Appraisal Bias and Emotion Dispositions Are Risk Factors for Depression and Generalized Anxiety: Empirical Evidence

Appraisal theory of emotion predicts that appraisal biases may generate stable emotion dispositions, which can ultimately lead to affective disorders. One example is the habitual underestimation of one’s potential to cope with adverse events, which favors frequent experiences of sadness and worry and therefore increases the risk for development of depression and generalized anxiety disorders. To examine the relationships between these variables as potential risk factors, in Study 1, we used appraisal and emotion questions in the Swiss Household Panel (SHP), a nationwide representative sample, and analyzed data for N = 4,859 participants in one annual survey wave (Wave 14, SHP 2012) via theory-based hierarchical regressions. Path analysis of the nomological network linking frequent experiences of depression and anxiety to the emotion dispositions of sadness and worry, and measures of perceived coping potential (appraisal bias) supports the theoretical predictions and further identifies the effects of important background variables such as personality, motivation, and life events. Discriminant analysis shows that these predictors allow correct classification of close to 70% of the participants with elevated risk. In Study 2, we used established validated instruments to assess the risk for depression and anxiety disorders, as well as a recently validated scenario method to assess appraisal bias and emotion disposition in a survey with N = 152 students. The results correspond to the theoretical predictions and largely confirm the findings with the household survey. The results of both studies demonstrate the utility of using current emotion theory to provide new vistas for research on risk factors for affective disorders and to inform the development of appropriate interventions to reduce the level of risk.


Implication assessment
Note: All rated on a 10-point scale from "never" to "always"   Figure A1

Path diagram showing the results (selected predictors) of three hierarchical regression analyses: 1) Appraisal bias "Coping potential" on background variable classes a-d; 2) "Sadness/Worry" emotion disposition on appraisal bias and a-d background variable classes; and 3) Dependent variable (Depression/Anxiety risk factor for mood disorder on emotion disposition, appraisal bias, and a-d background variable classes)
. Paths with standardized beta weights (all ps < .001) > .10 are shown and marked with the respective beta weights. Variance explained (R 2 ) is bolded.

Appendix 1 -Study 1
Excerpt from an unpublished Technical report: Assessing the psychometric properties of the psychological scales in the Swiss Household Panel

Development of a new Sense of Control Scale from existing items in the SHP
Analyses were performed on a subsample aged between 18 to 65 year of the 14th wave (2012)  Theoretical background Sense of control/coping potential is defined as the extent to which individuals believe that their destiny is controlled by themselves and their own decisions or by external forces over which they do not have any power. Individuals who believe more strongly that they control their own destiny are more likely to develop a feeling of self-efficacy that is beneficial for their overall subjective well-being. This is of major importance to assess the sense of control in nationwide longitudinal survey because it gauge the extent to which individual adapt to predictable or unpredictable life events or life transitions as well as the capacity individuals have to adapt to changing living conditions that are particularly measured in such survey. In addition, such measure of adaptation allows providing information in what way individuals adapt or whether they experience lack of adaptation capabilities. These lacks of resources can lead to certain forms of vulnerabilities in some subpopulation or subgroups that are important to target.
Six items measure a very general personal perception of the self. Some items measure in how far respondents believe that their destiny is controlled by themselves and their own decisions or by external forces over which they do not have any power. Individuals who believe more strongly that they control their own destiny are more likely to develop a feeling of self-efficacy. In addition, several items assessed the Personal mastery as well as the Perceived constraints. Three items come from Pearlin and Schooler (1978) and three items come from Lachman and Weaver (1998). Below the table sum up the items used in the SHP to assess one's control and coping potential.

Description of the items
An exhaustive presentation of the items, the formulation of the questions, the labels and name of the variables are displayed in tabular form below. According to Lachman and Weaver (1998), these items measure two dimensions of Sense of control and coping potential: Personal mastery and Perceived constraints.

Analyses
The model fit of this two component model was performed with the Lavaan program in the R package. As shown in Table 1  A second set of analyses was conducted based on a preliminary explanatory factor analysis and a theoretical perspective based on established concepts in the domain and the wording of the items. One item (P12C108) was removed because of the ambiguity of the formulation. This revised model postulates the following five theoretical subscales to measure sense of control and coping potential: 1. Determination: Doing everything set in my mind; Find a way to succeed 2. Self-efficacy: Capacity to overcome unexpected problem; Capacity to choose between two possibilities.

Lack of agency conviction:
Incapacity to make plans because of unpredictability; Little influence on life events 4. Power: What I want is in my hands; What will happen depends on me.

Lack of self-esteem:
Feeling of uselessness; Feeling of self-satisfaction; Feeling of being pushed in my life.  We combined the items in each subscale (mean values corrected for direction) to obtain 5 subscale variables for the control/coping construct which were used in study 1 for an exploration of the overall data structure. We computed a principal components analysis of these 5 subscale variables to obtain an indication of the underlying dimensionality. The analysis of the data in Study 1 yielded only a single factor with an Eigenvalue > 1 explaining 42.1% of the variance. We then computed the mean of the subscales loading highly (> .5) on this single factor, the determination, self-esteem, and power subscales, to create a composite appraisal bias variable of low self-perceived control/coping potential (inverted scale). A reliability analysis yielded a Cronbach's alpha of .66 for the mean of three combined subscale variables (based on seven individual items), indicating an acceptable level of reliability.

Appendix 2 -Study 2
Given the limitation of items imposed by the feasibility of the web survey conducted for Study 2, we reduced the number of 11 items in five subscales to assess control and coping ability beliefs to 5 items.
Based on a number of criteria, independently of the subscales identified earlier, we chose the following 5 items (controlcoping) for inclusion in the study: We computed two scale values, one based on the four items positively loading on the first factor and one using five items and reversing the value of the item loading on the second factor. As the two scales correlate to r=.97, we decided to use the homogeneous four-item scale for further analyses. A reliability analysis yielded Cronbach's alpha = .66 for these four items.