Attitude toward the ad and brand choice.

Provo, UT: Association for Consumer Research, 152-9.

Gabriel Biehal (Ph.D., Stanford University) is an Associate Professor of Marketing at the University of Maryland, College Park, Maryland.

Debra Stephens (Ph.D., University of Chicago) is an Assistant Professor of Marketing at Villanova University, Villanova, Pennsylvania.

Eleonora Curlo, a doctoral candidate at the University of Maryland, is now Assistant Professor of Marketing, Instituto de This study examines how attitude toward the ad, Aad, affects brand choice. We investigate the applicability of two contrasting perspectives--an independent or direct effects model and a mediated or indirect effects model. In a lab setting we manipulate (1) a brand's ad picture (good/bad) to affect its Aad and (2) when subjects give their Aad ratings (before/after choice). Choice is operationalized as selecting the best brand based on information contained in a set of ads. The results support the presence of an independent effect of Aad on brand choice. There were no significant timing effects, which suggests that subjects formed Aad during choice without prompting. Directions for future research and implications for advertisers are discussed.

Introduction

A strong managerial relevance to advertisers, coupled with the well-defined theoretical background of multiattribute attitude models (Fishbein and Ajzen 1975), has generated considerable research into attitude toward the ad (Aad). Research shows that Aad is affected by brand or nonbrand processing set (Hastak and Olson 1989; Madden, Allen and Twible 1988; Gardner 1985; Homer 1990), ad exposure level (Burke and Edell 1986; Cox and Cox 1988), message involvement (Park and Young 1986; Muehling and Laczniak 1988), the cognitive (Hastak and Olson 1989) and affective responses generated during ad exposure (Burke and Edell 1989; Batra and Ray 1986; Machleit and Wilson 1988), and ad message quality and content (Hastak and Olson 1989; Burton and Lichtenstein 1988). In turn, Aad has been found to be related to attitude toward the brand, AB (Mitchell 1986; Gardner 1985; Homer 1990; Muehling and Laczniak 1988; Stayman and Aaker 1988), advertised deal (Burton and Lichtenstein 1988), the likelihood a brand is considered (Moore and Hutchinson 1983), ad recall (Zinkhan and Fornell 1989), purchase intention (Mitchell and Olson 1981), and attitude towards the act of buying the brand (Mitchell 1986).

So far, however, there has been surprisingly little research that directly examines the Aad - brand choice relationship. This is an important issue for advertisers, who should be concerned that Aad affects not only AB and purchase intention, as has been found, but also brand choice (Shimp 1981). While intention and behavior, e.g., brand choice, may often be closely related, this is not always the case (Fishbein and Ajzen 1975). Further, as we shall argue, there are sufficient processing differences between intentions and brand choice that we need to be careful about accepting Aad - intentions results as indicative of the Aad - choice relationship. In particular, while existing Aad - intentions research shows support for indirect Aad effects (e.g., MacKenzie, Lutz and Belch 1986; Homer 1990), we examine the possibility that Aad may also directly affect brand choice. This constitutes the primary research issue addressed by our paper.

Our second concern is more methodological. Recently researchers have tried to better understand the antecedents of Aad (Homer 1990; MacKenzie, Lutz and Belch 1986). For example, MacKenzie and Lutz (1989) specify several Aad determinants, e.g., ad credibility and attitude toward the advertiser. Integrative models of ad information processing have also been described (MacInnis and Jaworski 1989). Despite these important contributions, it is still unclear when Aad formation actually occurs.

Advertisers need reassurance that Aad formation naturally precedes choice. While this is commonly assumed, so far the only evidence is correlational and based on attitude toward the act of purchase, not choice (Mitchell 1986). Yet research shows that consumers can choose without forming an overall AB (Loken and Hoverstad 1985; Chattopadhyay and Alba 1988). This makes the formation of Aad, its impact on AB and choice and its relevance to advertisers more problematic. These considerations led us to manipulate when subjects gave their Aad evaluations. If subjects asked to form Aad before making a brand choice make similar choices to unprompted subjects, this would suggest that the latter also formed Aad during choice processing. Thus, the secondary research issue examined here focuses on whether or not Aad is formed prior to brand choice.

In sum, this study examines (a) how Aad affects brand choice, and whether or not its effect is direct or mediated by AB, and (b) when Aad is formed during the choice process. We do this in a lab choice experiment. Consistent with much brand choice research, we operationalize choice as the selection of a best brand from a set of alternatives described in ads (Bettman 1979). To affect Aad we vary the ad picture for one brand to be good or bad. While researchers generally agree that ads are processed holistically, several studies have successfully used such manipulations (Chattopadhyay and Nedungadi 1990; Mitchell 1986; Tsal 1985; Mitchell and Olson 1981). We also manipulate when subjects give their Aad ratings, i.e., before/after choice. Preexisting product knowledge and brand and ad attitudes are controlled by using hypothetical brands.

Conceptual Issues

Attitude Toward the Ad

Attitude toward the ad (Aad) has been defined as a "predisposition to respond in a favorable or unfavorable manner to a particular advertising stimulus during a particular exposure situation" (MacKenzie, Lutz and Belch 1986, p. 130). Aad may contain both affective reactions, e.g., ad-created feelings of happiness, and evaluations, e.g., of an ad's credibility or informativeness (Baker and Lutz 1988).

This definition does not specify the components of the ad stimulus upon which Aad is based. The contents of the ad copy (brand attribute information), the headline, the creative platform (use of humor and other appeals to support delivery of the message), and ad images or pictures presumably all contribute to forming Aad (e.g., Edell and Staelin 1983; Baker and Lutz 1988). Pictures may communicate much information about the advertised brand (Dickson et al. 1986). Presumably the consumer will consider this information when forming an overall brand attitude (Tsal 1985). Mitchell (1986) has shown that Aad is based on the entire contents of the ad and not just the picture.

Attitude Toward the Ad and Brand Choice

The primary research issue in this study centers on how Aad affects brand choice. Existing research, while not directly examining this issue, provides some useful insights. It has identified two ways that Aad may affect consumers' information processing. The first is the indirect effects model, where Aad has an impact on AB through affect transfer, and AB affects intentions. Thus AB, which includes beliefs formed from ad brand attribute information and inferences based on ad picture content (Gardner 1985; Mitchell and Olson 1981), mediates the impact of Aad on intentions, i.e., there is no direct Aad-intention link. This seems to be the more popular view (Shimp 1981; Mitchell 1986; MacKenzie and Lutz 1989; Machleit and Wilson 1988; Mitchell and Olson 1981). One version of this model is the reciprocal mediation model in which a bidirectional relationship is proposed between Aad and AB, and the latter then determines intention. Alternative models propose a single, direct path from Aad to AB (MacKenzie, Lutz and Belch 1986).

The second view proposes that both Aad and AB have direct, separate influences on intentions. MacKenzie, Lutz and Belch (1986) call this the independent influences model. Gresham and Shimp (1985) also propose this model, but they add a bidirectional relationship between Aad and AB, as in the reciprocal mediation model. Thus, in their model Aad has a direct influence on intentions and an indirect effect through AB. AB operates in an analogous manner.

So far research supports the mediated effects of Aad on intentions. For example, in a comparison of four models, one of which included an independent influences (direct effects) model, MacKenzie, Lutz and Belch (1986) found strongest support for the dual mediating hypothesis (indirect effects) model. Homer (1990) replicated this finding.

However, existing research has focused on how Aad affects AB and intentions, not brand choice. This distinction is …