The connection between social anxiety, core fears and judgement

The connection between social anxiety, core

fears and judgement

Bachelorthesis – Laura Sattler

Student number: 10536701

Institution:

University of Amsterdam

Supervisor:

Jamie Elsey

Words: 4620

Tables/Figures: 502

Abstract: 152

While it is common knowledge that social anxiety is the fear of being judged by others, little is known about individuals with this condition judging other people. This study aims to investigate if a high level of social anxiety correlates with more critical judgement of others in

negative social situations. In Experiment 1 (N = 91), participants judged 12 written vignettes; in Experiment 2 (N = 42) short film fragments were used. Both experiments showed equal numbers of positive and negative examples. The level of social anxiety was assessed through the Social Phobia Inventory and the Negative Self Portrayal Scale. High social anxiety scores could be associated with more negative judgement in negative situations in Experiment 1. Those results could not be replicated in Experiment 2. Future research should aim to make the

results more robust. It is possible that new approaches for treatment for social anxiety could then be opened.

Contents

Abstract ... i Contents ... ii Acknowledgments ... iii Introduction ... 1 Methods ... 5 Participants ... 5 Materials ... 6 Procedure ... 8 Results ... 9 Discussion ... 13 References ... 17

First of all I would like to thank my dad, mum and stepfather as they support me constantly in my ideas and dreams. Without them I wouldn’t be studying in Amsterdam. Secondly, I want

to thank my unbelievably amazing friends – foremost Lotta, Lena und Claudia. Those three held up against my constant worrying and complaining and gave me new energy, motivation and hope every single time I was close to giving up. No matter what, I knew they had words

of encouragement for me. Of course there are so many more people who helped with this thesis. Thank you to Jamie, my supervisor, for his ideas, feedback, time and patience and to my research group for their input and company. Lastly, a huge and special thankyou to Andy

Field who unknowingly helped me through this thesis with his incredibly well written book

Introduction

Social anxiety disorder (SAD), or social phobia (SP), is a mental disorder where the sufferer is afraid of social situations in which one could potentially be judged by others. This fear is persistent and the situations will be avoided or endured with great fear or intense distress (DSM-V, 2013). Around 2.3% of the European population older than 14 years are diagnosed with SAD based on DSM-IV-TR or ICD-10 criteria (Wittchen et al., 2011). In other words, 10.1 million Europeans from the age of 14 years and upwards experience this mental disorder daily.

Social phobia is described in the DSM-V (2013) with different criteria that include the minimum length of the condition and the main symptoms. It is important that those symptoms are not due to other mental or physical conditions. Also, the fear or anxiety is disproportional to the threat and dysfunctional in everyday life. That means that a social phobic actively avoids or endures the feared situation with extreme distress. The situation in question can differ from person to person – some might fear telephone calls and some find that less

distressing than meeting others in person. Every situation where one can be potentially judged by others can provoke immense fear and anxiety in a person. That can lead to people not going to a doctor’s appointment because the thought that the doctor might judge them is too stressful. Social, cultural and personal contexts should always be considered by making the diagnosis.

A range of psychological and pharmacological therapies are available to treat social anxiety (Stein & Stein, 2008). Stein and Stein also state that even though the effect sizes are good, sometimes up to 60% of the participants in studies do not respond to pharmacological interventions after eight to 12 weeks. Even though there are already good interventions for social anxiety/phobia there is still need to research new ideas to make those interventions even more effective and to help more people with this condition (Moscovitch, 2009).

The current interventions to treat SAD are reasonably effective but there is, as Moscovitch (2009) stated, clearly room for improvement. One approach to improve psychological interventions could be to understand the disorder itself and the mechanisms behind it better. To do so, several models of social phobia have been developed.

A well-known and empirically implemented model of social anxiety is the Cognitive

Model (Clark & Wells, 1995). Adapted from Clark (2001), it is still often used to explain the

processes within SAD. Figure 1 shows a process-model, originally from Clark & Wells and adapted by Clark (1995).

Figure 1. An adapted model from Clark & Wells (1995) of the processes that take place when

a person with SAD/SP enters a feared social situation (Clark, 2001, Figure 18.1)

Clark (2001) describes all parts of Figure 1 as follows: The model shows how a

situation activates an assumption which therefore leads to perceived social danger. In reaction to that, safety behaviours and somatic and cognitive symptoms (e.g. sweating, rehearsing possible conversations) get activated. Those last three components are all serving the

processing of the self as a social object. The relationship with the perceived social danger is mutual and in constant exchange. That means the more one realizes it is an object in a social situation the higher the perceived danger is. In turn, safety behaviours and symptoms are triggered. Safety behaviours are small rituals, gestures or thoughts that a person uses to calm down and to prevent the ‘danger’. Those can actually work against the initial intention and worsen the symptoms. Here is where the process starts again and gets renewed over and over with every new thought and experience. In that way the experienced distress, fear or anxiety can get to a point where it is disproportional to the actual situation or threat and dysfunctional in everyday life. The model also shows the role not just of the social situation alone but also the individual assumptions that trigger the perceived social danger. Those assumptions can differ per person while the final outcome is the same – they fear the social situation. The difference lies in the reasons for the fear.

Moscovitch (2009) calls those personal reasons core fears. He indicates that people with social anxiety or social phobia have different core fears. In his theory he names four core fears. These are: the fear of (1) perceived flaws in social skills and behaviour, (2) perceived flaws in controlling and concealing internal feelings of anxiety, (3) perceived flaws in

physical appearance and (4) perceived characterological flaws. Later, Moscovitch and Huyder (2011) reduced those to three core fears naming them the concerns about (a) social

competence, (b) physical appearance and (c) signs of anxiety. One of the ideas behind their research is to find new ways to make the existing interventions more helpful through focussing more on the specific core fears for each individual.

Most of the research concerning SAD, such as Moscovitch and Huyder (2011), is about how people with SAD are afraid of being judged negatively by others. There are different theories about how people come to the conclusion that they are being judged in the first place, for example through their perception of others. If a person fears being judged on specific aspects it means they have a clear vision of what the social norms or expectations in

certain social situations are. Those are based on earlier personal experiences and maintained through interpretation bias of the social phobic (Antony, Rowa, Liss, Swallow, & Swinson, 2005; Taylor & Alden, 2010).

To interpret behaviour of others correctly, one must be able to accurately decode and reason about someone else’s intentions, beliefs, emotions, and desires. This capacity is known as having a Theory of Mind (ToM) (Wellmann. 1990). Sabbagh (2004) suggests that this ability depends on two components. Firstly, it is important that the mental state of others can be coded from cues such as facial expressions. This is called ‘decoding’. The second skill is ‘reasoning’: this involves being able to make predictions about future behaviour based on those decoded mental states. Recent studies suggest that people who suffer from SAD have difficulties with the ‘decoding’ part of ToM (Hezel & MacNally, 2014; Washburn, Wilson, Roes, Rnic, & Harkness, 2015). Based on their findings, both studies conclude that the ‘decoding’ error could be what leads to the unrealistic estimation in the socially anxious person of how other people evaluate and judge them. This, in turn, feeds the individual fear of being negatively judged by others.

Clark’s (2001) cognitive model of social phobia, Moscovitch and Hyders (2011) core fears and ToM’s ‘decoding’ error (Hezel & MacNally, 2014) could relate to one another and explain why individuals that are socially anxious are also more alert of the behaviour of others. This could result from these individuals having an attention bias to their own insecurities and threats (Bar-Haim, Lamy, Pergamin, Bakermans-Kranenburg & van

IJzendoorn, 2007; Cisler, Bacon, & Williams, 2009). Due to the ‘decoding’ error they might interpret others’ thoughts or intentions incorrectly and judge them on how they would judge the behaviour in themselves – as socially anxious.

As there was no SAD research to be found about how individuals with this condition judge others this study aims to make a start. As operationalization for social anxiety, two questionnaires were used. One of them is specifically aimed to also assess the core fears.

Furthermore participants were asked to score people in socially positive and negative situations.

The main hypothesis is that people with high social anxiety judge others more critically and therefore more negatively if they make mistakes in social situations. Those mistakes are visible signs of anxiety, discomfort or awkwardness. This would be supported by a negative correlation between high scores on the social anxiety questionnaires and low scores for the assessment task in the negative examples. Further explanation about the materials will be in the method section.

If this relationship does indeed exist it would be interesting to see if their own

insecurities make a difference in the judgment. Therefore a second hypothesis is that socially anxious individuals judge others more critically in aspects of their own core fears.For example: if a socially anxious individual is concerned about showing signs of anxiety, then that person might judge someone else more negatively if he or she stutters slightly for a moment. This will also be tested in case of significant results for hypothesis 1. Therefore we expect that high scores for the core fears will negatively correlate with the according

assessment scores.

Methods Participants

Experiment 1. The first sample consists of 91 first year psychology students who

participated during the mandatory test week at the University of Amsterdam (UvA). Age ranged from 18 to 25 years (M = 19.83, SD = 1.67). There were 70 females (77.8%) and 20 males (22.2%).

Experiment 2. This group consists of 42 UvA psychology undergraduates, aged from

(85.71%). Participants received a subject point in exchange for their participation. Table 1 gives an overview of age and number of participants of both samples.

As the questionnaires were translated beforehand from English to Dutch, all of the participants either had Dutch as their first language or were fluent in it as a second language. All participants filled in the questions completely and there was no dropout.

Table 1. Age and number of Participants (N) for both groups incl. male/female distribution.

Group Age (mean / SD) N (total) N female (%) N male (%) Experiment 1 19.83 / 1.67 91* 70 (77.8) 20 (22.2) Experiment 2 20.9 / 2.68 42 36 (85.71) 6 (14.29)

*Note. One person did not assign to the binary choice given for gender.

Materials

Social Phobia Inventory (SPIN). The participants filled in the Social Phobia

Inventory, SPIN, from Connor et al. (2000). This questionnaire consists of 17 statements that assess how much the described problems bothered the participant in the last week. Examples are: “7. Sweating in front of people causes me distress.” and “11. I avoid having to give speeches.” The scores range from 0 (=not at all) to 4 (= extremely). The sum of all scores describes a rough indication of the severity of social anxiety. A score under 20 means none, 21-30 points means mild and 31-40 moderate. 41-50 is severe and all above that very severe. The statements of the SPIN include three fear-domains of social anxiety disorder – avoidance, fear, and physiological arousal. Connor et al. found a Cronbach’s α of 0.94 for the full scale. Their results also showed adequate test-retest reliability with significant (p≤0.0001) Spearman correlation coefficients between 0.78 and 0.89 in different groups. The SPIN was shown to the participants in Dutch. The translation was provided by Boelen and Reijntjes (2009). In their research Cronbach’s α Dutch version of the SPIN was 0.93.

Negative Self Portrayal Scale (NSPS). The Negative Self Potrayal Scale (NSPS) was

developed by Moscovitch and Huyder (2011) to assess people’s concerns about the supposed core fears in social anxiety (Moscovitch, 2009). Each of the items covers one of the core fears

(social competence, showing signs of anxiety, and physical appearance). The score shows the level of how much one fears that others will see them showing or experiencing the described words/phrases in a potentially anxiety-provoking situation. To do so there are 27 words or phrases to be scored from 1 (=not at all concerned) to 5 (=extremely concerned). Included words are e.g., 3. boring, 7. blushing and 17. humourless. Phrases are e.g., “8. speaking with a trembling voice” or “22. having a bad hair day”. The sum of the items of each category shows the level of concern about those. The sum of all items indicates the global level of concern. In all cases a lower score resembles less concerns. The psychometric properties for the NSPS of Moscovitch and Huyder are good for test-retest reliability, r = 0.75. A Cronbach’s α of 0.96 demonstrated a very good internal consistency. Also the NSPS was translated and filled out in Dutch. Firstly, four translations were provided by four Dutch-speaking undergraduates. The supervisor compiled those and checked them for face validity and alternate translations along with a psychologically trained Dutch native speaker.

The assessment-task. Participants in both experiments were asked to judge others. In

Experiment 1 the participants were given short written vignettes about social situations. These vignettes consisted of three sentences where the last sentence evaluates if the person in

question is confident or not, e.g. positive: You are at a friend’s birthday party. One of the

other guests is asked to make a toast. When the guest delivers the speech, you notice that her voice is even and controlled for its whole duration. A negative example would be: You are waiting for some friends in a bar. As you wait, you watch another group at a table arguing. When one of them makes a point, she cannot express herself and the others ignore her.

There were twelve out of 24 of those situations presented – randomly 6 positive and 6 negative. After each situation participants scored how much they strongly disagreed (=1) or strongly agreed (=7) with the statements that the person in the clip is an anxious person, a physically attractive person and a socially competent person. They also scored how they

thought other people would judge this person. Here the range was from 1 (=very negatively) to 7 (=very positively).

In Experiment 2 the participants were shown eight short video fragments. In each clip there were one or two individuals in different social situations. The settings varied from night clubs to a family dinner to a first introduction after a sport match or a consultation with attorneys. The order of positive and negative clips was randomized. It was always specified in the assessment questions which person in the scene was to be assessed. Four of the clips displayed confident people and four clips showed people with obvious signs of discomfort and anxiety as a result of the situation. The scoring was the same as with the vignettes.

The statement scores for the positive examples where combined in ‘global positive scores’ or accordingly ‘global negative scores’ without splitting them into the core fear aspects. Scores of how ‘others’ would assess this person were also combined for the positive and the negative situations. The ‘global scores’ and ‘others scores’ were used to test the main hypothesis through correlating the positive and negative scores individually with the SPIN score and the over-all NSPS score.

If the results show support for hypothesis 1, hypothesis 2 will also be tested. Participant’s own NSPS core fear scores will be correlated with the core fear scores they assessed to the individuals in the vignettes/video fragments.

Procedure

Experiment 1. In this sample all parts were taken at a computer using the program

“Qualtrics” to present the assessment task and the questionnaires. Beforehand, the students signed an informed consent for the whole test week. This event took place at the university. As participants started this study, a short introduction text was presented. After that, they began with the assessment task. A short explanation of the task was followed by the first of twelve situations. Each situation was presented separately from the others, along with the

questions to score. It was not possible to go back to earlier vignettes, because we wanted an intuitive judgement. The assessment was followed by the SPIN and then the NSPS. Within the questionnaires, the participants were able to go back and forth between the questions. They couldn’t go back to the questionnaire after finishing it. After finishing, a short debriefing showed up on the screen about the hypothesis of the study.

Experiment 2. This experiment took place in the research rooms of the UvA. There

was no time limit set in this experiment. All participants were required to read the information brochure and signing an informed consent before taking part. The brochure did not include an explanation of the hypotheses or expected results. They were informed of this after

completing the study. There was no division into groups of gender or manipulation. Each participant was placed in a separate room at a computer after signing the

informed consent. The order was the same as in Experiment 1 – first the assessment task and then the SPIN and NSPS – and it was also taken using “Qualtrics”. Different from Experiment 1, there were no vignettes but 8 short film fragments to judge. This time, the fragments and the statements that had to be scored were shown separately. It was not possible to watch the fragment multiple times or go back to another fragment again for the same reasons discussed above. After finishing all questionnaires and tasks there was a short debriefing and the participants got one subject point as reward.

Results

Manipulation check. To check if the suggested positive and negative situations were

indeed contrasted enough, a paired-sample t-test is executed in both samples. The variables tested are the global judgement score for the positive videos against the global score for the negative videos. A value was counted as significant with an α < 0.05. In Experiment 1, participants scored the positive situations higher (M = 97.2, SE = 0.93) than the negative situations (M = 65.96, SE = 0.68). The difference of 31.24, BCa 95% CI [28.92, 33.56] was

significant, t(90) = 26.77, p < 0.001 with a large effect of r = 0.94. In Experiment 2,

participants scored the positive videos higher (M = 70.45, SE = 0.73) than the negative videos (M = 38.45, SE = 0.77). Again, the difference of 32, BCa 95% CI [29.77, 34.23], was

significant, t(41) = 28.95, p < 0.001. There was an even larger effect size of r = 0.98.

The same test was executed in both samples for the estimated score how others would judge the person in question for the positive and the negative videos/situations. Again, the average person scored higher on the positive ones (M = 22.38, SE = 0.41) than on the negative (M = 16.74, SE = 0.48) in Experiment 2 as well as in Experiment 1, Mpos = 34.08,

SEpos = 0.29 and Mneg = 21.99, SEneg = 0.33. In the sample of Experiment 1 the difference was

12.09, BCa 95% CI [11.23, 12.94] and significant, t(90) = 28.1, p < 0.001. Also the difference in Experiment 2 of 5.64, BCa 95% BI [4.29, 6.99] was significant, t(41) = 8.44, p < 0.001, with a large effect, r = 0.8. Here the effect size was again large, r = 0.9. That suggests that our separation into positive and negative vignettes and videos successfully provoked positive and negative ratings in both experiments. Table 2 and Table 3 show the mean, the standard deviation and the range of scores given for both samples.

Table 2. Mean, standard deviation and range of scores for SPIN and NSPS in both samples.

SPIN NSPS

Experiment 1 Experiment 2 Experiment 1 Experiment 2

M 12.41 13.1 50.7 47.93

SD 10.81 9.20 18.12 15.38

range 0 – 52 1 – 36 27 – 102 27 – 90

Table 3. Mean, standard deviation and range of scores for the global scores of the assessment task in both samples.

Experiment 1 (N = 91) Experiment 2 (N = 42) M SD range M SD range Global positive 97.20 8.84 74 – 126 70.45 4.74 59 – 81 Global negative 65.96 6.51 51 – 83 38.45 5.01 28 – 49 Positive others 34.08 2.73 25 – 41 22.38 2.70 15 – 27 Negative others 21.99 3.15 15 – 32 16.74 3.08 9 – 23

Hypothesis 1. To test the main hypothesis, if there is a relationship between the scores

used. These were calculated between the SPIN and the global scores of the assessment task for the negative situations as well as between the NSPS and those global scores. To control if the findings were specific to the negative situations or a reflection of the view of others, the analysis was extended to the questionnaires with own positive global scores and the global positive and negative scores of how they expected others would judge.

There were indeed significant correlations in the Experiment 1 which support hypothesis 1. The SPIN is significantly negatively correlated with the global negative assessment score, r = -0.27, p = 0.010, BCa 95% CI [-0.431, -0.088]. Also the NSPS

correlates with the global negative assessment score, r = -0.24, p = 0.02, BCa 95% CI [-0.411, -0.070]. The rest of the correlations didn’t show a significant correlation. All pairs with the according correlation coefficients, significance and the 95% confidence intervals of the bootstrapping are shown in Table 4.

Table 4. Correlation results of experiment 1 (N =91) analyses of all pairs with bootstrap 95% correlation interval. Pair Correlation coefficient (Pearson’s r) Significance (2-tailed) p BCa 95% CI upper lower SPIN – global positive -0.091 0.389 -0.282 0.126 SPIN – global negative -0.270* 0.010 -0.431 -0.088 SPIN – positive others -0.016 0.879 -0.255 0.214 SPIN – negative others -0.071 0.505 -0.286 0.189 NSPS – global positive 0.006 0.951 -0.214 0.224 NSPS – global negative -0.243** 0.020 -0.411 -0.070 NSPS – positive others 0.078 0.456 -0.156 0.294 NSPS – negative others -0.102 0.336 -0.311 0.109

Note. *A significant correlation at the 0.01 level (2-tailed). ** A significant correlation at the 0.05 level (2-tailed).

For experiment 2, due to the low number of participants, histograms were used to measure skewedness and kurtosis for the SPIN and the NSPS to control whether the data were normally distributed. The z-scores for both questionnaires were just above 1.96 which

indicates a slightly abnormal distribution. To control for this, Pearson’s correlation was used as well as Spearman’s rho. There were no significant correlations in any of the pairs

regardless of the coefficient used. Table 5 includes all pairs with the according correlation coefficients, significance and the 95 % confidence intervals of the bias corrected and

accelerated bootstrap. Neither the SPIN nor the NSPS were significantly correlated to any of the assessment results. We could not replicate the support for hypothesis 1 in Experiment 2.

Table 5. Correlation results of experiment 2 (N = 42) analyses of all pairs with bootstrap 95% correlation interval. Pair Correlation coefficient Pearson’s r (Spearman’s rho) Significance (2-tailed) p BCa 95% CI upper lower SPIN – global positive -0.003 (0.013) 0.984 (0.933) -0.325 (-0.347) 0.312 (0.351) SPIN – global negative -0.029 (-0.034) 0.853 (0.829) -0.295 (-0.334) 0.250 (0.271) SPIN – positive others -0.217 (-0.240) 0.168 (0.125) -0.534 (-0.556) 0.104 (0.128) SPIN – negative others 0.088 (0.117) 0.580 (0.461) -0.199 (-0.214) 0.399 (0.443) NSPS – global positive 0.150 (0.121) 0.345 (0.445) -0.188 (-0.273) 0.486 (0.462) NSPS – global negative 0.024 (0.009) 0.878 (0.954) -0.317 (-0.325) 0.331 (0.335) NSPS – positive others -0.112 (0.127) 0.481 (0.424) -0.405 (-0.427) 0.166 (0.198) NSPS – negative others -0.032 (0.054) 0.841 (0.734) -0.290 (-0.241) 0.278 (0.368)

Hypothesis 2. As the expected significant relationships were indeed found in

experiment 1, the correlations between the NSPS core fear scores and the equivalent assessment scores for the negative vignettes were executed in this sample as well to see if

these scores relate. A significantly negative correlation was found between the NSPS score for signs of anxiety and the anxiety scoring in the negative examples, r = -0.273, p = 0.009, BCa 95% CI [-0.201, -0.113]. Even with a Bonferroni correction to prevent false positives,

PCrit = 0.02, this result would still be significant. In table 6 the pairs are listed for with the

correlation coefficient, the significance and the bootstrap correlation interval of 95%. This is only partially supportive of hypothesis 2 as there were no significant correlations for the other core fears.

Table 6. Correlation results for experiment 1 (N =91) analyses of all pairs with bootstrap 95% correlation interval. Pair Correlation coefficient (Pearson’s r) Significance (2-tailed) p BCa 95% CI upper lower NSPS anxiety – negative anxiety -0.273* 0.009 -0.431 -0.113

NSPS SC – negative SC -0.165 0.117 -0.334 0.019

NSPS PA – negative PA -0.086 0.416 -0.262 0.093

Note. *A significant correlation at the 0.01 level (2-tailed).

Discussion

This research was performed to assess whether people who have higher levels of social anxiety judge other people more critically in potentially negative social situations.

This hypothesis (1) could be supported by the results of Experiment 1 where the situations were represented by vignettes. In this experiment significantly negative correlations could be found between high scores of anxiety and low, i.e. negative, scores for the person in question (Table 4). This indicates that the more anxious a person is the more negatively they assess someone else. Unfortunately the attempt to replicate these results did not succeed in Experiment 2 as there were no significant correlations at all (Table 5). Still, it seems that there is solid indication for that people who are socially anxious judge others more critically in negative situations, as shown in the more controlled Experiment 1. As expected, there was no indication that this also happens in positive situations. The second hypothesis, whether the negative judgements are related to own core fears, was only tested in Experiment 1 since there

was no indication for this in Experiment 2. The only significant correlation found here was regarding the concern about showing signs of anxiety. There were no correlations found between the level of anxiety and the estimate how others might judge the person. That shows that the opinion of the anxious person is not influenced by how they think others might judge the person showed in the situation. The implications of our findings, possible explanations and limitations of the study are discussed below.

As stated earlier, the idea that led to the hypothesis was based on the decoding-error of

ToM (Washburn et al., 2015) and attention bias (Cisler et al., 2009). While the present study

wasn’t able to cover whether the judging of others was actually due to this error, there was an indication that it might be connected to the attention bias. The significant correlation of the core fear concerning showing signs of anxiety between the NSPS score and the assessment could be a result of the fixation of the individual with their own insecurities. Of course visible signs of anxiety are easier to notice in comparison to the other two core fears as the perception of physical appearance is subjective and social competence could be dependent on the

situation. Why there were no correlations for the other core fears should be explicitly explored in future research.

The findings of this study could lead to a new approach of research and treatment for social anxiety. Up until now, most studies and therefore interventions in this field rest on the fact that individuals with SAD are afraid of being judged. While this is true and important, the indication that they are also judging others is new. Most psychological treatments are based on the fear of being negatively evaluated by others (Hoffman, 2007). Strategies such as shifting the internal attention to other people are used to show that the feared thing (e.g. blushing, stuttering) is also common in others. However, if one is already judging others, this approach might be misguided and alternatives should be explored.

To see if the findings of Experiment 1 are robust and can lay a base for future research it is necessary to highlight the limitations. Experiment 2 was the first attempt to support the

hypothesis even further but failed. This is not necessarily, if at all, due to the hypothesis not being true. A more realistic explanation would be the difference in the manipulations – vignettes versus video fragments. Both showed positive and negative situations which could, as the manipulation check showed, be distinguished by the participants. However, the

vignettes were more controlled. They had straight counterparts and were directed on the specific core fears. The film fragments were, on the other hand, more random. They showed everyday situations and the positive and negative situations were not counterparts at all. There is a possibility that the participants knew the movies and therefore judged based on the

character they knew and already had an opinion on rather than based on their own

insecurities. Also, participants could have known the actors from other appearances which could have biased their judgement. In a written context, as in the vignettes, the person to be judged was much more anonymous. This could result in more correlations to one’s own insecurities or level of anxieties. Concerning the core fears in the video clips (Experiment 2), it is possible that they don’t provide a clear enough distinction between the positive and negative examples. The “positive” person could just as well be as physically attractive as the “negative” person. As stated earlier, the vignettes were specifically designed to match a certain core fear.

Another issue is the population of both samples. As shown in Table 1, the distribution of gender between the samples is not actually representable for the general population. There are far less male participants than female participants in both samples. Even though the sex ratio (f:m) for SAD is 2.5 (Wittchen et al., 2011), the consequence is that the results cannot be generalized to an actual broader population. This was expected as the samples both consist of psychology undergraduates – a population with mostly females.

With these limitations in mind there are a couple of points future researches regarding this topic should pay attention to:

a) Controlled manipulation: In this study, the controlled manipulation – the vignettes in Experiment 1 - showed support for the hypothesis. Experiment 2, on the other hand, did not. Future studies could possibly work with just written or indeed different media sources to counter-control for this bias.

b) Diverse sample: The current population of both samples is quite homogeneous. It consists of young adults who all study psychology. Furthermore, the distribution of gender is not representable for a broader population. That is why the conclusion can’t be generalized. Future researches regarding this topic should aim to get a more representative sample to be able to increase external validity.

Of course there are other areas to explore if the hypothesis proves to be consistent. First of all it should be assessed how judging others has effects on socializing and the symptoms of individuals who suffer from SAD. Secondly, it should be investigated whether there are implications for interventions to treat SAD and if so, what kind of changes could be made to current interventions.

Summary. All together this study showed that with a controlled stimulus significant

negative correlations could be found between high levels of social anxiety and negative evaluation of others. This supports hypothesis 1, that socially anxious people are more judgemental. Regardless of the current limitations these findings could with more support indicate a new view of SAD and future treatments.

References

American Psychiatric Association (2013). Diagnostic and Statistical Manual of Mental

Disorderrs. 5th edn. Arlington, VA: American Psychiatric Association.

Antony, M. M., Rowa, K., Liss, A., Swallow, S. R., & Swinson, R. P. (2005). Social

Comparison Processes in Social Phobia. Behavior Therapy, 36, 65-75.

Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2007). Threat-related attentional bias in anxious and non-anxious individuals: A

meta-analytic study. Psychological Bulletin, 133, 1-24.

Boelen, P. A. & Reijntjes, A. (2009). Intolerance of uncertainty and social anxiety. Journal of Anxiety Disorders. 23, 130-135.

Clark, D. (2001). A cognitive perspective on social phobia. International Handbook of Social

Anxiety: Concepts, Research and Interventions Relating to the Self and Shyness.

Edited by W. Ray Crozier and Lynn E. Alden.

Clark, D. M. & Wells, A. (1995). A cognitive model of social phobia. In R. Heimberg, M. Liebbowitz, D. a. Hope, & F. R. Schneider (eds.), Social phobia: Diagnosis,

assessment and treatment. (pp. 69-93). New York: Guilford Press.

Cisler, J. M., Bacon, A. K., & Williams N. L. (2009). Phenomenological characteristics of

attentional biases towards threat: A critical review. Cognitive Therapy and Research,

33, 221-234.

Connor, K. M., Davidson, J. R. T., Churchill, L. E., Sherwood, A., Weisler, R. H., & Foa, E. (2000). Psychometric properties of the Social Phobia Inventory (SPIN). The British Journal of Psychiatry, 176:4, 379-386.

Hezel, D. M., & McNally, R. J. (2014). Theory of Mind Impairments in Social Anxiety. Behavior Therapy, 45, 530-540.

Hofmann, S. G. (2007). Cognitive Factors that Maintain Social Anxiety Disorder: a

Comprehensive Model and its Treatment Implications. Cognitive Behaviour Therapy,

36:4, 193-209.

Moscovitch, D., A. (2009). What Is the Core Fear in Social Phobia? A New Model to

Facilitate Individualized Case Conceptualisation and Treatment. Cognitive and

Behavioral Practice, 16, 123-134.

Moscovitch, D., A., & Huyder, V.(2011). The Negative Self-Portrayal Scale: Development,

Validation, and Application to Social Anxiety. Behavior Therapy, 42, 183-169.

Sabbagh, M., A. (2004). Understanding orbitofrontal contributions to theory-of-mind

reasoning: Implications for autism. Brain and Cognition, 55, 209-2019.

Stein, M. B., & Stein, D. J. (2008). Social Anxiety Disorder. The Lancet, 371(9618), 1115-1125.

Taylor, C. T., & Alden, L. E. (2010). Safety behaviors and judgmental biases in social

anxiety disorder. Behaviour Research and Therapy, 48, 226-237.

Washburn, D., Wilson, G., Roes, M., Rnic. K., & Harkness, K. L. (2015). Theory of Mind in

social anxiety disorder, depression, and comorbid conditions. Journal of Anxiety

Disorders, 37, 71-77.

Wellmann, H. M. (1990). The child’s Theory of mind. Cambridge, MA: MIT Press.

Wittchen, H. U., Jacobi, F., Rehm, J., Gustavsson, A., Svensson, M., Jönsson, B., Olesen, J., Allgulander, C., Alonso, J., Faravelli, C., Fratiglioni, L., Jennum, P., Lieb, R.,

Mearcker, A., van Os, J., Preisig, M., Salvador-Carulla, L., Simon, R., & Steinhausen, H.-C. (2011). The size and burden of mental disorders and other disorders of the brain

in Europe 2010. European Neuropsychopharmacology, 21, 655-679.

World Health Organization (1993). International Statistical Classification of Disease and