Emotion in the face of culture : the relationship between culture, emotion-antecedent appraisal, and the non-verbal expression of pride and shame in Ghana

Bachelor Thesis Social Psychology

Emotion in the Face of Culture:

The Relationship Between Culture, Emotion-Antecedent Appraisal, and the Non-Verbal Expression of Pride and Shame in Ghana

Olivia Pich

Supervisor: Milena Feldkamp

Student number: 10598502 Date of submission: 21-08-2017

Abstract

This study investigated the relationship between culture, non-verbal emotion expression, and the appraisal of agency. 28 high-school teachers from Ghana talked about pride and shame-evoking events with an imagined acquaintance while being filmed and recorded. Non-verbal expressions were assessed by coding their frequency according to a coding scheme. Verbal agency appraisals were analysed by assessing the frequency of external and internal appraisal terms based on a codebook. Appraisal terms related to groups, third persons, or deities were coded as external, terms related to the self as internal. Additionally, emotion ownership (group or self) was assessed by counting the frequency of terms appraising the group or the self as emotion owner. Cultural orientation was measured with the cultural value scale. Results indicate that pride-evoking events were more frequently appraised as external; shame-evoking events more frequently as internal. This was related to the high prevalence of references to god in the pride condition. Emotion ownership was more often attributed to the self than the group. Also, pride and shame corresponded to a number of distinct non-verbal expressions. Unlike in the verbal analysis, there was no relationship between non-verbal expressions and agency appraisals, indicating that non-verbal expressions did not correspond to distinct patterns of agency appraisals. The results illustrate meaningful differences between the current and prior studies in the expression of pride and shame and associated appraisal processes.

Emotion in the Face of Culture: The Relationship Between Culture, Emotion-Antecedent Appraisal, and the Non-Verbal Expression of Pride and Shame in Ghana

In cultures around the world, non-verbal expressions are of central importance for social interaction (Argyle, 2007), especially when it comes to the communication of emotional states (Ekman, 1993). Yet, the question to what extent and in which ways cultural factors influence non-verbal emotion expression remains a subject of debate until today (Russell, 1994). A number of studies in recent years indicate that emotion-antecedent appraisals could play an important role in shaping non-verbal expression in different cultural contexts (e.g. De Leersnyder, Boiger & Mesquita, 2013; Scherer, 2009). Because appraisal tendencies have been shown to differ widely across cultures, cross-cultural differences in emotion expression might be linked to cultural values which influence individuals’ appraisals of situations and internal states, thereby shaping their affective experiences and responses (Scherer, 1997).

However, the relationship between culture, emotion-antecedent appraisals, and non-verbal emotion expression has been little investigated. In order to answer the question in which aspects cultures are similar and in which aspects they differ in the non-verbal expression of emotions, research needs to be conducted with samples from diverse cultural backgrounds. Yet, most emotion research so far has been conducted in quite a limited number of mostly industrialised countries in the West and East Asia. In contrast, relatively little is known about non-verbal expressions and appraisal processes in other cultural spheres, such as Africa. Therefore, the present study investigated the relationship between culture, non-verbal emotion expression, and

emotion-antecedent appraisal in Ghana, a country that has been largely neglected by emotion research so far, and to assess whether prior findings concerning non-verbal expressions as well as appraisal

processes can be generalised to the Ghanaian context. Thereby, the current study was concerned with pride and shame and investigated whether these emotions correspond to distinctively different appraisal patterns. The focus lay thereby on the appraisal dimension of agency, which was divided

into two categories: External and internal agency. Furthermore, it was investigated whether the ownership of pride and shame would be appraised as belonging to the group or the self, and whether this would differ between pride and shame.

Universal Aspects of Non-Verbal Expressions

Non-verbal expressions consist of facial expressions but also body postures and gestures (Argyle, 2007). Evidence for the universal aspects of emotion-expression comes from a variety of studies (e.g. Izard, 1972; Ekman, 1993), amongst the most influential of which is the research of Ekman and Friesen (1969; 1971). Their work contributed to the recognition of six basic emotions, namely happiness, sadness, disgust, anger, fear, and surprise, which have been linked to specific non-verbal expressions, and are universally expressed and recognised by members of all cultures.

This early research on non-verbal expressivity provided important evidence for the notion of emotion-specific expressions (Ekman, 1993). This means that a particular set of expressions

corresponds to one distinct emotion and is recognised as such by other individuals (Ekman, 1993). The emotion specificity of non-verbal expressions has been further extended in recent years by studies indicating that also a range of other emotions possess universal aspects in expression and recognition.

A study by Tracy and Robins (2008) gave important evidence for the universality of pride by showing that pride is characterised by specific non-verbal expressions which are universally

recognised. The researchers demonstrated that participants from different Western countries, as well as members of an isolated West-African tribe were able to correctly judge faces as proud. If pride is universally recognised, it is likely that there will also be universal aspects in the expression of pride. This was supported by a later study by Tracy and Matsumoto (2008) in which the researchers demonstrated that sighted as well as congenitally blind athletes from more than 30 different countries showed the same characteristic expressions after a victory. In the course of the same

study, Tracy and Matsumoto (2008) showed that also shame has a genuine non-verbal expression, which was expressed by sighted as well as congenitally blind athletes as a reaction to defeat. This indicates that shame and pride correspond to specific and distinctively different facial expression patterns. Table 1 shows a summary of the facial expressions that have been found to be associated with pride and shame in Tracy and Matsumoto’s (2008) study and their adaption in the current research.

Despite the evidence pointing to universal aspects of emotions and facial expressions, numerous studies have indicated that there are various cross-cultural differences in emotion processes, such as emotional antecedents and physiological reactions, as well as non-verbal expressions (Mesquita, Frijda, & Scherer, 1997). These differences have been linked to appraisal processes that influence the way in which people perceive and evaluate emotion-eliciting events,

Expressions Tracy & Matsumoto (2008) Current study

Shame Chest narrow

Shoulders slumped Closed posture

Pride Head tilt back

Arms out Arms raised Chest expanded Torso out Smile Fists clenched Head back Arms open & up Body expansion

Controlled smile Small smile Wide smile X

Table 1. Characteristic Non-Verbal Expressions of Shame and Pride, Identified by Tracy

which lead to the emergence of specific emotions and in turn are shaped by cultural norms and value systems.

The Role of Culture and Emotion-Antecedent Appraisals in Non-Verbal Emotion Expression

According to appraisal theory, emotions are highly complex, multi-facetted phenomena which result from appraisal processes of events with respect to a person’s relevant goals, motives, and concerns (Sander, Grandjean, Kaiser, Wehrle, & Scherer, 2007). Appraisal theory posits that affective appraisals trigger conscious feelings, autonomic processes, as well as motor responses (Frijda, 1993). Facial expressions can therefore be seen as the result of a response pattern based on different appraisal outcomes which are translated into a particular expression. Importantly, appraisal theory holds that different appraisal patterns correspond to distinctively different emotions (Frijda, 1993). In line with this, Sander et al. (2007) hypothesised that observers who judge another person’s facial expression would base their judgement on the inference of the target person’s appraisal. The researchers assumed that an observer could infer the person’s appraisal processes from cues of cognitive activity in the face, such as the direction of gaze. Results showed that recognition rates were higher when people saw facial cues that they associated with specific appraisal processes. This supports the assumption that facial expressions indicate certain evaluative mental processes which are linked to specific affective states. Thus, if facial expressions are constituted by appraisal-processes, then different facial expressions are likely to correspond to different patterns of appraisals.

Furthermore, appraisal theory holds that evaluations of emotion-eliciting events are highly subjective, so that the same event can result in the emergence of different emotions in different individuals (Scherer, 1997). This can be linked to a vast amount of research indicating that there are various differences in norm and value structures, motivational processes, and self-concepts across cultures, pointing to great cultural variety in appraisal processes (e.g. Gudykunst et al., 1996;

Matsumoto, Yoo, & Fontaine, 2008). Because culture equips individuals with implicit theories about how to see the world, it provides a framework which guides people’s appraisals of the nature of situations (Matsumoto, 2008). Cultural differences in emotion processes have been linked to distinct cultural orientations which can be divided into individualistic and collectivistic orientations (Hofstede, 1980). Individualism is the dominant cultural orientation in Western countries, where people tend to see themselves as independent entities whose self-concept is clearly distinguishable from their social environment (Gudykunst et al., 1996). Collectivism, on the other hand, is more prevalent in East Asia, South-America and Africa. Individuals from collectivistic cultures define themselves mainly in terms of their social roles and group-membership. Their self-concept is an interdependent one and is intertwined with their social milieu. Members of collectivistic cultures tend to see their own goals, beliefs, and needs bound to those of their in-group. Therefore, emotions have less individual and greater interpersonal significance (Oyserman, Coon & Kemmelmeier, 2002).

Because cultural orientations have important implications for interpersonal relationships, they influence how people appraise others’ intentions and actions and in turn shape the way in which individuals express emotions non-verbally (Matsumoto, 2006). This was illustrated in a seminal study by Friesen (1972) which examined the spontaneous non-verbal expressions of Japanese and Americans. The participants watched a distressing film in two conditions: Alone and in the presence of a researcher. There were no differences in facial expressions between the participants when they watched the film alone. However, Japanese and Americans behaved differently when the researcher was present. While Americans continued to express their negative emotions freely, the Japanese participants tended to suppress their overt negative expressions and smiled more frequently instead. These striking results illustrate how culture shapes the non-verbal expression of emotions by defining norms about emotion regulation and affective expression in the presence of others.

Display Rules and Emotion Regulation

Cultural variability in emotion expression has been linked to specific display rules, which correspond to socially learned norms about appropriate and inappropriate display of emotions in different social contexts (Ekman, 1972). Display rules correspond to different emotion regulation strategies, which attune emotional experiences and expressions in a way that renders them

appropriate in a particular cultural setting (De Leersnyder, Boiger & Mesquita, 2013). There are two main types of emotion regulation strategies: namely antecedence-focussed regulation, which occurs early in the emotion process, before an emotion has fully emerged, and response-focussed

regulation, which takes place only when the emotion is already completely elicited (Gross, 1999). By employing a response-focussed strategy, individuals alter the experience or expression of an emotion. To do so one can suppress or amplify cognitions or expressive behaviour, including facial expressions. A person applying an antecedence-focused regulation strategy, seeks out or modifies certain aspects of situations and/or cognitions which promote a particular desired emotion or prevent the occurrence of an undesired one. De Leersnyder et al. (2013) extend this theory by positing that emotion-antecedent appraisal is a form of antecedence-focused emotion regulation. The researchers introduce the concept of cultural regulation which can be understood as a process in which emotional experience is being aligned with cultural norms, motives, and values. In contrast to emotion regulation, cultural regulation is not a deliberate, conscious process but takes place mainly automatically. This illustrates that cultural values provide a background against which events are appraised and which shapes the nature of the appraisals as well as the subsequent affective responses, including the non-verbal expression of emotions.

Cultural Models of Agency

In recent years, psychologists have increasingly focused on sociocultural models to explain variability in emotion processes and behaviour across cultures (Markus, Uchida, Omoregie,

Townsend, & Kitayama, 2006). Cultural models are defined as a set of culturally institutionalised beliefs about actions, motives, and communal practices. Cultural modes of agency differ between individualistic and collectivistic cultures and can be categorised into disjoint and conjoint models (Markus & Kitayama, 2003). In individualistic cultures, disjoint models of agency dominate. Individuals see actions as caused by independent, autonomous agents and are disjoint from others’ actions. Therefore, emotions tend to be appraised as belonging solely to the individual (Mesquita & Markus, 2003). By contrast, conjoint models of agency, dominant in collectivistic cultures, attribute the cause of an action to a collective of interdependent agents whose operations are intertwined with and influenced by the actions of others. The ownership of emotions is therefore mainly appraised as group based (Mesquita & Walker, 2003).

Studies suggest that cultural models can be linked to chronicle appraisal biases, which lead to differences in appraisal tendencies (Scherer & Brosch, 2009). This was confirmed by a seminal study, named Intercultural Study on Emotional Antecedents and Reactions (ISAER), in which participants from 37 countries took part (Scherer, 1997). Subjects filled in self-report measures regarding various emotions and appraisal dimensions. The results suggest that there is great consistency with regard to emotion specific appraisal patterns in different countries, pointing to universal appraisal mechanisms. However, there were considerable differences in appraisal

tendencies across cultures. Respondents from Latin-America would appraise events as less immoral than participants from other regions, while African respondents, when confronted with the same events, would appraise them as more immoral and externally caused.

The agency of emotion-eliciting events can be appraised as internal, thus as caused by the individual, or as external, indicating that the situation leading to a certain emotion is caused by something outside the self, such as a third person, one’s group, or deities (Scherer, 1997). Because emotions differ in terms of the situations that evoke their emergence (Scherer, 2009), it is

Especially more complex emotions, such as pride and shame, have been shown to differ across cultures.

Self-Evaluative Emotions: Pride and Shame

Pride and shame are termed evaluative emotions because they emerge as a result of self-evaluative processes and are closely intertwined with an individual’s self-concept (Mesquita, & Karasawa, 2004). As outlined above, cultural orientations shape the structure of a person’s self-concept. Due to its centrality in the emergence of pride and shame, they are likely to be highly influenced by cultural orientations.

A factor that might influence the emotion regulation of pride and shame is the degree of desirability of these emotions. In collectivistic contexts, self-critical emotions such as shame have been found to be more desirable than in individualistic cultures (Furukawa, Tangney & Higashibara, 2011). The reason for this can be found in the social functions of shame in collectivistic contexts. According to Wong and Tsai (2007), the expression of shame communicates an individual’s acknowledgement of social rules and the regret of trespassing these rules. Accepting shame and openly displaying it facilitates the reintegration of an individual into the in-group and can be seen as a tool for social cohesion. It can therefore be hypothesised that people from collectivistic cultures appraise the agency for shameful events as internal and see themselves as the owner of the emotion.

The opposite is the case for pride. According to Tracy and Robins (2004), pride is generally viewed as a less desirable emotion in collectivistic cultures and is more often experienced in response to accomplishments of one’s group than to individual accomplishments. This is in line with a study by Stipek (1998) in which members of a collectivistic culture attributed the agency of a pride-evoking event to the whole of their in-group (external agency) rather than appraising it as an individual accomplishment (internal agency). It was concluded that although pride in itself is generally not a desirable emotion in collectivistic cultures, it is favourable if it includes others.

When ascribed to the in-group as a whole, pride can function as a means to strengthen social cohesion (Wong & Tsai, 2007). Therefore, in collectivistic cultures, the agency of a pride-eliciting event is likely to be appraised as external and the ownership of pride is likely to be attributed to an individual’s in-group.

The results of ISAER have indicated that there is considerable variability in emotion and appraisal processes within the cluster of collectivistic nations (Scherer, 1997). Yet, the majority of research investigating emotion processes in collectivistic cultures has been conducted solely with people form East-Asian countries, like Japan or South Korea, which are rich and industrialised. By contrast, very little emotion research so far has taken place in collectivistic countries in other world regions, like Africa. Considering that circa 16.5 percent of the world’s population lives on the African continent (United Nations, 2017), more research in Africa is needed in order to deepen our understanding of emotion processes around the world. In order to account for this, the current study was conducted in a collectivistic culture in Africa.

Furthermore, most studies about emotion-antecedent appraisals and emotion expression have exclusively employed self-report measures. However, it is questionable whether introspection alone is an appropriate measure, because people might not be used to, or simply not good at, reflecting on their emotions and appraisal processes. Also, self-report measures can only assess information in retrospect, which might lead to the loss of valuable information that is not properly recalled. In order to deepen the understanding of appraisal processes as well as non-verbal

expressions, both verbal and non-verbal information should be assessed. The current study

accounted for this by employing observational measures and can thereby provide insights into both cognitive processes (in the form of appraisals) and verbal as well as non-verbal emotion expression.

The Current Study

The current study is part of a larger research project about affective communication across cultures. Next to Ghana, the project has also involved data collection in Canada, Tanzania, India, Turkey, and the Netherlands, and is still ongoing. In the present study, a group of teachers from an urban region in Ghana was asked to speak about a pride- and a shame-eliciting event with an

imagined acquaintance. In order to assess their non-verbal expressions, participants were filmed and the frequency of expressions was coded. Agency appraisals were assessed by recording participants’ speech and counting the frequency of external and internal appraisals, as well as terms referring to emotion ownership. Cultural orientation was measured with the cultural value scale (CVSCALE) (Yoo, Donthu, & Lenartowicz, 2011).

Importantly, because very little emotion research so far has been conducted in Ghana (or West-Africa for that matter), the overall character of the present study was highly exploratory. Prior research gives rise to expectations about the results of the current study, but it is unknown whether the theories outlined above can really be applied to the Ghanaian context. Therefore, this study is a mix of theory driven and exploratory elements, which is reflected in the research questions and expectations.

The current study investigated three hypotheses:

I. Pride and shame correspond to different categories of agency appraisals and to different attributions of the emotion ownership.

• It was expected that the pride-inducing event would correspond to a higher frequency of external than internal appraisal of agency and that it would be higher in pride than in shame. Also, it was expected that the ownership of pride would be more frequently attributed to the group.

• Furthermore, it was expected that the shame-eliciting event corresponds to a higher frequency of internal than external appraisals of agency, and that internal appraisals

would be more frequent in shame than in pride. Additionally, the agency of pride was expected to be more frequently attributed to the self.

II. Pride and shame each correspond to a set of distinct non-verbal expressions.

• For the theory driven part of the analysis, it was tested whether shame and pride correspond to the same characteristic expressions found by Tracy and Matsumoto (2008). Shame would thus correspond to a closed posture; (narrowed chest and slumped shoulders); pride would correspond to a tilting back of the head, stretched out and raised arms, smile, and expansion of the body (narrowed chest, torso out)

• For the exploratory part, it was investigated whether there are other characteristic pride and shame expressions than those identified by Tracy and Matsumoto (2008).

III. External and internal appraisals of agency correspond to distinctively different sets of non-verbal expressions.

• It was expected that expressions which are associated with shame correspond to internal agency appraisals, whereas expressions associated with pride correspond to external agency appraisals.

Method Participants

The participants were 28 high-school teachers, 21 men and seven women with a mean age of 37.8 years from Accra, Ghana. Ghana is a multi-ethnic country with several local languages spoken by different tribal groups, so that not all Ghanaians are fluent in English (Ghana Statistical Service, 2015). However, all participants in the current study were fluent English speakers, as this was an important prerequisite to ensure that they were able to comprehend the task descriptions and questionnaires, which were provided in English. 15 participants had another mother tongue than

English and 10 had several mother-tongues. All participants stated to be religious, 24 were

christians and one muslim. The participants were recruited by psychology students of the university of Ghana, who approached local high-schools and provided information about the current research for interested teachers. The participants received 10 Cedi (ca. 2€) and a pen as a reward for their participation.

Procedure

The current study consisted of two parts: In the first part, participants’ verbal and non-verbal emotion expression was assessed. In the second part, participants filled in self report measures concerning their cultural orientation (see Appendix 1 for the complete data collection manual).

Participants sat on a chair in a well-lit room in front of a desk. A video camera was placed on the table, so that the participants were filmed from the front, and face as well as upper body were clearly visible. First, participants were asked to read an information brochure carefully and to fill in an informed consent. Afterwards, the researcher was instructing the participants about the procedure of the research, without mentioning that the study was about emotion expression to prevent priming effects.

After the general instruction, the researcher switched on the camera and microphone and ran a trial session to familiarise the participants with the research procedure. The researcher emphasised that the participants were asked not to exceed the talking time of three minutes, which was indicated by an hourglass. Subsequently, the researcher left the room.

In the first part of the study, participants were presented with two emotion-eliciting vignettes, each describing a shame and a pride-eliciting situation. Participants were instructed to read the vignette text carefully and to imagine the situation at hand happening to them. After reading the instructions, participants turned the hourglass that was placed next to them, and were subsequently simulating a video call, in which they told their imaginative acquaintance about the

event. After each vignette, participants filled in a short questionnaire, which served as a manipulation check.

In the second part of the study, after the participants went through all vignettes and

questionnaires, they were asked to call in the researcher who turned off the camera and gave them further instructions. Participants were then asked to fill in a questionnaire about their cultural orientation. After completing this questionnaire, participants were thanked and debriefed but asked not to reveal the purpose of the study to their coworkers before they had had participated in the study themselves, in order to prevent the priming of subsequent participants.

Materials

In the present study, the observation of participants’ naturalistic verbal and non-verbal emotion expression was combined with self-report measures. Participants’ verbal and non-verbal behaviour was only coded during the interaction with the imagined acquaintance, not while participants were reading and filling in the questionnaires.

Pride and shame were elicited with the help of two different vignettes which contained the description of either a pride or a shame inducing event. The vignettes were pilot-tested and

counterbalanced, so that subjects received both vignettes in varying order. An example of the shame inducing vignette is the following:

• You are attending a teachers meeting hosted by the principal and an important government member is visiting your school. The government official asks a question and you answer the question with enthusiasm. It turns out to be the wrong answer and the official quickly turns away from you. Your colleagues start giggling.!

In order to assess non-verbal emotion expression, participants were filmed with small video cameras. The video footage was subsequently coded with an adapted coding scheme (see Appendix 2) based on Ekman and Rosenberg’s (1997) Facial Action Coding System (FACS). The coding

scheme was altered so that the large amount of facial Action Units used in the FACS were merged resulting in a smaller amount of expressions which are easier to code. Expressions were assessed in different parts of the face (e.g. upper and lower face), and head and body postures as well as

gestures were taken into account. The frequency of every expression was coded in the pride and in the shame inducing scenario.

Next to the vignettes, the current study made use of self-report measures. A short

questionnaire was presented after each vignette, which assessed participants’ affective experiences during each of the emotion-inducing scenarios and served as a manipulation check.

Verbal emotion expression was assessed by recording participant’s speech patterns with external microphones. The verbal content was coded with a self developed codebook (see Appendix 3) and processed with the Computer Assisted Textual Markup and Analysis (CATMA) programme. For the purpose of this study, the frequency and sort of agency appraisals, as well as emotion ownership attribution, and the frequency of references to god were assessed and used in the subsequent analysis.

Cultural orientation was measured with the cultural values scale (CVSCALE), which assesses cultural values on an individual level on five different dimensions (Yoo, Donthu, & Lenartcowicz, 2011). In the current research, only the dimension individualism/collectivism was used in the further analysis. The collectivism sub-scale consists of six items that were scored on a 7-point scale from 1 (strongly disagree) to 7 (strongly agree), with a minimum score of 6 and a maximum score of 42. A higher score indicates that an individual has a more collectivistic orientation. Examples of the items are: “Individuals should sacrifice self-interest for the group”; “Group welfare is more important than individual rewards”. The collectivism sub-scale is widely used for assessing cultural orientation and has shown a good reliability across different studies (e.g. Chan, Yim, & Lam, 2010; Soares et al., 2007), between α=.61 and α= .93.

Results

Importantly, because of the small sample size and the overall highly exploratory character of this study, the results are not conclusive. They can, however, be used as a starting point for further research.

Three of the 28 participants had to be excluded from the analysis: Two did not follow the instructions correctly and in one case there were administrative problems.

Cohen’s κ was performed in order to assess inter-rater reliability for the coding of the non-verbal expressions in the pride and shame scenarios. The results indicate that there was a substantial agreement between the two raters, κ = .690, p < .001.

Because there was only one group of participants in the current study, a comparison of scores on the collectivism sub-scale of the CVSCALE between groups was not possible. However, the total mean score of M = 31.5 (SD = 5.12) on a scale with an individual maximum score of 42 indicates that the participants of the current study were oriented more towards collectivism than individualism.

Pride, Shame and the Appraisal of Agency

To test the hypothesis that pride and shame correspond to different types of agency

appraisals, a 2x2 repeated measures factorial ANOVA was performed in order to compare the main effects of emotion category (two levels: shame, pride) and appraisal category (two levels: external, internal), as well as their interaction effect on the frequency of agency appraisals.

The results showed a trend towards significance for the main effect of emotion category (shame, pride) on appraisal frequency, F(1, 22) = 3.61, p = .070,

η

main effect of appraisal type (external, internal) on the frequency of appraisals, F(1,22) = 2.77, p = . 110,

η

internal appraisals of agency differed in frequency depending on emotion category, indicating that shame and pride correspond to different types of agency appraisals. As Figure 1 illustrates, the mean frequency of the external appraisal of agency is higher in pride (M = 1.52, SD = 2.89), than in shame (M = 0.26, SD = 0.54), whereas the frequency of the internal appraisal of agency is higher in shame (M = 0.48, SD = 0.90), than in pride (M = 0.22, SD = 0.52).

In order to gain more insights into the nature of the interaction effect between appraisal and emotion category, a number of exploratory analyses were conducted.

First, it was tested whether references to god were related to the higher prevalence of external agency appraisals in the pride scenario. This additional analysis was performed because it became apparent during coding of the verbal emotion expressions that participants talked overall very frequently about god, especially in the pride scenario. It seemed therefore worthwhile to explore the role of references to god in order to deepen insights into the interaction effect between emotion category and appraisal type. For this end, a simple linear regression was performed in

!18 Emotion Pride Shame Mean Frequency 1.50 1.25 1.00 .75 .50 .25 internal external Agency Page 1

which the frequency of references to god was used to predict the frequency of external agency appraisals in the pride scenario. References to god were found to significantly predict external agency appraisals in the pride scenario, R2 _{= 84%, F(1,21) = 107.87, p < .001, indicating that}

references to god can indeed be linked to the higher prevalence of external agency appraisals in the pride scenario.

To account for the remaining unexplained variability, the role of emotion ownership was explored. Ownership of emotion was divided into group and self ownership and measured by counting the frequency of terms relating to the self or group as owner of the particular emotion. For self ownership, first-person pronouns relating to the self as the owner of the emotion were coded (e.g. I felt ashamed). For group ownership, group reflexive pronouns relating to the group as the owner of the emotion were coded (e.g. We can really be proud).

In order to assess the combined impact of references to god and group ownership on the high prevalence of external appraisals in the pride scenario, a multiple regression was performed in which references to god and group ownership served as predictors for the frequency of external agency appraisals in the pride scenario. The results are summarised in Table 2.

Table 2. Linear Model of Predictors of the Frequency of External Agency Appraisals in the Pride Scenario Step 1 _R2 _{= .84 b SE B β p} Constant .37 (-.57, .65) .25 .901 Reference to god (pride) 1.45 (1.10, 1.64) .13 .92 <.001 Step 2 _R2 _{= .90} b SE B β p Constant -.20 (-.38, .62) .24 .616 Reference to god (pride) 1.55 (1.30, 1.80) .12 1.04 <.001 Ownership group (pride) -.73 (-1.20, -.29) .22 -.28 .003

It was further explored whether the differences in external and internal appraisal of pride and shame are related to the ownership of the emotion. A 2x2 repeated measures factorial ANOVA was performed in order to investigate this question. The main effects of emotion category (two levels: shame, pride) and ownership category (two levels: group, self), as well as their interaction effect on the frequency of ownership related verbal terms were compared. It was expected that pride would correspond to a higher frequency of group ownership terms, while shame would correspond to a higher frequency of self ownership terms. As Figure 2 illustrates, there was a significant main effect of ownership category, F(1,22) = 77.70, p < .001,

η

ownership related terms was significantly higher for self (M = 4.90, SE = 0.51) than for group (M = 0.22, SE = 0.36). The results showed no significant main effect of emotion category, F(1, 22) = 2.02, p = .169,

η

emotion category and ownership category, F(1,22) = 4.41, p = .684,

η

and shame do not correspond to different ownership categories.

!20 Emotion Pride Shame Me an Fre q u enc y 6 5 4 3 2 1 0 Group Self Ownership Page 1

Figure 2. Main effect of ownership category on the mean frequency of ownership related terms

The Non-Verbal Expression of Pride and Shame

For the theory-driven part of the analysis, the nine expressions, which had been linked to pride and shame in Tracy and Matsumoto’s (2008) study, were matched with the expressions from the coding scheme employed in the current study (see Table 1). Because the expression “fists” was not featured in our codebook, it was not taken into the analysis. Furthermore, the expressions “arms out” and “arms raised” in Tracy and Matsumoto’s (2008) study were represented by ”open and up arms” in the current study. Also, “chest expanded” and “torso out” were represented by “body expansion”, and chest narrowed and “shoulders slumped” were represented by “closed posture”. Because the coding scheme of the current study distinguished between three different types of smiles whereas Tracy and Matsumoto (2008) summarised all smiles in one category, the expression “smile” from Tracy and Matsumoto’s (2008) study was represented by controlled smile, small smile and wide smile in the current study.

Paired sample t-tests were performed for the seven expressions. Only one of them, wide smile t(22) = 4.18, p < .001, r = 0.70, showed a significant difference between the two emotions, and was greater in pride (M = 5.30, SD = 2.85) than in shame (M = 1.35, SD = 1.80) (see Figure 3).

For the exploratory part of the analysis, all expressions that were coded less than 15% of the time in both the pride and the shame scenario were excluded from the analysis. This resulted in 36 expressions, which were sorted according to different face and body regions and are summarised in Table 3. For each of the expressions, a paired samples t-test was performed to establish which frequencies differed significantly from each other in shame and pride. Of the 36 tested expressions, five were significantly different or showed a trend towards significance in pride and shame,

respectively (see Figure 4).

Also results that only showed a trend towards significance were included in the analysis because, given the small sample size, the chance to encounter a type II error was high, which means that p-values that fall marginally above the demarcation criterium of p < .05 are still likely to be

indications of an effect, which could be detected in a bigger sample (Batterham & Hopkins, 2006). In order to gain more information of the importance of the effect, effect sizes were calculated for the five expressions. All lay between r =.40 and r = .70, which corresponds to medium to large effects.

One expression differed significantly in the upper face region, namely fixed gaze t(22) = -2.91, p = .007, r = .53, which was greater in shame (M = 1.78, SD = 2.22) than in pride (M = 0.39,

SD = 0.72). Furthermore, there was a trend towards significance of furrowed brows, t(22) = -1.99, p

= .059, r = .40, which was also greater in shame (M = 5.39, SD = 5.90) than in pride (M = 3.40, SD = 4.10). In the lower face, wide smile t(22) = 4.18, p < .00, r = .70, was more often present in pride than in shame, whereas lip press, t(22) = 4.18, p = .036, r = .43, was more frequently displayed in shame (M = 2.40, SD = 3.18) compared to pride (M = 1.40, SD = 1.82). Also, there was a trend towards significance of small smile t(22) = 41.97, p = .061, r = .40, which was greater in pride (M = 0.60, SD = 1.10) than in shame (M = 0.09, SD = 0.29).

Expression Shame Pride

Upper Face Eyebrow raise .96 .84 Oblique brows .42 .36 Furrowed brows .83 .64 Gaze far .71 .56 Gaze down .96 .92 Gaze up .58 .52 Fixed gaze .63 .24 Head/Gaze to side .42 .92 Closed eyes .67 .60

Direct eye contact .96 .80

!22

Lower Face Wide smile .54 .80 Small smile .08 .28 Controlled smile .21 .32 Lip press .71 .60 Lip lick .59 .56 Lip suck .22 Lip pucker .19 .22 Head movements Head down .75 .80 Head back up .59 .44 Head bounce .25 .28 Nod .71 .72 Head tilt .96 .92 Head turn .96 .84 Head forward .37 .60 Body posture Cosed posture .25 .08

Open and up arms .08 .20

Sit up .17 .20 Clasp hands .34 .34 Fidgeting .50 .48 Shoulder shrug .13 .20 Crossed arms .21 .16 Lean forward .50 .44 Lean back .50 .28 Listening gestures

Pause & passive listening .79 .64

!24

Closed posture Body expansion Open & up arms Controlled smile Small smile Wide smile Head back 5.00 4.00 3.00 2.00 1.00 0.00 pride shame Emotion Page 1

Figure 3. Theory driven analysis: Mean frequency levels of non-verbal expressions in pride and

shame. Adapted from "The spontaneous expression of pride and shame: Evidence for biologically innate nonverbal displays." by Tracy and Matsumoto, 2008, Proceedings of the National Academy of Sciences, 105(33), 11655-11660. * = p<.05.

Expression Lipp press Fixed gaze Furrowed brows Small smile Wide smile Me an Fre q u enc y 6.00 5.00 4.00 3.00 2.00 1.00 0.00 Pride Shame Emotion Page 1

Figure 4. Exploratory Analysis: Mean frequency levels of non-verbal expressions in pride and

shame. * = p<.05; ** = p<.07 (trend towards significance).

M ea n F re qu en cy Expression * * * * _** **

In order to deepen insights into the nature of the relationship between emotion category and expression category a 2x5 repeated measures factorial ANOVA with simple contrasts was

performed in which the main effects of emotion category (two levels: shame, pride) and appraisal category (five levels: wide smile, small smile, fixed gaze, furrowed brows, lip press), as well as their interaction effect on expression frequency was assessed.

Maulchy’s test showed that the assumption of sphericity had been violated for the main effect of expression category, 𝝌2 _{(9) = 47.54, p < .001, and the interaction effect of emotion}

category x expression category, _𝝌2 _{(9) = 54.10, p < .001. Greenhouse-Geyser estimates of sphericity}

were used to correct the degrees of freedom for the main effect of expression category (ε = .53) and the interaction effect of emotion category x expression category (ε = .52).

There was no significant main effect of emotion category on the frequency of expressions,

F(1, 22) = 0.001, p =.976, η2_{p = .0, which indicates that there is no higher overall frequency of}

expressions in one of the two emotions. However, there was a significant main effect of expression category, F(2.116, 46.551) = 9.65, p <.001,

η

frequency of occurrence. This did neither confirm nor go against expectations, as this part of the analysis was mainly exploratory. Planned contrasts were performed for the five levels of expression category: Within the pride relevant behaviours, wide smile was significantly different from small smile; within the shame relevant behaviours furrowed brows was significantly more frequent than fixed gaze and lip press; lip press was more frequent than fixed gaze.

Also, as already indicated by the paired samples t-tests in the first part of the analysis, there was a significant interaction effect between emotion category and expression category, F(2.082, 45.808) = 11.593, p <.001,

η

expression category were different on the two levels of emotion category, shame and pride (see Figure 4).

Non-Verbal Expression and the Appraisal of Agency

In order to test the hypothesis that the non-verbal expressions associated with pride and shame correspond to specific patterns of appraisal, several simple linear regressions were

performed. Only expressions that had been shown to be significantly higher in pride or in shame were taken into the analysis because it was expected that they would correspond to different frequencies of external and internal agency appraisals. It was expected that expressions associated with shame would correspond to internal appraisals and expressions associated with pride would correspond to external appraisals. External and internal appraisals were used to predict each of the five expressions that differed between pride and shame, displayed in Figure 3. The results of the regression analysis are summarised in Table 3. Contrary to the hypothesis, the only significant result was obtained in internal appraisals, which was found to significantly predict small smile, R2 ₌

!26 Expression Lip press Furrowed brows Fixed gaze Small smile Wide smile Me an Fre q u enc y 6.00 5.00 4.00 3.00 2.00 1.00 .00 Pride Shame Emotion Page 1

Figure 4. The main effect of expression category and the interaction effect between

32%; F(1,21) = 9.84, p = .005, indicating that there is variability within the expressions which is not driven by agency appraisals.

Table 5. Linear Models of Internal and External Agency as Predictors of Non-Verbal

Expressions

Internal External

Discussion

This study examined the relationship between culture, the emotion-antecedent appraisal of agency and the non-verbal expression of pride and shame in Ghana.

In line with expectations, the scores of the CVSCALE revealed that the participants of the present study were more collectivistic than individualistic in their cultural orientation. The

manipulation check indicated that the manipulation was effective. The results provided mixed evidence for the three underlying hypotheses.

b SE B β p b SE B β p Wide smile -1.56 (-.64,.85) 2.04 -.17 .453 0.11 0.36 (-.03,.19) .07 .763 Small smile 0.85 (.29,1.42) 0.27 .57 .005 0.08 0.06 .30 .167 Fixed gaze -0.20 (-.32,.91) 0.54 -.08 .708 -1.00 0.71 (-.48,.48) -.29 .174 Fur-rowed brows -0.91 (-3.87, 2.02) 1.42 -.14 .524 -2.61 1.90 (-60,1.34) -.29 .183 Lip press -0.68 (-2.35, .98) 0.80 -.18 .402 0.55 1.12 (-.07-.19) .11 .682

Pride, Shame and the Appraisal of Agency and Emotion Ownership

First, the findings support the hypothesis that shame and pride correspond to different appraisal categories. In line with expectations, shame was found to correspond to the internal appraisal of agency, which was manifested in a higher frequency of verbal references to internal agency appraisals. Pride was linked to the external appraisal of agency, which resulted in a higher frequency of verbal references to external agency appraisals.

The difference in the frequencies of external agency appraisals between pride and shame could be attributed to the fact that participants were more frequently referring to god when they talked about the pride-eliciting event than when they talked about the shame-eliciting one. These findings suggest that participants only attributed the agency for positive, pride-causing events to god but not the agency for negative events which lead to the experience of shame. These results might be linked to participants’ religious beliefs, as religion plays a central role in the lives of the majority of Ghanaians (Graveling, 2010). All participants in the current study indicated that they were religious and all but one stated to be christian. Anthropological studies indicate that the appraisal of agency of events to supernatural forces such as deities or spirits has always been an important aspect of traditional African belief systems (Pokimica, Addai, Takyi, 2011). It might be that this tendency subsists in a form that is adapted to the new monotheist religions, so that events that elicit pride are appraised as caused by god, who, in the Christian and Islamic belief systems, is conceptualised as benevolent and kind. This, however, is purely speculative and should be

addressed in future research by comparing the agency appraisals of shame- and pride-eliciting events between members of different religions. The findings highlight that there might be other meaningful cultural dimensions next to cultural orientation, such as religious belief, that constitute differences in appraisal and emotion processes within the cluster of collectivistic cultures. This shows that more research is needed which directly compares members of different collectivistic cultures to see where differences and similarities lie in this diverse group.

The findings also raise the question whether the differences found between pride and shame are emotion-specific or valence-driven. Furthermore, it is unclear whether the findings only apply to self-conscious emotions or whether an effect can also be found within the cluster of basic emotions. Because the current study compared two self-conscious emotions, one of which was negative and one positive, this is unclear at the present point. These questions have to be addressed in future research by comparing appraisal patterns between self-conscious as well as basic emotions and by assessing whether similar differences in appraisal patterns can be found between several positive or negative emotions.

Contrary to expectations, there was no difference in the attribution of emotion ownership between shame and pride. In fact, pride as well as shame were more often appraised as owned by the self than the group. These findings are interesting because they contrast results of prior studies, predominantly conducted in collectivistic countries in East Asia, which found that pride, in

particular, is more often appraised as group based.

A possible explanation for these results is that pride was operationalised in a culture fair way, so that the pride-eliciting event could be interpreted as “good for yourself” as well as “good for the group”. It is therefore possible that the participants of the current study, who had a

predominantly collectivistic orientation, interpreted the pride scenario as an event that is beneficial for the whole of their in-group. This could have made it more socially acceptable to appraise the ownership of pride as owned by the self. However, due to the way in which pride was

operationalised in the current study, it remains unclear whether the unexpected results were due to differences in the attribution of the emotion ownership between the participants of the current and prior studies, or whether the findings were caused by participants’ interpretation of the nature of the pride-eliciting event. This could be addressed in future research by including a second group with an individualistic cultural orientation in the study and comparing the attribution of emotion ownership between the two groups.

The Non-Verbal Expression of Pride and Shame

The results supported the hypothesis that pride and shame are expressed by a number of distinct non-verbal expressions. Pride and shame differed in five out of the 36 tested expressions. Pride was thereby characterised by a wide smile and small smile, which showed a higher frequency in the pride- compared to the shame-eliciting scenario; shame was characterised by a fixed gaze, lip press, and furrowed brows, which was manifested in a higher frequency of these expressions in the shame scenario compared to the pride scenario. These sets of expressions did largely not correspond to the ones that had been identified as being characteristic for pride and shame by Tracy and

Matsumoto (2008), on whose research the theory driven part of the current study was based. Only the pride expressions wide and small smile were in line with Tracy and Matsumoto’s (2008)

findings. However, all other pride expressions did not correspond to the current findings; neither did the shame expressions.

The discrepancy between the findings by Tracy and Matsumoto (2008) and the current study could be due to the fact that Tracy and Matsumoto observed pride and shame expressions in a real-life situation, whereas the present study elicited pride and shame in an artificial setting. It is thus possible that participants did express their emotions differently in the context of the present research, in which they were only imagining the emotion-eliciting situation, than they would have done in a real life situation. Although the manipulation check did indicate that participants felt and expressed shame and pride in the respective situations, it is nonetheless conceivable that there are differences in non-verbal expressivity in an artificial and in a real life context. However, it is rather unlikely that the sets of expressions associated with pride and shame would differ to such a great extent, solely based on the differences between artificial and real-life settings.

Another factor that could have contributed to the discrepancies is the nature of the events that were used to elicit pride and shame in the current study and the research by Tracy and Matsumoto (2008). Whereas shame and pride were evoked as a result of success and failure in a

physical competition in Tracy and Matsumoto’s (2008) study, they resulted from social success and failure in the present study. It could therefore be argued that Tracy and Matsumoto (2008) tested a different aspect of pride and shame, which might be expressed by a different set of expressions. Thus, the feeling of pride that is evoked by a victory in a sports competition might be related to expressions that emphasise physical dominance in a more primal sense (hence body expansion, raised arms etc.), whereas pride evoked by social appreciation might be characterised by a different set of expressions, which is more adapted to the interpersonal aspect of the situation. Equally, shame as a reaction of physical defeat might correspond to expressions that signal physical

subordination (such as visibly slumped shoulders), while shame that is caused by social humiliation would be expressed in a different, more socially adapted manner. To test this assumption, future research should directly compare pride and shame expressions which are caused by success and

failure in a sport-related competition and in an interpersonal context.

Non-Verbal Expression and the Appraisal of Agency

The findings did not support the hypothesis that there is a direct relationship between the non-verbal expression of pride and shame and the verbal appraisal of agency. Contrary to

expectations, none of the expressions that had been connected to shame in the course of the analysis were associated with internal agency appraisals. Similarly, none of the expressions linked to pride were associated with external agency appraisals. Only one expression, associated with pride (small smile) was, unexpectedly, found to be linked to the internal appraisal of agency. While this suggests that there is variability within the cluster of expressions, it indicates that the non-verbal expressions of pride and shame cannot be directly connected to a specific pattern of verbal agency appraisals.

A possible explanation for these results might be the fact that the current study did not measure the agency appraisals of pride and shame directly, but the communication about

(a) the communication of one’s appraisals differs from the actual appraisal patterns which are altered in the process of communication, for example due to social desirability, so that there is no link between the communication of participants’ agency appraisals and their nonverbal expressions; and (b) if there was a direct relationship between the characteristic pride and shame expressions and the appraisal of the agency of pride and shame, this could possibly not be measured through the indirect link between non-verbal expressions and participants’ agency appraisals of the emotion-eliciting events. In order to clarify this point, future research would have to measure the appraisal of agency of pride and shame directly and assess whether this can then be linked to a specific set of non-verbal expressions. However, it is very difficult to measure the direct appraisal of the agency of emotions with observational measures, because participants can only report about their appraisals concerning the emotions in retrospect. This could subsequently bias the results because participants might not correctly recall their appraisals or might not be able to analyse them accurately enough.

Another important point that should be mentioned is the role of individual differences both in non-verbal expressivity and appraisals. Due to the highly subjective nature of appraisal

processes, emotional reactions can differ substantially between individuals, even when the emotion-eliciting event is the same (Scherer, 2009). Research indicates that specific emotions are to a large extent caused by a person’s subjective appraisals of a given situation which in turn depend on factors such as personality traits. It is possible that the high degree of individual variability in both the appraisal of agency and the non-verbal expression of pride and shame made it difficult to detect any systematic effects. In order to clarify this point, a follow-up study could be conducted in which, next to cultural orientation, also personality traits would be measured and related to specific agency appraisals, such as external and internal agency.

Limitations

The major limitation of the current study was the small sample size which impaired

statistical power and made it difficult to detect smaller effects. Therefore, the present results are not suited to provide clear or strong evidence. However, in the light of the overall exploratory character of this study, which protruded into little researched terrain, the results can serve as an early step in a new direction of research.

Related to this point, there was an unequal number of male and female participants in the current study, so that only one quarter of participants were women. This could have lead to confounding effects because there might be meaningful differences in non-verbal expressions and appraisal processes between men and women. Prior research indicates, for example, that men show less non-verbal expressivity than women (e.g. Sokolov, Krüger, Enck, Krägeloh-Mann, & Pavlova, 2011). This impairs the generalisability of the present findings. However, due to the small sample size, the influence of gender on the results of the current study is not clearly identifiable. A follow-up study should address this issue by including a higher number of participants and an equal distribution of men and women in the study.

Another limitation were the circumstances of data collection in the Ghanaian schools. Often, there was no electric light so that the video footage was at times quite dark and facial expressions were difficult to code. Also, the majority of schoolrooms in which the research was conducted did not have doors, and were therefore rather noisy. These and other disturbances, for example in the form of teachers or students entering the room, were a source of distraction for the participants. These issues should be addressed in future research by selecting different places for interviewing participants, which can provide a calmer environment. However, this might be difficult to

implement in a developing country like Ghana where facilities with good research conditions can be difficult to obtain.

Furthermore, the current study consisted of only one group of participants from a single country and could therefore not compare effects between cultures. The influence of culture could therefore only be tested relative to findings from prior studies, which makes it difficult to evaluate the effects of cultural orientation in relation to the present findings. This should be addressed by conducting a follow-up study in which participants from two different cultures are directly compared.

Conclusion

Although the present study was conducted on a very small scale and the results should be interpreted cautiously, the findings point to important aspects of the relationship between specific emotions, emotion-antecedent-appraisal, and non-verbal emotion expression and can serve as a starting point for future research.

The current study supports the notion of distinct, characteristic non-verbal expressions of pride and shame. Furthermore, the findings illustrate the importance of agency appraisal processes in the emergence of emotions by (a) indicating that pride and shame correspond to distinct

categories of agency appraisals and (b) by suggesting that there is a relationship between shame and pride and the appraisal of agency. This highlights the important role that appraisal processes play in shaping emotions and their non-verbal manifestations.

At the same time, the findings might purport to important differences in the non-verbal expression of emotions between different collectivistic cultures, which appear to contradict prior research about the universality of pride and shame expressions, as well as theories concerning the relationship between cultural models of agency and the emergence of emotions. This highlights the need to expand the current scope of research to include participants from more diverse world regions and cultural backgrounds, beyond the Western world and East Asia.

References

Argyle, M. (2013). Bodily communication. Abingdon, UK: Routledge.

Batterham, A. M., & Hopkins, W. G. (2006). Making meaningful inferences about magnitudes.

International journal of sports physiology and performance, 1(1), 50-57.

Chan, K. W., Yim, C. K., & Lam, S. S. (2010). Is customer participation in value creation a double-edged sword? Evidence from professional financial services across cultures. Journal of marketing, 74(3), 48-64.

Cohn, J. F., Schmidt, K., Gross, R., & Ekman, P. (2002). Individual differences in facial expression: Stability over time, relation to self-reported emotion, and ability to inform person

identification. In Proceedings of the 4th IEEE International Conference on Multimodal

Interfaces (p. 491). IEEE Computer Society.

De Leersnyder, J., Boiger, M., & Mesquita, B. (2013). Cultural regulation of emotion: Individual, relational, and structural sources. Frontiers in psychology, 4, doi: 10.3389/fpsyg.

2013.00055.

Ekman, P. (1993). Facial expression and emotion. American psychologist, 48(4), 384.

Ekman, P. (1992). Facial expressions of emotion: New findings, new questions. Psychological

science, 3(1), 34-38.

Ekman, P., & Friesen, W. V. (1971). Constants across cultures in the face and emotion. Journal of

personality and social psychology, 17(2), 124.

Ekman, P., & Friesen, W. V. (1969). The repertoire of nonverbal behavior: Categories, origins, usage, and coding. semiotica, 1(1), 49-98.

Ekman, P., & Rosenberg, E. L. (1997). What the face reveals: Basic and applied studies of

spontaneous expression using the Facial Action Coding System (FACS). Oxford University Press, USA.

Elfenbein, H. A., & Ambady, N. (2002). On the universality and cultural specificity of emotion recognition: a meta-analysis. Psychological bulletin, 128(2), 203.

Frijda, N. H. (1993). The place of appraisal in emotion. Cognition & Emotion, 7(3-4), 357-387. Furukawa, E., Tangney, J., & Higashibara, F. (2012). Cross-cultural continuities and discontinuities

in shame, guilt, and pride: A study of children residing in Japan, Korea and the USA. Self

and Identity, 11(1), 90-113.

Ghana Statistical Service, 2015. Ghana population trends.

Graveling, E. (2010). ‘That is not religion, that is the gods’: Ways of conceiving religious practices in rural Ghana. Culture and Religion, 11(1), 31-50.

Gross, J. J. (1999). Emotion regulation: Past, present, future. Cognition & Emotion, 13(5), 551-573. Gudykunst, W. B., Matsumoto, Y., Ting-Toomey, S. T. E. L. L. A., Nishida, T., Kim, K., &

Heyman, S. (1996). The influence of cultural individualism-collectivism, self construals, and individual values on communication styles across cultures. Human communication research,

22(4), 510-543.

Hofstede, G. (1980). Motivation, leadership, and organization: do American theories apply abroad?.

Organizational dynamics, 9(1), 42-63. Kaiser, S., & Wehrle, T. (2001). Facial expressions as

indicators of appraisal processes. Appraisal processes in emotion: Theory, methods,

research, 285-300.

Izard, C. E., Bartlett, E. S., & Marshall, A. G. (1972). Patterns of emotions. Acad. Press Kitayama, S., Markus, H. R., & Matsumoto, H. (1995). Culture, self, and emotion: A cultural

perspective on" self-conscious" emotions.

Kitayama, S., & Uchida, Y. (2005, March). Interdependent agency: An alternative system for action. In Cultural and social behavior: The Ontario symposium (Vol. 10, pp. 137-164).

Matsumoto, D. (2006). Culture and nonverbal behavior. Handbook of nonverbal communication, 219-235.

Matsumoto, D., Yoo, S. H., & Fontaine, J. (2008). Mapping expressive differences around the world: The relationship between emotional display rules and individualism versus collectivism. Journal of cross-cultural psychology, 39(1), 55-74.

Markus, H. R., & Kitayama, S. (2003, January). Models of agency: Sociocultural diversity in the construction of action. In Nebraska symposium on motivation (Vol. 49, pp.1-58). Markus, H. R., Uchida, Y., Omoregie, H., Townsend, S. S., & Kitayama, S. (2006). Going for the

gold: Models of agency in Japanese and American contexts. Psychological Science, 17(2), 103-112.

Mesquita, B., & Karasawa, M. (2004). Self-conscious emotions as dynamic cultural processes.

Psychological Inquiry, 15(2), 161-166.

Mesquita, B., & Walker, R. (2003). Cultural differences in emotions: A context for interpreting emotional experiences. Behaviour Research and Therapy, 41(7), 777-793.

Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and

colletivism: evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128, 3-72.

Pokimica, J., Addai, I., & Takyi, B. K. (2012). Religion and subjective well-being in Ghana. Social

Indicators Research, 106(1), 61-79.

Russell, J. A. (1994). Is there universal recognition of emotion from facial expressions? A review of the cross-cultural studies. Psychological bulletin, 115(1), 102.

Sander, D., Grandjean, D., Kaiser, S., Wehrle, T., & Scherer, K. R. (2007). Interaction effects of perceived gaze direction and dynamic facial expression: Evidence for appraisal theories of emotion. European Journal of Cognitive Psychology, 19(3), 470-480.

Scherer, K. R. (2009). Profiles of emotion-antecedent appraisal: Testing theoretical predictions across cultures. Cognition & Emotion, 11(2), 113-150.

Scherer, K. R. (1997). The role of culture in emotion-antecedent appraisal. Journal of personality

and social psychology, 73(5), 902.

Scherer, K. R., & Brosch, T. (2009). Culture-specific appraisal biases contribute to emotion dispositions. European Journal of Personality, 23(3), 265-288.

Soares, A. M., Farhangmehr, M., & Shoham, A. (2007). Hofstede's dimensions of culture in international marketing studies. Journal of business research, 60(3), 277-284.

Sokolov, A. A., Krüger, S., Enck, P., Krägeloh-Mann, I., & Pavlova, M. A. (2011). Gender affects body language reading. Frontiers in psychology, 2, 317-348.

Stipek, D. (1998). Differences between Americans and Chinese in the circumstances evoking pride, shame, and guilt. Journal of Cross-Cultural Psychology, 29(5), 616-629.

Tracy, J. L., & Matsumoto, D. (2008). The spontaneous expression of pride and shame: Evidence for biologically innate nonverbal displays. Proceedings of the National Academy of

Sciences, 105(33), 11655-11660.

Tracy, J. L., & Robins, R. W. (2008). The nonverbal expression of pride: evidence for cross-cultural recognition. Journal of personality and social psychology, 94(3), 516.

Tracy, J. L., & Robins, R. W. (2004). " Putting the Self Into Self-Conscious Emotions: A Theoretical Model". Psychological Inquiry, 15(2), 103-125.

United Nations, 2017. World population prospects, the 2017 revision. United Nations: New York. Wong, Y., & Tsai, J. (2007). Cultural models of shame and guilt. The self-conscious emotions:

Theory and research, 209-223.

Yoo, B., Donthu, N., & Lenartowicz, T. (2011). Measuring Hofstede's five dimensions of cultural values at the individual level: Development and validation of CVSCALE. Journal of

International Consumer Marketing, 23(3-4), 193-210.

Appendices

Appendix 1: Data Collection Manual Material

• Camera & charger (or extra battery)

• _{Tripod (or something to position camera on)}

• _{Microphone (in case the camera does not record sound well, test in advance)} • 40 numbered papers ( one for each participant)

• X printed informed consent & info brochures • X printed questionnaires

• Hourglass of 3 minutes • _{Glass of water}

• _{2 blue pens (to fill out informed consent & participant payment sheet)} • Table & chair

• A quiet, well-light room • SD cards for camera)

Setup

• _{The subject sits on a chair in front of a desk with the laptop in front of him or her. The} camera is behind the laptop or next to the laptop, so that the subject is filmed from the front. Face and upper body are clearly visible – zoom in if necessary.

• Tips:

o Make sure the camera is more or less on eye level. If the camera is too high, the subject is filmed from above. In this case it is difficult to judge if the person is for example reading from the laptop screen or looking down because of sadness or shame.

o Take into consideration that subjects sometimes move or change body position while talking (eg when talking about pride, some stretch and sit more upright, when talking about sadness they sink in). In all conditions subject should be clearly visible, so camera position needs to allow room for these changes (eg not zoomed in too much).