2.5. Calidad de software y análisis de los modelos de calidad
2.5.7. Estándar ISO/IEC 9126
The first aim of this study was to examine the relationships between perfectionism dimensions and pathological worry in adults with elevated
perfectionism and GAD who presented for perfectionism treatment. Dimensions of perfectionism as measured by CM, PS and CPQ scores each significantly predicted
pathological worry and remained significant after accounting for gender, anxiety and depression. These findings uniquely contribute to the literature, as previous research has not examined the associations between perfectionism dimensions and
pathological worry in individuals with GAD after controlling for gender, anxiety and depression.
The finding of CM being significantly related to pathological worry after accounting for anxiety and depression supports theoretical links between concern over mistakes and worry (Flett, Madorsky, et al., 2002). It also concurs with the findings of previous studies utilising student samples (Kawamura et al., 2001; Santanello & Gardner, 2007; Stoeber & Joormann, 2001). The current finding provides evidence that the relationship between CM and pathological worry in individuals with GAD is important of its own accord and not an artefact of the relationships CM has with anxiety and depression (Antony, Purdon, et al., 1998; Enns et al., 2001; Kawamura et al., 2001). Importantly, the current finding generalises to clinical populations of individuals with elevated perfectionism and GAD.
Furthermore, this is the first study to find that PS has a significant relationship with pathological worry in a clinical sample. This finding provides empirical support for the theorised links between elevated personal standards and pathological worry (Flett, Madorsky, et al., 2002; Pratt et al., 1997). As this finding emerged in a clinical sample but not in student samples, this may reflect PS being differentially associated with pathological worry in non-clinical and clinical samples (Kawamura et al., 2001; Santanello & Gardner, 2007; Stoeber & Joormann, 2001). The current finding is also noteworthy as it is only the third finding of PS being significantly related to anxiety symptoms in a clinical sample. While Iketani et al.
(2002a; 2002b) found that PS was related to panic disorder with agoraphobia symptoms, most studies have reported that PS was not significantly related to anxiety symptoms (Egan et al., 2011).
The current finding of PS being significantly related to pathological worry in individuals with GAD indicates that PS is not a purely positive construct as has been put forward by Stoeber and Otto (2006). Stoeber and Otto (2006) argued that
particularly when maladaptive components of perfectionism were controlled (CM, DA, PE, PC, SPP), the positive achievement striving dimension of perfectionism (PS, O, SOP) is a positive construct associated with positive characteristics.
However, in the current study, CM was statistically controlled by being entered as a predictor in the hierarchical linear regression analysis, but the significant relationship between PS and pathological worry still emerged. This significant association also remained after controlling for gender, depression and anxiety. This suggested that a unique relationship between PS and pathological worry was present that cannot be attributed to the relationship reported between positive achievement striving and depression (Hewitt & Flett, 1993). The current findings add to those of Iketani et al. (2002a; 2002b) in demonstrating that PS has a significant relationship with certain anxiety symptoms in clinical samples. Together with studies showing PS to be consistently associated with eating disorder symptoms (Bardone-Cone et al., 2007; Bardone-Cone et al., 2008), as well as studies showing that positive achievement striving components such as SOP are associated with depression (Hewitt & Flett, 1993; Norman et al., 1998), this adds support for positive achievement striving perfectionism being maladaptive in some circumstances (Egan et al., 2012).
The significant association between CPQ scores and pathological worry further contributes to the literature as it demonstrates for the first time that clinical
perfectionism is significantly associated with pathological worry in a clinical sample. This relationship remained after controlling for gender, anxiety and depression, which supports the relationship between clinical perfectionism and pathological worry being important of its own accord. This finding is consistent with the theoretical links between clinical perfectionism and adverse consequences such as worry, anxiety and stress (Shafran et al. 2002; Shafran et al., 2010) and provides support for the validity of the CPQ as a measure of clinical perfectionism (Fairburn et al., 2003b). This finding also extends those of Chang and Sanna (2012) by
highlighting that clinical perfectionism is not just significantly related to anxiety and stress, but also has a unique relationship with pathological worry and that this relationship is present in a clinical sample with GAD.
Collectively, the findings of perfectionism as measured by CM, PS and CPQ scores significantly predicting pathological worry after controlling for gender,
anxiety and depression add to the growing body of research supporting perfectionism being a process which occurs across psychological disorders (Egan et al., 2011). These findings provide a rationale for Study II of this thesis and future studies to investigate whether perfectionism treatments can decrease GAD symptoms in addition to the symptoms of other psychological disorders. The current findings also provide justification for future studies to investigate whether adding a treatment module targeting perfectionist standards and concern over mistakes can improve the efficacy of psychological treatments based on current models of GAD (Dugas et al., 1998; Wells, 1995; 1999).
In the current study, a significant correlation between DA and pathological worry did not emerge. This differs from previous studies that found this relationship in student samples (Kawamura et al., 2001; Santanello & Gardner, 2007; Stoeber &
Joormann, 2001). There are many reasons why these non-significant findings may have occurred. First, the previous studies examining the relationships between Frost et al.’s (1990) perfectionism dimensions and pathological worry only explored the associations between composite constructs of perfectionism (i.e., MEC, CM+DA) and pathological worry (Kawamura et al., 2001; Santanello & Gardner, 2007; Stoeber & Joormann, 2001). It is therefore possible that the significant association between the composite construct of perfectionism and pathological worry reported in student samples is an artefact of the significant association between CM and
pathological worry (Kawamura et al., 2001; Santanello & Gardner, 2007; Stoeber & Joormann, 2001). If this is the case, this would be congruent with the current
findings of CM being a significant predictor of pathological worry. A second possibility is that due to all adults in this sample having elevated perfectionism, the data may have been affected by restriction of range on the study variables, which may have weakened the associations between the predictors and the dependent variable (Tabachnick & Fidell, 2007). A third possibility is that due to the sample size, a Type II error may have occurred where the study had insufficient power to detect a relationship between DA and pathological worry in this sample,even if one is present in the population. The non-significant zero-order correlation then resulted in DA being excluded from the first hierarchical linear regression model. A Type II error is quite likely given that the zero-order correlation between DA and
pathological worry had a significance value of p = .07 (Tabachnick & Fidell, 2007). Future studies need to examine the relationships between these perfectionism dimensions and pathological worry in a larger sample of individuals with GAD and include participants with a wider range of perfectionism scores to overcome the limitations of the current study.
The second aim of this study was to examine whether perfectionism
dimensions significantly predicted a principal diagnosis of GAD in a clinical sample of adults with elevated perfectionism and a range of diagnoses who presented for perfectionism treatment. The findings revealed that DA was a significant positive predictor of a principal diagnosis of GAD. Interestingly, CM, PS and CPQ scores were not significant predictors of a principal diagnosis of GAD. The significant predictive utility of DA is consistent with this perfectionism dimension being an important construct in GAD, which again supports perfectionism being a
transdiagnostic process (Egan et al., 2011). As DA has been found to be a significant construct in obsessive-compulsive disorder and social phobia (Antony, Purdon, et al., 1998; Rheaume et al., 1995), the current finding of DA being a significant positive predictor of a principal diagnosis of GAD may indicate a cognitive process that is shared across these disorders. Nevertheless, the specific function of DA in a principal diagnosis of GAD requires clarification in future research as it is unusual that DA did not significantly predict pathological worry, yet it was a significant predictor of a principal diagnosis of GAD. These findings can be reconciled by again proposing that the non-significant correlation between DA and pathological worry arose due to a Type II error (Tabachnick & Fidell, 2007); however, future research is needed to clarify this.
It is also unusual that CM, PS and CPQ scores were not significantly related to a principal diagnosis of GAD given that each of these dimensions of perfectionism significantly predicted pathological worry.One possibility for the non-significant zero-order correlations between these perfectionism dimensions and pathological worry is a restriction of range on the study variables. Additionally, as 71 per cent of the adults in this sample met the criteria for a principal diagnosis of GAD, the
sample may not have been sufficiently diverse for these variables to arise as
significant predictors (Tabachnick & Fidell, 2007). Future investigation is required using samples that have greater diversity.
Some additional limitations of this study require discussion. First, as all participants had elevated perfectionism, findings of CM, PS and CPQ scores being significant predictors of pathological worry can only be generalised to adults with elevated perfectionism and GAD; whereas the finding of DA being a significant predictor of a principal diagnosis of GAD can only be generalised to adults with elevated perfectionism. Additionally, this study did not incorporate a non-clinical control group, which prevented the perfectionism levels of individuals with GAD from being compared to those of non-clinical controls. Future research needs to include a sample with GAD and a sample of non-clinical controls. Furthermore, while this study controlled for gender, anxiety and depression, it did not control for other constructs associated with pathological worry and GAD, such as intolerance of uncertainty and meta-beliefs about worry (Dugas et al., 1997; Dugas et al., 1998; LaDouceur et al., 1998; LaDouceur, 2000; Wells, 1995; 1999; Wells & Carter, 1999). Future studies in clinical samples need to explore whether perfectionism dimensions account for additional variance in GAD symptomatology to that accounted for by these established constructs (Dugas et al., 1998; Wells, 1995; 1999). Finally, the cross-sectional design of this study prevents inferences about the directions of effect. Future research needs to utilise prospective designs to examine whether dimensions of perfectionism temporally precede pathological worry and GAD.
Nonetheless, the current study has still provided a significant contribution to the literature by demonstrating that significant relationships exist between CM, PS,
CPQ scores and pathological worry in individuals with elevated perfectionism and GAD. Additionally, this study has revealed that DA can significantly predict a principal diagnosis of GAD in a clinical sample with elevated perfectionism and a range of diagnoses. These findings support perfectionism being a transdiagnostic process and provide a rationale for Study II of this thesis and other studies to
investigate whether perfectionism interventions can reduce the symptoms of GAD in addition to the symptoms of other psychological disorders (Bieling, Summerfeldt, et al., 2004; Egan et al., 2011).