Perceived Causal Relationship (PCR) between stress and control for the trait Neuroticism : a preliminary study approaching a new way to look at Post Traumatic Stress Disorder (PTSD)

(1)

1

Bachelor thesis

Perceived Causal Relationships (PCR) between stress and control for the trait Neuroticism: a preliminary study approaching a new way to look at P ost Traumatic Stress Disorder (PTSD)

Name: Bonita van Es Studentnumber: 10745491 Mentor: Henk Cremers Year: 2017

Wordcount abstract: 220 Wordcount: 5484

(2)

2

Index

Abstract ... 3 PTSD and Neuroticism... 4 Coping and PCR ... 5 Method ... 7 Participants ... 7 Material ... 7 Procedure ... 7 Statistical Analysis ... 7 Results ... 9 Discussion ... 13 Closing Comments ... 15 References ... 16 Appendix ... 18 I. Adjusted PSS ... 18 II. PCR questionnaire ... 20

III. Factor Analysis PSS 2016-2017 ... 25

IV. Reliability ... 26

V. Factor analysis PSS PCR 2016-2017 ... 27

VI. Reliability ... 28

VII. Factor analysis PCR 2016-2017 ... 29

(3)

3

Abstract

Post-traumatic stress disorder (PTSD) is by definition a disorder of stress. While recovery rates for current therapies are good, little is known about the processes that mediate this change. This

study focuses on a possible patient characteristic for PTSD which could help improve therapy outcome, namely, perceived causal relations (PCR) between stress and control, and a vulnerability factor for PTSD, neuroticism. College undergraduates completed both an adjusted

perceived stress scale (PSS) and a PCR questionnaire to measure their level of stress and their PCR between stress and control. Neuroticism (N), a trait associated with PTSD, both as a risk factor and a moderator, was measured using the Dutch instrument Five Personality Factors Test

(5PFT). Based on the theory of network analysis, the in and out degree for control was calculated and plotted in a network with stress. Compared with low-N individuals, high-N

individuals perceived more stress and reported PCR flowing from stress to control. This suggests that control is perceived to be affected by stress. Literature points in the other direction, that stress is observed to be affected by control. This disparity between observed and perceived could be a lead for unravelling the individual characteristics which influence therapy. Instead of focussing on what is observed, therapy could be tailored to the individual, based on

(4)

4

Perceived Causal Relationships (PCR) between stress

and control for the trait Neuroticism: a preliminary

study approaching a new way to look at PTSD

Post-traumatic Stress Disorder ( PTSD) is an illness of both psyche and body and by default a disorder of stress. Symptoms may include intrusive images, traumatic nightmares, exaggerated startle responses and self-destructive behaviour ( DSM 5, 2016). These symptoms are also visible on personality questionnaires. People with PTSD score distinctively high on neuroticism (Jaksie, Brajkovic, Ivezic, Topic and Jakovljevic, 2012). Therefore neuroticism can be considered a marker for PTSD. Both on a behavioural and biological level is PTSD linked to stress. Alterations in brain structures and pathways responsible for a stress response and memory are found in PTSD patients (McFarlane, Atchison & Yehuda 1997, Kozlovsky, Matar, Kaplan, Kotler, Zohar & Cohen 2007) resulting in, but not limited to: heightened levels of stress sensitivity and increased feelings of stress. Patients often describe a perceived causal

relationship between their symptoms. Intrusive images for example, are attributed as a cause of consecutive depressive symptoms rather than the depression being a stand-alone complex. How people view the relations between their symptoms influences not only the development of their disorder but also therapy outcome (Frewen, Schmittmann, Bringmann & Borsboom, 2013). For a disorder such as PTSD, which exists of many different symptom clusters, mapping the symptoms relations at an idiographic level could help improve clinical aid and quality of life.

In this article we propose a new vantage point for PTSD. We investigate the coherence of perceived causal relationships between stress and control and compare this to the dimension of neuroticism. We first introduce PTSD and neuroticism. Secondly we explain the theory behind coping and perceived causal relationships, and then derive our hypotheses concerning the PTSD specific PCR between stress and control

PTSD and Neuroticism

The life prevalence of PTSD varies between the 2.1%-2.7%, depending on gender, nationality and method of study (Atwoli, Stein, Koenen & McLaughlin, 2015- Perkonigg, Kessler, Storz & Wittchen, 2000- Stein,Walker, Hazen & Forde, 1997). PTSD exerts its influence in all assets of daily life. Patients who suffer from PTSD are more likely to have other psychiatric diagnoses, have higher rates of unhealthy behaviour and are more at risk for suicide. Moreover do they experience overall less productivity, impairment in relationships and difficulties in parenting (Holderman, 2009). Apart from the problems for the individual and his immediate environment, the costs for society are substantial as well. Holderman (2009) estimated the costs for PTSD veteran patients accessing the American healthcare system during the first two post-deployment years. The costs range from $5,904 until $10,298 per individual, excluding

additional costs from unhealthy behaviour, comorbid diagnoses and productivity loss. It is clear that PTSD should be considered a serious and expensive illness, both for the individual as for society.

Currently pharmacotherapy, Eye Movement Desensitisation and Reprocessing (EMDR) and (trauma specific) Cognitive Behavioural Therapy with exposure are the recommended options for treating PTSD (Trimbos, 2013). Psychotherapy is the preferred choice since it is slightly more effective and has a lower drop-out rate than pharmacotherapy (Trimbos 2013). Overall the direct effects of psychotherapy for PTSD are substantial. Large effect sizes (1.11) moderate drop-out rates (21.1%) and a 67% recovery rate post-treatment are reported (Bradley, Greene, Russ, Dutra & Westen 2005). Although these recovery rates exceed recovery rates of other disorders, it is far from perfect. The mechanisms explaining why the therapies work remain veiled. In order to increase the recovery rates and reduce the drop-out rates,

(5)

5

unravelling the characteristics that lead to change is a crucial step. Scholar Huibers (2015) proposes to look at individual patient characteristics rather than continuing the debate about existing mediators. Each process of change during a therapy differs per individual. Determining what patient factors contribute to a successful therapy outcome is an efficient way of increasing therapy effectiveness.

A useful patient characteristic could be personality type. It can give you information where to focus your therapy and help form expectations (Harkness & Lillienfeld, 1997) Moreover does personality type also influence how a disease progresses (Watson, Haviland, Greer, Davidson & Bliss, 1999). Apart from the personality disorders, no disorder has a specific personality type, rather some personality traits are common. Personality traits themselves don’t instigate behaviour, rather they predispose to certain thoughts, strategies and actions.

PTSD has been linked to the personality trait of neuroticism, which can be described as a mix of anger, anxiety and depression (Costa & McCrae, 1987). Neuroticism has been found to be both a predictor and moderator for PTSD and symptomatic aftermath after a trauma (Bramsen,

Dirkzwager & van der Ploeg,2000, Caska & Renshaw, 2011) It is both a risk factor for the development of PTSD as it is adaptive during times of stress. People who score high on

neuroticism tend to view life more pessimistic and are sooner afraid of something. This overall increased pessimism is adaptive when they have to make decisions in a negative mood state. Neurotic people make decision faster when they feel bad than when they feel good. Their observed control is better than less neurotic people during negative mood states(Tamir & Robinson, 2004). Yet it is unclear whether neurotic people also view themselves to be more capable and more in control. This a matter of coping and the perception of causality of your own actions.

Coping and PCR

While the underlying mechanisms of stress are the same for all of us, we differ in the way we perceive and deal with stress. The perception of stress is mediated partially by our

personality. Dealing with stress on the other hand is done through coping. Coping is a term used to summarise the different strategies we use to combat stress. Coping serves two functions: solving the presence of the stressor (problem-solving) and internally easing the tension caused by the treat ( emotion-focused) (Solomon, Mikulincer & Avitzur, 1988). Often people use a mix, the ratio varying according to how the problem at hand is being appraised. In general, situations with more controllability elicit problem-focused coping and situations with less controllability emotion-focused coping (Folkman and Lazarus, 1980).

The appraisal of situations and the distinction between a ‘stressor’ and a ‘normal’ event differ per person. Folkman (1984) describes two forms of appraisal, first and secondary. First appraisal of a stressor depends on situational factors (timing, novelty of the event),

commitments ( what is at stake?) and general believe about control ( internal or external locus). Secondary appraisal is an evaluation of the coping resources and options at hand. These are physical, social, psychological but also material assets. The appraisal of a situation entails the controllability that you feel over the situation.

How people determine the level of controllability depends not only on the problem at hand but also on how capable they view themselves to be. Individuals themselves perceive their behaviour and physiological/psychological symptoms as being causally related ( Frewen, Allen, Lanius & Neufeld, 2012). While these causal relations might not exist, the perception of them influences their consecutive thoughts and actions (Brogan & Hevey, 2009). Related to PTSD, the perception of control is an important factor influencing stress and coping.

Perceived Causal Relations (PCR)scaling is a technique which can be used to capture this relationship between the perception of control and stress/coping. By using PCR you can map the way the patient believes their symptoms influence each other (Frewen, Schmittmann,

Bringmann & Borsboom, 2013). For example if you would like to know if intrusive images (II) cause stress (S), you could ask questions such as “How much do you feel your problems with (II) cause your problems with (S)?” and vice versa. The idea is that the correlation between

(6)

6

frequency of (II) and the frequency of (S) should be a positive function of the PCR(II)(s) and PCR(s)(II). That is, the stronger the direction of one PCR, the higher the correlation between the frequencies of (II) and (S). Network-analyses can be used to visualise these connections.

Mapping the PCR between stress and control could be relevant for the treatment of PTSD. Specific PCR between stress and control could be linked to specific forms of coping. By mapping how people view stress ( and consecutively the relationship with control plus the selection of a specific coping style) you could tap into the core of the stress. If you could identify the PCR of stress and control for a patient, you could tailor the therapy to the specific individual needs. If the patient views lack of control as the causal factor of stress, you could help him gain more control over his life. If it is the other way around you could focus on relaxation therapy, for example.

Evidence of specific coping strategies for PTSD exist. Overall, patients who suffer from PTSD prefer to use emotional coping rather than problem solving coping ( Dudek & Kroniarek, 2003, Karsoft, Armour, Elklit & Solomon, 2015). Furthermore, chronic PTSD is also linked to a more internal locus of control ( Karsoft, Armour, Elklit & Solomon, 2015 ). However, as Solomon, Mikulincer and Avitzur (1988) point out, this relationship is mediated by emotion-focused coping. The correlation between an internal locus of control and PTSD disappears when you control for emotion-focused coping. Emotion-focused coping itself is a broad subject.

In the introduction we have addressed two possible patient characteristics of PTSD , the PCR of stress and control based on coping style and neuroticism. With these patient

characteristics it might be possible to detect a specific coping style in the form of a PCR in a non-clinical group who match on personality traits with PTSD patients. Based on the above

description of coping styles we have hypothesised a PCR between stress and control specific for PTSD. In our research we’ve given a sample of first year psychology students a personality questionnaire (5PFT), an adaptation of the Perceived Stress Scale (PSS) by Cohen, Kamark and Mermelstein(1983) and a PCR questionnaire. Using network analysis, a network was for both high and low scoring N individuals, to illustrate the indegree and outdegree of both control and stress, as based on the PSS and PCR questions. We expected to find a positive relationship between neuroticism and stress , the higher the score on neuroticism, the higher the stress. Additionally, we expected that people who have a high out degree of control would score high on neuroticism Finally, we expected to find a PCR running from control to stress, which would mediate the relationship between neuroticism and stress. The score on neuroticism was measured by the 5PFT and served as an indication of PTSD.

In exploratory analyses have we looked at the mean difference between the in and out degree of control and the relationship with neuroticism. Furthermore have we analysed the in and out degree of control in two separate regression analyses instead of a multiple regression, because of expected multicollinearity.

(7)

7

Method

Participants

A total of 374 undergraduate psychology students from the University of Amsterdam (UvA) in Amsterdam, North-Holland, the Netherlands, participated in this study for partial mandatory course credits. They received no other compensation for their participation.

Material

All participants were asked to fill in both the adjusted Perceived Stress Scale (PSS) [appendix I] and a consequent Perceived Causal Relations (PCR) questionnaire [appendix II]. The adjusted PSS was based on the original PSS of Cohen, Kamarck and Mermelstein (1983). It contains ten items, which measure to what extent the subject had felt: control, upset, nervousness, anger and irritation during the last month. A reliability analysis was conducted to justify the use of the adjusted PSS, the scale showed good reliability (α.=81) [appendix IV].

The PCR questionnaire measures in 20 consecutive items how those five

abovementioned factors depend on and influence each other. It measures whether the subject perceives one of the factors to influence the others. For example, question 21 states “ Last month you were […] able to control your irritations and you had the feeling that you had everything under control. Explain in what way these factors influenced each other. Because I was irritated, I could control everything […]” The first blank would be filled with the answer of question 7 from the PSS. The second blank can then be filled in by choosing one of the five answers, ranging from ‘a lot less’ to ‘a lot more’. The technique of PCR is further explained in the article of Frewen, Allen, Lanius and Neufeld (2012).

The Dutch Five Personality Factors Test (5PFT) from Elshout and Akkerman (1973) was conducted to obtain the measure of neuroticism. The testst consists of 70 items, 14 for each of the five personality factors, extraversion, neuroticism, openness to experience, agreeableness and conscientiousness. Subjects rate whether these items describing a form of behaviour are applicable to them on a seven point Likert scale. Then the total scores per personality factor are obtained and transformed into a stanine score, ranging from 1 until 9. For example, having a total between 14 and 21 would be a stanine score of 1. These stanine scores serve as an indication of how well the personality factor is represented within the individual. The overall reliability of the 5PFT is good (α=.77-.81) (Smits, Dolan, Vorst, Wicherts & Timmerman, 2011).

Procedure

The students were required to fill in both questionnaires during the mandatory ‘test week’ at the University of Amsterdam. Students were placed together in a computer lab and asked to fill in all the questionnaires that were presented to them during a three hour lasting session. One group of students (n=91) took the tests during the testweek of 2016, the other group (n=283) during the testweek of 2017. The students made both the PSS and the PCR questionnaire tests on the computer, always in the same order. First the PSS and then the adjusted PCR questionnaire. The 5PFT was conducted somewhere during the three hour sessions at random interval.

Statistical Analysis

To test our first hypothesis that a higher stress score was related to a higher Neuroticism score we performed a multiple regression. The measurement of stress was obtained by adding all the responses of the PSS items up to a total. These analyses were based on the participants of both years who completed both the PSS items and the Big Five questionnaire (n=251).

Two mediation analyses were conducted to test our second hypothesis that the

relationship between stress and neuroticism is mediated by control. The factor stress was again measured by taking the total PSS score. The factor control was created by calculating the out and in degree of control. A separate analysis was run per degree in order to test whether control was perceived to influence stress or stress was perceived to influence control. Calculating out and in-degrees are ways of analysing a network ( Frewen, Schmittmann, Bringmann & Borsboom,

(8)

8

2013). When illustrated in a network, the out degree of a factor (x) can be seen as the total sum of all the arrows leaving the factor (x). Thus the out degree denotes the total direct effect from factor x to all other factors a,b,c, etc. With the in degree this is the exact opposite. The in degree of factor (x) is the total direct effect from all factors a,b,c etc. on factor (x). It is the total sum of all arrows entering factor (x). In the present study, scores for question 23 until 26 from the PCR questionnaire were added to find the out-degree of control. These questions were all formulated in such a way that the answer would indicate the effect of control on other factors. For example question 23: “Because I had everything […] under control, I felt […] upset. The more the total score of the out degree deviated from zero, the more influence control had on other emotional states. A positive score [1, 2] meant a positive effect of control on negative emotions, variance in control increased the intensity of negative emotional states. A negative score [-1, -2] meant a negative effect of control on negative emotions, variance in control decreased the intensity of negative emotional states. The in degree was created by adding the scores for questions 13, 17,21 and 30 of the PCR questionnaire together. These questions were all formulated in such a way that the answer would indicate the effect of one of the five negative emotional states from the PSS on the feeling of control. For example question 13: ”Because I was […] upset, I felt like I had everything […] under control”. The more the score deviated from zero, the more influenced control was by the negative emotional states. A positive score [1, 2] meant a positive effect of negative emotional states on control, feeling stressed increased the feeling of control. A negative score [-1, -2] meant a negative effect of negative emotional states on control, feeling stressed decreased the feeling of control.

Our third hypothesis entailed that there was a specific PCR between stress and control for people who score high on neuroticism. Namely a PCR between stress and control flowing from control to stress. In order to test this hypothesis we conducted another multiple

regression, between the in degree, out degree of control and neuroticism. SPSS (version 24) was used for all of the above analyses. For the mediation analyses an extra tool ( PROCESS) by

(9)

9

Results

Relationship between stress and Neuroticism

A linear multiple regression was conducted to establish the relationship between stress as measured by the total PSS score and the Big Five. Neuroticism predicted the presence of stress F (5,251)=14.51, p=<.001. Neuroticism explained 23% of the variance in stress. Overall did people who scored higher on Neuroticism also score higher on stress b=0.44, SE= 0.17

t(251)= 7,42, p= <.001. This relationship is illustrated in figure 1. There was no significant

relationship between stress and the other four personality factors: extraversion, openness, conscientiousness and agreeableness.

PCR analysis

A multiple regression was conducted to examine the relationship between neuroticism and the in/out degree of control. While the model including both predictors was significant,

F(2,249)=16,54, p=<.001, only the in degree of control was a valid predictor, b=-0.31, CI

95%[-0.34, -0.15], t(251)=-5,04, p=<.001. The more emotional stress had an influence on their perceived control, the more neurotic participants were [figure 2]. The out degree of control showed no relationship with neuroticism, p= 0.15.

Figure 2 relationship between mean in degree of control and neuroticism

Figure 1 Relationship between stress and Neuroticism Note: mean stress scores have been used

(10)

10

Mediation Analyses

Two mediation analyses as modelled by the PROCESS tool were used to establish the influence of control on the relationship between neuroticism and stress. There was a significant direct effect of neuroticism on stress, b=1, CI 95% [0.69, 1.31], t(251)= 6.32, p= <.001. People who scored higher neuroticism experienced more stress. Furthermore did an increase in neuroticism mean an increase in the out degree of control, b=0.29, CI 95% [0.08, 0.52],

t(251)=2.64, p= 0.009. For the people who were more neurotic, control had a greater effect on

negative emotions. Moreover, people who had a higher out degree of control, control having a greater effect on negative emotional states, experienced more stress, b=0.42, CI 95% [0.25, 0.59],

t(251)=4.90, p =<.001. Even after controlling for the out degree of control, there was still a

significant indirect effect of neuroticism on stress, b=0.12, CI 95% [0.03, 0.27].

There was a significant direct effect of neuroticism on stress, b=1, CI 95% [0.69, 1.31],

t(251)= 6.32, p= <.001. People who scored higher neuroticism experienced more stress.

Furthermore did an increase in neuroticism mean a decrease in the in degree of control, b=-0.43,

CI 95% [-0.57, -0.27], t(251)= -5,56, p= <.001. The more neurotic people were, the less negative

emotional states influenced control. Moreover, people who had a higher in degree of control, negative emotional states having more effect on control, experienced less stress, b= -0.58, CI 95% [-0.82, -0.33], t(251)= -4.63, p= <.001. Even after controlling for the in degree of control, neuroticism still had a significant indirect effect on stress, b= 0.25, CI 95%[0.12, 0.42].

Figure 3 Mediation analysis neuroticism and stress via out degree of control

(11)

11

Network Analysis

In R we have composed two networks in order to illustrate the relationship between stress and control. Using the package qgraph, the figures 5 and 6 were created (Epskamp,2014). Red arrows represent negative correlations, green arrows positive. The wider and more

saturated an edge, the stronger the correlation. Figure 5 displays the network for individuals who score low on neuroticism, a stanine score of either 1 or 2 (n=25). Except for the positive relationship between upset-control and anger-control, all emotions are negatively correlated. An increase in one emotion is not related to an increase in one of the other emotions. When we compare figure 5 to figure 6, the network of individuals who score high on neuroticism, a stanine score 8 or 9, we notice a stark difference. All negative emotional states are positively correlated. An increase in one of the emotional states is related to an increase in a different emotional state. Moreover do all negative emotional states correlate negatively with control. An increase in control is related to a decrease of negative emotional states and a decrease in negative emotional states is related to an increase of control. The strength of the influence of negative emotions within their group is stronger than the strength of the negative emotional states-control relationship.

Figure 5 Network of the 20 PCR items for Individuals with a Neuroticism score of 1 or 2 [Low]

(12)

12

Exploratory analyses: two simple regressions

While the multiple regression showed no relationship between the out degree of control and neuroticism, the mediation analysis did. In order to further explore these contradictory results we have conducted a simple bivariate correlation between the in and out degree of control and two simple regression between the in degree of control and neuroticism and the out degree of control and neuroticism, to see whether the factors were correlated. The in degree and out degree of control showed no signs of multicollinearity in the initial multiple regression, VIF= 1,065, Tolerance= 0.939. Nor did the factors correlate highly, r= -0.29, p= <.001. However, when conducting two separate simple regressions, both predictors and models were significant. In degree, F(1,250)=30,87, p= <.001, b=-0.33, CI 95% [-0.35, -0.17], t= -5,56, p= <.001. Out degree, F(1,250)= 6,99, p= 0.009, b=0.17, CI 95% [0.02, 0.16], t= 2,64, p=0.009.

Exploratory analyses: mean differences regression

In order to get a more clear picture of the relationship between neuroticism and control, we have conducted a simple regression between neuroticism and the mean difference between the absolute out degree and in degree scores. The absolute scores were created by adding the scores for questions 23 through 26 [ out degree] and questions 13, 17, 21 and 30 [in degree]. While adding these scores, the negative answers [ -1, -2] were substituted for positives [1,2]. Then the mean difference was calculated by subtracting the absolute in degree from the absolute out degree. This way the interpretation of the mean difference is as follows, the higher the mean the difference, the more control has an influence on other factors. The lower the mean

difference, the more other factors influence control. By substituting the negative answers for positives, the direction of the influence was removed, only leaving the in and out degree of control.

A simple regression was conducted in order to see whether there was an relationship between the in/out degree of control and neuroticism. The model nor the predictor was

significant. There was no relationship between neuroticism and control, F(1,250)= 1,2, p= 0.27,

b= -0.07, CI 95% [-0.13, 0.04], t= -1.1, p= 0.27.

Figure 6 simple regression mean differences in and out degree control and neuroticism

(13)

13

Discussion

This study examined the application of PCR between stress and control as a patient characteristic for people who score high on neuroticism and therefore possibly resembled PTSD characteristics. We investigated whether neuroticism was related to perceived stress, if there was a specific PCR between stress and control for participants who scored high on neuroticism and whether the in or out degree of control was a mediator of the relationship between stress and neuroticism.

Support was found for both the relationship between stress and neuroticism and the mediating role of control. The more stress people experienced, the more neurotic they were. This was in line with earlier research (Bolger & Schilling,1991, Suls, Green & Hilis, 1998). However, as the mediation analysis pointed out, the stress people experienced was also

influenced by the importance of control. First off, the more neurotic participants were, the more control influenced other factors [high out degree], the more stress they experienced. Secondly, the more neurotic participants were, the lesser the impact of other factors on control, [high in degree], the less stress they experienced. This was in line with our hypothesis, control being an important perceived factor in consequent stress for people who score high on neuroticism. Similarly, these findings were in line with other research (Schneider, 2004, Gunthert, Cohen & Armeli, 1999). For example, Schneider (2004) ) found a relationship between neuroticism, experienced stress, negative appraisals and consecutive coping. As in our research he asked undergraduate students of the non-clinical population to participate. The level of neuroticism could predict whether people would have negative appraisals. Negative appraisals ( I do not have the resources to deal with this stressor) were related to negative emotional experiences and poor task performance. Negative appraisals mediated the relationship between neuroticism and negative emotional experiences. Or in other words, how you perceived your controllability influenced your stress vulnerability and perceived stress. This link we also found in path a of our mediation analysis of out degree of control. The more neurotic people were, the more influence control had on emotional negative states (out degree control). Since the relationship was

positive, b= 0.42, it meant that experiencing a lack of control caused a greater impact of negative emotional states, thus increasing the stress. This is also illustrated in figure 6 in the negative relationship between control and the emotional states.

Furthermore did figure 5 and figure 6 illustrate the difference between low neuroticism participants and high neuroticism participants. In the low N network there were strong negative correlations between all emotional states, feeling one state cancelled the other out. In high N individuals this was the exact opposite, all negative emotional states were positively correlated, they appeared together. This was in line with our first analysis, high N participants experience more stress ( negative emotional states) than low N participants. Contradictory to the mediation analyses, did figure 5 and 6 show an inverse relationship of the in and out degree of control for neuroticism. Figure 5 showed saturated thick lines flowing from control to the emotional states, relative to small thin lines coming in, displaying a high out degree. In figure 6 the in and outflow of control was about the same, but in all ways were the arrows slimmer than in figure 5. This meant less distinctive in and out degree for people who scored high on neuroticism. This was partially in line with our second analysis, the regression of the in and out degree and

neuroticism, which stated that the out degree was no valid predictor for the level of neuroticism. However figure 6 did not display a high in degree either, this was not in line with our outcome of the multiple regression. Since figure 5 displayed a high out degree , rather than the high in degree as was expected from the output of the mediation analysis, it becomes questionable which results are most sound. The data used for the network analyses was smaller, only 10% of the total participants. However since we couldn’t find the high out degree for our target group ( participants who scored high on neuroticism) we could not justify our third hypothesis, that there was a specific PCR flowing from control to stress for participants who scored high on Neuroticism.

Additionally, the insignificant relationship between the out degree of control and neuroticism itself in the multiple regression and exploratory analysis remains a point of

(14)

14

criticism. Whereas in the mediation analyses the hypothesised relationship was found, in the multiple regression it was not. In the mediation analyses the out degree of control had a positive relationship with both neuroticism and stress, the more neurotic you were, the more important control seemed, the more the lack of it caused stress. Along the same lines, the more neurotic you were, the less control was perceived to be influenced by other factors and the more control was perceived to be influenced by other factors, the less stress you experienced. In the multiple regression the negative relationship between neuroticism and in degree of control was found,

b=-0.31, p= <.001. However, while the positive trend between neuroticism and out degree of

control was also illustrated, it was not significant, b= 0.09, p=0.15. We assumed it must have been multicollinearity, a high correlation between the two factors, however, multicollinearity diagnostics and further exploratory tests yielded contradictory results. When in a multiple regression one of the two factors becomes insignificant, violating expectations but without multicollinearity, it could be that the effect is too small, and perhaps, indeed insignificant. The probability that a variable is significant is a function of several things, magnitude of the effect, the variance of the variable itself, the amount of data, to name a few. Whether variables are correlated is important, but it doesn’t override these other conditions. The exploratory

regression, analysing the relationship of the mean differences of in/out degree and neuroticism supported this criticism. When we left out the positive or negative influence, and solely looked at whether having an in or out degree of control was related to neuroticism, all effects disappeared. Since the sample size was identical, it becomes questionable whether the effects found in the multiple regression for the in degree of control are justifiable.

Moreover, the effects found between in/out degree of control and the other factors of neuroticism and stress are weak. Although they are significant in the mediation analyses, they are small. Significance is easily acquired within a large sample. Only neuroticism has a strong effect. What axioms of neuroticism are responsible for this effect remains unclear. The same can be said about control. There are many different forms of coping and appraisals (Folkman and Lazarus, 1980). Which forms which emphasise control and which forms which emphasise the influence of emotions are responsible for the discovered effects remain veiled. More research dedicated to illustrating the effects of these specific forms has to be conducted.

Finally, an equally important point of criticism is that the relationship between

neuroticism, negative appraisals(control) and stress differs between the non-clinical and PTSD population. Engelhard and van den Hout (2004) looked at the correlations between pre-trauma neuroticism, negative appraisals of intrusions and severity of PTSD symptoms in a sample of pregnant women. While all the three variables correlated, the association between negative appraisals and PTSD symptoms remained almost unchanged when controlled for by

neuroticism. Neuroticism on its own explained some variance in PTSD but as an earlier article from Engelhard, van den Hout and Kindt (2002) states, this was probably due to overlap in arousal symptoms and not due to some inherent factor. Neuroticism isn’t a part of PTSD, it is a stand-alone personality factor which happens to correlate quite high with PTSD. Thus, while PTSD patients may score high on neuroticism, they are substantially different from people without a diagnosis of PTSD who score high on neuroticism ( Engelhard & van den Hout, 2004). For the present research a convenience sample was used. Undergraduate psychology students aged 19-22, all without a PTSD diagnosis. All of the hypothesis were based upon literature regarding PTSD patients, with the underlying assumption that participants who scored high on neuroticism were comparable to the samples used in the existing literature. This might be another reason why the proposed relation between the specific PCR for stress and control and neuroticism had small effects and wasn’t supported by all analyses.

(15)

15

Closing Comments

In conclusion, individuals who suffer from PTSD are both a burden to themselves as to society and their direct surroundings. While therapy recovery rates are promising, they are far from perfect. The present research proposed a two new patient characteristics for the treatment of PTSD, based on the theory of coping and personality research. The perceived causal

relationship between stress and control and neuroticism. By using an undergraduate psychology sample, a relationship was found between perceived causal relations of stress and control and neuroticism. However, neuroticism in a non-clinical population and a PTSD population differ, therefore it is hoped that additional studies will follow this present research and further explore these perceived causal relationships and the consequences for clinical significance.

(16)

16

References

Amir, M., Kaplan, Z., Efroni, R., Levine, Y., Benjamin, J., & Kotler, M. (1997). Coping styles in post-traumatic stress disorder (PTSD) patients. Personality and Individual

differences, 23(3), 399-405.

Atwoli, L., Stein, D. J., Koenen, K. C., & McLaughlin, K. A. (2015). Epidemiology of

posttraumatic stress disorder: prevalence, correlates and consequences. Current

opinion in psychiatry, 28(4), 307.

Bolger, N., & Schilling, E. A. (1991). Personality and the problems of everyday life: The role of neuroticism in exposure and reactivity to daily stressors. Journal of personality, 59(3), 355-386.

Bradley, R., Greene, J., Russ, E., Dutra, L., & Westen, D. (2005). A multidimensional meta- analysis of psychotherapy for PTSD. American journal of Psychiatry, 162(2), 214- 227.

Bramsen, I., Dirkzwager, A. J., & Van der Ploeg, H. M. (2000). Predeployment personality traits and exposure to trauma as predictors of posttraumatic stress symptoms: a prospective study of former peacekeepers. American Journal of Psychiatry,

157(7), 1115-1119.

Brogan A, Hevey D. The structure of the causal attribution belief network of patients with obesity. British Journal of Health Psychology. 2009; 14: 35–48.

Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping strategies: a theoretically based approach. Journal of personality and social psychology, 56(2), 267.

Caska, C. M., & Renshaw, K. D. (2013). Personality traits as moderators of the

associations between deployment experiences and PTSD symptoms in OEF/OIF service members. Anxiety, Stress & Coping, 26(1), 36-51.

Cloitre, M., Stolbach, B. C., Herman, J. L., Kolk, B. V. D., Pynoos, R., Wang, J., & Petkova, E. (2009). A developmental approach to complex PTSD: Childhood and adult cumulative trauma as predictors of symptom complexity. Journal of traumatic

stress, 22(5), 399-408.

Cloitre, M., Stolbach, B. C., Herman, J. L., Kolk, B. V. D., Pynoos, R., Wang, J., & Petkova, E. (2009). A developmental approach to complex PTSD: Childhood and adult cumulative trauma as predictors of symptom complexity. Journal of traumatic stress, 22(5), 399-408.s

Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress.

Journal of Health and Social Behavior, 385–396.

Costa, P. T., & McCrae, R. R. (1987). Neuroticism, somatic complaints, and disease: is the bark worse than the bite?. Journal of personality, 55(2), 299-316.

Dudek, B., & Koniarek, J. (2003). Coping style and the development of posttraumatic stress disorder symptoms. Polish Psychological Bulletin.

Engelhard, I. M., van den Hout, M. A., & Kindt, M. (2003). The relationship between neuroticism, pre-traumatic stress, and post-traumatic stress: A prospective study. Personality and

Individual differences, 35(2), 381-388.

Engelhard, I. M., & van den Hout, M. A. (2004). Pretrauma neuroticism, negative appraisals of intrusions, and severity of PTSD symptoms. Journal of Psychopathology and Behavioral

Assessment, 26(3), 181-183.

Epskamp, S. (2014) Network Model Selection Using qgraph 1.3. Retrieved from:

http://psychosystems.org/network-model-selection-using-qgraph-1-3-10/

Foa, E. B., Keane, T. M., Friedman, M. J., & Cohen, J. A. (Eds.). (2008). Effective treatments

for PTSD: practice guidelines from the International Society for Traumatic Stress Studies. Guilford Press.

Folkman, S. (1984). Personal control and stress and coping processes: a theoretical analysis. Journal of personality and social psychology, 46(4), 839.

(17)

17

Folkman, S., & Lazarus, R. S. (1980). An analysis of coping in a middle-aged community sample. Journal of health and social behavior, 219-239.

Folkman, S., & Lazarus, R. S. (1985). If it changes it must be a process: study of emotion and coping during three stages of a college examination. Journal of personality

and social psychology, 48(1), 150.

Freeman, D., Pugh, K., & Garety, P. (2008). Jumping to conclusions and paranoid ideation in the general population. Schizophrenia research, 102(1), 254-260.

Frewen P. A, Allen S. A, Lanius R. A, Neufeld R. W. J. Measuring Perceived Symptom Causal Relations: Examination with Posttraumatic Stress, Depression, and Anxiety

Symptomatology. Assessment. 2012; 19: 480–493.

Frewen, P. A., Schmittmann, V. D., Bringmann, L. F., & Borsboom, D. (2013). Perceived causal relations between anxiety, posttraumatic stress and depression: extension to

moderation, mediation, and network analysis. European Journal of Psychotraumatology,

4(0), 1

Gunthert, K. C., Cohen, L. H., & Armeli, S. (1999). The role of neuroticism in daily stress and coping. Journal of personality and social psychology, 77(5), 1087.

Harkness, A. R., & Lilienfeld, S. O. (1997). Individual differences science for treatment planning: Personality traits. Psychological Assessment, 9(4), 349.

Holdeman, T. C. (2009). Invisible wounds of war: Psychological and cognitive injuries, their consequences, and services to assist recovery. Psychiatric Services, 60(2), 273-273.

Huibers, M. (2015). Voorbij het oordeel van de dodo. Tijdschrift voor Psychotherapie,

41(3), 174-186.

Jakšić, N., Brajković, L., Ivezić, E., Topić, R., & Jakovljević, M. (2012). The role of

personality traits in posttraumatic stress disorder (PTSD). Psychiatria Danubina, 24(3.), 256-266

Karstoft, K. I., Armour, C., Elklit, A., & Solomon, Z. (2015). The role of locus of control and coping style in predicting longitudinal PTSD-trajectories after combat exposure.

Journal of anxiety disorders, 32, 89-94.

Kozlovsky, N., Matar, M. A., Kaplan, Z., Kotler, M., Zohar, J., & Cohen, H. (2007). Long-term down-regulation of BDNF mRNA in rat hippocampal CA1 subregion correlates with PTSD-like behavioural stress response. International Journal of

Neuropsychopharmacology, 10(6), 741-758.

Lawrence, J. W., & Fauerbach, J. A. (2003). Personality, coping, chronic stress, social support and PTSD symptoms among adult burn survivors: a path analysis. Journal of Burn Care & Research, 24(1), 63-72.

McFarlane, A. C., Atchison, M., & Yehuda, R. (1997). The acute stress response

following motor vehicle accidents and its relation to PTSD. Annals of the New York

Academy of Sciences, 821(1), 437-441.

Perkonigg, A., Kessler, R. C., Storz, S., & Wittchen, H. U. (2000). Traumatic events and post‐traumatic stress disorder in the community: prevalence, risk factors and comorbidity. Acta psychiatrica scandinavica, 101(1), 46-59.

Schneider, T. R. (2004). The role of neuroticism on psychological and physiological stress responses. Journal of Experimental Social Psychology, 40(6), 795-804.

Shalev, A. Y., Peri, T., Canetti, L., & Schreiber, S. (1996). Predictors of PTSD in injured trauma survivors: a prospective study. The American Journal of Psychiatry,

153(2), 219.

Stein, M. B., Walker, J. R., Hazen, A. L., & Forde, D. R. (1997). Full and partial

posttraumatic stress disorder: findings from a community survey. American

journal of psychiatry, 154(8), 1114-1119.

Suls, J., Green, P., & Hillis, S. (1998). Emotional reactivity to everyday problems, affective inertia, and neuroticism. Personality and Social Psychology Bulletin, 24(2), 127-136.

(18)

18

negative affect, and evaluative processing. Journal of personality and social psychology,

87(6), 913.

Trimbos (2013) Richtlijnen GGZ voor PTSS. Retreived from: https://assets-sites.trimbos.nl/docs/cc5b0f4e-9398-452d-8792-1453a38bf0a8.pdf

Walker, E. A., Katon, W., Russo, J., Ciechanowski, P., Newman, E., & Wagner, A. W. (2003). Health care costs associated with posttraumatic stress disorder symptoms in women. Archives of General Psychiatry, 60(4), 369-374.

Watson, M., Haviland, J. S., Greer, S., Davidson, J., & Bliss, J. M. (1999). Influence of psychological response on survival in breast cancer: a population-based cohort study. The Lancet,

354(9187), 1331-1336.

Appendix

I. Adjusted PSS

TW48 H Cremers- Stress Scale Test

E De vragen in deze vragenlijst gaan over uw gevoelens en gedachtes gedurende de laatste maand. We vragen u aan te geven in hoeverre de situatie op u van toepassing was door één van de

antwoordopties te selecteren.

Q1 Hoe vaak was u de afgelopen maand van streek omdat er iets onverwachts gebeurde?  Nooit (1)

 Bijna nooit (2)  Soms (3)  Vaak (4)  Heel vaak (5)

Q2 Hoe vaak had u de afgelopen maand het gevoel dat u geen controle had over de belangrijke dingen in uw leven ?  Nooit (1)  Bijna nooit (2)  Soms (3)  Vaak (4)  Heel vaak (5)

Q3 Hoe vaak voelde u zich de afgelopen maand nerveus of gespannen?  Nooit (1)

(19)

19

Q4 Hoe vaak heeft u zich de afgelopen maand zeker gevoeld over uw vermogen om met persoonlijke problemen om te kunnen gaan?

 Nooit (1)  Bijna nooit (2)  Soms (3)  Vaak (4)  Heel vaak (5)

Q5 Hoe vaak had u de afgelopen maand het gevoel dat de dingen verliepen zoals u wilde?  Nooit (1)

Q6 Hoe vaak had u de afgelopen maand het gevoel dat u niet om kon gaan met alle dingen die u moest doen?  Nooit (1)  Bijna nooit (2)  Soms (3)  Vaak (4)  Heel vaak (5)

Q7 Hoe vaak was u de afgelopen maand in staat om irritaties in uw leven onder controle te houden?  Nooit (1)

Q8 Hoe vaak had u de afgelopen maand het gevoel dat u alles onder controle had?  Nooit (1)

(20)

20

Q9 Hoe vaak was u de afgelopen maand boos om dingen die buiten uw controle lagen?  Nooit (1)

Q10 Hoe vaak had u de afgelopen maand het gevoel dat problemen zich zo hoog opstapelde dat u ze niet aan kon?

 Nooit (1)  Bijna nooit (2)  Soms (3)  Vaak (4)  Heel vaak (5)

II. PCR questionnaire

E Dankuwel voor het invullen van het eerste gedeelte van deze vragenlijst. We willen u nu graag een paar vragen stellen over de relatie die u ervaart tussen de antwoorden die u zojuist heeft gegeven. Q11 U voelde zich de afgelopen maand van streek omdat er iets onverwachts gebeurde en u voelde zich nerveus of gespannen. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik van streek was, was ik ... nerveus of gespannen.

 Veel minder (-2)  Minder (-1)

 Niet meer of minder (0)  Meer (1)

 Veel meer (2)

Q12 U voelde zich de afgelopen van streek omdat er iets onverwachts gebeurde en u was in staat om irritaties in uw leven onder controle te houden. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik van streek was, was ik ... geïrriteerd.

(21)

21

Q13 U voelde zich de afgelopen maand van streek omdat er iets onverwachts gebeurde en u had het gevoel dat u alles onder controle had. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik van streek was, had ik ... alles onder controle.

 Veel meer (2)

Q14 U voelde zich de afgelopen maand van streek omdat er iets onverwachts gebeurde en u was om dingen die buiten uw controle lagen. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik van streek was, was ik ... boos.

 Veel meer (2)

Q15 U voelde zich de afgelopen maand nerveus of gespannen en u was van streek omdat er iets onverwachts gebeurde. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik nerveus of gespannen was, was ik ... van streek.

 Veel meer (2)

Q16 U voelde zich de afgelopen maand nerveus of gespannen en u was in staat om irritaties in uw leven onder controle te houden. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik nerveus of gespannen was, was ik ... geïrriteerd.

(22)

22

Q17 U voelde zich de afgelopen maand nerveus of gespannen en u had het gevoel dat u alles onder controle had. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik nerveus of gespannen was, had ik ... alles onder controle.

 Veel meer (2)

Q18 U voelde zich de afgelopen maand nerveus of gespannen en u was boos omdat dingen buiten uw controle lagen. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik nerveus of gespannen was, was ik ... boos.

 Veel meer (2)

Q19 U was de afgelopen maand in staat om irritaties in uw leven onder controle te houden en u was van streek omdat er iets onverwachts gebeurde. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik geïrriteerd was, was ik ... van streek.

 Veel meer (2)

Q20 U was de afgelopen maand in staat om irritaties in uw leven onder controle te houden en u voelde zich nerveus of gespannen. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik geïrriteerd was, was ik ... nerveus of gespannen.

(23)

23

Q21 U was de afgelopen maand in staat om irritaties in uw leven onder controle te houden en u het gevoel dat u alles onder controle had. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik geïrriteerd was, had ik ... alles onder controle.

 Veel meer (2)

Q22 U was de afgelopen in staat om irritaties in uw leven onder controle te houden en u was } boos om dingen die buiten uw controle lagen. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik geïrriteerd was, was ik ... boos.

 Veel meer (2)

Q23 U had de afgelopen maand het gevoel dat u alles onder controle had en u was van streek omdat er iets onverwachts gebeurde. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik alles onder controle had, was ik ... van streek.

 Veel meer (2)

Q24 U had de afgelopen maand het gevoel dat u alles onder controle had en u voelde zich nerveus of gespannen. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik alles onder controle had, was ik ... nerveus of gespannen.

(24)

24

Q25 U had de afgelopen maand het gevoel dat u alles onder controle had en u was in staat om irritaties in uw leven onder controle te houden. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik alles onder controle had, was ik ... geïrriteerd.

 Veel meer (2)

Q26 U had de afgelopen maand het gevoel dat u alles onder controle had en u was boos om dingen die buiten uw controle lagen. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik alles onder controle had, was ik ... boos.

 Veel meer (2)

Q27 U was de afgelopen maand boos om dingen die buiten uw controle lagen en u was van streek omdat er iets onverwachts gebeurde. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik boos was, was ik ... van streek.

 Veel meer (2)

Q28 U was de afgelopen boos om dingen die buiten uw controle lagen en u was nerveus of

gespannen. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik boos was, was ik ... nerveus of gespannen.

(25)

25

Q29 U was de afgelopen maand boos om dingen die buiten uw controle lagen en u was in staat om irritaties in uw leven onder controle te houden. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik boos was, was ik ... geïrriteerd.

 Veel meer (2)

Q30 U was de afgelopen boos om dingen die buiten uw controle lagen en u had het gevoel dat u alles onder controle had. Geef aan in welke mate dit elkaar beïnvloedde. Doordat ik boos was, had ik ... alles onder controle.

 Veel meer (2)

E Dit is het einde van de vragenlijst. Hartelijk bedankt voor uw deelname!

III. Factor Analysis PSS 2016-2017

A principal axis factor analysis was conducted on the 10 items of the Perceived Stress Scale with varimax rotation. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis. KMO=0.85. Bartlett’s test identified the correlation matrix as being an identity matrix, x2_{(45)=1422,44, p<.001. The correlation matrix determinant (0.021) indicated that there was no}

sign of multicollinearity. An initial analysis was run to identify eigenvalues for each factor in the data. Two values had eigenvalues over Kaiser’s criterion of 1 and explained 61,25% of the variance. The scree plot identified 1 clear inflexion. We retained 2 factors since those were the only ones who initially passed both Kaiser’s criterion ( >1) and were visible on the scree

plot[figure 1]. The reproduced correlations table shows 19 (42%) residuals with absolute values greater than 0.05, this poses a concern about the fit of the data to our model. Table 1 shows the factor loadings after rotation. The items that cluster on the same factors suggest that factor 1 represents feelings related to feeling no control and factor 2 represents feelings related to being in control

(26)

26

IV. Reliability

A consecutive reliability analysis was conducted. Overall the PSS questionnaire had a good reliability Cronbach’s α=.81. If Item Q7omgescoord would be deleted the alpha would increase to (.82).

Table 1 Rotated Component Matrix

Factor 1 Factor 2 Q10 .779 Q2 .746 Q6 .744 Q3 .737 Q1 .677 -.341 Q9 .671 Q7omgescoord .738 Q5omgescoord .445 .711 Q4omgescoord .337 .703 Q8omgescoord .522 .614 Figure 2 screeplot PSS

(27)

27

V. Factor analysis PSS PCR 2016-2017

A principal axis factor analysis was conducted on the 10 items of the Perceived Stress Scale and the 20 items of the Perceived Causal Relations with varimax rotation. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis. KMO=0.90. Bartlett’s test identified the correlation matrix as being an identity matrix, x2_{(435)=7181,89, p<.001. The correlation}

matrix determinant (2.442E-9) indicated that multicollinearity was present. An initial analysis was run to identify eigenvalues for each factor in the data. Eight values had eigenvalues over Kaiser’s criterion of 1 and explained 75,14% of the variance. The scree plot identified 2 clear inflexions. We retained 8 factors since those were the only ones who initially passed both Kaiser’s criterion ( >1) and were clearly visible on the scree plot [figure 2]. The reproduced correlations table shows 45 (10%) residuals with absolute values greater than 0.05, this poses no concern about the fit of the data to our model.

Table 2 shows the factor loadings after rotation. The items that cluster on the same factors suggest that, after looking at the content of the questionnaires, that:

Factor 1 represents the out degree of being ‘van streek’. Factor 2 represents the out degree of control.

Factor 3 represents the out degree of anger.

Factor 4 represents the out degree of ‘nerveus & gespannen’. Factor 5 represents the reversed score items.

Factor 6 represents the out degree of irritation ( missing question 21) Factor 7 represents being unable to cope and feeling overwhelmed.

Factor 8 represents how certain feelings influence control, partially the in degree of control

(28)

28 Table 2 Rotated component matrix

* note: all loadings under the .4 have been determined weak and are not shown in the matrix.

VI. Reliability

A consecutive reliability analysis was conducted. Overall the combined PSS_PCR questionnaire had a good reliability Cronbach’s α=.88. If Items Q13,Q17 ,Q30 would be deleted the alpha would increase to (.89), (.89), (.89) respectively.

Q13 loads negatively on factor one, being upset. It is the only one of the questions flowing from Upset to (X) that doesn’t have to do with an emotion but the feeling of control. The same is true for Q17. It should be related to ‘nerveus en gespannen’ but it is the only question flowing from ‘nerveus en gespannen’ that doesn’t have to do with an emotion but the feeling of control. This is also true for Q30. It should be related to anger but it is the only question flowing from anger that doesn’t have to do with an emotion but the feeling of control.

Factor

1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6 Factor 7 Factor 8

Q11 .839 Q12 .812 Q13 -.764 Q14 .764 Q1 .760 Q25 .890 Q26 .850 Q24 .841 Q23 .834 Q29 .807 Q27 .775 Q28 .718 Q9 .660 .420 Q16 .766 Q18 .748 Q15 .739 Q17 -.653 .490 Q3 .565 .559 Q5omgescoord .832 Q7omgescoord .786 Q4omgescoord .771 Q8omgescoord .747 Q19 .831 Q20 .808 Q22 .754 Q10 .742 Q6 .731 Q2 .605 Q21 .863 Q30 -.521 .543

(29)

29

VII. Factor analysis PCR 2016-2017

A principal axis factor analysis was conducted on the 20 items of the Perceived Causal Relations scale with varimax rotation. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis. KMO=0.87. Bartlett’s test identified the correlation matrix as being an identity matrix, x2_{(190)=4893,59, p<.001. The correlation matrix determinant (1,532E-6) indicated that}

multicollinearity was present. An initial analysis was run to identify eigenvalues for each factor in the data. Six values had eigenvalues over Kaiser’s criterion of 1 and explained 78,54% of the variance. The scree plot [figure 3] identified 1 clear inflexions and 2 ambiguous ones. We retained 6 factors since those were the ones who initially passed both Kaiser’s criterion ( >1) and were to be expected from looking at the content of the PCR scale. The reproduced

correlations table shows 34 (17%) residuals with absolute values greater than 0.05, this poses no concern about the fit of the data to our model.

Table 3 shows the factor loadings after rotation. The items that cluster on the same factors suggest that, after looking at the content of the questionnaires, that:

Factor 1 represents the out degree of control

Factor 2 represents the out degree of being upset (van streek)

Factor 3 represents the out degree of being nervous and tense ( nerveus en gespannen) Factor 4 represents the out degree of being angry

Factor 5 represents the out degree of being irritated

Factor 6 represents a rest category, the in degree of control. Table 3: Rotated Component Matrix

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6

Q25 .895 Q24 .871 Q26 .854 Q23 .852 Q11 .851 Q12 .835 Q13 -.738 Q14 .770 Q16 .788 Q15 .783 Q18 .766 Q17 -.663 .498 Q29 .819 Q27 .779 Q28 .752 Q30 -.556 .545 Q19 .829 Q20 .811 Q22 .754 Q21 -.325 .865

(30)

30 Figure 4 Screeplot PCR

VIII. Reliability

A consecutive reliability analysis was conducted. Overall the PCR scale had a good reliability Cronbach’s α=.82. If Items Q13,Q17 ,Q30 would be deleted the alpha would increase to (.85), (.84), (.84) respectively. This can be explained by looking at table 3. Q13, Q17 and Q30 make up factor 6, the in degree of control. The rest of the questionnaire measures an out degree. A subsequent analysis per scale/factor has also been conducted.

Factor 1 (control) Cronbach’s α=.92

Factor 2 (Being Upset) Cronbach’s α=.30 if Q13 deleted, Cronbach’s α=.91 Factor 3 (Being Nervous and Tense) Cronbach’s α=.36 if Q17 deleted, Cronbach’s α=.87 Factor 4 (Anger) Cronbach’s α=.48 if Q30 deleted, Cronbach’s α=.89 Factor 5 (Irritation) Cronbach’s α=.47 if Q21 deleted, Cronbach’s α=.81 Factor 6 ( in degree of control) Cronbach’s α=.62