Bachelor Thesis Social Psychology
High or Low Context? General and Affective Communication of Ghanaians and
the Influence of Collectivism and Christianity
Jara Figalist
Supervisor: Milena Feldkamp, Disa Sauter Studentnr: 10550720
Amount of words: 7276, Abstract: 186 Date: 22th December 2016
2 Index
Abstract ... 3
Introduction ... Error! Bookmark not defined. Individualism-Collectivism and its Relation to High and Low Context General Communication ... 5
Low and High Context Affective Communication ... 6
A Criticism on Prior Research Findings ... 7
The Role of Religion in Affective Communication ... 9
Purpose of the Study ... 10
Methods ... 12 Participants ... 12 Procedure ... 13 Measures ... 14 Data Analysis ... 16 Results ... 16 Manipulationchecks ... 17
General Affective Communication in Ghana ... 18
Collectivism as Predictor for HC/LC (affective) Communication ... 18
Display of High or Low Context Characteristics in (affective) Communication ... 20
Christianity as Moderator ... 21
Differences in Affective Communication depending on Scenario ... 22
Conclusions and Discussion ... 22
Literature ... 28
Appendix A ... 31
3 Abstract
The present study examined whether Ghana shows a more high or low context general and affective communication style. Thereby, focus was put on collectivism as potential predictor for these. Additionally, it was investigated whether Ghana shows more
characteristics of a HC or LC (affective) communication style. Responding to prior criticism on collectivism as predictor, Christianity was offered as possible alternative explanation for the emergence of a LC (affective) communication style in Ghana. General communication style preferences were assessed by having 27 Ghanaian school teachers fill in self-report measures, thereby focusing on indirect and dramatic communication corresponding to
HC/LC. Affective verbal communication was assessed by analyzing speech-samples, thereby counting frequencies of valenced and specific emotion words which corresponded to HC/LC. Hypotheses were only partly confirmed. Collectivism did not serve as predictor for general, but partly for affective communication (valenced emotion words). Generally, Ghanaians preferred characteristics of a LC (affective) communication style (except for dramatic communication). Contrary to predictions, Christianity did not seem to moderate any of the relationships. Concluding, these findings suggest that Ghana may not fit so neatly into the HC/LC (affective) communication construct as expected.
4
After a loved person has passed away, one individual might feel the urge to verbally express his pain and sadness in a direct and open manner, whereas another communicates his feelings quietly and indirectly to the outer world.
People differ extensively in how they communicate emotions. While basic emotions like happiness and sadness are supposed to be universal (Niedenthal, Krauth-Gruber & Ric, 2006), there is less known about the universality of affective communication. Affective communication describes the expression of positive or negative emotions about things, situations, the self and other people (Gudykunst & Ting-Toomey, 1988). Thereby, affective communication itself is subject to the world’s complexity and seen as social process and, as such, it is influenced by traditions, religion and value systems of the culture in which it was found (Mitamura, Leu, Campos, Boccagno & Tugade, 2014). Interestingly, there seems to be a dearth of literature concerning affective communication across cultures since the main research focus in this area has been placed on general communication (Gudykunst & Ting-Toomey, 1988). Hence, stating that general and affective communication might share underlying patterns, the assumptions made for general communication could also apply for affective communication. However, if they do not work the same way, a distinct conceptual framework for affective communication is needed.
One attempt to capture general patterns regarding similarities and differences between cultures was the work undertaken by Geert Hofstede (Voronov & Singer, 2002). The most utilized dimension of his cultural dimensions’ theory is individualism-collectivism, defined as the “degree to which people in a society are integrated into groups” (Hofstede, 1984).
Individualistic cultures prefer a loosely-knit social framework in which individual goals and needs are of supreme importance and the predominant norm is to affirm the importance of the individual (Heine, 2012). Their counterpart, collectivism describes the preference for a tightly-knit social framework, thereby emphasizing the relatedness to the own in-group, cooperation and the predominant norm of maintaining group harmony (Eid & Diener, 2001).
5
Individualism-Collectivism and its Relation to High and Low Context General Communication
Prior research suggested that Individualism-collectivism (I-C) directly affect communication because it influences the norms that guide behavior and leads to the
preference of a certain communication style over others (Kapoor, Hughes, Baldwin & Blue, 2003). Hall introduced the idea of high context (HC) versus low context (LC) cultures in order to understand their basic differences in communication style (1976). Whereas prior research suggests that a HC communication style is predominant in collectivistic countries, individualistic countries favor a LC communication style (Gudykunst, Matsumoto, Ting-Toomey, Nishida, Kim & Heyman, 1996). In HC communication, greater emphasis is placed on the nonverbal aspects of communication and the internal meaning is usually embedded deep in the information, that is to say, not everything is explicitly stated or verbally expressed. For instance, Park and Kim found Japanese and Chinese students preferred a rather indirect, vague and understated way of communication (2008).In cultures like these, the listener is expected to understand the “unsaid” thanks to the context. Open communication is mainly avoided since it could endanger the goal of group harmony and discomfort or hurt in-group members (Gudykunst & Ting-Toomey, 1988). Since the indirectness of communication was often considered as highly essential and representative for a HC communication style, this study focuses solely on this factor (Park & Kim, 2008).
Low-context communication is just the opposite, “the mass of information is vested in
the explicit code’’ (Hall, 1976). Thus, it includes the use of explicit and direct messages in
which the meanings are contained mainly in the transmitted messages, leaving barely nothing to the context. For instance, prior research indicates that European-American’s favor a direct, precise, dramatic and open communication (Gudykunst & Ting-Toomey, 1988). Also, the tendency for dramatic communication is found as one of the factors that most distinguish a LC from a HC culture (Oguri & Gudykunst, 2002). Thus, differences in dramatic
6
communication of two countries would point to substantial discrepancies in underlying expressivity and intensity. Due to these reasons, the present study focuses solely on the factor of dramatic communication. Looking at general HC/LC communication style, the question remains how an affective HC/LC communication would look like.
High or Low Context Affective Communication Styles
Would people communicate emotions in line with the HC/LC communication style assumptions just outlined? Generally, people adopting a LC communication style are expected to communicate in ways that are consistent with their feelings (Gudykunst et al., 1996). This is, they might tend to talk openly about their happiness, thereby communicating their feelings with directness and a disposition for drama (Kittler, Rygl & Mackinnon, 2011). The verbal expression of specific emotion terms such as “happy” instead of a vague “I am good” fits well with the equivalent factors of preciseness and directness of the general LC communication style. Also, since personal feelings and their free expression affirm the importance of the individual compared to the group, people with a LC-communication style might show a greater overall expressivity in their affective communication (Matsumoto et al., 2008).
People adopting a HC communication style, in contrast, are expected to communicate in ways that conceal and mask their true feelings (Gudykunst et al., 1996). Since personal feelings and their free expression are relatively less important compared to their interpersonal meaning, emotions might be communicated in an ambiguous and rather vague manner. For instance, direct and open naming of an emotion such as sadness might cause discomfort in the listener and is therefore preferably communicated using understated and valenced emotion terms. “I am a bit down” implies a different intensity and verbal expressivity than “I am sad” and, supposedly, leaves more to the context.
Since many non-verbal aspects of affective communication are rather signs than intentional signals of an emotion, it seems difficult to derive different affective
7
communication styles from it (Fussel, 2002). Also, in order to create a rather broad picture of affective communication, verbal expression takes the greatest part in communication
behaviors and therefore seems best applicable for this purpose and study (Fussel, 2002). Also, across the whole study, information regarding affective communication is obtained by
focusing on the basic emotions happiness and sadness. Although the qualitative differences between these two emotions cannot be overlooked, this study does not aim to illuminate these differences. Rather, happiness and sadness are chosen due to their representative character for positive and negative basic emotions. Accordingly, both emotions are taken together in the analyses, assuming they do not differ in terms of frequency.
A Criticism on Prior Research Findings
Taking a head count around the world, about 80% of the world’s population is considered as collectivistic (Heine, 2012). As noted earlier, prior research findings suggest that I-C can predict communication style. Thus, a great part of the world should then adopt a rather HC (affective) communication style, thereby serving the underlying goal of
maintaining group harmony. But - can it be that simple?
In the following, two considerations are discussed which point to the possible inconsistencies of the argumentation outlined above. First, the I-C concept and its
applicability to HC/LC communication is scrutinized. Although this reasoning is partly based on research findings, the generalization from collectivism to a HC communication style has often been made too rapidly (Kapoor et al., 2003). Considering the world’s diversity, a black-white distinction into individualistic and collectivistic countries seems deceptive and hides dangerous pitfalls for overgeneralization. Triandis warned that “their (individualism and
collectivism) wide applicability also represents a danger, like the man with a hammer who uses it at every opportunity, if we do not sharpen their meaning, we can overuse the
8
often assumed, the construct also loses some of its validity to serve as a predictor for (affective) communication styles across cultures.
A culture may incorporate and combine elements of individualism and collectivism as well as characteristics of HC and LC (affective) communication (Nishimura et al., 2008). For instance, Kapoor and colleagues found that, even if collectivistic Indians engaged in more indirect communication than Americans, thereby confirming collectivism as predictor for HC, they also reported more dramatic communication, what rather suits a LC communication style (2003). Also, Nishimura and colleagues analyzed HC/LC communication styles and found that the individualistic Finland showed a preference for indirect communication norms, suiting a rather HC communication style (2008). Thus, collectivism might not always be an adequate predictor for (affective) communication styles.
The second consideration refers to the underlying goal that a HC communication style serves: maintain and guarantee group harmony. Assuming the I-C concept is valid, and the main goal in collectivistic countries is group harmony - does it necessarily result in a HC (affective) communication style? Presumably, there are several alternative possibilities for protecting group harmony which do not involve indirect communication and hiding one’s true feelings. Perhaps - a country is collectivistic, but that fact does not necessarily predict the emergence of an HC affective communication style. For instance, the Mexican culture has a script of simpatía that promotes group harmony through the open and rather extreme and dramatic expression and communication of emotions (Triandis, 1995). The open and direct communication of emotions is seen as attempt to create closeness and relatedness within the group and is interpreted as social caring (Triandis, 1995). Even though Mexico is a country with many collectivistic tendencies and values, the described characteristic of affective communication resembles more with a LC communication style. Showing a preference for dramatic emotion expression, evidence also suggested that Arabic countries do not fit as neatly as might assumed within a HC affective communication style (Fernández, Carrera,
9 Páez & Sánchez, 2008).
Thus, this study aims to address the inconsistencies found in prior research and investigates if collectivism actually represents an adequate predictor for HC/LC (affective) communication styles. Additionally, more information about HC/LC(affective)
communication in a different collectivistic country is needed, since research findings until now have mainly focused on East Asia. Actually, the geographical imbalance of research findings could offer an alternative explanation for the contradictory findings. Maybe, it is not collectivism but the religious and philosophical background of East Asian countries that explains the found preference for a HC communication style (Kapoor et al., 2003).
The Role of Religion in Affective Communication
In anthropological research, religion and ethical philosophies of a culture are already
considered to be one of the most essential factors which have pervasive consequences for the shaping of emotion norms and communication behavior (Fiske, 2002; Salleh, 2005). The preference for a HC communication style in East Asian countries might be the product of the predominant philosophical system of Confucianism (Niedenthal et al., 2006).
Confucianism is deeply interwoven into East Asian’s value system and thereby impacts emotion norms (Park & Kim, 2008). Since Confucius placed strong importance on the values of harmony, humility and moderation, openness and expressiveness in
communication or disclosure of emotions are highly devalued. Also, any form of dramatic emotional expression is avoided since it is not in line with humility. For instance, a person who communicates loudly and openly his feelings is considered “light” according to Confucius (Nishida, 1996). By verbally expressing emotions in an indirect manner, East Asians prevent the embarrassment of rejection or the disagreement of the communication partner and subsequent loss of face (Park & Kim, 2008). Thus, eventually, the indirectness of HC affective communication style seems to facilitate the adherence to Confucian values.
10
Interestingly, the predominant religion in individualistic countries is Christianity (“Christianity”, 2015). Perhaps, Christianity also serves as a more specific and rich explanation for the emergence of a LC communication style than solely considering
individualism itself as an adequate origin. Christianity is thought to be a major determining factor in the development of the western concept of the individual. Emotions became a personal characteristic and part of the individual’s identity. The communication of emotions and their free expression reaffirm the importance of the individual compared to social relationships and is considered desirable according to Christian values (Sampson, 2000). Also, since the responsibility for fulfillment in life is placed upon the individual self,
competition is promoted and enhanced. To defend one’s own values might result in an open, direct and dramatic communication style, fostering emotional expressivity (Matsumoto et al., 2008).
Purpose of the Study
Research is needed that disentangle I-C and religion and analyzes a country that falls in between these two observations. Ghana is highly collectivistic but the population is 80% Christian. Thereby, Christian values seem to be important and deeply interwoven into everyday life of Ghanaians (Dzokoto & Adams, 2007).
Given the lack of research attention paid to African countries, there is fairly little known about Ghanaians general and affective communication. However, one finding of prior
research revealed that moderation is prized in communication by people of several African countries (e.g. Nigeria, Zimbabwe), (Kim-Prieto & Eid, 2004). Assuming a geographical similarity between these countries this could also hold true for Ghana, mirroring results found in other collectivistic countries in East Asia. Oppositely, Leets found that while Asians were more sensitive to the communication context and found indirect racist speech as most harmful, African-Americans relied more on the explicit, direct messages and less on the
11
context for their evaluation of racist speech (2003). These findings were interpreted as a tendency of African-Americas to prefer low context communication. Concerning affective communication, African countries apparently show a greater openness in the expression of emotions than East Asian countries (Kim-Prieto & Eid, 2004). Anyhow, for the rest research is scarce and Ghana remains a “blind-spot”.
Due to that reason, the purpose of the present study is to throw light on general and affective communication behavior of Ghanaians and to start filling the gap concerning psychological research conducted in Africa. Concretely, focus is put on the question whether Ghanaians adopt a more high or low context (affective) communication style. Accordingly, the preference for a HC (affective) communication style would be in line with the assumption that collectivism can serve as predictor and mirror prior findings. The preference for a LC (affective) communication style would rather point to the influence of other factors in the emergence of such, such as for example Christianity. Hypotheses are divided into three sections.
First, the study will investigate how much Ghana’s general and affective
communication style is in line with the classical theory concerning HC/LC communication. The classical theory is defined by the assumption that collectivism serves as predictor for a HC (affective) communication style. As already outlined, focus has been placed on two factors: indirect and dramatic communication (HC/LC). These factors are obtained by self-report measures. The use of self-self-report has been increasingly criticized since self-self-report studies are often biased by the person's feelings, social desirability and researcher demands (Paulhus & Vazire, 2007). This criticism applies even more to affective communication, since it appears to be inherently difficult to talk about emotions and to have adequate insight and self-reflection. Based on this concern, affective communication was measured via an
observational method. Thus, in order to examine whether Ghanaians affective communication is in line with the classical theory, the focus was placed on the use of valenced and specific
12
emotion terms (HC/LC). The first hypothesis states: Collectivism predicts the use of a HC (affective) communication style. Specifically, a positive relationship between collectivism and indirect communication and the use of valenced emotion words, a negative relationship with dramatic communication and the use of specific emotion words.
Secondly, how much Ghana shows the typical characteristics of a HC or LC
(affective) communication style will be investigated, leaving the role of collectivism aside. Accordingly, dramatic communication and the use of specific emotion words would point to a rather LC, indirect communication and the use of valenced emotion words to a HC (affective) communication style. It was hypothesized as follows: Ghanaians will be more likely to use (affective) communication deemed low context than high context. Specifically, a higher use of dramatic communication and specific emotion words than indirect communication and valenced emotion words was predicted.
Thirdly, the influence of Christianity as a possible factor of influence for LC affective communication style is taken into account. Maybe, the establishment of the Christian religion in Ghana and Christian values during colonialization has left its traces on the Ghanaians and led to the adoption of a rather LC affective communication style. Thus, the third hypothesis examines whether high Christianity can serve as an intensifier of the relationship of
collectivism and LC affective communication styles. Also, it might degrade the strength of the relationship of collectivism and a HC affective communication styles.
Methods Participants
The sample in this study consisted of 26 participants of Ghanaian origin and citizenship. Seven female and 19 male Ghanaians with age range between 22 and 59 (M = 39.5, SD = 9.7) participated. All were teachers at different schools in Ghana and received a small amount of money for their participation.
13 Procedure
Participants were welcomed in a quiet, well-lit room and asked to have a seat. They sat on a chair in front of a desk, were filmed from the front and received the instructions in paper form. The assistant tried to ensure a pleasant and comfortable atmosphere and allowed the participant to read the information brochure and fill in the informed consent. After this, the assistant explained the instructions for the procedure but did not direct subjects towards talking about emotions in order to prevent bias (participants were debriefed after the study was finished). Participants were asked to imagine skyping with an acquaintance and telling him/her in their own words about the event they had just read about. Thereby, it was
emphasized that there are no right or wrong answers but that they should talk as they usually do. Also, they could talk as much as they wanted but should not exceed three minutes of the hourglass per scenario. In order to get used to the procedure, every participant completed a practice trial with the assistant present in case of questions. For the four subsequent scenarios, the assistant left the participant alone and came back after the last trial was completed. The scenarios consisted of four short stories about an event that was supposed to trigger the emotions shame, sadness, happiness or pride, respectively (this was neither openly mentioned nor written in the instructions). Since in this study affective communication was
operationalized by the emotions happiness and sadness (for the scenarios, see Appendix A), pride and shame are considered irrelevant for this purpose and left aside. After reading each scenario, the participant turned the hourglass and started talking to the imagined acquaintance via skype about the event. After each scenario, participants answered a few questions about the experience of any emotion (if yes, specify it) during the call, the strength of the
experienced emotion and its expression, using rating scales ranging from 1 = very weak to 10 = very strong. Afterwards, the camera was turned off and the participants continued with parts 2 and 3 of the questionnaires (about personal beliefs and communication styles (2) and
14
demographic information (3)). The participant then filled in the participant form and were thanked for their participation.
Measures
Cultural Value Scale. In order to be able to measure and compare general cultural
patterns and values, the renewed Cultural Value Scale (CVScale) was used (Yoo, Donthu, & Lenartowicz, 2011). The CVScale intends to measure Hofstede’s five cultural dimensions on an individual level with a 26-item scale. Prior research for the CVScale indicates reasonable validity and reliability for all dimensions (α = 0.78 - 0.91) and sample and across-national generalizability, (Mazanec, Crotts, Gursoy, & Lu, 2015). The dimensions include: power-distance, Individualism vs. Collectivism, Masculinity vs. Femininity, Uncertainty Avoidance and Long-Term-Orientation vs. Short-term-normative orientation. However, since the present study solely focuses on the influence of collectivism on (affective)
communication, only the factor Individualism-Collectivism was included for data analysis. Participants filled in the questionnaire and could respond on a 7-point Likert scale ranging from 1 = I strongly disagree to 7 = I strongly agree. They were asked to base their answers on what they actually believed to be true rather on what they think they should choose or what they would like to be true.
General Communication. In order to assess general communication of Ghanaians, Gudykunst’s questionnaire about High and Low context communication was included as measurement (1996). Previous research indicates reasonable validity and reliability of the questionnaire α = 0,77 - 0,86, (Kapoor et al., 2003). The questionnaire consists of the following 8 sub-scales: participants’ perceptions of their ability to infer others meanings, inter- personal sensitivity, use of feelings to guide behavior, openness in and initiation of communication with others, preciseness in communication and positive perceptions of
15
questionnaire: indirect communication (e.g.” I communicate in an indirect fashion”) and dramatic communication (e.g. „I use a lot of colorful words when I talk”). High context communication was operationalized in terms of higher scores in indirect communication, whereas low context communication by higher scores on the dramatic communication factor. Participants answered 80 items on a 7-point Likert scale ranging from 1 = strongly disagree to 7 = strongly agree.
Affective Communication. In order to assess affective communication, the content of the
verbal messages was analyzed. First, the speech content of the happiness and sadness scenario was literally transcribed and afterwards coded with a computer software for qualitative
research, CATMA (Computer Aided Textual Markup & Analysis) and a corresponding Codebook (see Appendix B). In this study, the focus was placed on the coding of two
categories of verbal messages: the use of specific and valenced emotion terms. The frequency of specific as well as valenced emotion terms was counted. Specific emotion terms were defined as words directly and explicitly referring to the emotions happiness and sadness. This included verbal expressions like “I am happy” and “his sadness was obvious to all”,
corresponding to a LC affective communication style. Valenced emotion-terms on the other hand are defined as words directly referring to happiness and sadness but only in terms of valence. This includes verbal expressions such as “I feel good” or “I feel down” what corresponds more to people with a rather HC communication style (Gudykunst et al. 1996). References to God. It is possible, Christianity has been a determining factor in the adoption of a LC affective communication style in Ghana. In order to measure in how far Christianity has influenced affective communication style of Ghanaians, verbal references to God were counted. Thereby, only references were counted which were directly related to the expression of the emotion (“I was so happy and I thank God for that!”). Additionally, a higher rate of references to God could go hand in hand with a higher use of specific rather than
16
valenced emotion-terms since the prediction is that a higher rate of references to God is related to LC affective communication behavior.
Data Analysis
Four linear regressions were conducted in order to examine whether collectivism serves as predictor for a HC/LC (affective) communication style (H1). Hereby, collectivism represents the independent variable; indirect communication, dramatic communication, frequency of valenced and specific emotion terms the dependent variables.
Focusing on HC/LC affective communication, the means of valenced and specific emotion terms were compared by conducting a paired-samples-t-test. For general HC/LC communication, the means of indirect and dramatic communication were each compared to a point estimate of the means of an external study that used the same measurements, thereby conducting a one-sample-t-test (H2).
To test whether Christianity might serve as an intensifier of the relationship between collectivism and a LC affective communication styles, two moderation analyses were carried out. Thereby, collectivism represented the independent variable, frequency of valenced emotion terms and specific emotion terms the dependent variables and Christianity served as moderator (H3).
Results
The research included 27 participants in total, from whom four participants were excluded from analysis because they did not understand the instructions in the happiness or sadness scenario. Due to this, the data from only 23 participants was included in the analyses. Also, based on the fact that sample size was really small, it was not possible to determine whether the data was normally distributed or not. The assumption of normality will therefore be treated as non-violated for all analyses.
17
Christian, demographical data was used to proof it. Indeed, more than 96% of the participants indicated to consider themselves as Christians.
Manipulationchecks
In order to investigate if the manipulation of the emotions happiness and sadness actually worked, people were asked which emotion they had experienced and specify the experienced emotion. These scores were subdivided into 3 categories: Happiness and sadness,
respectively, were considered to be the target emotions. Congruent emotions were defined by having the same valence and being in the same emotion field with the target emotion (elated, down). The category incongruent was defined by emotions that did not resemble nor in valence nor in emotion field happiness and sadness (surprise, anxious).
Table 1
Percentage of the three Emotion Categories, differentiated for the Happiness and Sadness Scenario
Target Congruent Incongruent
Happiness 48% 37% 3,7%
Sadness 63% 18,5% 7,4%
In both scenarios three participants did not fill in the questionnaire, what accounted for 11% in total for both emotions. Looking at the percentages in Table 1 approximately 85% of the participants experienced happiness or a very similar emotion and around 81% experienced sadness or a very similar emotion. Based on these observations the manipulation of the emotions is considered successful. Also, looking at the intensity of the experienced emotions, the manipulation seems to be successful. Participants indicated on a Likert-Scale ranging from 1 (very weak) to 10 (very strong) how intense they experienced the emotion. The mean values of both scenarios showed a rather high average for happiness (M = 7.81, SD = 3.51) and sadness (M = 7.23, SD = 3.10). In order to be sure that participants understood the research procedure in English, the third manipulation check concerned their English level. On a scale
18
from 1 (I do not agree at all) until 10 (I strongly agree), they indicated that they did not have difficulties understanding the instructions in English (M = 8.73, SD = 2.03) nor talking about the events in English (M = 8.85, SD = 1.85).
General Affective Communication in Ghana
Given the general lack of conducted research concerning affective communication in an African country it seems interesting to look at Ghanaians affective communication in general as illustrated in Figure 1. However, for the purpose of this study, the frequencies of specific and valenced emotion terms, were assessed, see Figure 1.
Figure 1. Disclosed affective communication in happiness and sadness scenario (together)
Collectivism as Predictor for HC/LC (affective) Communication Style
H1. Assumptions of linearity, homoscedasticity and independence of residuals were met
for data of all regressions. Also, no significant outliers were found for the data of valenced emotion terms, indirect communication and dramatic communication. The variable specific emotion terms showed one significant outlier (participant 12) that did not fall between the range of -3 and 3, but indicated a value of 3.35. Comparing raw and transformed results with
19
the removed outlier did not show a significant difference. Consequently, it was assumed that the outlier did not affect the interpretation of the results and raw results are reported.
Tabel 2
Four Simple Regression Analyses with Collectivism as Predictor
B t p
Indirect Com.* .012 .06 .953
Dramatic Com.* -.051 -.25 .802
Valenced ET** .434 2.21 .038
Specific ET** .031 1.45 .886
Note. *Com = Communication; **ET = Emotion terms
Concerning general communication, no significant results were found, see Tabel 2. Collectivism did neither significantly predict the use of indirect nor of dramatic
communication. Thus, higher scores on collectivism are neither related to a higher use of indirect communication nor of lower use of dramatic communication. These findings do not support hypothesis 1, which assumed a significant positive/ negative relationship between these variables, respectively.
Collectivism significantly predicts the use of valenced emotion terms, see Tabel 2, suggesting that people who score higher on collectivism will also use more valenced emotion terms. Additionally, a significant regression equation was found F (1,21) = 4.88, p = .038,
R2 = .188. This means that collectivism accounts for 18,8% of the variation in the use of valenced emotion terms. This finding supports hypothesis 1, which predicted a positive relationship. However, collectivism did not significantly predict the use of specific emotion terms, see Tabel 2. Thus, higher scores on collectivism are not related to a lower use of specific emotion terms hereby not supporting the hypothesis, which predicted a negative relationship between these variables.
20
Display of High or Low Context Characteristics in (affective) Communication
H2. There were several outliers found for the dependent variable valenced emotion terms
(participants 10, 27, 25, 13) and one for specific emotion terms (participant 12). In order to see if these outliers significantly affect the interpretation of the results, a non-parametric Wilcoxon-Signed-Rank Test was conducted. Results indicate that the outliers do not
significantly affect the final result. Still, in order to account for the possibility that the mean values are distorted by the outliers, z-value and p-value of the Wilcoxon-Signed-Rank Test are additionally reported. However, with such a small sample it is rather difficult to determine who an outlier is and who not. The paired-samples-t-test revealed that, on average,
participants did use significantly more specific (M = 5.30, SE = 3.66) than valenced emotion terms (M = 1.30, SE = 1.77) while talking about happiness and sadness. The difference, 4.0, BCa 95% CI [2.32, 5.68], was significant t(22) = 4.94, p < .001 and adds support to
hypothesis 2. Wilcoxon-Signed-Rank Test indicated a Z-value of -3.36 and p <.001, confirming the results.
Since the raw scores of the two variables indirect communication and dramatic communication are difficult to interpret without a comparison group, external mean scores served as foundation for a comparison. Kapoor and colleagues conducted a study in India and the United States (US), using the same questionnaire of Gudykunst as in this study. A one-sample-t-test was conducted to determine if a significant difference existed between mean values of both collectivistic countries, Ghana and India. Applying a rather conservative rule, the lower bound of the confidence interval was used as India’s mean for both variables. Results revealed that Indians scored higher on indirect communication (M = 43.53, SD = 7.5) than Ghana (M = 35.46, SD =9.90). This difference was significant, t(21) = -3.63, p = .002. Carefully interpreting these rather explorative results, there is evidence provided that Indians use significantly more indirect communication than Ghanaians, hereby adding support to hypothesis 2. However, Indians (M = 60.34, SD = 7.8) scored significantly higher on
21
dramatic communication than Ghanaians, (M = 54.67, SD = 11.89), t(26) = -2.5, p = .019. Contrary to prior results, these did not confirm prior predictions.
Interestingly, also Kapoor and colleagues who conducted the study expressed surprise looking at India’s rather high values of dramatic communication compared to an
individualistic country, the US (2003). Due to that reason, an additional one-sample-t-test was conducted, comparing means of dramatic communication of Ghana and the US. Results revealed that the individualistic US (M = 54.45, SD = 9.4) did not score higher on dramatic communication than Ghana, (M = 54.67, SD = 11.89), indicating that the difference was not significant, t(26) = .095, p = .462. Looking at the almost identical mean values, these findings suggest that the US and Ghana did not differ in their use of dramatic communication even though they differ clearly in terms of I-C.
Christianity as Moderator
H3. All assumptions of a moderation analysis were met. Overall the found results were not
significant as depicted in Tabel 3. Tabel 3
Two Multiple Regressions for the Test of Moderation Effects
DV (N = 23) B se t p
Valenced ET -.11 .031 -0.37 -.717
Specific ET -.003 .024 0.14 .888
Note. DV = dependent variables; ET = Emotion Terms
The results did not confirm hypothesis 3. Tabel 3 indicates that there was neither a significant interaction effect found for valenced emotion terms nor for specific emotion terms, suggesting that References to God did not serve as moderator for the relationship between
22
Differences in Affective Communication depending on Scenario
The use of specific emotion terms in the happiness and sadness scenario were counted together, assuming that these emotions would not differ in terms of frequency. The same applies to the use of valenced emotion terms. To test this assumption, a paired-samples-t-test. On average, participants did use significantly more specific emotion terms in the happiness (M = 3.38 SE = 2.2) than in the sadness scenario (M = 1.71 SE = 2.66), t(23) = 2.6, p = .016. However, the difference was not significant for the use of valenced emotion words, t(22) = -1.46, p = .157. Apparently, the use of specific emotion terms seems to be dependent on the kind of emotion participants talked about. Further implications of these results will be commented in the discussion.
Conclusions and Discussion
The purpose of this study was to investigate whether Ghanaians show a high or low context (affective) communication style, whereby collectivism and Christianity served as potential factors of influence. Overall, the hypotheses were only partially confirmed.
The first hypothesis aimed to test the classical theory proposing that collectivism would predict the use of a HC (affective) communication style. Results concerning affective communication revealed that collectivism serves as predictor for the use of valenced but not for specific emotion terms. Thus, people who are more collectivistic also seem to use more valenced emotion terms. In this case, collectivism seemed to be related to the use of a HC affective communication style, what mirrors prior findings who found a similar relation but with general HC communication style. This suggests that Ghanaians who internalized
collectivistic values more intensely also prefer a rather implicit and indirect verbal expression of happiness and sadness. However, the use of specific emotion terms does not seem to be related to collectivism hereby contradicting the hypothesis which assumed a negative
23
predictor for a HC communication style in this study (neither for indirect nor for dramatic communication). It is interesting to compare these findings with those of Gudykunst (1996), who found that collectivistic countries such as Japan and Korea tend to communicate in an indirect manner and avoid dramatic communication. These findings could also be replicated with Asian samples (Park & Kim, 2008). However, the preference for an indirect
communication style by Asians may better be explained by the predominant religions and philosophies such as Confucianism. An African country like Ghana might embrace collectivistic values but has found a different way of defending them than via indirect communication and the avoidance of dramatic communication. To summarize, since
collectivism seems to partly predict an affective communication style, this does not hold true for general HC/LC communication in this study. It may be the case that affective and general communication do differ in their underlying working mechanisms and that affective but not general communication in Ghana is related to collectivism. Considering the dearth of research in this area, further studies are needed to get deeper insight in the underlying mechanism of affective communication. However, the fact that no relationship was found for general communication supports prior criticism that questioned the role of collectivism as a potential predictor for HC/LC communication style.
The second hypothesis aimed to test in how far Ghana shows more characteristics of a high or low context (affective) communication style. Results mostly confirmed the
hypothesis. Concerning affective communication, Ghanaians seem to favour specific over valenced emotion words when talking about happiness and sadness. These results add evidence to the hypothesis, pointing to the preference for LC affective communication characteristics.
Furthermore, results for general communication of a comparison with another collectivistic country such as India revealed that Ghanaians tend to communicate in a more direct manner than do Indians. Accordingly, also this finding suggests that Ghanaians show a preference for
24
the characteristics of a LC communication style. However, contrary to prediction, Ghanaians reported less dramatic communication than Indians. Interestingly, Kapoor and colleagues who conducted the study also expressed surprise looking at India’s rather high values of dramatic communication compared to an individualistic country such as the United States (2003). Due to that reason, a further comparison of Ghana and the US was included. Results revealed that both countries do not differ in their use of dramatic communication. Although Ghana uses less dramatic communication than India, the comparison to the individualistic low context US gives reason for the interpretation that Ghanaians tendency for dramatic communication deemed to be more representative for a LC than HC communication style. Since the data for this comparison was extracted from an external source, these findings might motivate future research to conduct a real comparison study between Ghana and an individualistic country like the US or a collectivistic country in East Asia. Generally speaking, these findings suggest that Ghanaians show similar preferences in general as well as in affective communication, namely for rather LC characteristics. That is, the HC/LC communication concept might also apply to affective communication. Anyhow, further research needs to found these findings. In answer to the latter hypothesis, Christianity was not found to be of importance as an intensifier or degrader of the relationship between collectivism and a LC/HC affective communication style. A possible alternative explanation for the non-significant results concerns the measurement used. Presumably, the operationalization of Christianity lacked construct validity. What this study aimed to measure was how far the assimilation of Christian values influenced the emergence of a LC affective communication style. However, perhaps, counting the frequency of references to God does not reflect this assimilation. Since religion plays an active role in everyday life in Ghana (Dzokoto & Adams, 2007), references to God may depict more a hollow set phrase than an adequate representation of the strength of Christian value assimilation. Although simple references to God which were not related to expressions of happiness or sadness were not counted, this relationship remains disputable.
25
Considering the general dearth of research concerning the relationship of affective
communication and religion, adequate and valid measurements are still rare. One possibility would be to add another questionnaire about Christian norms which would address the link between Christianity and emotion norms. Considering the potential role of religion as factor of influence for affective communication, future research is needed to address the limitations of this study.
This study was subject to several general limitations. The emotions of happiness and sadness were taken together, assuming no difference in frequency in the use of valenced and specific emotion words in the scenarios. However, this assumption was not confirmed by the results. Apparently, affective communication style is dependent on the kind of emotion Ghanaians talk about. Ghanaians did use more specific emotion words when talking about happiness than about sadness, however, in the use of valenced emotion words no difference was found. These findings could be explained by looking at Africa-specific norms for positive and negative emotions. Kim-Prieto and Eid found a surprisingly strong aversion for negative emotions in Sub-Saharan countries (2004). It may be the case that the use of specific emotion words while talking about a negative emotion as sadness is considered inappropriate by Ghanaian society and emotion norms. The direct and explicit naming of sadness could cause discomfort in the communication partner and consequently be avoided. Valenced emotion words on the other hand already implicate a certain moderation itself whereby emotion norms might not be violated. Thus, conclusions of the present study have to be interpreted with caution. Future research could address these limitations by conducting the same research about affective communication but separately looking at differences between happiness and sadness.
One possible limitation concerns the use of English as the language of instruction. Given that Ghana has adopted English as the lingua franca after colonialization, it seems rather intuitive to assume that it is deeply manifested in people’s minds and communication
26
behavior. However, Ghana is a multilingual country with more than 12 national and actively spoken languages. Since different languages create different cultural scripts for the
communication and expression of emotions, language could have had a strong influence on the development of a HC/LC affective communication style (Guerini, 2007). The impact of colonization and the imposition of Christian values and emotion norms on the one hand and old African traditions on the other hand, has led to a complex reality in Ghana. Maybe the use of two languages in everyday life results in the development of different communication styles for each language, whereby the indigenous languages serve as more intuitive option for affective communication. Dzokoto, Opare-Henaku and Kpobi found that Ghanaians relied strongly on using somatic references when talking about emotions (2013). These somatic references could only be found when they talked in their indigenous languages, but not when talking in English. Apparently, in order to adequately communicate emotions and relate to what the other person feels, somatic references are necessary. Even if English does provide opportunities for somatic references in its emotional lexicon such as having “butterflies in one’s stomach”, these metaphors are distinct from labels for emotion. For instance, Ghanaians express deep sadness in the indigenous language Ewe by referring to a feeling of something “pressing on the heart”, but they probably would not find an equivalent in English for such a feeling (Dzokoto et al., 2013).
The current study reveals that there is initial evidence that Ghana does not fit as well as might be assumed into the HC (affective) communication classification. It may be the case that Ghanaians express their emotions in line with a LC communication while talking in English but that they would communicate them differently in their traditional African
language. They use of somatic references for the expression of emotions could also be labeled as a form of indirect communication, since specific and direct emotion words are not
explicitly mentioned. However, strong emotional statements, even communicated in an indirect manner via somatic references and metaphors do not appear to be in line with the
27
form of indirect communication found in East Asian countries. Rather it might be a typical characteristic of African or Ghanaian affective communication (Dzokoto & Adams, 2007). Future research could account for this alternative explanation by conducting a study and using African traditional languages as the language of instruction.
In practice, this indicates that one might rethink the general applicability of constructs such as HC/LC communication and accept that the dimensions found in prior literature do not fit well for every country as neatly as we would like them to. Although, further research is needed in order to confirm this preliminary conclusion. As Hall suggested, probably all cultures incorporate both elements of HC and LC (affective) communication (1976). Possibly, there are plenty more determining elements influencing general and affective communication which do not fit into the boxes that have been studied so far. Looking at Ghana’s complex multilingual reality and its unique history, it seems to be of great interest to look at the specifics of the culture via an emic perspective.
Generally, studying cultures from within offers deeper specific understandings of the nuances and characteristics and may even lead to new insight into more general patterns. This is not to disparage etic frameworks in general, but to strive for a better integration of etic and emic research approaches. Eventually, by integrating these two approaches it may be possible to sketch a more sophisticated and complex but also more real picture of general and affective communication in Ghana.
28 Literature
Christianity 2015: Religious Diversity and Personal Contact. (2015). International Bulletin of
Mission Research, 39(1), 28-29. doi:10.1177/239693931503900108
Dzokoto, V., & Adams, G. (2007). Analyzing Ghanaian emotions through narrative: A textual analysis of Ama Ata Aidoo’s novel changes. Journal of Black Psychology, 33(1), 94-112.
Dzokoto, V. A., Opare-Henaku, A., & Kpobi, L. A. (2013). Somatic referencing and psychologisation in emotion narratives: A USA–Ghana comparison. Psychology &
Developing Societies, 25(2), 311-331.
Eid, M., & Diener, E. (2001). Norms for experiencing emotions in different cultures: Inter- and intranational differences. Journal of Personality and Social Psychology, 81(5),
869-885.
Fernández, I., Carrera, P., Páez, D., & Sánchez, F. (2008). Interdependent Self-Construal, Competitive Attitudes, Culture And Emotional Reactions On Sadness. Psychologia,
51(3), 214-234.
Fiske, A. P. (2002). Using individualism and collectivism to compare cultures-A critique of the validity and measurement of the constructs: Comment on Oyserman et al. (2002). Psychological Bulletin, 128(1), 78-88.
Fussell, S. R. (2002). The Verbal Communication of Emotions: introduction and overview. The verbal communication of emotions: interdisciplinary perspectives, 1-16. Gudykunst, W. B., Matsumoto, Y., Ting-Toomey, S., 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.
Gudykunst, W. B., & Mody, B. (2002). Handbook of international and intercultural
29
Gudykunst, W. B., & Ting-Toomey, S. (1988). Culture and affective communication. The
American Behavioral Scientist, 31(3), 384.
Hall, E. (1976). Beyond culture. New York: Doubleday. Heine, S. J. (2012). Cultural psychology. 1st and 2nd ed.
Hofstede, G. (1984). Cultural dimensions in management and planning. Asia Pacific journal
of management, 1(2), 81-99.
Kapoor, S., Hughes, P. C., Baldwin, J. R., & Blue, J. (2003). The relationship of
individualism–collectivism and self-construal’s to communication styles in India and the United States. International Journal of Intercultural Relations, 27(6), 683-700. Kim-Prieto, C., & Eid, M. (2004). Norms for experiencing emotions. Journal of Happiness
Studies, 5(3), 241-268.
Kuppens, P., Realo, A., & Diener, E. (2008). The role of positive and negative emotions in life satisfaction judgment across nations. Journal of Personality and Social
Psychology, 95(1), 66-75.
Mazanec, J. A., Crotts, J. C., Gursoy, D., & Lu, L. (2015). Homogeneity versus heterogeneity of cultural values: An item-response theoretical approach applying Hofstede's cultural dimensions in a single nation. Tourism Management, 48, 299-304.
Niedenthal, P. M., Krauth-Gruber, S., & Ric, F. (2006). Psychology of emotion:
Interpersonal, experiential, and cognitive approaches. Psychology Press.
Nishimura, S., Nevgi, A., & Tella, S. (2008). Communication style and cultural features in high/low context communication cultures: A case study of Finland, Japan and India. In Proceedings of a Subject-Didactic Symposium, Finland, Helsinki.
Oguri, M., & Gudykunst, W. B. (2002). The influence of self construals and communication styles on sojourners’ psychological and sociocultural adjustment. International
Journal of Intercultural Relations, 26(5), 577-593.
30
communication styles among Asian American and European American college students. Cultural Diversity and Ethnic Minority Psychology, 14(1), 47.
Sampson, E. E. (2000). Reinterpreting individualism and collectivism: Their religious roots and monologic versus dialogic person-other relationship. American Psychologist, 55(12), 1425-1432.
Voronov M. & Jefferson A. Singer (2002) The Myth of Individualism-Collectivism: A Critical Review, The Journal of Social Psychology, 142:4, 461-480.
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.
31 Appendix A
Happiness and Sadness Scenario used in the Procedure
Happiness: “Your favorite brother just came by to tell you that he and his wife had a baby. This baby is the first child of the two of them and they were very much looking forward for it. After a quick delivery, the child was born healthy this morning. You can hardly wait to see it.”
Sadness: “Your sister just came by to let you know that a very dear cousin of yours died. This cousin had been your best friend since childhood and you used to do many nice things together. Your cousin was not very old, but had been seriously ill recently. You realize that you will miss your cousin a lot.
32 Appendix B
CODEBOOK for Analysis of Verbal Affective Communication
Variable Description Instruction Name in SPSS
Amount of general information disclosed
How much general information
(emotional & non-emotional information together) does person share?
Count number of words
(per scenario and across all scenarios)
DISC_TOT (sum of all 4 scenarios)
DISC_JY (# words joy scenario)
DISC_PD (# words pride scenario)
DISC_SD (# words sadness scenario)
DISC_SH (# words shame scenario)
Length How long (in seconds) does subject talk? 1
st word until last LENGTH_TOT (sum of seconds spoken in all 4 scenarios)
LENGTH_JY/SD/PD/SH Speech Rate* How fast/slow does
subject talk? How many words does subject disclose per minute?
Divide number of words by length (per scenario and across all scenarios)
RATE_TOT
RATE_JY/SD/PD/SH Organization At which point in time
does the subject reveal the first affective information?
Note down second at which 1st verbal affective information unit is mentioned ORG_TOT ORG_JY/SD/PD/SH Amount of affective information disclosed
How much affective information does subject reveal?
Count number of verbal affective information units (eg specific emotion terms, valenced-only terms, appraisals, emotion-related actions) VBINFOA_TOT (sum of verbal affective information units) VBINFOA_JY/SD/PD/SH (per scenario) Relative amount of affective information disclosed*
How much affective information does subject reveal controlling for the amount of time (s)he talks? Divide number of verbal affective information units by number of words • Eg VBINFOA_TO T : DISC_TOT • Eg. VBINFOA_PD : DISC_PD REL_VBINFOA_TOT REL_VBINFOA_JY/SD/PD/SH Amount of specific emotion terms disclosed
How many specific emotion words does subject mention?
• Definition:
Count number of specific emotion terms (per scenario and across all
SPECEMO_TOT
33 Words directly referring
to affective states (note: can be different types of words
• Eg. Sadly, the
government has decided to abandon financing • I am happy/sad, worried/confiden t, angry/pleased, keen/unintereste d
• His fear was
obvious to all • I was overcome with joy scenarios) Relative amount of specific emotion terms disclosed*
How many specific emotion words does subject mention controlling for the amount of time (s)he talks?
Count number of specific emotion terms (per scenario and across all
scenarios) and divide by number of words SPECEMO_TOT SPECEMO_JY/SD/PD/SH Amount of valenced-only emotion terms disclosed
How many valenced-only emotion words does subject mention?
• Words directly referring to affective state but only in terms of valence • eg. I feel good, I
feel down
Count number of valenced-only emotion words (per scenario and across all scenarios) VAGEMO_TOT VAGEMO_JY/SD/PD/SH Relative amount of valenced-only emotion terms disclosed* Amount of
appraisals How many appraisals does subject use? • Definition: Evaluation of significance of event (and attribution of the cause of those events) • Appraisals occur
after the event, before the
Count number of
34 experience of an emotion or emac. • Eg. something great just happened, bad news, it was expected
• Please mark with CATMA which type of appraisal was made (more details bellow) Relative amount
of appraisals* REL_APP_TOT REL_APP_JY/SD/PD/SH
Amount of
emotion-related actions
How many emotion-related actions does subject use? • Definition: Actions and emotional responses indirectly referring to affective states, eg physiological reaction, motor expression, action tendencies • EMACS occur as a consequence (after) of an experienced emotion • Eg. I am blushing, I wanted to punch someone, I wanted to run away Count number of emotion-related actions EMAC_TOT EMAC_JY/SD/PD/SH Relative amount of emotion-related actions * REL_EMAC_TOT REL_EMAC_JY/SD/PD/SH Owner of
emotion (self) How often does subject speak of himself as the owner of the emotion? • eg. I feel great,
something horrible
happened to me,
35 I wanted to run away Owner of the emotion (3rd person)
How often does subject speak of others as the owner of the emotion?
• eg. one surely
feels great in such a situation
• Note: only count when emotion is a result of event (receiving a praise; being shamed publicly) so the fact that your boss is happy with your work does not count!
OTHRD_TOT
Owner of
emotion (group) How often does subject speak of a group including him/herself as the owner of the
emotion? • Eg we were delighted by the news OGROUP_TOT General attempt to involve interaction partner
How often does the subject make an attempt to reach out, get a response from the imagined receiver of the message?
• Eg. Do you have
any children yourself? INVOLVE_GEN Attempt to emotionally involve interaction partner
How often does the subject make an attempt to emotionally reach out, get an emotional
response from the imagined receiver of the message?
• Eg. wouldn’t you
feel the same?
• Eg. how would
you feel?
