'Influence of Dark Triad Traits on Health Behaviour in Affective Disorders' (AsPredicted #77913)
Author(s) This pre-registration is currently anonymous to enable blind peer-review. It has 4 authors.
Pre-registered on 10/25/2021 07:19 AM (PT)
1) Have any data been collected for this study already? No, no data have been collected for this study yet.
2) What's the main question being asked or hypothesis being tested in this study? The main question of this study is:
Is there an association between affective disorders (versus mentally healthy individuals) and health behaviours (substance abuse, diet, physical activity, use of health services, sleep, mental health), and is this association moderated by the individual manifestation of the Dark Triad traits (psychopathy, Machiavellianism, narcissism)?
The following hypotheses (H) will be investigated:
H1: We expect individuals with affective disorders to show less protective health behaviours compared to mentally healthy individuals. Further, we expect this relationship to be significant positively moderated by the characteristics of the Dark Triad, i.e. it is stronger for individuals with high scores in the Dark Triad compared to individuals with low scores in the Dark Triad.
H2: Individuals with affective disorders show more substance use compared to mentally healthy individuals. This relationship is positively moderated by the Dark Triad traits.
H3: Individuals with affective disorders show less physical activity compared to mentally healthy individuals. This relationship is positively moderated by the Dark Triad traits.
H4: Individuals with affective disorders show poorer dietary behaviour compared to mentally healthy individuals. This relationship is positively moderated by the Dark Triad traits.
H6: Individuals with affective disorders show poorer sleep quality compared to mentally healthy individuals. This relationship is positively moderated by the Dark Triad traits.
H7: Individuals with affective disorders show less use of health services compared to mentally healthy individuals. This relationship is positively moderated by the Dark Triad traits.
H8: Individuals with affective disorders show poorer mental health compared to mentally healthy individuals. This relationship is positively moderated by the Dark Triad traits.
3) Describe the key dependent variable(s) specifying how they will be measured. This study is essentially correlative. All key variables are listed below.
Substance use, physical activity, diet and use of health services. These health behaviours will be measured with the German version of Multidimensional Health Behaviour Inventory (MHBI; Kulbok et al., 1999), consisting of 36 items. Subjects answer on a five-point Likert scale, ranging from (1) = "never" to (5) = "always".
Diet. Dietary behaviour will be measured with the German version of the Mediterranean Diet Score (MEDAS; Hebestreit et al., 2017), which consists of 14 items. Items are either presented dichotomously or in frequencies of the eating habit asked (e.g., 3-4x per week).
Sleep. Sleeping behaviour will be measured with the German version of Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989), which consists of 19 items. Subjects answer on a four-point Likert scale, ranging from (1) „not during the past month" to (4) „three or more times a week".
Perceived Sensitivity to Medicines. Perceived sensitivity to medicines as a part of use of health services will be measured with the German version of the Perceived Sensitivity to Medicines Scale (PSM-Scale; Horne et al., 2013), which consists of 5 items. Subjects answer on a five-point Likert scale, ranging from (1) = "strongly disagree" to (5) = "strongly agree".
Vaccination attitude. Vaccination attitudes as a part of use of health services will be measured with the German version of Vaccination Attitudes Examination Scale (VAX-Scale; Martin & Petrie, 2017), which consists of 12 items. Subjects answer on a six-point Likert scale, ranging from (1) = "strongly disagree" to (6) = "strongly agree".
Mental health. Mental health status will be measured with the German version of the Patient Health Questionnaire (PHQ-9; Löwe et al., 2004), which consists of 9 items. Subjects answer on a four-point Likert scale, ranging from (0) = "not at all" to (3) = "almost everyday".
Dark Triad. The Dark Triad will be measured with the German version of the Short Dark Triad questionnaire (SD3; Jones & Paulhus, 2014), which consists of 27 items. Subjects answer on a five-point Likert scale, ranging from (1) = "strongly disagree" to (5) = "strongly agree".
For the questionnaires, responses will be averaged over items to build (sub-) scale scores. Specifically, a composite score will be calculated for the Dark Triad traits.
4) How many and which conditions will participants be assigned to? Participants will be assigned to two conditions (affective disorders vs. mentally healthy controls).
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. Before we examine the hypotheses, we will compute bivariate correlations between all items in this study. All correlation analyses will be calculated with Bonferroni adjustments for the family of predictors (α = .05 divided by 3 (Dark Triad traits) and 10 (health behaviours)). To examine the causal relationship between the Dark Triad traits, health behaviours and affective disorders (H1), a structural equation model will be calculated. Further, single moderation analyses will be administered to determine the effects of affective disorders and the single traits of the Dark Triad on each health behaviour domain (H2-H8). The range in which moderators show significant interactions is determined using Johnson-Neyman intervals. All analyses will also differentiate by type of affective disorder (bipolar disorder vs. depression). We will use 95% BCa bootstrapping confidence intervals based on 2000 samples for the interpretation of affected analyses if assumptions regarding normality, homoscedasticity, or (plausible) outliers are violated. All hypotheses will be tested two-tailed.
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. Only people within the age of 18-90 will be tested. Outliers will be considered in further analyses. Data will only be excluded if they are not plausible or realistic (e.g. participants skipped parts of the questionnaires). Plausible data (e.g. high narcissism scores) will not be excluded.
7) How many observations will be collected or what will determine sample size? No need to justify decision, but be precise about exactly how the number will be determined. Due to a lack of related research, we assume a small effect size (path coefficient = 0.1). With an α-level of 5%, a power of 80%, 10 manifest variables, and one latent construct (health behaviour), a power analysis for structural equation models yields a sample size of 1100. We expect 15% of the participants to drop out of the study, thus the planned number of study participants will be 1265 (1100 + 15% drop-outs). However, if more data are available, they will also be included in all analyses.
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?)
