recognition test of memory cou ld have been used. However, i n a s tudy wi th only four groups of objects to be remembered, a recognition test is quite l ikely to produce ceil ing effects. In our previous work wi th memory i n O C D , recall tests were sufficient to detect effects that were quite s trong (Radomsky & R a c h m a n , 1999; Radomsky, R a c h m a n & H a m m o n d , i n press). Overal l D i scuss ion The three studies described above represent a very pre l iminary investigation of compulsive ordering and arranging behaviour. The first s tudy resulted i n the development of the Symmetry, Order ing and Ar rang ing Quest ionnaire (SOAQ), a self-report measure wi th very good in i t i a l psychometr ic properties, i nc lud ing very h igh inter-i tem and test-retest reliabil i ty, a stable one-factor solut ion, and very good convergent a n d divergent validity. The second study provided addi t ional information on the 65 validi ty of the S O A Q , and also established that preferences for order are common i n normal adults . The th i rd s tudy showed that while anxiety responding i n compulsive ordering and arranging is consistent w i th other subtypes of O C D , the beliefs and cognitions associated wi th ordering and arranging may be somewhat different. A supplementary f inding of this th i rd s tudy was that no memory effects were observed, l ikely resul t ing from self-focussed attention occurr ing i n response to a provocation c o m m o n i n social anxiety disorder experiments. This last f inding provides addi t ional support for the use of both ecologically val id s t imul i A N D provocations i n the s tudy of memory i n associat ion wi th anxiety disorders (see Radomsky & R a c h m a n , 1999; or Radomsky, R a c h m a n 8s H a m m o n d , i n press). Of course, these investigations have only part ial ly answered several important questions about ordering and arranging behaviour and have raised many new ones. Whi le there are some good reasons to expect a single factor associated wi th measures of ordering and arranging behaviour, it is possible that a future val idat ion of the S O A Q us ing c l in ica l populat ions might prove differently. In addi t ion to reasons to expect a single factor l isted above (e.g., c l in ica l impressions, possible lack of threat-related cognitions, use of undergraduate val idat ion sample, etc.), there are neuropsychological models of O C D (see Tal l is , 1997 for a review) that might implicate ordering and arranging behaviour as a basic prepared response to 66 disorder. A s such , it wou ld be difficult to argue that ordering and arranging behaviour shou ld be mult ifactorial . It is reasonable to assume that there are adaptive preferences for order, as wel l as for cleanliness, for stability, for confidence, etc. We might also assume that a s imi lar ly adaptive uneasiness occurs wi th disorder, contaminat ion, uncer ta inty and even for the basic movements of reptiles and insects (possible carriers of disease). These tendencies wou ld result i n an in ternal negative reinforcement of si tuations where, for example, order replaced disorder, and are l ikely at least part ial ly accounted for by genetic and neuropsychological factors. The finding that a l l but two par t ic ipants indicated feeling more relaxed i n ordered surroundings than i n disordered sur roundings i n the photograph sorting study lends support to a basic calmness and ease i n associat ion wi th an orderly environment. Of course, a genetic a n d / o r neuropsychological analysis of prepared responses to disorder is not w i th in the scope of this investigation; however future studies of this nature - or even the use of tachistoscopic studies of compar isons between reactions to ordered and disordered scenes - might provide an indica t ion of how basic these processes are to h u m a n functioning. The current investigation d id not examine possible gender differences i n preferences for order. Certa inly stereotypic descriptions of undergraduate males tend to indicate that they are m u c h less "tidy" than undergraduate females. The development of the S O A Q enables controlled 67 research on gender differences i n this behaviour. These differences were not a focus of the current studies because of our interest i n examin ing the basic structure and function of the behaviour but wou ld certainly be easy to examine i n future work of this type. One issue common to compulsive ordering and arranging that was not assessed i n the current set of investigations was menta l ordering or pr ior i t iz ing. It is not u n c o m m o n for patients wi th these concerns to report that they also engage i n a great deal of mental ordering or pr ior i t iz ing before they can begin the phys ica l ordering and arranging of their sur roundings . A n investigation of this more complex mental aspect of this behaviour wou ld follow wel l from the investigations reported above. Also par t icular ly relevant to compulsive ordering and arranging behaviour are the concepts of obsessive-compulsive personali ty disorder (OCPD) and of perfectionism. It is not u n c o m m o n to hear reports of patients w i th O C P D who engage i n this type of behaviour, par t icular ly those who exhibi t a great deal of perfectionism. Order ing and arranging behaviour has also been associated wi th eating disorders (Matsunaga et a l . , 1999) (another area where perfectionism often plays a cr i t ical role). It follows that a perfectionistic style (associated wi th both O C P D and wi th eating disorders) wou ld be related to ordering and arranging behaviour. We often associate a sense of things being 'just right ' wi th things being 'perfect'. The role of compulsive ordering and arranging behaviour w i l l l ikely be 6 8 different, depending on the personal significance it holds for the ind iv idua l . In an eating disorder, for example, it might be important for the person to order their food - or even to impose an order on their eating meals; i n body dysmorphic disorder, the order and symmetry wou ld l ikely be associated wi th the person's appearance; etc. It w o u l d not be surpr i s ing to find that the form of ordering a n d arranging behaviour i n associat ion wi th these other disorders a n d personali ty styles is m u c h like that observed i n obsessive-compulsive disorder; however there are l ikely to be great differences i n the function of this behaviour i n these other groups. More specifically, ordering and arranging i n the context of O C D might serve the function of a n external search for in ternal feelings or cues of cer ta inly and safely, whereas i n O C P D and perfectionism, it might serve the function of providing external and objective indicat ions of correctness. Of course, there is overlap here as there is between the two disorders, but a proper funct ional analysis of ordering and arranging behaviour wi th in the disorder i n quest ion is l ikely to be most fruitful i n explaining its nature and function. These issues are empir ica l i n nature and would be best addressed by experimental analyses of ordering and arranging behaviour i n different no rma l and c l in i ca l populat ions. Another relevant issue here is the dis t inct ion between categorical and d imens iona l conceptualizat ions of this behaviour. In these studies, 69 compuls ive ordering and arranging have been d iscussed i n terms of a d imens iona l approach, a l though it is possible to consider the behaviour from a categorical perspective. We found it helpful to conceptualize the problem along a c o n t i n u u m i n order to help us reconcile both no rma l and abnormal ordering and arranging behaviour. Of course, it is posible to make this d is t inc t ion a categorical one, but we feel that there is a great degree of variance i n the amount of order that different people prefer and that this is best represented by a d imens ional model . Other useful follow-ups, i n addi t ion to the ones suggested above, include possible investigations of the relat ionships between some of the factors that have been established i n other subtypes of O C D (e.g., inflated responsibil i ty, thought-act ion fusion, neutral izat ion, etc.) and behaviours associated wi th compulsive orderers and arrangers. One wou ld expect that i n addi t ion to these factors, a need for control a n d / o r certainty might be par t icular ly relevant to ordering and arranging and provocations w h i c h result i n a lack of control or i n uncer ta inty might lead to amplif ied ordering and arranging behaviour i n these individuals . Before these projects can be under taken though, there is a need for a repl icat ion of the above findings i n a c l in ica l populat ion. Add i t iona l behavioural experiments that elicit the uneasiness demonstrated here through plac ing part icipants i n disordered surroundings are l ikely to be quite fruitful. There wou ld also be m u c h to be learned from a careful 70 examinat ion of the strategies and rules by w h i c h these people arrange their sur roundings . What are the v i sua l and non-v isua l cues that indicate that sufficient order has been achieved? Are they s imi lar or different from those of no rma l par t ic ipants? Investigations of these issues w i l l l ikely provide addi t ional information about the nature of the behaviour both i n no rma l adul ts and i n populat ions of people who are greatly t roubled by their need for things to be ordered, arranged, symmetr ica l and 'just right'. The work reported here is very pre l iminary but has already provided some basis for unders tand compulsive ordering and arranging problems i n O C D , i n other disorders, and i n the normal adul t populat ion. Ul t imately , the largest gains i n our unders tanding of this behaviour w i l l be from a testable model . It seems l ikely that such a model of ordering and arranging behaviour w i th in the context of O C D wi l l include explanations of the no rma l rewarding value of ordered surroundings , a need for certainty, the connect ion between external order and internal sensations of things being 'just right', and the apparent lack of associated threat-relevant cognit ions. Once s u c h a model has been proposed, perhaps along cognitive behavioural l ines, careful hypothesis testing wi l l hopefully result i n subsequent modifications of in i t i a l conceptualizat ions of this behaviour and i n a better unders tanding of the nature and function of compulsive ordering and arranging i n no rma l and abnormal h u m a n behaviour. 71 Of course, issues relevant to ordering and arranging occur i n many other discipl ines . In different governments, cul tures and societies, the amount of 'order' i n the hierarchical s tructure of the society varies a great deal. One has only to consider the cul ture of the mil i tary to appreciate the role of 'order' on dai ly life. Addi t ional ly , i n art (architecture, paint ing, dance, etc.) there is varying emphasis on order and symmetry. These more pol i t ical , cu l tu ra l and phi losophica l issues, while not examined directly i n this dissertat ion, a l l represent related and relevant areas of ordering and arranging preferences i n our lives. 72 References Amer i can Psychiat r ic Associa t ion. (1994). Diagnostic and stat ist ical m a n u a l of menta l disorders (4 t h Ed.) . Washington , D C : A m e r i c a n Psychia t r ic Press. Baer , L . (1994). Factor analysis of symptom subtypes of obsessive-compuls ive disorder and their relation to personali ty and tic disorders. J o u r n a l of C l i n i c a l Psychiatry, 55, 18-23. Bar low, D . H . (1988). Anxiety and its disorders. NY: Gui l ford . Beck, A . T . , & Steer, R .A. (1990). Beck Anxiety Inventory M a n u a l . Toronto: Psychological Corporat ion. Beck, A . T . , Steer, R . A . , 8s Garb in , M . G . (1996). Beck Depress ion Inventory M a n u a l ( 2 n d ed.). San Antonio: Psychological Corporat ion. Chambless , D . , Caputo , G . , Bright , P., 8s Gallagher, R. (1984). Assessment of fear i n agoraphobics: The Body Sensations Quest ionnaire a n d the Agoraphobic Cognit ions Quest ionnaire. J o u r n a l of Consu l t i ng and C l i n i c a l Psychology, 52, 1090-1097. Craske , M . G . (1999). Anxiety disorders. Boulder : Westview Press. de Si lva , P. , 8& R a c h m a n , S. (1998). Obsessive compuls ive disorder: The facts ( 2 n d Ed.). Oxford: Oxford Univers i ty Press. F a i r b u r n , C . G . , Shafran, R. , 8& Cooper, Z. (1999). A cognitive behavioura l theory of anorexia nervosa. Behaviour Research 8& Therapy, 37, 1-13. 73 Flament , M . F . , Whi taker , A . , Rapoport, J . L . , Davies, M . , Berg , C , Ka l ikow, K . , Sceery, W. , & Shaffer, D . (1988). Obsessive-compulsive disorder i n adolescence: A n epidemiological study. J o u r n a l of the A m e r i c a n • Academy of C h i l d and Adolescent Psychiatry, 27, 764-771 . Foa, E . B . , Kozak, M . J . , Sa lkovskis , P . M . , Coles, M . E . , 8B A m i r , N . (1998). The val idat ion of a new obsessive-compulsive disorder scale: The Obsessive-Compuls ive Inventory. Psychological Assessment , 10, 206-214. Gibbs , N . A . (1996). Noncl in ica l populat ions i n research on obsessive-compulsive disorder: A cr i t ica l review. C l i n i c a l Psychology Review, 16, 729-773. Hollander , E . (1993). Obsessive-compulsive-related disorders. Washington , D C : Amer i can Psychiatr ic Press. Hotel l ing, H . (1940). The selection of variates for use i n predict ion wi th some comments on the general problem of nuisance parameters. A n n a l s of Mathemat ica l Statist ics, !J_, 271-283. L e c k m a n , J . F . , Grice , D . E . , Boa rdman , J . , Zhang , H . , Vitale , A . , B o n d i , C , Alsobrook, J . , Peterson, B . S . , Cohen , D . J . , Rasmussen , S .A. , Goodman , W . K . , McDougle , C . J . , 8B Pauls , D . L . (1997). Symptoms of obsessive-compulsive disorder. Amer i can J o u r n a l of Psychiatry, 154, 911-917. Matsunaga , H . , K i r i ike , N . , Iwasaki, Y . , Miya ta , A . , Yamagami , S., 8& Kaye, W . H . (1999). C l i n i c a l characterist ics i n patients wi th anorexia nervosa and obsessive-compulsive disorder. Psychological Medic ine , 29, 407-414. Meyer, V . (1966). The modification of expectations i n cases wi th obsessional r i tuals . Behaviour Research 66 Therapy, 4, 273-280. M u r i s , P. , Merke lbach , H . , 8B C lavan , M . (1997). A b n o r m a l a n d no rma l compuls ions . Behaviour Research 8B Therapy, 35, 249-252. Pigott, T .A. (1998). Obsessive-compulsive disorder: Symptom overview and epidemiology. Bu l l e t in of the Menninger C l in i c , 62, A 4 - A 3 2 . R a c h m a n , S. (1997). A cognitive theory of obsessions. Behaviour Research 8B Therapy, 35, 793-802. Rachman , S. (1998). A cognitive theory of obsessions: Elaborat ions . Behaviour Research 8B Therapy, 36. Rachman , S., 8B de Si lva , P. (1978). A b n o r m a l and no rma l obsessions. Behaviour Research 8B Therapy, 16, 233-248. R a c h m a n , S., 8B Hodgson, R . J . (1980). Obsessions and compuls ions . N J : Prent ice-Hal l . Rachman , S., 8B Taylor, S. (1993). Analyses of Claus t rophobia . J o u r n a l of Anxie ty Disorders , 7, 281-291. 75 Radomsky , A . S . , 8B Rachman , S. (1999). Memory bias i n obsessive-compuls ive disorder (OCD). Behaviour Research 8B Therapy, 37(7), 605-618. Radomsky, A . S . , Rachman , S., 8B H a m m o n d , D . (in press). Responsibi l i ty , confidence and memory i n compulsive checking. Behaviour Research 8& Therapy. Radomsky, A . S . , Rachman , S., Thordarson, D .S . , Mc l saac , H . K . , 8B Teachman, B . A . (in press). The Claus t rophobia Quest ionnaire (CLQ). J o u r n a l of Anxiety Disorders. Rapee, R . M . , 8B Heimberg, R . G . (1997). A cognitive-behavioral model of anxiety i n social phobia . Behaviour Research 8B Therapy, 35, 741 -756. Rasmussen , S., 8B E i sen , J . (1992). The epidemiology and c l in ica l features of obsessive-compulsive disorder. Psychiatr ic C l in i c s of Nor th Amer ica , 15, 743-758. Rodin , J . (1977). Research on eating behavior and obesity: Where does it fit i n personali ty and social psychology? Personality and Socia l Psychology Bu l l e t in , 3, 333-355. Sa lkovskis , P . M . (1985). Obsessional-compulsive problems: A cognitive-behavioural analysis . Behaviour Research 8B Therapy, 27, 677-682. Sa lkovskis , P . M . , 8B Har r i son , J . (1984). A b n o r m a l and no rma l obsessions: A repl icat ion. Behaviour Research 8B Therapy, 22, 549-552. Sasson, Y . , Zohar , J . , Chopra , M . , Lust ig , M . , Iancu, I., 8s Hendler , T. (1997). Epidemiology of obsessive-compulsive disorder: A wor ld view. J o u r n a l of C l i n i c a l Psychiatry, 58, 7-10. Summerfeldt , L . J . , Richter , M . A . , Antony, M . M . , 8B Swinson , R .P . (1999). Symptom structure i n obsessive-compulsive disorder: A confirmatory factor-analytic study. Behaviour Research 8B Therapy, 37, 297-311 . Tal l is , F . (1997). The neuropsychology of obsessive-compulsive disorder: A review and considerat ion of c l in ica l impl icat ions . B r i t i s h J o u r n a l of C l i n i c a l Psychology, 36, 3-20. Thordarson , D . S . , Radomsky, A . S . , Rachman , S., Sawchuk , C , 8s Shafran, R. (1993). The Vancouver Obsessive Compuls ive Inventory. Vancouver : R a c h m a n Fear Lab, U B C . Turner , S . M . , Beidel , D . C , D a n c u , C . V . , 8B Stanley, M . A . (1989). A n empir ical ly derived inventory to measure social fears and anxiety: The Socia l Phobia and Anxiety Inventory. Psychological Assessment: A J o u r n a l of Consu l t ing and C l i n i c a l Psychology, 1_, 35-40. Val len i -Bas i le , L . A . , Gar r i son , C . Z . , J a c k s o n , K . L . , Waller , J . L . , M c K e o w n , R . E . , Addy, C . L . , 8s Cuffe, S.P. (1994). Frequency of obsessive-compuls ive disorder i n a communi ty sample of young adolescents. J o u r n a l of the Amer i can Academy of C h i l d 8s Adolescent Psychiatry, 33, 782-791 . 77 Wells , A . (1995). Meta-cognit ion and worry: A cognitive model of generalized anxiety disorder. Behavioura l and Cognitive Psychotherapy. 23, 301-320. Wi lhe lm, S. (2000). Personal communica t ion . Wolpe, J . (1958). Psychotherapy by reciprocal inh ib i t ion . Stanford: Stanford Univers i ty Press. Woody, S.R. (1996). Effects of attention on anxiety levels on social performance of individuals wi th social phobia. J o u r n a l of A b n o r m a l Psychology. 105. 61-69. 78 Appendix I - Items from Study 1, Phase 1 1. I feel upset if my furniture is not always i n exactly the same posi t ion. 2. Other people th ink I spend too m u c h time ordering and arranging my belongings. 3. It is essential that I arrange my clothing i n a par t icular and specific way. 4. I a m more at ease when my belongings are "just right". 5. I cannot go to sleep unless my belongings have been arranged properly. 6. I mus t keep my papers, receipts, documents , etc. organized according to a specific set of rules. 7. Order ing and arranging take u p so m u c h time that I a m prevented from doing important activities i n my daily life. 8. It is important that my belongings are placed i n a symmetr ica l and evenly dis t r ibuted way. 9. W h e n I a m anxious , I find myself arranging and ordering my sur roundings m u c h more than u sua l . 10. The wa l l hangings (pictures, posters, etc.) i n my home mus t be exactly even or straight. 11. If someone accidental ly dis turbs my belongings - however slightly, I become bothered or upset. 12.1 feel compelled to arrange my possessions u n t i l it feels "just right". 79 13. When I think that my belongings are out of place, I am uncomfortable or anxious. 14.1 cannot leave my desk or workspace until it has been arranged so that it is "just right". 15. When I put things away, I feel compelled to do it carefully and precisely. 16. My main problem is that I spend too m u c h time arranging my belongings 17. Putting objects in order helps to calm me down. 18. The furniture in my home must be in exactly the "right" spot. 19.1 feel calm and relaxed only when objects around me are organized and placed correctly. 20.1 feel compelled to arrange cans or boxes of food on my kitchen shelves in a specific way. 21. When I see that my belongings are out of place, I become anxious until I can arrange them properly. 22.1 feel compelled to arrange objects so that they are balanced and evenly spaced. 23. Ordering and arranging take up so much of my time that I am often late for appointments, events, etc. 24.1 am m u c h more concerned about ordering and arranging than is necessary. 80 25.1 ins is t that people not touch my belongings because they might move them out of place. 26.1 feel c a l m / a t ease only when my surroundings are neat a n d tidy. 27. Even when my home is messy, I keep things organized according to a specific set of rules. 28. Th ings i n m y home have a proper and exact place. 29. If I cannot have a l l of my belongings i n a properly arranged manner , I feel that something awful might happen. 30.1 cannot concentrate unless things are i n the right place. 31 . It is sometimes difficult to tell whether or not I have arranged things properly. 32.1 don't l ike to d is turb objects once they are properly arranged. 33. Mess iness i n my home always makes me uncomfortable or anxious . 34. Cook ing items i n my k i tchen (e.g. pots, pans, measur ing cups) mus t be arranged i n a par t icular way. 8 1 Appendix 2 - The S O A Q Please circle a number from 0 to 4 to indicate how much you agree with each statement: Not at all Sligh Moder tly ately Very Extre mely 1. I feel upset if my furniture or other possessions are not always in exactly the same position. 0 1 2 3 4 2. Other people think I spend too much time ordering and arranging my belongings. 0 1 2 3 4 3. It is essential that I arrange my clothing in a particular and specific way. 0 1 2 3 4 4. I am more at ease when my belongings are "just right". 0 1 2 3 4 5. I must keep my papers, receipts, documents, etc. organized according to a specific set of rules. 0 1 2 3 4 6. It is important that my belongings are placed in a symmetrical and evenly distributed way. 0 1 2 3 4 7. If someone accidentally disturbs my belongings -however slightly, I become bothered or upset. 0 1 2 3 4 8. I feel compelled to arrange my possessions until it feels "just right". 0 1 2 3 4 9. When I think that my belongings are out of place, I am uncomfortable or anxious. 0 1 2 3 4 10. When I put things away, I feel compelled to do it carefully and precisely. 0 1 2 3 4 11. The furniture in my home must be in exactly the "right" spot. 0 1 2 3 4 12.1 feel calm and relaxed only when objects around me are organized and placed correctly. 0 1 2 3 4 13.1 feel compelled to arrange cans or boxes of food on my kitchen shelves in a specific way. 0 1 2 3 4 14. When I see that my belongings are out of place, I become anxious until I can arrange them properly. 0 1 2 3 4 15.1 feel compelled to arrange objects so that they are balanced and evenly spaced. 0 1 2 3 4 16.1 feel calm/at ease only when my surroundings are neat and tidy. 0 1 2 3 4 17. Even when my home is messy, I keep things organized according to a specific set of rules. 0 1 2 3 4 18. Things in my home have a proper and exact place. 0 1 2 3 4 19.1 cannot concentrate unless things are in the right 0 1 2 3 4 place. 20.1 don't like to disturb objects once they are properly arranged. 0 1 2 3 4