Incorporating emotions as antecedents and mediators in Theory of Reasoned Action (TRA) model

Theory of Reasoned Action (TRA) is widely used in different decision making situations. The TRA model measures behavioral intentions and predicts a certain behavior on the basis of these behavioral intentions. Literature has questioned the predictive ability of TRA model in terms of behavior prediction in a decision making situation as there are number of factors which may influence the decision making process. One of these factors is identified as “emotions” experienced during the decision making process. The paper makes an attempt to improve the TRA model by incorporating the emotions at two different points in the model i.e. emotions as antecedents and emotions as mediators between attitude and behavioral intention. It is proposed that this inclusion of emotions in the TRA model will improve the overall predictability power of the model.


Introduction
Theory of Reasoned Action (TRA) is proposed to explicate the process through which human behavior is developed or performed. It posits that behavior of a person is dependent upon the attitudes, beliefs, and intentions. It assumes that intention is the most important determinant of the behavior. This intention is formed after considering attitude of self and others regarding the behavior. It also assumes that all other exogenous factors operate through the model. TRA has been widely used in various settings including health, education, consumer behavior, brand choice etc. Various researchers have attempted to improve both the measurement of intentions and forecasting ability of the model into actual behavior. The concept has also been extended and modified to understand specific behaviors related to health and education. Models like protection motivation theory, prevention motivation theory, and health belief model are rooted in the concepts proposed by TRA.
Although the power of the model to measure intentions has been good but its ability to predict behavior based on the measured intentions has been questioned on a number of occasions. It has been accepted in literature that external factors like time, resources, promotional campaigns etc. may influence consumer decision at the time of purchase which Business Review -Volume 9 Number 2 July -December 2014 41 can cause variation in the predictive ability of the model whereby behavioral intention is converted into actual behavior. These limitation of the model is also accepted by Ajzen (1991) and he consequently introduced an improved version of the model i.e. Theory of Planned Behavior (TPB).
The proposed paper is an attempt to improve the TRA model by incorporating various dimensions of emotions at different points of the model consequently improving both the measurement of behavior intention and predictability of the behavior. Although affect has been incorporated in the original model of TRA in the form of evaluation of beliefs, but the authors interchangeably used the terms of evaluation and affect, hence, masking the discriminant effect of human emotions in predicting the actual behavior. This fact, probably, has reduced the power of the model and has created an error in the measurement of behavioral intentions. It is hypothesized that the addition of these components will help both the academicians and the practitioners in better understanding of how human behaviors are formed. This will consequently improve their ability to predict the behavior. These changes will also help in determining the effective strategies of creating interventions for required changes in the human behavior. The paper is organized in the following manner. First part of the paper gives brief overview of literature of TRA and emotions. Second section then provides the theoretical discussion on improved model of TRA.

Literature Review
Recent developments in the field of emotions clearly demonstrate that emotions have different forms and intensity and that these emotions can be discriminated quantitatively. This is suggested based on the multiple reasons. Firstly, it is argued that affect, emotions, and moods are distinct categories and need individual treatment while predicting human behaviours Ekman and Davidson (1994). Secondly, emotions have four distinct dimensions (conceptual breadth, possibility versus probability, dynamic appraisal versus static and temporal focus) that can help researchers in truly predicting human behavior (Henning, Thurau, & Feiereisen, 2012). Thirdly, emotions have significant role in consumer decision making process that make it imperative for market researchers to use consider set of emotions while designing and implementing studies (Gardner, 1985).Hence it can be said that affect has different facets, it can appear in different forms like enjoyment of performing a behavior, attachment with a product, guilt, anger, association with referents, and putting in efforts to gather information. Hence, the interplay of affect dimension on human behavior can be captured at various points within the existing model of TRA.

Theory of Reasoned Action
The Theory of Reasoned Action (TRA) (Ajzen & Fishbein, 1977) explores the relationship between volitional behaviors attitudes, and subjective norms through the creation of behavior intentions. The theory posits that individual's intentions are the best determinants of behavior in question. Behavioral Intention (BI) measures the motivation of an individual to perform a given behavior, hence higher the BI higher will be the probability of performing that behavior. BI is function of person's attitude towards performing that behavior and his consideration of approval or disapproval of other people (subjective norm). The theory assumesthat all exogenous factors operate through the model's constructs and have no independent effect on the behavior In TRA attitude is defined as person's overall evaluation about the subject behavior. The model suggests that attitude is comprised of two factors: salient beliefs about the https://ir.iba.edu.pk/businessreview/vol9/iss2/4 DOI: https://doi.org/10. 54784/1990-6587.1270 behavior (Bi) and the positive or negative evaluation of the outcome of the behavior (Ei). This evaluation is the affective component of the model. The product of belief and evaluation is then summated to calculate overall attitude of the person towards that behavior i.e.
Subjective norm reflects the social pressure that an individual takes into consideration that is exerted from his important referents while considering to perform the subject behavior. Like attitude it is also a summation of product of the perceptions about the beliefs of others (BOi) and the motivation to comply(MCi) with the important othersi.e.
Subjective Norms = ∑BO i xMC i TRA has been applied in many settings to measure various behaviors like coupon usage (Shimp & Kavas, 1984) and moral behaviors (Vallerand, Deshaies, Cuerrier, Pelletier, & Mongeau, 1992). There have been quite a few meta analyses conducted so far to understand TRA (Cooke & French, 2008;Hausenblas, Carron, & Mack, 1997;Sheppard, Hartwick, & Warshaw, 1988).The model appears to predict consumer intentions quite well and it also provides a relatively simple basis for strategizing behavioral change attempts. The simplicity of the model has been well accepted and the model has been found to be robust under various conditions. The model was improved by (Ajzen & Fishbein, 1980) when he incorporated the self efficacy concept to improve the overall predictability of the model. That model came to be known as Theory of Planned Behavior or TPB. However, recently various researchers have begun to question the simplicity of the TRA model. Few researchers have challenged the separation of subjective norms and attitude constructs (Oliver & Bearden, 1985). They argue that there exists crossover effect between the two constructs, hence SN construct should not be measured separately rather it should be made part of the attitude construct. Another emerging area is the way emotions are measured in the traditional TRA model. It is argued that affective component is not being measured effectively and its effects are being masked by the current methods of measurement (Allen, Machleit, & Kleine, 1992;Henning et al., 2012).
Theory of Reasoned Action is widely accepted to predict human behavior but like all other theories, it is also not free from criticism. Firstly, Ajzen and fellow researchers favored the idea that human behavior is dependent upon attitudes, beliefs, and intentions. This assumptive definition of human behavior was criticized mainly on limiting the predictive ability of TRA model such that human behavior is modeled by factors like time, resources, and promotional campaigns as well. Secondly, another development in the field of understanding human behavior is the inclusion of emotions that was not properly included in the initial modeling of TRA. Likewise, it is believed that affect has different set of facets such as; attachment with a product, guilt, anger, association, and putting in some efforts to gather information. It is concluded through philosophical assumptions that researchers use BI to predict behavior through TRA, resulting in high degree of errors in predicting human behavior. This calls for an improvement in TRA, specifically in terms of predicting human behavior by incorporating emotional aspect in the model.

Emotional Aspect of TRA
As mentioned earlier, TRA measures attitudes as a product of beliefs and evaluations. According to Fishbein and Ajzen (1975) evaluations tell the directionality or the valence of the attitude (like/dislike, positive or negative aspects) or it displays the importance https://ir.iba.edu.pk/businessreview/vol9/iss2/4 DOI: https://doi.org/10. 54784/1990-6587.1270 of the behavior when measured on important/not important scales. They posit that the terms "evaluation" and "affect" are synonym. They based their argument on the premise that it is difficult to empirically distinguish between the emotion and evaluation. However recent advances in consumer research has shown that the two constructs are empirically different (Henning et al., 2012).
According to Ekman and Davidson (1994) affect, emotions and moods are three distinct categories. Affect is a broader term encompassing all types of emotions whereas emotions are directional always associated with an object and are associated with actions and are of short duration while moods have longer duration, less intense and are not typically associated with action and are not under the control or non-intentional in nature. The experiential view of consumer behavior (Hirschman & Holbrook, 1982) suggests that the traditional information processing view is useful for studying utilitarian products and situations while the affective responses are required when studying hedonic consumptions situations. Henning et al. (2012) presented four aspects of emotions that makes them distinct from evaluations: conceptual breadth, possibility versus probability, dynamic appraisals versus static predispositions, and temporal focus. Conceptual breadth refers to the wide variety of human emotions that a person can experience; Probability vs. possibility highlights that attitudes are measured in terms of probability whereas emotions are highly sensitive and can be felt even when there is slight probability of an outcome.
TRA assumes attitudes as predisposition and static in nature; on the contrary emotions are dynamic in nature and change according to the context. Emotions exist at pre, during and post stages of decision making. This reflects the temporal aspect of emotions.
Literature also highlights the impact of emotions on consumer decision making (Gardner, 1985). Goldberg and Gorn (1987) found that positive mood increases therecall and information processing ability of the consumer. Batra and Stayman (1990) analyzed impact of consumer emotions and attitudes on advertisement. Their findings supported the concept that mood had a direct effect on brand attitude. Positive moodnot only reduces the level of negativity about the brand by reducing negative thoughts but also reduces cognitive elaboration. According to them happy mood makes processing 'more heuristic than systematic'. In their article Allen et al. (1992) posit that emotion extends beyond attitude and encompasses a richer and more diverse domain of phenomenological experience. According to (Batra & Ahtola, 1991) consumer purchase of goods can be attributed to either affective or instrumental reasons, while Allen et al. (1992) found that emotions explain behaviors over and above attitudes. Emotions appear to have evolved as drivers of behavior because of their approach/avoidance function hence they also have the ability mobilize energy (Zeelenberg, Nelissen, Breugelmans, & Pieters, 2008). In the similar vein Allen et al. (1992)research showed that emotions have a direct effect on behavior which is over and above the effect of attitudes. Similar types of results were also obtained by Perugini and Bagozzi (2001) who find that inclusion of emotions increase variance explanation of behavior significantly.
In the conclusion of above discussion it can be said that use of BI to predict behavior through TRA is devoid of emotional contents and it may result in high degree of errors in predicting human behavior. This calls for improvement in TRA. Since the focus of TRA is on creation of behavior intention not the after effects of behavior, the proposed model will also focus on only two impacts of emotion which are critical during the process of decision https://ir.iba.edu.pk/businessreview/vol9/iss2/4 DOI: https://doi.org/10. 54784/1990-6587.1270 making: emotions as antecedents; and emotions as mediator of attitude and behavior intention.

Emotions as Antecedents
Antecedent is any event, circumstance, cause or happening that occurs before the event under study. When we argue that emotions are an antecedent to attitude or behavior it implies that emotions exist before the attitudes or behaviors are formed. Specifically in the case of TRA it implies that emotions exist before the consumer starts the process of evaluation. As mentioned earlier, the literature highlights that emotions may exist even before the process of decision making starts (Henning et al., 2012).
Literature on emotions proposes that emotion can be treated as an antecedent to attitude (Allen et al., 1992). According to Kempf (1999) emotions are antecedents of attitude for hedonic based products, similar arguments were also presented byBatra and Stayman (1990) who, while using theory of ELM, highlighted that affect is an antecedent of attitude. This aspect also gets support from the stream of literature that argues that emotions are drivers of behaviors by virtue of their approach avoidance behavior (Henning et al., 2012).
According to the literature mood could be one of the antecedents to the process of evaluation (Gardner, 1985). According to her the term mood is used in literature in a very broad term and encompasses all preconceived thinking before making a decision. Since, TRA assumes that all the information exist at the start of the process of evaluation, hence when we consider emotions as an antecedent mean it has an effect on the information processing ability of the consumer. Research has highlighted that mood has following effects: it improves information recall and processing ability (Goldberg & Gorn, 1987), it has an impact on brand attitudes and brand elaboration (Batra & Stayman, 1990), and it reduces size of evoked set (Zeelenberg et al., 2008). It was also found that in situation of increased time pressures and high demanding tasks people may tend to rely more on their moods (Pham, Cohen, Pracejus, & Hughes, 2001).
The concept of "feeling is for doing" argues that people may take decisions without considering detailed information. It shows that decision makers sometimes follow their gut feeling when deciding. While the literature on "How do I feel about it?" heuristic (Schwarz & Clore, 1988) argues that people may choose their decision based on their feelings towards the target. They choose their preference based on the intensity of these feelings(G. J. Gorn, Goldberg, & Basu, 1993;Levine, Wyer Jr, & Schwarz, 1994). People may additionally infer the strength of their preferences by monitoring the intensity of these feelings; that is, the level of arousal elicited by the target (Gorn, Pham, & Sin, 2001). According to the affect-asinformation framework (Schwarz & Clore, 1996), people rely on their feelings because they perceive these feelings to contain valuable judgmental information. Using the concept of affect as information, Pham et al. (2001) argued that feelings contain important information and people tend to listen to these while making decisions. Moreover, according to Maddux, Gosselin, Leary, and Tangney (2002)our emotional state, that is either we are calm or distressed, effects our self efficacy beliefs and hence our decision making. Summarizing the above discussion it can be clearly seen that emotions are an antecedent to evaluation in TRA. Proposition 1: Emotions are an antecedent to evaluation in TRA.
https: //ir.iba.edu.pk/businessreview/vol9/iss2/4 DOI: https://doi.org/10.54784/1990//ir.iba.edu.pk/businessreview/vol9/iss2/4 DOI: https://doi.org/10.54784/ -6587.1270 Emotions as Mediators The Cognitive Appraisal Theory (Watson & Spence, 2007) posits that it is the cognitive appraisal of a situation that generates our emotions which inturn affects our behaviors. The theory tries to explain how various emotions are created and how these lead to different behavioral responses. Appraisal theory also argues that emotions are the results of comparison and evaluation of actual and desired states of a person . Personal stakes are a necessary condition for generating a emotional response in that the event either helps or hinders in performing the desired behavior.Watson and Spence (2007)has highlighted the following issues that appraisal theory addresses regarding situation and emotion: 1.
To understand the characteristics of the events that is being appraised 2.
To identify which emotions, if any, are being experienced 3.
To identify behavioral responses given to the emotions elicited.
They also showed that situation when appraised in terms of fairness or certainty people feel different types of emotions in each case. For example happiness and satisfaction may be experienced in situations of certainty and anger or outrage if situation is perceived as unfair. Hence various appraisals of situation will give rise to different emotions which inturn will create different behaviors. Regarding relationship between an event and emotions Bagozzi, Gopinath, and Nyer (1999) wrote in their paper:"a necessary condition for an emotional response to an event or happening is that a personhas a personal stake in it and at the same time judges the event or happening to facilitate orthwart this stake (p. 185)".
Hence, based upon the above arguments it can be said that emotions mediate between evaluation of a situation and behavior formed consequently.
Proposition 2: Emotions will mediate between attitude and behavioral intention

Conclusion
Emotions play critical role in creating behaviors in our everyday life. Emotions exist as an antecedent and as a mediator of human behavior. Understanding antecedent emotional states are critical for the marketing manager in order to connect the value of the product or services with the customer. Moods are a typical case of emotions as antecedents. In the case of service based business like retail stores, customer care centers, hotels environment could be made in such a way that has positive impact on customer's mood. Mood states are known to create or hinder shopping behavior, advertising effect, recall of the brand. Understanding of these mood states will help the brand manager. In case of products the antecedent emotional states to consumption refers to need arousal and emotional cues. Understanding such cues and emotional states may lead to better understanding of consumer decision making process alongwith purchase and consumption behavior. In certain situation, like a highly ethnocentric behavior or a highly religious behavior, antecedent emotions may be so strong that they will directly create behavior and no formal cognitive appraisal will take place. This could also be observed in mob or group behavior. In situations where emotions run high cognitive decision making stops.
Similarly if customer is involved in the situation of decision making or consumption various emotional factors come into play. For example in a high involvement decision situation where risk of wrong purchase will be high and confidence in the decision is low emotions of fear will guide the behavior, where as in high involvement situation where https://ir.iba.edu.pk/businessreview/vol9/iss2/4 DOI: https://doi.org/10. 54784/1990-6587.1270 outcome expectations are positive and related to happiness the decision will be guided by the feeling of love or positive emotion set. Hence, reflection of a situation after analysis will generate either a single or a set of emotions that will impact the behavior. Therefore, for the marketing manager understanding of various generic emotions or set of emotions will help us in managing the response of the customer, creation of the advertising message and in creating strong relationship of the customer with the brand.