1
DOES IT MATTER WHAT YOU TWEET AS AN ORGANIZATION?
The effect of obligation and gratitude inducing tweets on gender dimensions of
brand personality and the moderating role of self-monitoring
DRAFT VERSION
Master thesis, Msc BA, specialization Marketing Management University of Groningen, Faculty of Economics and Business
July 4 , 2011 IRENE WIERSMA Studentnummer: S1910590 Florakade 364 9713 ZK Groningen Tel: +31(0)6-12427300 e-mail: i.j.g.wiersma@student.rug.nl Supervisor / university: Prof. dr. B.M. Fennis Word count: 5996
2
DOES IT MATTER WHAT YOU TWEET AS AN ORGANIZATION?
ABSTRACT
This research contributes to the field of academic research on the medium Twitter. The present research examines the relationship between gratitude and obligation inducing messages on gender dimensions of a brand. It was also hypothesized that this relationship is moderated by the personal trait self-monitoring. The second dependent variable incorporated in this research was brand attachment. In an experiment participants were randomly assigned to two conditions: gratitude inducing tweets or obligation inducing tweets and were afterwards divided in high or low self-monitoring. The present findings show that there was one significant outcome; gratitude and obligation inducing tweets have a significant impact on gender dimensions of brand personality. These results demonstrate that by using gratitude inducing tweets, consumers perceive the brand personality as more feminine and by using obligation inducing tweets as more masculine.
3 TABLE OF CONTENT
1. INTRODUCTION ... 4
1.1 Literature overview ... 5
1.2 Impacts of gratitude and obligation ... 6
1.3 Gratitude and obligation in relation to brand attachment ... 6
1.4 Gratitude and obligation in relation to gender dimensions of brand personality ... 7
1.5 The influence of the personal trait self-monitoring ... 8
2 METHOD ... 10
2.1 Participants and design ... 10
2.2 Procedure ... 10
2.3 Independent variables ... 11
2.3.1 Gratitude and obligation ... 11
2.3.2 Self-monitoring ... 12
2.4 Dependent variables ... 12
2.4.1 Brand attachment ... 12
2.4.2 Gender dimensions of brand personality ... 13
2.5 Analytical procedures...13
3 RESULTS AND CONCLUSIONS ... 14
3.1 Results ... 14
3.2 Conclusion and discussion ... 15
3.3 Practical implications ... 17
3.4 Limitations ... 18
3.5 Implications for further research ... 19 REFERENCES
4
1. INTRODUCTION
Direct marketing is often used in research and in practice. Due to the increasing use of the Internet, this creates new opportunities for direct marketing (Hoffman & Novak, 1996; Rust, Lemon & Zeithaml 2004). Recently a new form of direct marketing has emerged through social media, like Twitter, Hyves and Facebook. One characteristic of direct marketing is that it can be applied in a personal way and it is beneficial for creating and maintaining a relationship (Fletcher &Peters, 1997). In this research the increasing use of Internet and direct marketing is combined, by executing an experiment on Twitter. Twitter is an online micro blog, launched in October 2006 with millions of users today (Java, Song & Tseng, 2007).
Through direct marketing organizations communicate with their consumers, the goal is a purchase or positive awareness (Kolyesnikovaa, Sullivan & Callison, 2010). In this research not purchase intention will be examined, but two other indicators, because they are interesting variables in the marketing field. Brand attachment can be used as a scale to measure purchase intentions; this is relevant because the ultimate goal of direct marketing is encouraging people to purchase (Lambert-Pandraud &Laurent, 2010; Whan Park, MacInnis, Priester, Eisingerich & Iacobucci, 2010).
Kolyesnikovaa et al. (2010) scrutinized whether obligation and gratitude messages in direct mail have different outcomes regarding purchase intentions. They provide strong evidence that gratitude messages result in a more positive evaluation than obligation messages. In their research they demonstrated that gratitude and obligation are two independent affective influencers. It would be interesting to investigate whether this also applies to Twitter.
Another indicator that can be used to designate purchase intention is a strong gender personality of a brand. Grohmann (2009) indicates that a strong feminine or masculine brand personality results in increased purchase intentions, this is especially the case at brand extensions. The second dependent variable, gender dimensions of brand personality, is incorporated in this research because it is also an interesting variable in this context. This variable is added because it is possible that the two different tweets have an effect on the gender dimensions of a brand personality.
5 Unfortunately none of these unique theories have been researched in relation to Twitter, this is interesting and scientific relevant because Twitter is a new medium to which direct marketing can be applied. Direct marketing by making use of Twitter can be used to send all the followers an offer or give information about the organization in a 140 word message. It is interesting to scrutinize whether the effect of gratitude messages will also result in a more positive attitude towards the company. In line with the research of Becherer and Richard (1978) it is intriguing to investigate if high self-monitoring individuals are more influenced by the tweets because Twitter is a situational factor and not personality related. This research also contributes with practical relevance, because organizations then can distinguish which kind of messages they should tweet to create positive awareness or purchase intention.
Following this line of reasoning, the central aim of the present study is to examine whether different tweets (obligation inducing or gratitude inducing) have different effects on the brand attachment or gender dimensions. In addition, the moderating role of the personal trait self-monitoring will be examined.
After this introduction the literature review presents the dimensions of two different messages, gratitude and obligation, and it will be discussed how the relation with brand attachment and gender dimensions can be moderated by self-monitoring. Definitions and effects between the four variables will be explained. Finally, the hypotheses tested under undergraduates of the University of Groningen will be presented and discussed.
1.1 Literature review
This part of the paper will elaborate on the literature of all the variables mentioned in the introduction. Finally the hypotheses will be given.
Before the used variables will be addressed, first direct marketing will be mentioned. Direct marketing is used extensively (Scovotti & Spiller, 2006) and due to that many definitions were created. The definition of Scovotti and Spiller (2006) will be used because it encompasses not only what direct marketing is but also what makes it unique:
6 Fletcher and Peters (1997) postulate that addressing the normative aspects of relationship development in direct marketing should encourage commitment and support of the consumers. It can be declared that direct mail marketing is beneficial when creating and maintaining a relationship. Another important subject that has risen in popularity is Twitter (Michelson & Macskassy, 2010). By using Twitter it is possible to reach the customers personal and thereby a perfect medium to exploit direct marketing (Palmer & Koenig, 2009). In this research both will be combined.
1.2 Impacts of gratitude and obligation
Direct marketing is used to create a relationship between organizations and consumers and in this relationship exchange of communication is necessary. Gouldner (1960) mentioned that gratitude and obligation are both independent predictors of the principle of reciprocity. The principle of reciprocity is investigated frequently in the field of social psychology, but not in the field of marketing. In addition, in most research the principle of reciprocity is scrutinized with incentives. For instance, if you give something to another individual, this individual then feels obliged to give something in return.
Kolyesnikovaa et al. (2010) states that it is possible to create the principle of reciprocity just by mentioning, this even results in affective responses (consumer attitude and behaviour). This will also be the case in this study, in the experiment no incentives will be given but the feeling of obligation and gratitude will be shown through the Twitter messages and as a consequence this will influence their attitudes. Earlier research proved that feelings in text have influence on individuals’ responses. Emmons and McCullough (2003) found that gratitude manipulation created an increase in positive affect. In line with this finding, Brown, Homer, and Inman (1998) found that strong negative feelings in an advertisement elicited negative feelings towards the brand. Kolyesnikovaa et al. (2010) define that after gratitude inducing message consumers may feel appreciation to the personnel for services provided. In contrast to an obligation inducing message, consumers may feel pressure to make a purchase in return for services received. It is already shown by Kolyesnikovaa et al. (2010) that gratitude messages by mail result in a more positive attitude towards the company than obligation-inducing direct mail. This information is relevant for my hypothesis because it is subsequently fascinating to examine if this effect will also arise at direct marketing by Twitter.
1.3 Gratitude and obligation in relation to brand attachment
7 and practical value difference between brand attachment and brand attitude. They postulate that both constructs have similarities and differences. The most important difference is that at brand attachment the bond that connects the brand with the individual is much more stronger than at brand attitude. According to Whan Park et al. (2010); “Bonds are stronger (1) as connections between the brand and the self-become closer and (2) as brand-related thoughts and memories become more prominent”. They also demonstrated that brand attachment is a better predictor of actual behaviour than brand attitude, therefore brand attachment will be used in this study. Consistent with attachment theory (Mikulincer & Shaver, 2007), this bond is exemplified by a rich and accessible memory network (or mental illustration) that involves thoughts and feelings about the brand and the brand’s relationship to the self. Therefore brand attachment would be a perfect dependent variable to investigate which type of Tweet has the most influence on brand attachment. Because earlier mentioned literature states that gratitude results in more positive outcomes and obligation in negative outcomes, it will be expected in this research that obligation inducing tweets will lead to a lower brand attachment and gratitude inducing tweets messages lead to an higher brand attachment. Regarding the above discussion the following hypothesis is compiled:
Hypothesis 1. Gratitude-inducing Twitter messages will be associated with higher attachment towards the brand, while obligation-inducing Twitter messages will be associated with a lower attachment towards the brand.
1.4 Gratitude and obligation in relation to gender dimensions of brand personality
8 different personality traits, measured in many scales, individuals can also associate masculine and feminine personality traits with brands. Masculine stereotypes are described as independent, arrogant, feels superior and greedy. Feminine stereotypes are described as helpful, whiny, emotional and fussy (Eagly &Mladinic, 1989). Deaux (1985) states that masculine can be best described as measures of dominance and self-assertion and feminine can be best described as nurturance and interpersonal warmth. Due to obligation being dominance inducing, we expect that it will be seen as masculine, whereas gratitude is observed to be feminine for the reason that gratitude implies thankfulness. To measure the extra dependent variable the following is hypothesised:
Hypothesis 2. Gratitude-inducing Twitter messages will be associated with higher feminine brand personality towards the brand, while obligation-inducing Twitter messages will be associated with higher masculine brand personality towards the brand.
1.5 The moderating influence of the personal trait self-monitoring
According to Snyder (1974) the self-monitoring individual is one who, out of a concern for social appropriateness, is particularly sensitive to the expression and self-presentation of others in social situations and uses these cues as guidelines for monitoring his own self-presentation.
9 therefore more impressionable for situational factors (Harrison & Price, 2003). Prislin and Kovrlija (1992) confirm as well that high self-monitors are strongly influenced by external expectations and social cues. In addition, the research of Swaminathan, Stilley and Ahluwalia (2008) postulate that, nervously attached individuals are more likely to be differentially influenced by brand personalities. Based on the above line of reasoning the following hypotheses will be suggested:
Hypothesis 3. Self-monitoring will moderate the relationship between gratitude and obligation inducing tweets and brand attachment.
Hypothesis 4. High self-monitors will score higher on brand attachment than low self-monitors.
Hypothesis 5. Self-monitoring will moderate the relationship between gratitude and obligation inducing tweets and gender dimensions.
Hypothesis 6. High self-monitors will score higher on gender dimensions than low self-monitors
In sum, six hypotheses are compiled. These hypotheses are visually represented in figure 1. If these hypotheses will be supported this would provide the first evidence that obligation and gratitude inducing Twitter messages have different influence on brand attachment or gender dimensions of brand personality and this is moderated by self-monitoring.
FIGURE 1 Proposed model Gratitude or obligation inducing tweets Brand attachment High or low
self-monitoring
10 2. METHOD
In the second chapter the method and procedures of the experiment are described. First the participants of the experiment will be revealed, as well as the design of the experiment. Then the procedure of the experiment will be declared. Subsequently the independent variables (gratitude, obligation and self-monitoring) will be explained. Chapter two ends with the explanation of the dependent variables: brand attachment and gender dimensions of brand personality.
2.1 Participants and design
Eighty undergraduate students of the University of Groningen (45 females, 35 males; M age = 22.4 and SD = 2.53) participated in exchange for a candy bar. The study used a 2 (gratitude inducing tweet vs. obligation inducing tweet) x 2 (high self-monitoring vs. low self-monitoring) between-subjects factorial design. Undergraduate students were randomly assigned to each of the cells.
2.2 Procedure
In this part of the method section the procedure of the experiment is delineated in detail. Two
examples of tweets were prepared for this study; the first contained a gratitude inducing Tweet and the second contained an obligation inducing Tweet.
self-11 monitoring were asked on the fourth and final page of the experiment. These variables were not clearly stated, only the scales are mentioned and how to answer them. This means that regarding the scale to measure self-monitoring, “self-monitoring” was not the title of the scale but only: “Are these sentences applicable to you?”
After completing the experiment the subjects were thoroughly debriefed, thanked for their participation, and dismissed. The undergraduates also had the chance to choose a candy bar.
2.3 Independent variables
In this section of chapter two the independent variables will be discussed. First gratitude and obligation will be mentioned and how these variables were exposed in the experiment. Then the moderator; self-monitoring will be clarified.
2.3.1 Gratitude and obligation
Gratitude and obligation were embedded in the second page of the experiment. To be sure the stimuli induced the accurate feelings (gratitude or obligation) a pre-test was necessary. In this pre-test twenty undergraduates were approached in the University canteen to view a total of six messages. After reading the messages the students had to rate each message in terms of feelings; either gratitude or obligation. These messages were judged based on the following items: grateful, recognized, valued, uncomfortable, obligated and irritated, adapted on the procedure of the research of Kolyesnikovaa et al. (2010). The first three items reflect a gratitude inducing message and the last three items an obligation inducing message. Based on the means and standard deviations scores, one message was defined as most suggestive of feelings of gratefulness and another message as most suggestive of obligatory feelings. After selecting these two messages, these were graphically designed to be appearing as a Tweet. The two tweets were displayed identically, except for the message, which was different. The most gratitude inducing message was: “Leuk dat je onze Twitter pagina bezoekt, heb je vragen, stel ze gerust!” and the most obligation inducing message was: “Wij hebben er alles aan gedaan om Coca Cola zo goed mogelijk te maken, nu is het aan de klanten om dit zo veel mogelijk te drinken!” For the background of the tweets the original Twitter account of Coca Cola was chosen. The brand Coca Cola is chosen because this brand is used in many academic research (Escalas & Battman, 2003; Fennis, 2008;
Louis & Lombart, 2010; Zhang & Sood, 2002). All the current tweets of Coca Cola were removed and a
12 2.3.2 Self-monitoring
In the fourth page of the experiment self-monitoring questions were prompted. The self-monitoring construct can be measured by a 25 item scale of Snyder (1974). In the research of Snyder and Gangestad (1986) the scale is adjusted to an 18 item scale due to the discovery that the old scale comprised a meaningful and interpretable causal variable with pervasive influences on social behavior, a variable reflected as a general self-monitoring factor. Hence the new scale of Snyder and Gangestad (1986) will be used in the experiment. All the respondents had to answer the questions concerning self-monitoring and afterwards the respondents were divided in high or low self-monitoring by splitting the database on the median (Median was 4.58). 18 statements were stated and the undergraduates had to answer these statements on a seven point Likert scale with anchors on (1) “Totally disagree” and (7) “Totally agree”. In appendix B the scale is translated into Dutch. A higher response on questions 4,5,6,8, 10,12,17, and 18 reflects high self-monitoring and a lower response on questions 1,2,3,7,9,11,13,14,15 and 16 also reflects high self-monitoring. The Cronbach’s alpha for this scale was 0.62. To clarify: the subjects were divided in high self-monitoring or low self-monitoring groups after answering the self-monitoring scale.
2.4 Dependent variables
In this subdivision of the method chapter the dependent variables will be explained. First the variable brand attachment and afterwards the dependent variable gender dimensions of brand personality will be elucidated.
2.4.1 Brand attachment
13 experiment the different means of the obligation group and the inducing group will be compared. In line with the research of Whan Park et al. (2010) the following four items are used to measure brand attachment. The first two questions refer to the component brand self-connection and the last two refer to the component: brand prominence.
- “To what extent is Coca Cola part of you and who you are?”
- “To what extent do you feel that you are personally connected to Coca Cola
- “To what extent are your thoughts and feelings toward Coca Cola often automatic, coming to mind seemingly on their own?”
- “To what extent do your thoughts and feelings toward Coca Cola come to you naturally and instantly?”
2.4.2 Gender dimensions of brand personality
To investigate to what extent the different Tweets effect the gender personality of the brand Coca Cola, the scale of Grohmann (2009) is used. Grohmann (2009) developed a scale which is two dimensional and measures masculine and feminine brand personality. Each dimension was measured by six
characteristics. To describe masculine personality the following items were used: adventurous,
aggressive, brave, daring, dominant and sturdy. To assess the female brand personality the next items were used: expressive tender feelings, fragile, graceful, sensitive, sweet and tender. After reading the Tweet, the undergraduates had to rank the brand Coca Cola on all the dimensions using a 9 point scale anchoring, by “not at all descriptive” (0) and “extremely descriptive” (8) (masculine and feminine traits
α= 0.84).This scale is translated in Dutch. Average item scores are used to reflect the two dimensions.
2.5 Analytical procedures
14 3. RESULTS AND CONCLUSIONS
In the final part of this paper the results and conclusions will be elaborated. First the findings of the experiment will be presented. After that the results will be discussed. Subsequently the conclusions and practical implications will be given. This chapter ends with limitations and suggestions for future research.
3.1 Results
Data were analysed using a 2 (obligation or gratitude twitter messages) × 2 (high or low self-monitoring) full factorial ANOVA. Two ANOVA’s were performed. In the first ANOVA the dependent variable was brand attachment and in the second ANOVA the dependent variable was gender dimensions. The analyses yielded the following results. Hypothesis 1 stated that gratitude-inducing Twitter messages would be associated with higher attachment towards the brand, while obligation-inducing Twitter messages would be associated with a lower attachment towards the brand. Hypothesis 1 was not supported (F(1,76)= .98, p = .33). It was also examined in hypothesis 3 if there would be a statistical difference between gratitude and obligation inducing Twitter messages and high self-monitors and low self-monitors in relation to brand attachment. This interaction effect was not significant (F(1,76)= 2.28, p = . 33). Hypothesis 4 predicted that high self-monitors would score higher on brand attachment than low self-monitors. The results showed no significant outcomes (F(1,76)= .04, p = . 84). Hence, H 4 cannot be confirmed. Table one shows the mean score and standard deviation per experimental condition.
TABLE 1
ANOVA DV: Brand attachment Self-monitoring High Low Mean SD Mean SD Twitter messages Obligation inducing 3,31 1,01 2,98 1,14 Gratitude inducing 2,68 1,40 3,11 0,89
15 In the second ANOVA, the independent variables were the same as in the first ANOVA, only the dependent variable was changed into gender dimensions. Hypothesis 2 stated that gratitude-inducing Twitter messages would be associated with higher feminine brand personality towards the brand, while obligation-inducing Twitter messages would be associated with higher masculine brand personality
towards the brand (F(1,76)=44.88, p < .01 = .37). Both gratitude inducing messages means are higher
than the obligation inducing messages (Mobligation inducing=3.85, SD 1.05 vs. Mgratitude inducing =5.48 SD =
1.03). Hypothesis 2 is thereby supported. In hypothesis 5 the interaction effect between gratitude and obligation messages and self-monitoring was examined. The outcomes were not significant (F(1,76)= .03, p = .87). Hypothesis 6 predicted that high self-monitors would score higher on gender dimensions than low self-monitors. The results showed that this main effect was not significant (F(1,76)= .08, p = . 79). In table two the mean score and standard deviation per condition are provided.
TABLE 2
ANOVA DV: Gender dimensions Self-monitoring High Low Mean SD Mean SD Twitter messages Obligation inducing 3.84 1.24 3.86 .71 Gratitude inducing 5.42 1.11 5.52 .99
Note: Higher values indicating more feminine brand personality and lower values indicating more masculine brand personality.
16
3.2 Conclusion and discussion
The goal of the current study was to examine whether two types of Twitter messages have a different effect on brand attachment and whether this relationship is moderated by self-monitoring. Kolyesnikovaa et al. (2010) revealed in their research that obligation inducing tweets and gratitude inducing tweets have a different effect on the attitude towards a company and therefore in this study it was hypothesized that these two messages would have a different impact on brand attachment. The present findings do not support the two main effects and the interaction. One possible reason why the results did not support hypotheses is that the experimental manipulations did not have the intended effects on the subjects.
In the research of Kolyesnikovaa et al. (2010) gratitude and obligation inducing messages gave significant result on brand attitude. In this research the effect of these messages is measured by brand attachment, although brand attachment and brand attitude share some similarities, this does not guarantee that both will give significant results. In addition, other research suggests that the brand attachment scale is differentiated from measures of satisfaction, involvement, and brand attitudes (Thomson, McInnis & Park, 2005). The findings of this study are contrary to the most recent findings on gratitude and obligation messages (Fung & Carstensen 2003; Kolyesnikovaa et al. 2010). Although Fung and Carstensen (2003) focus mainly on elderly people, in this study undergraduates participated in the experiment. This immediately shows a major difference between the current study and their study, and therefore the outcomes may differ.
Another difference between the study of Kolyesnikovaa et al. (2010) and the current study is that they used direct mail and in this study the medium was Twitter. A plausible explanation to this extent could be that Twitter is an open medium and direct mail a closed, more personal medium. Consumers can therefore react differently. However in the research of Fischer and Reuber (2011) it is suggested that Twitter conversations between companies and consumers can result in cognitive responses.
17 Another dependent variable in this research was gender dimensions. It was hypothesized that gratitude and obligation messages would result in a more feminine or masculine brand personality. This was hypothesized because of the fact that Grohmann (2009) states that individuals can associate masculine and feminine personality traits with brands. The findings of this second performed ANOVA supports this hypothesis but only for one main effect namely, between the gratitude and obligation tweets on gender dimensions. This result confirms current literature about the fact that personal traits can be linked to brands (Aaker, 1997; Grohmann, 2009).
The other main effect and the interaction were not significant. This is possible due to the fact that in both hypotheses self-monitoring was incorporated. In both ANOVA’s the role of the moderator self-monitoring was not significant. Kassarjian (1971) states that personality is not the only variable that influences the behaviour process of the consumers. This is rather a process of many interrelated variables that influence each other. Therefore it is not odd that including one personal trait in this research gives no significant outcomes.
Considering the above discussion, and although many outcomes proved to be insignificant, the current research findings forges new ground by illuminating the role of different Twitter messages on the gender dimensions of a brand.
3.3 Practical implications
18 This means that when a company has a female target group, they can use gratitude inducing tweets, because this results in a stronger feminine brand personality. For these companies it is recommended to appreciate consumers in their tweets. For example; “For our loyal customers, 10 per cent discount this week.” Or “It is great to have you as our customer, please tell us if you want us to improve our assortment”.
The outcomes of this research results in a practical implication for practitioners because now it is obvious how they have to structure their tweets and which words they can best use. This especially is an important finding regarding the fact that there is not much scientific research done in the medium Twitter yet. Twitter is an open medium and therefore it is important that every posted tweet is well thought through because it influences the brand personality of the brand. The research of Opoku, Caruana, Pitt, Berthon, Wahlstrom and Nel (2009) underlines this statement. They postulate that the words that are selected for online messages affect the strength of the message and associations with other personality dimensions. In other words, each word that a brand uses online has effect on the perceived personality dimensions.
3.4 Limitations
As with all research, this study has limitations. First, this study was done in an experimental setting. Translating experimental results can have different results than in real life setting (Borgatta & Bohrnstedt, 1974).
Another concern was that people can perceive the Twitter messages different than intended. Although a pre-test is used, participants can perceive the Twitter messages differently than intended. Unfortunately, I did not have any manipulation checks to follow up on these possibilities. Thus, there is no data to confirm whether or not participants did or did not see their tweets as obligation inducing or gratitude inducing.
A different limitation is that there is no distinction made between Twitter users and non-Twitter users. This may affect the results of the study. Although it is showed in the research of Romaniuk and Gaillard (2007) that having strong associations is not positively related to past usage of brand preference. Hence, not making a distinction between Twitter users and non-Twitter users is not an extremely deficiency in this research.
19 to change. He demonstrated it is possible to manipulate attitude but to change the values customers attach to the attributes of the brand is much more difficult. Although this experimental setting manipulated two Twitter messages, it could be possible that the manipulation could have no effect and therefore may result in no significant outcomes. The manipulation (two Twitter messages) was in that case not strong enough to influence the brand attachment.
3.5 Implications for further research
This is the first study examining the effects of gratitude and obligation inducing Twitter messages on brand attachment, gender dimensions and the moderating role of the personal trait self-monitoring variable. This study gives several implications for further research. Future research could investigate more personal traits which can influence the relationship between gender dimensions and gratitude and obligation tweets, for example sex role identity (Stern, Barak & Gould, 1987). If consumers sees themselves as highly feminine or highly masculine, this can have effect on recognizing these traits at a brand, this might be interesting to investigate.
In the current study individuals read one Twitter message of one brand (Coca Cola), in the future studies this can vary and be expanded more. For example study the differences between products and service brands and show more than one Tweet, this may result in more significant outcomes.
Another implication for further research is to replicate this research in a real life setting, in which individuals already follow a Twitter account of a brand and have to evaluate different Twitter messages. Although there were no significant results in the relationship between gratitude and obligation tweets and self-monitoring and brand attachment, this conceptual model can be adjusted for future research. It would be interesting, because of the fact that brand attachment was not significant in this research, brand attitude might be more appropriate to study. Or maybe other antecedents of brand attachment, for example brand name and symbols (Aaker 1997; Batra,Lehmann, & Singh 1993).
20 REFERENCES
Aaker, J. L. 1997. Dimensions of Brand Personality. Journal of Marketing Research, 34 (August), 347–56.
Allen, D.G, Weeks , K.P & Moffitt, K.R. 2005. Turnover intentions and voluntary turnover: The moderating roles of self-monitoring, locus of control, proactive personality, and risk aversion. Journal of Applied Psychology, 90, 980–990.
Bagozzi, B. 1975. Marketing as Exchange. Journal of Marketing, Vol. 39 (October 1975), pp. 32-39.
Batra, R. Lehmann, D. R., & Singh D. 1993. The Brand Personality Component of Brand Goodwill: Some Antecedents and Consequences. Brand Equity and Advertising , eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 83–96.
Becherer, R. C., & Richard, L. M. 1978. Self-monitoring as a moderator of consumer behavior. Journal of Consumer Research, 5, 159-162.
Borgatta, E. & Bohrnstedt, G. 1974. Some Limitations On Generalizability from Social Psychological Experiments. Sociological Methods & Research 3:1, 111-120
Brown, S.P. Homer, P.M. & Inman J.J. 1998. A meta-analysis of relationships between ad-evoked feelings and advertising responses. Journal of Marketing Research 35, no. 1: 114–26.
Deaux, K. 1985. Sex and Gender. Annual Review of Psychology, 36, 49–81.
Eagly, H. & Mladinic A. 1989. Gender Stereotypes and Attitudes Toward Women and Men. Personality and Social Psychology Bulletin 15:4 543-558
21 Escalas, J. & Bettman, J. 2003. You Are What They Eat: The Influence of Reference Groups on Consumers’ Connections to Brands. Journal of Consumer Psychology, 13(3), 339–348
Fennis, B. M. 2008. Branded Into Submission: Brand Attributes and Hierarchization Behavior in Same-Sex and Mixed-Same-Sex Dyads. Journal of Applied Social Psychology, 38: 1993–2009.
Fischer, E. & Reuber, R. 2011. Social interaction via new social media: (How) can interactions on Twitter affect effectual thinking and behavior? Journal of Business Venturing 26; 1–18
Fletcher, K. & Peters L. 1997. Trust and Direct Marketing Environments: a Consumer Perspective. Journal of Marketing Management, 13, 523-539
Fung, H.H., & Carstensen L.L. 2003. Sending memorable messages to the old: Age differences in preferences and memory for advertisements. Journal of Personal and Social Psychology 85, no. 1: 163– 78.
Golden, L., Allison, N. & Clee, M. 1977. The Role of Sex Role Self-Concept in Masculine and Feminine Product Perceptions. Advances in Consumer Research, Vol. 5, 599–605.
Gouldner, A.W. 1960. The norm of reciprocity: A preliminary statement. American Sociological Review 25, no. 2: 161–78.
Grohmann, B. 2009. Gender Dimensions of Brand Personality. Journal of Marketing Research Vol. XLVI (February 2009), 105–119
Harrison, D. A. & Price, K. H. 2003. Context and consistency in absenteeism: Studying social and dispositional influences across multiple settings. Human Resource Management Review, 13, 203–226.
22 Java, A., Song, X., Finin, T. & Tseng, B. 2007. Why we twitter: understanding micro blogging usage and communities. In Proc. of the 9th WebKDD and 1st SNA-KDD 2007 workshop on Web mining and social network analysis. ACM, 56-65.
Kassarjian, H. 1971. Personality and Consumer Behavior: A Review. Journal of Marketing Research, Vol. VIII , 409-18
Kolyesnikovaa, N. Sullivan S. & Callisonc C. 2010. Consumer affective responses to direct mail messages: The effect of gratitude and obligation. Journal of Marketing Communications , 1-17
Lambert-Pandraud, R. & Laurent, G. 2010. Why Do Older Consumers Buy Older Brands? The Role of Attachment and Declining Innovativeness. Journal of Marketing, Vol. 74 104–121
Louis, D. & Lombart, C. 2010. Impact of brand personality on three major relational consequences (trust, attachment, and commitment to the brand). Journal of Product & Brand Management 19/2; 114–130 Lutz, R.J. 1975. Changing Brand Attitudes Through Modification of Cognitive Structure. Journal of Consumer Research, 1, March, 49-59.
Michelson, M. & Macskassy, S. A. 2010. Discovering users’ topics of interest on twitter: A first look. In Proceedings of the Workshop on Analytics for Noisy, Unstructured Text Data (AND), Toronto, Canada. to appear.
Mikulincer, M. & Shaver, P. 2007. Attachment in Adulthood: Structure, Dynamics and Change. New York: The Guilford Press.
Opoku, R. Caruana, A.. Pitt L., Berthon P..Wahlstrom A & Nel. D. 2009. Online communication of brand personality A study of MBA programs of top business schools. Journal of General Management Vol. 35 No. 1 47- 64
23 Puri, R. 1996. Measuring and Modifying Consumer Impulsiveness: A Cost Benefit Accessibility Framework. Journal of Consumer Psychology. 5(2), 87-113
Romaniuk , J,. & Gaillard, E. 2007. The relationship between unique brand associations, brand usage and brand performance: analysis across eight categories. Journal Of Marketing Management, Vol. 23, No. 3-4, pp. 267-284
Rust, R. Lemon, K. & Zeithaml, V. 2004. Return on Marketing: Using Customer Equity to Focus Marketing
Strategy. Journal of Marketing 68(1)109-127
Scovotti, C. & Spiller, L. D. 2006. Revisiting the conceptual definition of direct marketing: perspectives from practitioners and scholars. The Marketing Management Journal, 16(2), 188-202
Snyder, M. 1974. Self-Monitoring of Expressive Behavior," Journal of Personality and Social Psychology. 10. 526-37.
---- --- & Gangestad, S. 1986. On the Nature of Self-Monitoring: Matters of Assessment, Matters of Validity. Journal of Personality and Social Psychology, W. 51, No. 1,125-139
Stern, B. Barak, B. & Gouldner, S. 1987. Sexual Identity Scale: A New Self-Assessment Measure. Social Sciences (General) / Women's Studies 17:9/10 503-519
Swaminathan, V., Stilley, K. & Ahluwalia, R. 2008. When Brand Personality Matters: The Moderating Role of Attachment Styles. Journal of Consumer Research, 35 (April), 985-1002.
Thomson, M., MacInnis, D. & Whan Park C. 2005. The Ties That Bind: Measuring the Strength of Consumers’ Emotional Attachments to Brands. Journal of Consumer Psychology, 15 (1), 77-91.
25 APPENDIX A SCALES AND TRANSLATIONS
1. Pre-test, Kolyesnikovaa N., Sullivan S., Doddb and Coy Callisonc (2010)
Gratitude Obligation Grateful = dankbaar* Recognized = erkend* Thankful = dankbaar Valued = gewaardeerd* Acknowledged = gewaardeerd Positive = positief Pleased = tevreden Honored = vereerd Uncomfortable = ongemakkelijk* Obligated = verplicht* Indebted = verschuldigd Reluctant = huiverig Guilty = schuldig Pressured = gedwongen Irritated = geïrriteerd* Owing = verschuldigd * = is used in the pre-test
2. Brand attachment scale Whan Park C., MacInnis D., Priester J.. Eisingerich, A., & Iacobucci D. (2010) 1To what extent is [Brand Name] part of you and who you are?
In hoeverre is het merk Coca Cola een deel van je en van wie je bent?
2 To what extent do you feel that you are personally connected to [Brand Name]? In welke mate voel je je persoonlijk verbonden met het merk Coca Cola?
3 To what extent are your thoughts and feelings toward [Brand Name] often automatic, coming to mind seemingly on their own?”
In hoeverre roept het merk Coca Cola onbewust gevoelens en gedachten bij je op die je moeiteloos schijnen binnen te schieten?
4“To what extent do your thoughts and feelings toward [Brand Name] come to you naturally and instantly?”
In hoeverre zijn je gedachten en gevoelens betreffende het merk Coca Cola vanzelfsprekend en spontaan?
3. Extra dependent variable; Gender dimensions of brand personality Grohmann (2009)
Masculine brand personality Feminine brand personality
1. adventurous = avontuurlijk 2. agressive = agressief 3. brave = moedig 4. daring =gewaagd 5. dominant =dominant 6. sturdy =stoer
1. expressive tender feelings= zachtmoedige
26 4. Moderator Self-monitoring 18 items scale of Snyder and Gangestad (1986)
I find it hard to imitate the behavior of other people.
Ik vind het moeilijk om het gedrag van andere mensen te imiteren.
At parties and social gatherings, I do not attempt to do or say things that others will like.
Op feestjes en bijeenkomsten, probeer ik geen dingen te zeggen of te doen die anderen leuk vinden. I can only argue for ideas which I already believe.
Ik kan alleen maar discussiëren over onderwerpen waar ik al achter sta.
I can make impromptu speeches even on topics about which I have almost no information. Ik kan onvoorbereid spreken over onderwerpen waar ik nauwelijks informatie over heb. I guess I put on a show to impress or entertain others.
Ik denk dat ik mijzelf anders voordoe om andere mensen te vermaken of om indruk te maken. I would probably make a good actor.
Ik zou waarschijnlijk een goede toneelspeler zijn. In a group of people I am rarely the center of attention.
In een groep mensen sta ik zelden in het middelpunt van de belangstelling.
In different situations and with different people, I often act like very different persons. Mijn gedrag is sterk afhankelijk van de situatie en het gezelschap waar ik me in bevind. I am not particularly good at making other people like me.
Ik ben er niet zo goed in ervoor te zorgen dat anderen me aardig vinden. I'm not always the person I appear to be.
Ik kom wel eens anders over dan dat ik werkelijk ben.
I would not change my opinions (or the way I do things) in order to please someone or win their favor. Ik zal niet van gedachten veranderen (of dingen anders doen) om iemand een plezier te doen of voor me te winnen.
I have considered being an entertainer. Ik heb overwogen om entertainer te worden.
I have never been good at games like charades or improvisational acting. Ik ben nooit goed geweest in spelletjes zoals Hints of geïmproviseerd acteren. I have trouble changing my behavior to suit different people and different situations.
Ik vind het moeilijk om mijn gedrag aan te passen aan diverse mensen in verschillende situaties. At a party I let others keep the jokes and stories going.
27 I feel a bit awkward in public and do not show up quite as well as I should.
Ik voel me ongemakkelijk in groepen mensen en daardoor kom ik niet over zoals ik echt ben. I can look anyone in the eye and tell a lie with a straight face (if for a right end).
Ik kan mensen recht in de ogen aankijken en zonder blikken of blozen een leugen om bestwil vertellen. I may deceive people by being friendly when I really dislike them.
28 APPENDIX B REPRESENTATION OF THE PRE-TEST
Beste student,
Bedankt dat je wilt deelnemen aan mijn onderzoek voor mijn Master Thesis. Dit onderzoek gaat over het beoordelen van korte Twitter berichten en zal ongeveer vijf minuten duren. Op de volgende pagina zal je een aantal Twitter berichten lezen, ik wil je vragen om deze berichten te evalueren aan de hand van een aantal woorden. Mocht je vragen hebben, stel ze gerust. De antwoorden worden anoniem verwerkt.
Heel erg bedankt voor je deelname!
29 1. “Wij hebben er alles aan gedaan om Coca Cola zo goed mogelijk te maken, nu is het aan de
klanten om dit zo veel mogelijk te drinken!” Door het lezen van dit bericht voelde ik mij:
Niet mee eens mee eens
Dankbaar Erkend Gewaardeerd Ongemakkelijk Verplicht Geïrriteerd
2. “Kijk nu naar onze nieuwe reclame op www.cocacola.com voordat deze vanavond wordt uitgezonden!”
Door het lezen van dit bericht voelde ik mij:
Niet mee eens mee eens
Dankbaar Erkend Gewaardeerd Ongemakkelijk Verplicht Geïrriteerd
3. “ Alleen vandaag Coca Cola voor de helft van de prijs! Koop Coca Cola zo snel mogelijk!” Door het lezen van dit bericht voelde ik mij:
Niet mee eens mee eens
Dankbaar Erkend Gewaardeerd Ongemakkelijk Verplicht Geïrriteerd
4. “Wij waarderen lekker weer, lekkere drankjes en goede klanten. Elk moment is een goed moment voor Coca Cola.”
Door het lezen van dit bericht voelde ik mij:
Niet mee eens mee eens
30 5. “We waarderen al onze klanten, daarom nu genieten van extra grote flessen Coca Cola, voor
dezelfde prijs. “
Door het lezen van dit bericht voelde ik mij:
Niet mee eens mee eens
Dankbaar Erkend Gewaardeerd Ongemakkelijk Verplicht Geïrriteerd
6. “Leuk dat je onze Twitter pagina bezoekt, heb je vragen, stel ze gerust!” Door het lezen van dit bericht voelde ik mij:
Niet mee eens mee eens
31 APPENDIX C REPRESENTATION OF THE EXPERIMENT
Beste student,
Bedankt dat je wilt deelnemen aan mijn onderzoek voor mijn Master Thesis! Dit onderzoek gaat over Twitter en zal ongeveer vijf minuten duren. Graag wil ik je vragen om eerst je leeftijd en geslacht in te vullen. Op de volgende pagina zal je een Twitter bericht lezen van het merk Coca Cola. Lees rustig deze Tweet en neem deze in je op. De vragen daarna zullen uitsluitend gaan over de evaluatie van de Tweet en het merk wat gebruikt is. Mocht je vragen hebben tijdens het experiment, stel ze gerust. De
antwoorden worden anoniem verwerkt.
Heel erg bedankt voor je deelname!
Irene Wiersma
1. Wat is je leeftijd? ………….
2. Wat is je geslacht?
33 Beantwoord de volgende vragen na het lezen van de Tweet:
Helemaal niet Volledig
1. In hoeverre is het merk Coca Cola een deel van je en van wie je bent?
2. In welke mate voel je je persoonlijk verbonden met het merk Coca Cola?
3. In hoeverre roept het merk Coca Cola onbewust gevoelens en gedachten bij je op, die moeiteloos schijnen binnen te schieten
4. In hoeverre zijn je gedachten en gevoelens betreffende het merk Coca Cola vanzelfsprekend en spontaan.
In welke mate vind je de volgende woorden passen bij de inhoud van de Tweet?
Totaal niet Totaal wel
1. Avontuurlijk 2. Agressief 3. Moedig 4. Gewaagd 5. Dominant 6. Stoer
7. Expressieve tedere gevoelens
8. Breekbaar
9. Sierlijk
10. Gevoelig
11. Zoet
34 Helemaal mee
oneens
Helemaal mee eens De volgende vragen gaan over jezelf. Geef aan in welke mate je het met de volgende stellingen eens bent.
1. Ik vind het moeilijk om het gedrag van andere mensen te imiteren. 2. Op feestjes en bijeenkomsten, probeer ik geen dingen te zeggen of te doen die anderen leuk vinden.
3. Ik kan alleen maar discussiëren over onderwerpen waar ik al achter sta.
4. Ik kan onvoorbereid spreken over onderwerpen waar ik nauwelijks informatie over heb.
5. Ik denk dat ik mijzelf anders voordoe om andere mensen te vermaken of om indruk te maken.
6. Ik zou waarschijnlijk een goede toneelspeler zijn.
7. In een groep mensen sta ik zelden in het middelpunt van de belangstelling.
8. Mijn gedrag is sterk afhankelijk van de situatie en het gezelschap waar ik me in bevind.
9. Ik ben er niet zo goed in ervoor te zorgen dat anderen me aardig vinden.
10. Ik kom wel eens anders over dan dat ik werkelijk ben.
11. Ik zal niet van gedachten veranderen (of dingen anders doen) om iemand een plezier te doen of voor me te winnen.
12. Ik heb overwogen om entertainer te worden .
13. Ik ben nooitgoed geweest in spelletjes zoals Hints of
geïmproviseerd acteren .
14.Ik vind het moeilijk om mijn gedrag aan te passen aan diverse mensen in verschillende situaties.
15. Op een feestje laat ik anderen de grappen maken en verhalen vertellen.
16. Ik voel me ongemakkelijk in groepen mensen en daardoor kom ik niet over zoals ik echt ben.
17.Ik kan mensen recht in de ogen aankijken en zonder blikken of blozen een leugen om bestwil vertellen.
