Self-construal and need-for-cognition effects on brand attitudes and purchase intentions in...

Brand advertising in the United States frequently includes claims of superiority relative to competing brands or the product category (Grewal et al. 1997). Evidence indicates that such comparative advertising approaches are increasingly used in other national markets as well (Jeon and Beatty

Donthu (1998), the first study designed exclusively to examine consumer responses to comparative (CP) advertising in multiple cultures, reported no difference in ad recall across the four countries tested. CP ads were better recalled in the United States, Canada, Britain, and India. Attitudes toward CP ads, however, were more negative for consumers in Britain and India, countries in which this approach was not widely used at the time of the study. Jeon and Beatty (2002) found that comparative versus noncomparative (NC) ads elicited more favorable brand attitudes and stronger purchase intentions in both the United States and Korea.

Paralleling limited research on cross-cultural differences is the relatively small number of studies on individual differences in consumer responses to CP versus NC advertising. Although several studies analyze the moderating effects of ad-related variables such as the intensity of CP ads (Donthu 1992), source credibility (Gotlieb and Sarel 1991), and one-sided versus two-sided message structures (Belch 1981; Etgar and Goodwin 1982), few have evaluated potential moderating effects of respondent-related (individual difference) variables. One notable exception is research by Dasgupta and Donthu (1994), which examined the impact of involvement as well as familiarity with the product category and style of processing.

Given the scarcity of comparative advertising research in both areas, this study looks at ways that individual difference factors influence brand attitudes and purchase intentions in response to CP versus NC advertising in two distinct national markets. More specifically, we examine the impact of an important culture-related, individual difference factor: independent versus interdependent self-construal (Markus and Kitayama 1991). We also include the individual difference construct, need for cognition (NFC) (Cacioppo and Petty 1982), in order to identify conditions under which self-construal matters. Finally, enhancing the study's reliability and external validity, samples are drawn from Thailand and the United States. These markets are known to differ on several economic and cultural dimensions, including the extent to which individual consumers tend to be characterized by independent or interdependent self-construal (Markus and Kitayama 1991). We begin by reviewing the relevant literature to develop research hypotheses.

CULTURE-RELATED SELF-CONSTRUAL AND COMPARATIVE AD PERSUASIVENESS

Individualism--collectivism (INDCOL; Hofstede 1990) is one of the most central dimensions in international marketing research (Aaker and Maheswaran 1997). Individualism refers to a social orientation in which individuals tend to see themselves as relatively autonomous and independent, whereas collectivism involves viewing oneself as an integral part of one or more groups (Triandis 1995). This cultural construct has attracted interest from researchers in several advertising areas, including ad theme persuasiveness (Han and Shavitt 1994; Wang and Chan 2001), ad humor (Alden, Hoyer, and Lee 1993), and ad cognitive response (Shavitt, Nelson, and Yuan 1997).

Focused on the individual level rather than the overall culture, self-construal (SC) has been used extensively as an alternative indicator of INDCOL (Gardner, Gabriel, and Lee 1999; Hong et al. 2000; Lee and Kacen 1999; Zhang, Mittal, and Feick 2002). SC refers to an individual's thoughts, feelings, and actions concerning relationships to others as well as the self as distinct from others (Singelis 1994). In individualist cultures, found predominantly in the West (Aaker and Maheswaran 1997), consumers tend to have an independent self-construal (INDSC), which emphasizes the separateness and uniqueness of the individual. In collectivist cultures, located predominantly in the East, consumers tend to have an interdependent self-construal (INTSC), which emphasizes social connectedness and group relations (Singelis 1994). Thus, INDSC suggests increased separation from social context, whereas INTSC suggests enhanced connectedness.

The normative expectations of people with INDSC include: being unique, expressing oneself, promoting one's own goals, and being direct in communication. Those with high INTSC, on the other hand, are expected to be more concerned with fitting in, promoting group goals, and using indirect communication. INDSC stresses the role of others as benchmarks for social comparison and self-evaluation. INTSC focuses on the role of others for self-definition via relationships in specific contexts. Whereas self-expression is crucial for INDSC, the ability to maintain harmony with social in-groups is critical for INTSC (Markus and Kitayama 1991).

In the United States, where INDSC predominates, assertive and competitive advertising themes are commonly used to make strong claims concerning a brand (Wang and Chan 2001). This practice appears to be less frequent in collectivist cultures, however, where INTSC predominates. Content analytic studies (cf. Han and Shavitt 1994; Hong, Muderrisoglu, and Zinkhan 1987; Mueller 1987) consistently find that ads in Western cultures tend to use themes related to individualism, competitiveness, and comparisons between brands, whereas ads in Asian cultures tend to use status symbols and indirect expression.

It may be that Asian brand managers are reluctant to cause a competitive brand to lose "face" or the projected image of one's self in a relational situation (Gudykunst 1998). Losing face is a critical issue in collectivist cultures. This is the case because face is fundamentally the social-self and is associated with respect, honor, status, and trust (Ting-Toomey and Kurogi 1998). In addition, because maintaining harmony with others is emphasized, the self-expressiveness associated with direct competitive approaches may be viewed as immature or "showing off" (Markus and Kitayama 1991). Reflecting similar cultural tendencies, higher levels of INDSC have been associated with dominating, competitive tactics and a direct confrontation style in the conflict management literature. On the other hand, higher levels of INTSC have been associated with avoiding, obliging, and compromising (e.g., Oetzel 1998a, 1998b).

In sum, American ads have long been created in such a way that the sponsoring brand is differentiated from its major competitors (Taylor, Wilson, and Miracle 1994). Furthermore, competition is encouraged and frontal attacks are considered natural in advertising (Hong, Muderrisoglu, and Zinkhan 1987). Hence, it is generally assumed that comparative message structures in advertising are not unusual or distasteful. In collectivist cultures, however, cooperation is a traditional value, and public, face-to-face brand confrontations are discouraged. As a result, comparative message structures are often referred to as confrontational (Lin 1993) and are discouraged. Thus, the "traditional wisdom" recommends avoidance of direct brand comparisons in markets characterized by higher INTSC and lower INDSC.

Prior to hypothesizing relations between self-construal and comparative versus noncomparative advertising, it is necessary to address a conceptual measurement issue. There has been extensive debate over whether self-construal is one-dimensional (Hofstede 1990; Markus and Kitayama 1991) or two-dimensional (Oyserman, Coon, and Kemmelmeir 2002). If SC is two-dimensional, then high INDSC would not be considered equivalent to low INTSC. Thus, without a single continuum, the two would not be equally and inversely related to other variables (Kagitcibasi 1997). Numerous studies, especially at the individual level, argue for the orthogonality and coexistence of these two SCs (e.g., Ryder, Alden, and Paulhus 2000; Singelis 1994; Trafimow, Triandis, and Goto 1991). In addition, a comprehensive review of more than 50 studies by Oyserman, Coon, and Kemmelmeir (2002) provides evidence that the two SC constructs may not be correlated.

If INDSC and INTSC are indeed orthogonal, then it is reasonable to assume that independent self-construal will generally have more impact on the persuasiveness of CP versus NC advertising. This is likely because the INDSC construct is more directly related to the nature of comparative advertising content, which emphasizes individual brand uniqueness and direct brand competition. Competition is a major defining feature of INDSC (Oyserman, Coon, and Kemmelmeir 2002). Consumers with high INDSC value competition (Kagitcibasi 1997) and emphasize the promotion of their own individual goals (Markus and Kitayama 1991). Consumers with low INDSC, on the other hand, do not highly value competition or the use of others for comparison purposes (Markus and Kitayama 1991). As a result, individuals with high INDSC are more likely to use competitive communication and persuasion tactics than individuals with low INDSC (Oetzel 1998a).

On the other hand, the core concept of interdependent self-construal, duty to the in-group, is likely to be less relevant to the comparative advertising approach. For consumers with high INTSC, cooperation within in-groups is important, but less likely with out-groups (Triandis 1995). CP advertising may attempt to incorporate in-group/outgroup affiliation techniques in positioning the brand (Shimp 2003) relative to the competition. However, such an approach may not strike a sufficiently salient and clear in-group chord to impact attitudes and intentions for high INTSC consumers. This is because the definition of in-group versus outgroup is likely to vary across and within cultures, and in some cases, the size of the "collective" may be small and too limited for advertisers to meaningfully use in their brand positioning strategies. If not, an ambiguous brand relationship may result and INTSC will exert little effect. For example, if an auto brand emphasizes its superiority on gas mileage compared with another brand, unless this attribute is a clear marker of in-group affiliation, its relation to the high INTSC consumer's in-group may remain ambiguous, and therefore irrelevant to persuasion.

In short, should two rather than one SC dimension be found, it will be necessary to separately analyze INDSC and INTSC. Furthermore, should the two-dimension approach be used, based on the foregoing rationale, it is likely that INDSC will be more accessible in the context of CP advertising exposure and therefore have a stronger moderating impact on the two outcome measures of interest in this study: attitudes toward the brand and purchase intentions. This leads to our first hypothesis:

H1: Given a two-dimensional SC solution, INDSC will exert a stronger moderating impact than INTSC on attitudes toward the brand and purchase intentions in comparative versus noncomparative advertising.

Despite the pervasive influence of local norms on consumer behavior (Steenkamp, Batra, and Alden 2003), it is probable that certain individual differences further moderate culture-related factors under given conditions (Aaker 2000). In the context of CP advertising, NFC (Cacioppo and Petty 1982) may well be one of those constructs. We now turn to this possible moderator prior to offering our next hypothesis.

NEED FOR COGNITION AND COMPARATIVE AD PERSUASIVENESS

Need for cognition (NFC), or the tendency to engage in and enjoy effortful thinking, is an individual difference factor affecting motivation (Cacioppo and Petty 1982). High-NFC consumers more frequently base their attitudes on evaluations of product attributes, whereas low-NFC consumers more often form attitudes using simple peripheral cues such as source attractiveness, source expertise, or the number of arguments rather than the quality of the arguments (Haugtvedt, Petty, and Cacioppo 1992). For example, Zhang (1996) finds that humorous advertising is more persuasive for low-NFC consumers than high-NFC consumers because humor is used as a peripheral cue by the former group to a greater degree. In addition, high-NFC consumers are less susceptible to message-framing effects than low-NFC consumers (Zhang and Buda 1999). Furthermore, high-NFC consumers are reported to recall more information (Kassin, Reddy, and Tulloch 1990), generate more issue/task-relevant thoughts (Lassiter, Briggs, and Slaw 1991), and demonstrate more consistency between beliefs and attitudes (Verplanken 1989). High-NFC consumers' attitudes also have longer persistence (Verplaken 1991), more resistance to counterattitudinal influence (Haugtvedt and Petty 1992), and more consistency between intentions and behaviors (Cacioppo et al. 1986).

In the current study, low- (versus high-) NFC consumers are likely to be less involved with the advertising message, and as a result, INDSC should play a stronger role in shaping brand attitudes and purchase intentions. Hence, message structures that are more congruent (versus incongruent) with their self-construal should be more persuasive. Support for this possibility comes from Aaker (2000), who found that culturally congruent ad themes led to more favorable attitudes only under conditions of low involvement. She argued that increased accessibility of culturally congruent versus culturally incongruent material was the underlying causal mechanism. In a similar vein, Han and Shavitt (1994) used low-involvement products such as detergent in their study, and found that culturally congruent advertising themes were more persuasive.

These findings suggest that, under low-involvement conditions, advertising message structures (e.g., CP versus NC), which are congruent with self-construal, will be more accessible. Because NFC has been found to influence involvement level (e.g., in the Elaboration Likelihood Model; Petty and Wegener 1998), it is reasonable to extend evidence from the involvement construct to the NFC construct. In particular, self-construal (SC) and message structure are likely to interact only for low-NFC consumers. Low-NFC consumers should regard congruity between their SC and ad message structure (CP versus NC) as a peripheral cue and form their evaluations based on this match. In sum, based on the above discussion, self-construal is expected to "matter" only for low-NFC consumers. Furthermore, following HI, we expect a significant interaction only for high- versus low-INDSC consumers (see Figure 1). Thus, we hypothesize:

H2: For low-NFC consumers with high (low) INDSC. CP ads will (will not) produce more positive brand attitudes and stronger purchase intentions than NC ads.

[FIGURE 1 OMITTED]

A different pattern of results is expected for high-NFC consumers. These consumers have a stronger desire to engage in effortful information processing (Cacioppo and Petty 1982). As a consequence, they are not expected to use congruity between SC and ad message structure as a peripheral cue. Instead, they should scrutinize the message argument to come up with their evaluation. That is, SC accessibility should be overridden by a desire to solve the problem, and CP ads should be more persuasive than NC ads. Several empirical findings support this hypothesis. First, Aaker (2000) reports that the superiority of culture-congruent ad themes was not observed under conditions of high involvement. Second, CP (versus NC) ads elicited more favorable brand attitudes when the products were both cognitively involving and affectively involving (Putrevu and Lord 1994). Third, CP ads were found to be more persuasive than NC ads when the level of involvement was high (Gotlieb and Sarel 1991). Finally, Jeon and Beatty's (2002) use of personal computers as the target products probably resulted in relatively high involvement. As noted, this study reported superior performance of CP ads over NC ads in both the United States and Korea.

Based on the foregoing review and Wilkie and Farris's (1975) argument that CP ads provide consumers with more factual information that assists rational brand choice, it is hypothesized that consumers with high NFC will be more strongly persuaded by CP ads, irrespective of whether they are high or low on INDSC.

H3: For high-NFC consumers, CP ads will produce more positive brand attitudes and stronger purchase intentions than NC ads, regardless of INDSC congruity.

METHOD

Participants

Two hundred and fourteen business students from a major western U.S. university and four hundred students from a major university in northeastern Thailand volunteered as research participants in the main study. Extra credit was provided for participation.

Stimuli Development

To develop ad stimuli, three pilot tests were conducted with college students from the same universities used in the main study. Those who participated in the pilot tests were excluded from the main study. From a larger list of product category alternatives, toothpaste was selected as the focal product following focus groups (pilot test 1) in both countries, indicating that the category was very relevant to their market groups. The second pilot test with 52 U.S. and 80 Thai college students found that taste/breath freshness, cleaning ability, whitening ability, and color were salient attributes for toothpaste in both cultures.

Following work by Barone and Miniard (1999), print ads were then developed featuring a general headline, which was followed by four attribute descriptions and a concluding remark (see Appendix A). The two print ads (CP and NC) contained the same information except that CP ads also indicated that the sponsoring brand was better on each attribute than the leading brand. A third pilot test was then conducted as a manipulation check of these stimuli. At least 75 % of the participants correctly identified the ad message structure as intended (CP as CP; NC as NC) in both countries. Given this high percentage of correct identification, manipulation of ad message structure appeared to work effectively in both locations. In addition, in both countries, the ads were viewed as credible, that is, Thailand, [[bar.X].sub.NC] = 4.46, [[bar.X].sub.CP] = 4.31; United States, [[bar.X].sub.NC] = 4.79, [[bar.X].sub.CP] = 4.62, significantly greater than four on a seven-point sum scale, p < .01.

Procedure

Manipulations were provided in survey booklet format. First, participants were informed of the study descriptions. Participants then read the ad copy and subsequently completed measures of two-item brand attitudes ([A.sub.b]) and one-item purchase intention (PI) drawn from Miniard et al. (1993). Next, participants completed a three-item ad involvement measure, followed by filler tasks, including questions regarding product familiarity and product involvement. Participants then filled out two individual difference measures: Singelis's (1994) 30-item SC scale and Cacioppo, Petty, and Kao's (1984) 18-item NFC scale. Finally, participants provided demographic data. All ad stimuli and measures were in English for U.S. participants and in Thai for Thai participants. All experimental materials were drafted in English, translated into Thai, and back-translated (Brislin 1980) to maximize functional and conceptual equivalence.

RESULTS

From the United States, 211 usable questionnaires ([[bar.X].sub.age] = 23; percentage female = 51%) and from Thailand, 395 usable questionnaires ([[bar.X].sub.age] = 20; percentage female = 69%) were included in the data analysis. Three U.S. questionnaires were excluded because of insufficient English ability and five Thai questionnaires were excluded due to missing data.

As expected, a two-dimensional rather than a one-dimensional solution best represented the self-construal construct. INDSC and INTSC exhibited an orthogonal relation in the U.S. sample ([r.sub.US] = .12, p > .1) and a positive relation in the Thai data ([r.sub.Thai] = .30, p < .01). Neither the Thai nor the U.S. data revealed significantly negative correlations between the two SCs. Thus, INDSC ([[alpha].sub.US] = .70 and [[alpha].sub.Thai] = .61) and INTSC ([[alpha].sub.US] = .65 and [[alpha].sub.Thai] = .71) were treated as two separate dimensions rather than opposite poles of a single construct. As a consequence, each participant received one score for INDSC and one score for INTSC. Participants were categorized as high or low on these measures using median splits. Participants were also classified as high versus low NFC ([[alpha].sub.US] = .87 and [[alpha].sub.Thai] = .75) based on a median split.

Because the dependent measures were conceptually and empirically related ([r.sub.Ab-PI, US] = .66 and [r.sub.Ab-PI, Thai], = .55), MANOVAs (multivariate analysis of variance) were used in the overall analysis to guard against type I error (Zhang 1996). MANOVAs were then followed by ANOVAs (analysis of variance) of each of the dependent measures. Given consistency for both dependent measures (i.e., [A.sub.b] and PI), only the [A.sub.b] ANOVA results are presented.

U.S. Sample Analysis

A full MANOVA was run with [A.sub.b] and PI as the dependent measures and ad message structure, INDSC, INTSC, and NFC as the independent variables. In the initial analysis, only INDSC moderated the effects of CP versus NC on brand attitudes and purchase intentions (message structure x INDSC, p < .01). This result supported H1. That is, INDSC exerted a stronger moderating impact on consumer responses to comparative versus noncomparative advertising than did INTSC. As a result, INTSC was dropped and the model was reestimated. Support for H2 and H3 required a significant threeway interaction effect of the three independent variables. The MANOVA results revealed a significant three-way interaction effect of ad message structure x INDSC x NFC, F(2, 200) = 3.80, p < .05. Mean contrasts produced predicted results for high- but not low-NFC consumers.

Specifically, regarding [A.sub.b] for low-NFC consumers with low INDSC, CP ads ([bar.X] = 5.06), as opposed to NC ads ([bar.X] = 4.25, p < .01), were more persuasive. But for low-NFC consumers with high INDSC, CP ads ([bar.X] = 3.93), as opposed to NC ads ([bar.X] = 4.85, p < .05), were less persuasive. The results of P1 closely mirrored those of [A.sub.b]. Although the cell means were statistically different, they were in the opposite direction of what was predicted. As a consequence, H2 was not supported.

For consumers with high NFC, INDSC did not have a moderating impact. Specifically, for high-NFC consumers with low INDSC, CP ads ([bar.X] = 4.80), as opposed to NC ads ([bar.X] = 3.92, p < .05), resulted in more positive brand attitudes. For high-NFC consumers with high INDSC, CP ads ([bar.X] = 5.15), as opposed to NC ads ([bar.X] = 4.18, p < .001), generated more positive attitudes. Hence, H3 was supported. The cell means and standard deviations for [A.sub.b] and PI are displayed in Table 1.

Thai Sample Analysis

As in the United States, the initial MANOVA results revealed no significant moderating impact of INTSC. Only INDSC moderated the effects of CP versus NC on brand attitudes and purchase intentions (message structure x INDSC, p < .01). H1 was thus supported and INTSC was dropped. The reestimated model revealed a significant three-way interaction effect of ad message structure x INDSC x NFC, F(2,374) = 7.94, p < .001. Overall, the Thai results closely mirror the U.S. resuits. Specifically, regarding [A.sub.b] (r = .73) for low-NFC consumers with low INDSC, CP ads ([bar.X] = 4.41), as opposed to NC ads ([bar.X] = 3.73, p < .001), produced more positive brand attitudes and stronger purchase intentions. But for low-NFC consumers with high INDSC, CP ads ([bar.X] = 3.85), as opposed to NC ads ([bar.X] = 4.28, p < .05), were less persuasive in terms of attitudes and intentions. Although significant, the cell means were in the opposite direction of what was predicted by H2.

For consumers with high NFC, INDSC did not have an impact. Specifically, for high-NFC consumers with low INDSC, CP ads ([bar.X] = 4.35), as opposed to NC ads ([bar.X] = 3.89, p < .05), were more persuasive. For high-NFC consumers with high INDSC, CP ads ([bar.X] = 4.49), as opposed to NC ads ([bar.X] = 3.71, p < .001), were also more persuasive. As a result, H3 was supported. Figures 2 and 3 graphically display the findings for H2 and H3.

[FIGURES 2 & 3 OMITTED]

Post Hoc Mediation Analysis

Although the results failed to support H2, a systematic pattern did occur in the opposite direction of what was expected. Furthermore, the fact that this pattern was uncovered in two distinct national markets using individual-level measures of the cultural factor (INDSC) enhances the reliability and external validity of this finding. In addition, extensive pretesting coupled with substantial theory and prior research on both key constructs (INDSC and NFC) enhances internal validity, minimizing the possibility that method variance is responsible for the observed outcomes.

Assuming that the results for H2 are valid, it is possible that differential levels of situational involvement for low-NFC consumers produced the unexpected effects. That is, situational involvement may have been higher when there was incongruit), between self-construal and ad message structure (high-INDSC/NC ad and low-INDSC/CP ad) than when there was congruity (high-INDSC/CP ad and low-INDSC/NC ad). Specifically, deviations from expectations (or alternatively, ad schema preferences and expectations) of the comparative ad message structure for consumers who were low on NFC may have been sufficiently strong to produce higher levels of situational involvement and overcome accessibility effects identified by Aaker (2000).

In place of accessibility effects, increased affect from more effortful processing required by low-NFC consumers to resolve the deviation from expectations may have produced comparative advertising's stronger effects for this group (Alden, Stayman, and Hoyer 1994; Maheswaran and Chaiken 1991; Meyers-Levy and Tybout 1989). Given that this possible mechanism is consistent with both theory and the observed results, mediation analysis was conducted to explore whether situational involvement played a role in the unexpected results for H2.

Applying Baron and Kenny's (1986) mediation analysis procedure, situational ad involvement was found to mediate the relation between the SC-incongruent ad message structure and [A.sub.b] or PI for low-NFC participants in both countries. Specifically, in the United States, the squared [beta] coefficient indicates that the effect ofad message structure x INDSC on [A.sub.b] dropped by 71% when ad involvement was included in the regression equation ([beta] {before inclusion} = -3.35, p [less than or equal to] .01; [beta] {after inclusion} = -1.79, p [less than or equal to] .05). In Thailand, the findings are consistent with the U.S. results. The effect of ad message structure x INDSC on [A.sub.b] dropped by 68% when ad involvement was included in the regression equation ([beta] {before inclusion} = -2.81, p [less than or equal to] .01; [beta] {after inclusion} = -1.60, p [less than or equal to] .05). For high-NFC participants, ad message structure x INDSC did not significantly affect involvement in either the United States ([beta] = .065, t = .14, p > .1) or Thailand ([beta] = .241, t = .77, p > .1). (See Table 2.)

DISCUSSION

This study investigated the moderating effects of self-construal (SC) and need for cognition (NFC) on brand attitudes and purchase intentions for comparative (CP) versus noncomparative (NC) advertising in two distinct national markets. Overall, results from both Thailand and the United States demonstrate that SC plays an important role in the effectiveness of comparative advertising. However, the role of SC was further moderated by NFC. Specifically, low-NFC consumers with low independent SC (INDSC) had more positive brand attitudes and stronger purchase intentions following exposure to comparative advertising in both countries. For low-NFC consumers with high independent SC, the noncomparative message structure was more persuasive. For high-NFC consumers, however, regardless of self-construal congruency, comparative ads were viewed as more persuasive than noncomparative ads. Investigating the unexpected findings for low-NFC consumers, mediation analyses indicated that higher situational involvement, which probably increased as a result of the incongruency between INDSC and ad message structure, enhanced brand attitudes and purchase intentions.

Limitations and Future Research

Before discussing research and managerial implications, it is important to note study limitations. First, in addition to brand attitude and purchase intentions, future studies should include upstream consumer responses to alternative message structures, such as recall and ad attitude. In addition, because the name of the competing brand was not explicitly mentioned, the intensity of the comparative ad treatment was relatively low. An indirect comparison was deliberately used to enhance external validity, given current legal restrictions on brand-to-brand comparisons in Thailand. Should such laws liberalize in the future, however, competitors may start using direct comparisons. If so, it will be important to repeat this study with more "intense" direct comparisons of actual brand names. Indeed, determining whether message structure variables (e.g., comparative ad intensity) interact with respondent-related variables (e.g., self-construal and NFC) would help broaden overall theory in this area.

In addition, this study focused on print ads. Other media are worth investigating as well. Because consumers can process print media at their own pace, the extent of elaboration may be influenced by levels of consumers' NFC. For television and radio commercials, in which processing opportunity is more limited, NFC may have less impact and self-construal may exert a stronger role. In other words, type of media may moderate individual difference influences.

Furthermore, certain features of this research may limit the generalizability of the findings. Research participants were exposed to prints ads that were not embedded in a book or a magazine. In addition, the viewing environment was relatively free of the "normal" clutter. Moreover, the stimulus contained copy only with no photographs. Having said this, it should be noted that ad credibility and perception of comparative versus noncomparative message structure were validated through extensive pretests in both countries. In addition, replication of our findings in two culturally and economically distinct countries further enhances the external validity of these results. Even so, given that the use of self-made stimuli, fictitious products, and student samples may limit external validity (see, e.g., Muehling and Kangun 1985; Rogers and Williams 1989), future research should validate our findings with real ads and nonstudent samples.

Another limitation is the fact that this study employs within-culture comparisons to test hypotheses. The issue of measurement invariance (i.e., whether cross-national differences in scale means derive from true differences between cultures on the underlying construct or systematic biases in the way people from different cultures respond to certain scale items; Steenkamp and Baumgartner 1998) is bypassed with this approach. The within-culture study design has been used by other researchers, such as Aaker (2000; see also Aaker and Williams 1998), but does not lend itself to between-culture comparisons. Multicountry studies in which measure invariance is considered in the design would allow direct comparisons.

Finally, in addition to NFC, testing of other factors related to processing motivation should be explored, such as use of single versus double or multiple ad exposures (Aaker 2000) and ending ad messages with rhetorical questions rather than simple statements (Petty and Wegener 1998). Other culture-related factors, including uncertainty avoidance (Hofstede 1990) also warrant attention. CP ads, as opposed to NC ads, provide more information regarding the sponsoring brand. Hence, CP ads may be more effective in high uncertainty avoidance cultures.

Research and Managerial Implications

This study contributes to existing advertising and consumer theory in several ways. First, although Jeon and Beatty (2002) and Donthu (1998) examined cultural effects on consumer responses to CP advertising, this study examines interactions between CP versus NC and culture at the individual level. As such, it is the first to specifically address and provide evidence of the impact of independent self-construal on two important indicators of CP ad persuasiveness--brand attitude and purchase intentions. Another contribution is the identification of need for cognition for the first time as an individual difference factor that moderates the influence of self-construal. In particular, self-construal interactions with ad message structure appear to depend, at least in part, on the elaboration of ad messages resulting from longer term, intrinsic motivation to solve problems.

This study also sheds light on psychological mechanisms underlying the apparent superiority of SC-incongruent ad message structure among low-NFC consumers. Specifically, the findings suggest that SC-incongruent ad message structures evoke higher levels of involvement from low-NFC consumers, and as a result, enhance brand attitudes and purchase intentions. This research, therefore, also contributes to the incongruity literature by demonstrating that ad content may differ from culture-based expectations and among certain consumers (specifically low-NFC segments), produce counterintuitive outcomes.

Finally, the fact that these results were replicated in two culturally and economically distinct national marketplaces (Thailand and the United States) further strengthens the study's contribution. A number of researchers (e.g., Cialdini et al. 1999; Gudykunst et al. 1996) have suggested that culture-related variables measured at the individual level (e.g., SC) complement traditional, aggregated culture-level constructs (e.g., individualist versus collectivist cultures) in explaining differences in outcome measures. However, few studies in marketing have demonstrated how such effects are likely to interact with alternative advertising message strategies. This is accomplished here in a Western, developed-country market and an Asian, developing-country market. Obtaining such similar results in such different environments suggests a high level of external validity and enhances confidence in the observed relationships between ad message structure, SC, and NFC.

Turning to managerial implications, this research provides a partial replication of Jeon and Beatty's (2002) findings. In their studies, CP ads were more persuasive than NC ads in both the United States (an individualist culture) and Korea (a collectivist culture). Our results are similar. That is, CP ads enhanced brand attitudes and purchase intentions relative to NC ads in both the United States and Thailand (a collectivist culture) for certain groups of consumers. This pattern suggests that comparative ads may not be regarded by all collectivists as negatively as previously thought. Indeed, our findings indicate that current marketing practices in Asia, which limit use of CP messages, may not be optimal. In addition, the fact that such similar results were found in both countries demonstrates that standardized advertising content may well be appropriate under certain conditions (Onkvisit and Shaw 1999). Furthermore, because message elaboration was found to influence the relative impact of CP versus NC ads, managers should consider the extent to which consumers are typically involved with the ad. When consumers are more likely to process the ad (e.g., when they are more involved with the media, when they have ample opportunity, or when they have in-depth knowledge) (MacInnis, Moorman, and Jaworski 1989), CP ads should have an advantage over NC ads.

APPENDIX A

Comparative Ad Copy

Why should you choose DentiFresh instead of the leading brand the next time you shop for toothpaste?

Taste

Recent tests conducted by a marketing research firm have shown that 9 out of 10 consumers prefer the refreshing taste of DentiFresh over the taste of the leading brand. You can be more confident about your fresh breath with DentiFresh than with the leading brand.

Cleaning Ability

Independent clinical tests document DentiFresh's cleaning ability--DentiFresh removes more plaque than the leading brand.

Powerful Whiteners

Another clinical test demonstrates that DentiFresh is 20% more effective than the leading brand in removing stains from your teeth.

Color

DentiFresh is available in a greater variety of colors than the leading brand, including natural white, green, blue, and pink.

DentiFresh: Toothpaste for Today's Consumers!

Noncomparative Ad Copy

Why should you choose DentiFresh the next time you shop for toothpaste?

Taste

Recent tests conducted by a marketing research firm found that 9 out of 10 consumers like the refreshing taste of DentiFresh. You can be confident about your fresh breath after using DentiFresh.

Cleaning Ability

Independent clinical tests document DentiFresh's cleaning ability--DentiFresh removes plaque!

Powerful Whiteners

Another clinical test demonstrates that DentiFresh is effective in removing stains from your teeth.

Color

DentiFresh is available in a variety of colors, including natural, white, green, blue, and pink.

DentiFresh: Toothpaste for Today's Consumers!

TABLE 1
Dependent Measures Across Conditions (Mean and SD)

United States

                                       Low
                                       NFC

                             Low                 High
                            INDSC                INDSC

Dependent measures    NC ad      CP ad     NC ad      CP ad

[A.sub.b]              4.25       5.06      4.85       3.93
                      (1.00)      (.96)    (1.31)     (1.36)
PI                     2.96       4.24      4.13       3.00
                      (1.40)     (1.42)    (1.62)     (1.92)

                                       High
                                       NFC

                             Low                 High
                            INDSC                INDSC

Dependent measures    NC ad      CP ad     NC ad      CP ad

[A.sub.b]              3.92       4.80      4.18       5.15
                      (1.34)     (1.60)    (1.32)     (1.09)
PI                     2.62       4.45      3.07       4.48
                      (1.17)     (1.96)    (1.63)     (1.77)

Thailand

                                       Low
                                       NFC

                             Low                 High
                            INDSC                INDSC

Dependent measures    NC ad      CP ad     NC ad      CP ad

[A.sub.b]              3.73       4.41      4.27       3.85
                       (.86)      (.79)     (.82)      (.91)
PI                     3.00       4.23      4.20       3.53
                      (1.25)     (1.11)    (1.25)     (1.56)

                                       High
                                       NFC

                             Low                 High
                            INDSC                INDSC

Dependent measures    NC ad      CP ad     NC ad      CP ad

[A.sub.b]              3.89       4.35      3.70       4.49
                       (.87)      (.74)    (1.15)      (.97)
PI                     3.13       3.98      3.53       4.16
                      (1.21)     (1.40)    (1.42      (1.46)

TABLE 2
Mediation Analysis (Standardized [beta] Coefficients)

United States

                                  2. Persuasiveness   3. Persuasiveness
               1. Involvement          measure             measure

[A.sub.b]
AMS (b) x      [beta] = -2.77,     [beta] = -3.35,
  INDSC             t = -3.40,          t = -4.19,
                    p <   .01           p <   .01
Involvement                                             [beta] = 60,
                                                             t =  7.42,
                                                             p <   .01
[R.sup.2]                 .105                .156                 .355
PI
AMS x          [beta] = -2.77,     [beta] = -3.03,
  INDSC             t = -3.40,          t = -3.77,
                    p <   .01           p <   .01
Involvement                                             [beta] =   .53,
                                                             t =  6.33,
                                                             p <   .01
[R.sup.2]                 .105                .127                 .284

Thailand

[A.sub.b]
AMS x          [beta] = -2.65,     [beta] =  2.81,
  INDSC             t = -4.69,          t = -4.97,
                    p <   .01           p <   .01
Involvement                                             [beta] =   .52,
                                                             t =  8.84,
                                                             p <   .01
[R.sup.2]                 .104                .121                 .273
PI
AMS x          [beta] = -2.65,     [beta] = -2.90,
  INDSC             t = -4.69,          t = -5.25,
                    p <   .01           p <   .01
Involvement                                             [beta] =   .57,
                                                             t =  9.90,
                                                             p <   .01
[R.sup.2]                 .104                .142                 .319

United States

                                        % of
              4. Persuasiveness    [[beta].sup.2]
                   measure            drop (a)

[A.sub.b]
AMS (b) x      [beta] = -1.79,
  INDSC             t = -2.49,
                    p <   .05
Involvement    [beta] =   .53,           71%
                    t =  6.30,
                    p <   .01
[R.sup.2]                 .401
PI
AMS x          [beta] = -1.74,
  INDSC             t = -2.30,
                    p <   .05
Involvement    [beta] =   .48,           67%
                    t =  5.31,
                    p <   .01
[R.sup.2]                 .322

Thailand

[A.sub.b]
AMS x          [beta] = -1.60,
  INDSC             t = -3.04,
                    p <   .01
Involvement    [beta] =   .47,           68%
                    t =  7.60,
                    p <   .01
[R.sup.2]                 .315
PI
AMS x          [beta] = -1.59,
  INDSC             t = -3.17,
                    p <   .01
Involvement    [beta] =   .50,           70%
                    t =  8.55,
                    p <   .01
[R.sup.2]                 .371

(a) % of [[beta].sup.2] drop: percentage of drop in squared [beta]
coefficient of ad format X INDSC after ad involvement is included
in the regression equation.

(b) AMS = ad message structure.

REFERENCES

Aaker, Jennifer L. (2000), "Accessibility or Diagnosticity? Disentangling the Influence of Culture on Persuasion Process and Attitude," Journal of Consumer Research, 26 (March), 340-357.

--, and Durairaj Maheswaran (1997), "The Effect of Cultural Orientation on Persuasion," Journal of Consumer Research, 24 (December), 315-328.

--, and Patti Williams (1998), "Empathy Versus Pride: The Influence of Emotional Appeals Across Cultures," Journal of Consumer Research, 25 (December), 241-261.

Alden, Dana L., Wayne D. Hoyer, and Chol Lee (1993), "Identifying Global and Culture-Specific Dimensions of Humor," Journal of Marketing, 57 (April), 64-75.

--, Douglas M. Stayman, and Wayne D. Hoyer (1994), "Evaluation Strategies of American and Thai Consumers," Psychology and Marketing, 11 (2), 145-162.

Baron, Reuben, and David Kenny (1986), "The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations," Journal of Personality and Social Psychology, 51 (6), 1173-1182.

Barone, Michael J., and Paul W. Miniard (1999), "How and When Factual Ad Claims Mislead Consumers: Examining the Deceptive Consequences of Copy X Copy Interactions for Partial Comparative Advertisements," Journal of Marketing Research, 36 (February), 58-74.

Belch, George E. (1981), "An Examination of Comparative and Noncomparative Television Commercials: The Effects of Claim Variation and Repetition on Cognitive Response and Message Acceptance," Journal of Marketing Research, 18 (August), 333-349.

Brislin, Richard W. (1980), "Translation and Content Analysis of Oral and Written Material," in Handbook of Cross-Cultural Psychology, vol. 2, Harry C. Triandis and John W. Berry, eds., Boston: Ally and Bacon, 389-444.

Cacioppo, John T., and Richard E. Petty (1982), "The Need for Cognition," Journal of Personality and Social Psychology, 42 (1), 116-131.

--, --, and Chuan Feng Kao (1984), "The Efficient Assessment of Need for Cognition," Journal of Personality Assessment, 48 (3), 306-307.

--, --, --, and Regina Rodriguez (1986), "Central and Peripheral Routes to Persuasion: An Individual Difference Perspective," Journal of Personality and Social Psychology, 51 (5), 1032-1043.

Cialdini, Robert B., Wilhelmina Wosinka, Daniel W. Barrett, Jonathan Burner, and Malgorzata Gurnik-Durose (1999), "Compliance with a Request in Two Cultures: The Differential Influence of Social Proof and Commitment/Consistency on Collectivists and Individualists," Personality and Social Psychology Bulletin, 25 (10), 1242-1253.

Dasgupta, Chanda Ghose, and Naveen Donthu (1994), "The Influence of Individual Difference on the Effectiveness of Comparative Advertising," Journal of Marketing--Theory and Practice, 2 (1), 42-53.

Donthu, Naveen (1992), "Comparative Advertising Intensity," Journal of Advertising Research, 32 (November/December), 54-61.

-- (1998), "A Cross-Country Investigation of Recall and Attitude Towards Comparative Advertising," Journal of Advertising, 27 (2), 111-122.

Etgar, Michael, and Stephen A. Goodwin (1982), "One-Sided Versus Two-Sided Comparative Message Appeals for New Brand Introductions," Journal of Consumer Research, 8 (March), 460-465.

Gardner, Wendi L., Shira Gabriel, and Angela Y. Lee (1999), "I Value Freedom, but We Value Relationships: Self-Construal Priming Mirrors Cultural Differences in Judgment," Psychological Science, 10 (4), 321-326.

Gotlieb, Jerry B., and Dan Sarel (1991), "Comparative Advertising Effectiveness: The Role of Involvement and Source Credibility," Journal of Advertising, 20 (March), 38-45.

Grewal, Dhruv, Sukumar Kavanoor, Edward F. Fern, Carolyn Costley, and James Barnes (1997), "Comparative Versus Noncomparative Advertising: A Meta-Analysis," Journal of Marketing, 61 (October), 1-15.

Gudykunst, William B. (1998), "Individualistic and Collectivistic Perspectives on Communication: An Introduction," International Journal of Intercultural Relations, 22 (2), 107-134.

--, Yuko Matsumoto, Stella Ting-Toomey, Tsukasa Nishida, Kwangsu Kim, and Sam Heyman (1996), "The Influence of Cultural Individualism-Collectivism, Self-Construals, and Individual Values on Communication Styles Across Cultures," Human Communication Research, 22 (4), 510-543.

Han, Sang-Pil, and Sharon Shavitt (1994), "Persuasion and Culture: Advertising Appeals in Individualistic and Collectivistic Societies," Journal of Experimental Social Psychology, 30 (4), 326-350.

Haugtvedt, Curtis P., and Richard E. Petty (1992), "Personality and Persuasion: Need for Cognition Moderates the Persistence and Resistance of Attitude Changes," Journal of Personality and Social Psychology, 63 (2), 308-319.

--, --, and John T. Cacioppo (1992), "Need for Cognition and Advertising: Understanding the Role of Personality Variables in Consumer Behavior," Journal of Consumer Psychology, 1 (3), 239-260.

Hofstede, Geert (1990), Culture and Organization: Software of the Mind, London: McGraw-Hill.

Hong, Jae W., Aydin Muderrisoglu, and George M. Zinkhan (1987), "Cultural Differences and Advertising Expression: A Comparative Content Analysis of Japanese and U.S. Magazine Advertising," Journal of Advertising, 16 (1), 55-63.

Hong, Ying-yi, Michael W. Morris, Chi-Yue Chiu, and Veronica Benet-Martinez (2000), "Multicultural Minds: A Dynamic Constructivist Approach to Culture and Cognition," American Psychologist. 55 (7), 709-720.

Jeon, Jung Ok, and Sharon E. Beatty (2002), "Comparative Advertising Effectiveness in Different National Cultures," Journal of Business Research, 55 (November), 907-913.

Kagitcibasi, Cigdem (1997), "Individualism and Collectivism," in Handbook of Cross-Cultural Psychology, vol. 3, Social Behavior and Applications, John W. Berry, Marshall H. Segall, and Cigdem Kagitcibasi, eds., Needham Heights, MA: Allyn and Bacon, 1-50.

Kassin, Saul M., M. E. Reddy, and W. F. Tulloch (1990), "Juror Interpretations of Ambiguous Evidence: The Need for Cognition, Presentation Order, and Persuasion," Law and Human Behavior, 14 (1), 43-55.

Lassiter, G. Daniel, Michael A. Briggs, and David R. Slaw (1991), "Need for Cognition: Causal Processing and Memory for Behavior," Personality and Social Psychology Bulletin, 17 (6), 694-700.

Lee, Julie A., and Jacqueline J. Kacen (1999), "The Relationship Between Independent and Interdependent Self-Concepts and Reasons for Purchase," Journal of Euromarketing, 8 (1/2), 83-99.

Lin, Carolyn A. (1993), "Cultural Differences in Message Strategies: A Comparison Between American and Japanese TV Commercials," Journal of Advertising Research, 33 (July/August), 40-48.

MacInnis, Deborah J., Christine Moorman, and Bernard J. Jaworski (1989), "Enhancing and Measuring Consumers' Motivation, Opportunity and Ability to Process Brand Information from Ads," Journal of Marketing, 55 (October), 32-53.

Maheswaran, Durairaj, and Shelly Chaiken (1991), "Promoting Systematic Processing in Low-Motivation Settings: Effect of Incongruent Information on Processing and Judgment," Journal of Personality and Social Psychology, 61 (1), 13-25.

Markus, Hazel Rose, and Shinobu Kitayama (1991), "Culture and the Self: Implications for Cognition, Emotion, and Motivation," Psychological Review, 98 (2), 224-253.

Meyers-Levy, Joan, and Alice M. Tybout (1989), "Schema Congruity as a Basis for Product Evaluation," Journal of Consumer Research, 16 (June), 39-54.

Miniard, Paul W., Randall L. Rose, Michael J. Barone, and Kenneth C. Manning (1993), "On the Need for Relative Measures When Assessing Comparative Advertising Effects," Journal of Advertising, 22 (September), 41-57.

Muehling, Darrel D., and Norman Kangun (1985), "The Multi-Dimensionality of Comparative Advertising: Implications for the Federal Trade Commission," Journal of Public Policy and Marketing, 4 (1), 112-128.

Mueller, Barbara (1987), "Reflection of Culture: An Analysis of Japanese and American Advertising Appeals," Journal of Advertising Research, 27 (June/July), 51-59.

Oetzel, John G. (1998a), "Culturally Homogeneous and Heterogeneous Groups: Explaining Communication Processes Through Individualism-Collectivism and Self-Construal," International Journal of Intercultural Relations, 22 (2), 135-161.

-- (1998b), "The Effects of Self-Construals and Ethnicity on Self-Reported Conflict Styles," Communication Reports, 11 (2), 133-144.

Onkvisit, Sak, and John J. Shaw (1999), "Standardized International Advertising: Some Research Issues and Implications," Journal of Advertising Research, 39 (November/December), 19-24.

Oyserman, Daphna, Heather M. Coon, and Markus Kemmelmeir (2002), "Rethinking Individualism and Collectivism: Evaluation of Theoretical Assumptions and Meta-Analyses," Psychological Bulletin, 128 (1), 3-72.

Petty, Richard E., and Duane T. Wegener (1998), "Attitude Change: Multiple Roles for Persuasion Variables," in Handbook of Social Psychology, Daniel T. Gilbert, Susan Fiske, and Gardner Lindzey, eds., Boston: McGraw-Hill, 323-390.

Putrevu, Sanjay, and Kenneth R. Lord (1994), "Comparative and Noncomparative Advertising: Attitudinal Effects Under Cognitive and Affective Involvement Conditions," Journal of Advertising, 23 (2), 77-90.

Rogers, John C., and Terrell G. Williams (1989), "Comparative Advertising Effectiveness: Practitioners' Perception Versus Academic Research Findings," Journal of Advertising Research, 29 (October/November), 22-37.

Ryder, Andrew G., Lynn E. Alden, and Delroy L. Paulhus (2000), "Is Acculturation Unidimensional or Bidimensional? A Head-to-Head Comparison in the Prediction of Personality, Self-Identity, and Adjustment," Journal of Personality and Social Psychology, 79 (1), 49-65.

Shavitt, Sharon, Michelle R. Nelson, and Rose Mei Len Yuan (1997), "Exploring Cross-Cultural Differences in Cognitive Responding to Ads," in Advances in Consumer Research, vol. 24, Merrie Brucks and Debbie MacInnis, eds., Provo, UT: Association for Consumer Research, 245-250.

Shimp, Terence A. (2003), Advertising, Promotion and Supplemental Aspects of Integrated Marketing Communications, Mason, OH: South-Western.

Singelis, Theodore M. (1994), "The Measurement of Independent and Interdependent Self-Construals," Personality and Social Psychology Bulletin, 20 (October), 580-591.

Steenkamp, Jan-Benedict E. M., Rajeev Batra, and Dana L. Alden (2003), "How Perceived Brand Globalness Creates Brand Value," Journal of International Business Studies, 33 (January), 35-47.

--, and Hans Baumgartner (1998), "Assessing Measurement Invariance in Cross-National Consumer Research," Journal of Consumer Research, 25 (June), 78-90.

Taylor, Charles R., R. Dale Wilson, and Gordon E. Miracle (1994), "The Impact of Brand Differentiating Messages on the Effectiveness of Korean Advertising," Journal of International Marketing, 2 (4), 31-52.

Ting-Toomey, Stella, and Atsuko Kurogi (1998), "Facework Competence in Intercultural Conflict: An Updated Face-Negotiation Theory," International Journal of Intercultural Relations, 22 (2), 187-225.

Trafimow, David, Harry C. Triandis, and Sharon G. Goto (1991), "Some Tests of the Distinction Between the Private Self and the Collective Self," Journal of Personality and Social Psychology, 60 (5), 649-655.

Triandis, Harry C. (1995), Individualism and Collectivism, Boulder, CO: Westview Press.

Verplanken, Bas (1989), "Involvement and Need for Cognition as Moderators of Beliefs-Attitude-Intention Consistency," British Journal of Social Psychology, 28 (2), 115-122.

-- (1991), "Persuasive Communication of Risk Information: A Test of Cue Versus Message Processing Effects in a Field Experiment," Personality and Social Psychology Bulletin, 17 (2), 188-193.

Wang, Cheng Lu, and Allan K. K. Chan (2001), "A Content Analysis of Connectedness Vs. Separateness Themes Used in US and PRC Print Advertisings," International Marketing Review, 18 (2), 145-160.

Wilkie, William L., and Paul Farris (1975), "Comparison Advertising: Problems and Potential," Journal of Marketing, 36 (October), 7-15.

Zhang, Yinlong, Vikas Mittal, and Lawrence Feick (2002), "To Choose or to Reject: It Is Up to Who I Am: The Impact of Self-Construal on Decision Strategy," in Asia Pacific Advances in Consumer Research, vol. 5, Rami Zwick and Tu Ping, eds., Provo, UT: Association for Consumer Research, 212-216.

Zhang, Yong (1996), "Responses to Humorous Advertising: The Moderating Effect of Need for Cognition," Journal of Advertising, 25 (1), 15-32.

--, and Richard Buda (1999), "Moderating Effects of Need for Cognition on Responses to Positively Versus Negatively Framed Advertising Messages," Journal of Advertising, 28 (2), 1-15.

Kawpong Polyorat (Ph.D., University of Hawaii at Manoa) is a lecturer of marketing in the Faculty of Management Science, Khonkaen University.

Dana L. Alden (Ph.D., University of Texas at Austin) is a professor of marketing, College of Business Administration, University of Hawaii at Manna.

Partial funding for this paper was provided by the Department of Marketing, College of Business Administration at the University of Hawaii at Manoa and the Department of Business Administration, Faculty of Management Science at Khonkaen University.