Discussion
The aim of the present work was to develop a questionnaire specifically for assessing myths in clinical psychology and psychotherapy and to test its factor structure, psychometric quality, as well as correlations with personality traits on two independent samples (general population, psychology students).
According to the principal component analysis, the FMKPP comprises three factors that can be described in terms of content with the subscales “myths about the effectiveness of psychotherapy,” “myths about mental disorders/processes in psychotherapy,” and “myths concerning the functioning of memory.” This structure shows a clear overlap with the differentiation of myths in the American context of psychotherapy discussed in the review article by Lilienfeld et al. [2013], with regard to “myths about effective interventions,” “myths about memory,” and “myths about the meaning of early experiences.” The two samples were mostly the same with respect to the maximum loadings on the respective factors. The subscales correlated weakly to moderately (r = 0.14–0.28, p ≤ 0.007; Table 3). As with previous studies [Swami et al., 2012; Kanning et al., 2013, 2014; Swami et al., 2016], this finding may indicate that there are differences in myth acceptance depending on the subject area. Since the development and testing of the questionnaire were performed on the same sample, it is also possible that these correlations were overestimated and are in the low range.
Out of all 20 misconceptions, the myth “Some people have a real photographic memory” (item 25) was most consistently accepted in both samples (mean = 4.72/4.53, SD = 1.14/1.13). Like a study in the USA [Patihis et al., 2014], which found 87.7% agreement with this misconception about the functioning of memory, the present study showed agreement rates (at least “somewhat agree”) of 89.2% (S1) or 83.7% (S2). The sociodemographic data of the student sample examined here (S2) is most comparable to the sample of Patihis et al. [2014] (students, 75% female, mean = 20 years). A high level of myth acceptance, as well as varying levels depending on the myth, can be explained in different ways, e.g., by the person’s experiences and attitudes, the type of studies or professional activity, or the myth’s popularity and media coverage [Lewandowsky et al., 2012; Lilienfeld et al., 2013; Appel and Schreiner, 2014]. The fact that “myths concerning the functioning of memory” are particularly widespread could be explained by the immediate concern and everyday relevance of memory processes, but also by the constant high media presence and controversy about the connection between memory and psychopathological processes [Crews, 1995; Appel and Schreiner, 2014]. Although misconceptions were found to be persistent [Lilienfeld et al., 2013; Otgaar et al., 2019], the present study found, consistent with previous research, that the mediation of relevant expertise – e.g., for those studying psychology – is associated with significantly lower myth acceptance [Standing and Huber, 2003; Taylor and Kowalski, 2004; Furnham and Hughes, 2014].
The item analysis indicated satisfactory to good psychometric properties. Three items showed low discriminative power in one sample each (S1: item 17; S2: items 6 and 15) (rit < 0.30) [Lienert and Raatz, 1994]. Regarding item 17 (“A very high intelligence quotient raises the risk of mental disorders”), from a scientific point of view there are also isolated divergent findings. In a study with over 10,000 adolescents [Keyes et al., 2017], the presence of a mental disorder showed no correlation or a negative one with IQ (among the disorders were ADHD, substance use disorder, and various anxiety disorders; the exception was a positive correlation with depression). With regard to the psychopathological severity across all disorders, greater severity was associated with lower fluid intelligence [Keyes et al., 2017]. A recent study [Karpinski et al., 2018] found, however, that a very high IQ (above the 98th percentile) is associated with a higher rate of ADHD, autism spectrum disorders, and anxiety and depressive disorders; but interpretation of the result should take into consideration the study’s significant methodological limitations (e.g., self-reported diagnoses), and no direct (causal) connection (“intelligence increases the risk”) has been demonstrated. The fact that item 17 cannot currently be indisputably evaluated scientifically in the form of such a sweeping statement could also explain the low factor loadings and the low discriminative power of the item, which should be checked in further studies and might lead to exclusion of the item.
Item 6 (“Disclosing the diagnosis ... harms the therapeutic alliance”) showed the lowest agreement of all the statements, especially among psychology students, with comparably high variance. Furthermore, the explanation of item 6 might be more heavily corrected and internalized by those studying psychology, due to its practical relevance compared to the other statements. With regard to item 15 (“Patients with a mental disorder, e.g., schizophrenia, are often violent”), it was striking that, contrary to the other items (except item 22), the student sample on average agreed more often than did participants from the general population. In addition, there was an unexpected positive correlation between myth acceptance and the number of semesters of study (r = 0.26, p < 0.001), with a comparatively low average number of semesters overall (mean = 3.6, SD = 2.2), which could indicate that the students had acquired some knowledge of mental disorders, which, however, could also have led to assumptions that are incorrect from a scientific standpoint.
Reliabilities were within the acceptable range for the FMKPP total scale and for the “myths about the effectiveness of psychotherapy” subscale (ω = 0.70–0.75). The subscales “myths about mental disorders/processes in psychotherapy” (ω = 0.61/0.50) and “myths concerning the functioning of memory” (ω = 0.69/0.63) yielded low values and should be checked in further studies. The low reliability of the subscale “myths about mental disorders/processes in psychotherapy” in S2 can be explained, among other things, by the fact that the two ambiguous items (6 and 13) were assigned to the second factor, to the disadvantage of S2 (content fit and higher loadings in S1). This factor also appeared to be less consistent in content, since it covers myths about both characteristics of mental disorders and processes of psychotherapy (therapeutic alliance, side effects).
Regarding correlations with personality traits, our findings indicate a lack of connection with “Openness.” The association with personality traits may also be determined by the specific content of the myths. While Swami et al. [2016] identified a negative relationship between “Openness” and a scientific myth (r= –0.21, p < 0.001), Swami et al. [2012] found no relationship to everyday (psychological) myths. Moreover, the operationalization of the construct “Openness” could also explain the differences between our study’s findings and those of Swami et al. [2016]. While we used the BFI short version, which assesses the imaginative and aesthetic components of Openness, Swami et al. [2016] studied “Openness” with the BFI more comprehensively, including the intellectual curiosity component, which was also used to explain the negative relationship with acceptance of the myth. It should also be further investigated to what extent the content and/or, for example, the form of presentation is important (more detailed description of a scenario [Swami et al., 2016]).
Consistent with the findings of Patihis et al. [2014], this first psychometric test of the FMKPP presents positive correlations between absorption (the disposition to pay more attention to sensory/imaginative experiences) and the total score as well as the subscale “myths concerning the functioning of memory” of the FMKPP (r = 0.18–0.45, p < 0.003). In previous studies, absorption was related, for one thing, to suggestibility [Eisen and Carlson, 1998], which might explain an association with greater myth acceptance in general. On the other hand, people with a higher level of absorption had greater hypnotic responsivity and more frequent reports of childhood abuse [Eisen and Carlson, 1998; Roche and McConkey, 1990]. Both areas are directly related to the factor “myths concerning the functioning of memory” of the FMKPP.
Also as expected, positive correlations were found between the FMKPP and the UIS (total FMKPP: r = 0.20–0.21, p < 0.018), which could indicate that people with higher myth acceptance have a greater need for safety and control over events and the future. Comparison of the subscales showed a differentiated picture. An inability to act associated with intolerance of uncertainty (e.g., “It paralyzes me to have to act,” low self-esteem) was, for example, positively correlated with “myths about mental disorders/processes of psychotherapy,” which could be explained by the overlapping content of the two subscales (self-esteem issues, causes of mental problems). Considering the study design (construction and validation for the same sample in each case), the (already weak) correlations may have been overestimated so that the correlations must be interpreted carefully and should only be regarded as preliminary indications. To be able to derive more reliable statements, including on (different) functions of myths, it will be necessary to conduct further studies with more representative samples and investigation of additional variables (e.g., fearfulness or situational use of myths).
Posts
