Validation of the Self - Discrepancies Scale ( S - DS ) A tool to investigate the Self in clinical and research settings

instrument, especially with regards to indices related to the socially prescribed self. A Self-Discrepancies Scale 5 Convergence with other measures was particularly strong for the first, idiographic questionnaire. The authors have consequently advised researchers to rely on the idiographic measures of self-discrepancies in clinical and personality research. Another attempt to overcome the limits of Higgins’s Self-discrepancy questionnaire has been formulated by Hardin and Lakin (2009): the Integrated Self-Discrepancy Index (ISDI). The ISDI assesses ideal and socially prescribed self-discrepancies from the participants’ own standpoint and from the standpoint participants attribute to a significant other. It comprises two components. In the idiographic component, participants are requested to list up to five characteristics that best describe each of the four target selves (ideal and socially prescribed selves, from either their own standpoint or the standpoint of a significant other). After generating traits for each self, in the nomothetic component, participants are shown a list of 100 traits from which they can choose to complete (if fewer than five attributes were listed) or modify their lists. Finally, participants have to rate each trait on a 5-point scale, indicating the extent to which each of the traits listed actually describes their ideal or socially prescribed self. Self-discrepancy scores are the average of the ratings of the five attributes generated for each of the self-states. Hardin and Lakin reported two studies conducted on undergraduate samples showing good convergent validity of the ISDI. In addition, they observed that socially prescribed self-discrepancies were more specifically related to agitation but not to dejection (after partialling out other selfdiscrepancies), whereas ideal self-discrepancies were uniquely related to dejection but not to agitation (also after partialling out other self-discrepancies). Given the qualities reported by Hardin and Lakin (2009), the ISDI seems very promising: It captures participants’ idiographic self-discrepancies, it is accessible to a diverse sample of participants (as it does not require an extended vocabulary about personality traits), A Self-Discrepancies Scale 6 and it can be objectively scored. Yet, it leaves unanswered at least two questions. First, it is unclear how the averaged discrepancy scores, stemming from the individual trait ratings, relate to the overall phenomenological feeling from the individual of that very discrepancy. In other words, how would the scores of the ISDI relate to an abstract measure such as that proposed by Watson, Bryant and Thrash (2010)? A second question is the relation between the perceived discrepancy and the psychological distress it may induce. Indeed, some individuals might perceive large self-discrepancies, but accept them and develop healthy selfcompassion (Neff, 2009), while others might experience great distress to similar or even weaker self-discrepancies. One might suspect that the relation between self-discrepancy and depression or anxiety might be stronger for the distress elicited by the discrepancy than for the extent of the discrepancy itself. Carver and Scheier (1990) noted that self-discrepancies could lead to anxiety and depression under two conditions. One is that the rate of reduction of the discrepancy is perceived as too slow by the individual. The second is goal conflict. Goal conflict can occur between more than one ideal self-state, or between an ideal and a socially prescribed self-state. Kelly et al. (2015) has even proposed that it is this discrepancy between ideals, or between ideals and social prescriptions, that maintains actual-ideal discrepancies. The common point of these two conditions is that distress results from the discrepancies. With these considerations in mind, we endeavor to develop a self-discrepancies questionnaire that would merge the qualities and advantages of those developed by Watson, Bryant and Thrash (2010) and by Hardin and Lakin (2009). In addition, we desired to have a questionnaire that would be practical and useful in clinical settings, i.e., informative about the idiosyncratic self-representations of the clients and easy (and short) to administer. We created the SelfDiscrepancies Scale (S-DS), by simplifying the ISDI and by adding scales evaluating the overall feeling of discrepancy and the associated distress. In order to simplify the ISDI, A Self-Discrepancies Scale 7 we deleted the assessment of the discrepancies from the perspective of another person, as it does not seem to add much predictive power in terms of emotional vulnerability, such as depression or anxiety. We also directly presented a non-restrictive list of possible traits directly at the start of the questionnaire (not at a second step, like in the ISDI), as previous clinical experience has shown that this order of presentation made the instructions clearer for the participants and reduced the completion time. Another way in which the S-DS is different from the ISDI is that each self in the S-DS is defined not only in terms of the desired traits that best characterize it, but also with the undesired traits that negatively define it (i.e., the traits in opposition to the self in question). Again, our clinical experience suggests that negative formulation might be particularly relevant for people experiencing self-dejection (Leary & Baumeister, 2000). Concretely, the S-DS consists of two parts, one that defines the ideal self and that estimates the discrepancy between the ideal self and the actual self, and one that defines the socially prescribed self and that estimates the discrepancy between the socially prescribed self and the actual self. The composition of the S-DS is fully described in the method section. The aims of the present study are to establish the psychometric characteristics of the SDS as well as to investigate the relations between average scores and global judgments of discrepancies and distress. In addition, we investigated whether the distress elicited by a discrepancy between two self-states is a better predictor of depression or anxiety than the extent of that discrepancy itself. Finally, we desired to examine the discriminant validity of the S-DS in a clinical sample. Therefore, we completed a validation study in a large community sample. We complemented that study with an additional data collection in a clinical sample composed of clients seeking psychotherapy help for diagnosed mood or