Advances in Consumer Research Volume 17, 1990 Pages 755-660
THE EFFECTS OF SITUATIONAL AND INTRINSIC SOURCES OF PERSONAL RELEVANCE ON BRAND CHOICE DECISIONS
Simeon Chow, University of South Carolina
Richard L. Celsi, California State University
Robin Abel, University of South Carolina
In the present study, the effects of subjects' intrinsic sources of personal relevance (ISPR) and situational sources of personal relevance (SSPR) with using fragrances (perfumes and/or colognes) on their brand choice intentions were examined through the use of multivariate profile analysis of repeated measures. ISPR and SSPR, individually as well as together, were found to have significant influences on individual brand choice intentions. These results emphasize the importance of examining both variables concurrently in consumer research.
Consumer researchers have reached a general consensus that views the involvement construct as the degree of personal relevance an individual perceives in a concept such as a product or product related behavior (Petty, Cacioppo, and Schumann 1983; Zaichkowsky 1985). Personal relevance is defined as "the perceived linkage between an individual's needs, goals, and values and his/her product knowledge (i.e., product attributes and benefits)" (Celsi and Olson 1988, p. 211). Key to this perspective is that the overall amount of personal relevance an individual experiences or feels at any particular point in time is determined by situational and intra-personal factors (Houston and Rothschild 1978). Celsi and Olson (1988) defined these factors as situational and intrinsic sources of personal relevance.
Situational sources of personal relevance (SSPR) are specific cues or contingencies in a consumer's immediate environment that are perceived as personally relevant in the context of that particular situation. For example, price promotions, the presence of significant others, or the perceived importance of an event create contingencies in consumers' immediate environments that may activate personally relevant goals or values such as "be thrifty" or "look good" (Celsi and Olson 1988). SSPR are transitory and decline in relevance when the personal goal is achieved. In contrast, intrinsic sources of personal relevance (ISPR) are "relatively stable, enduring structures of personally relevant knowledge which are stored in memory and concern objects or behaviors that are intrinsically relevant to the individual. For instance, wine connoisseurs and computer buffs tend to perceive consumption activities associated with these products as personally relevant across situations" (Celsi and Olson 1988, p. 212).
To our knowledge only two studies have specifically considered the combined effects of situational and intrinsic factors on consumer behavior. First, Richins and Bloch (1986) examined the effects of situational and enduring involvement (i.e., SSPR and ISPR, respectively) on consumer information acquisition behavior. Then, Celsi and Olson (1988) demonstrated the effects of SSPR and ISPR on consumers' attention and comprehension processes. Both, Celsi and Olson (1988) and Richins and Bloch (1986) demonstrated that situational and intrinsic factors combined to affect consumers' behaviors and concluded that future consumer research should examine the effects of both factors on consumers' thought processes and behaviors.
Therefore, in the present study, our objective is to extend this stream of research by examining the effects of LSSPR and ISPR on consumers' brand choice intentions.
As discussed above, ISPR and SSPR have been demonstrated to affect consumers' attention to and search for product information as well as their elaboration of that information. However, to our knowledge, no study has examined the combined effects of SSPR and ISPR on brand choice decisions. For example, consider the product class of fragrances (perfumes and colognes) which is characterized by many competitive brands and low brand switching barriers. Do women who use fragrances have a single brand of perfume and/or cologne that they wear in most or all circumstances, or, does the perceived level of SSPR in a situation affect their choice of brands? Does the level of a woman's ISPR with fragrances also affect her choice of brands? Moreover, do SSPR and ISPR interact? For instance, do women with low levels of ISPR with fragrances tend to perceive usage situations differently than women with higher levels of ISPR?
We would expect the answers to these questions to be yes. The situation should affect brand (e.g., fragrance) choice to the extent that the situation is perceived by the individual to be personally relevant and leads to the activation of goals such as the maintenance or enhancement of one's self-image. ISPR should also affect brand choice across situational frames by heightening the perceived instrumental value of the product class to enhance, accentuate, or maintain one's self-image. Finally, we would expect to observe an interaction between ISPR and SSPR since women with different levels of ISPR with the product class of fragrances are likely to have different perceptions of the same situation. This interaction would most likely occur when individuals are confronted with situations independently judged to be of moderate or ambiguous importance. For example, in a study concerning the sending of greeting cards in various situations, Walker (1988) found that women with high levels of ISPR tended to perceive moderate situations (e.g., sending a greeting card to a non-significant other on a common occasion) as more personally relevant than did women with lower levels of ISPR.
In summary, we propose that SSPR and ISPR will demonstrate significant effects on individuals' brand choice intentions. Furthermore, we propose that SSPR and ISPR will interactively affect individuals' brand choice intentions. In the following study, we empirically test the above propositions.
A study was designed to examine the effects of SSPR and ISPR on women's brand choice behavior in the product class of fragrances. A questionnaire was given to 450 subjects to determine their level of ISPR with fragrances and brand usage intentions under different situations. A total of 200 subjects responded (response rate of 44%), of which 55 subjects were dropped due to missing data. The analysis to-be reported here is based on a sample of 145 respondents, representing a usable response rate of 32%.
The subject population was comprised of undergraduate and graduate university students from a large southern university. The subjects' ages ranged from 17 to 27.
Measuring Intrinsic Sources of Personal Relevance
In the present context, ISPR refers specifically to the self-relevant meanings stored in long-term memory that individuals associate with fragrance. Thus, measures of ISPR must capture or reflect the personal relevance individuals perceive in perfume/cologne. In this study, the subjects' ISPR with fragrance was measured with Zaichkowsky's Personal Involvement Inventory (cf. Zaichkowsky 1985). This scale was selected because it was developed to measure "a person's perceived personal relevance of the object based on inherent needs, values, and interests" (p. 342). This scale contains 20 7-point items each labeled with bipolar adjectives (e.g., important/unimportant; irrelevant/relevant; essential/unessential). Subjects were asked to think about what colognes/perfumes meant to them and then to complete the scales. By use of the Cronbach alpha, the internal consistency of this multi-item scale was found to be quite good with a reliability of .97. Moreover, the item-to-total correlations ranged from a low of .61 to a high of .87.
Situational Sources of Personal Relevance
Each subject was presented with 13 situations that varied in their levels of SSPR relative to the product class of fragrances. A previous focus group was conducted and used as the basis for generating usage situations for the questionnaire. Most situations revolved around typical activities of student life.
Ranking of the Situations. Three female college students served as a panel of judges and ranked-ordered the 13 situations on the criterion of importance of the situation for wearing a perfume/cologne. Three of the situations gave divergent rankings across the panel and were dropped from further analysis. For the remaining 10 situations, an interjudge reliability of .76 was achieved.
The situations were then divided into three groups representing differing levels of SSPR (low, moderate, and high). The following are examples of low SSPR (L-SSPR), moderate SSPR (M-SSPR), and high SSPR situations, respectively:
L-SSPR You are going to participate in a physical activity, like volleyball, at the beach.
M-SSPR You are going to the mall with a group of friends, both male and female. There are NO boyfriend/girlfriend relationships in the group.
H-SSPR You plan to study for exams with a group of female classmates at the library. A few members of the group and yourself have made plans to go to Five Points (an area with a number of college bars) after studying and w ill be leaving directly from the library.
The Dependent Measures: Shifts In Usage Intentions
In the questionnaire each respondent was asked to list her three (or fewer if she does not own at least three) most preferred brands of fragrances that she owns and the percentage of time that she uses each of them. Each of the subjects three brands were then rank-ordered according to their percent usage, where the most used brand was ranked number one, the second was ranked number two, and the third was ranked number three. This rank ordering of each subject's brands produced a baseline or "average" likelihood usage for that respondent which served as a point of comparison. In addition, each brand of fragrance listed by the subjects was classified into four categories (1 = high, 4 = low) representing the "prestige" image of the brand. Membership to each "prestige" category was determined by consensus opinion of two industry observers (the first author and an executive from an international cosmetic company) and was consistent with instrument variables such as price and exclusiveness of the distribution of each brand.
In the questionnaire, each subject was presented with each usage situation and was required to state which if any of her listed fragrances she would mostly likely use in that situation and in what order. Each response made by a subject in each situation was then compared to the subject's baseline to derive whether, in a given situation, the respondent would shift upwards or downwards with respect to fragrance prestige. A negative shift of value -1 was assigned to situations when the respondent selects a lower prestige brand over a higher prestige brand. Conversely, a positive shift of value 1 was assigned to situations when the respondent selects a higher prestige brand over a lower image brand. If there is no change in response with respect to the baseline, then the shift has value zero. Finally if the respondent chooses to wear no fragrance in a situation, the shift value assigned is -2. For example, consider the case where the baseline is (2,1,3), where the respondent's most often used brand has prestige value 2, her second most often used brand is of higher prestige than her most often used brand, and her third most often used brand is lower in prestige than her first brand. If her response to a situation is (3,2,1), then we would consider this a negative shift because in this situation she has stated an intention to use a lower prestige brand over her usual brand. If the response is- (1,2,3), we would consider this a positive shift because the respondent has stated that she is more likely to use a higher image brand over her usual brand of prestige value 2. Similarly, we would consider a response of (2,3,1) as a shift of -1.
For each respondent, an average shift for each set of situations is computed. That is, for every respondent j, we compute L-SSPRj, M-SSPRj, and HSSPRj, where L-SSPRj is the mean shift across situations in the L-SSPR group, M-SSPRj is the mean shift across situations in the M-SSPR group, and H-SSPRj is the mean shift across situations in the H-SSPR group, respectively.
The purpose of this research is to compare the effect of ISPR and SSPR on behavioral intentions. Subjects are given a questionnaire which they were asked to complete and return. Each subject is presented with all usage situations and asked to choose from her 3 listed fragrances, which ones she would most likely use in that situation. A research design employing multivariate profile analysis of repeated measures on different situations is chosen to control for individual differences between respondents. In particular, differences between respondents could be quite large relative to differences in ISPR and SSPR effects and, therefore, we hope to account for this within subject variations by the selection of this design. To minimize the carry-over effects across situations, the order in which the situations are presented to the respondent is counterbalanced.
METHOD OF ANALYSIS
Since each respondent has repeated measures of brand usage intentions as measured by shifts from her baseline, Winer (1971) suggested treating the trials factor (differing levels of SSPR) as a multivariate profile and conducting contrasts across the estimates for the profile effects. The repeated measures model used in this analysis is presented below, where the situation (L-SSPR, M-SSPR, HSSPR) is a fixed factor and the subjects are random:
Xij = ai + biISPRj + ei(j) where
i = situation treatment subscript (L-SSPR, M-SSPR, H-SSPR)
j = respondent subscript, j=1,...,145
Xij = shift in brand usage intentions for respondent j within situation treatment i
ISPRj = level of intrinsic sources of personal relevance for respondent j
ai = effect of situation treatment i
bi = effect of ISPR nested within situation i
ei(j) = random error for respondent j within situation group i.
The multivariate profile analysis of repeated measures computations for this model is carried out using the MGLH program included in the SYSTAT software package (Wilkinson 1984).
The research hypotheses are tested by a series of multivariate profile hypotheses. We begin by testing the ISPR main effects first, then the SSPR effects, then the ISPR-by-SSPR effects. The main effect of ISPR is tested with a univariate F statistic. For all of the tests involving situations, the test statistic is a multivariate F. Furthermore, hypotheses contrasting adjacent pairs of profiles are tested with univariate*F statistics. The homogeneity of covariance matrices hypothesis is supported by the data (Bartlett's Chi Square = 2.60 with 6 df).
The above model was estimated and provided the following results:
L-SSPR M-SSPR H-SSPR
ai -2.443 -2.182 -1.726
pi .011 .016 .013
adj R2 .133 .216 .163
The profiles corresponding to these estimates are shown in the Figure. These profiles indicate that the effect of ISPR on shifts in intentions differs low and high ISPR respondents behave in different ways depending on the relevance of the situation.
The first hypothesis we test is that ISPR would have a significant effect on brand choice intentions (i.e., bi is not zero for all i). This hypothesis is supported by the data (F1,143 = 44.34, p < .01). Referring back to the parameter estimates, we conclude that as the respondent's level of ISPR increases, she is more likely to shift towards more prestigious brands.
PROFILES OF SHIFTS IN BRAND CHOICE INTENTIONS AS A FUNCTION OF ISPR FOR THREE SSPR LEVELS
The next hypothesis we test is that the personal relevance subjects perceive in different situations (SSPR) affect their brand choice intentions. The results support this hypothesis with the multivariate SSPR effect found to be significant at the .02 level (Wilks' A = .94, F2,142 = 4.59, p < .02). From the Figure, we can see that as the perceived SSPR increases, the respondents state intentions to use more prestigious fragrances. The univariate F statistics for contrasting the adjacent profiles are F1,143 = .98, p < .33 and F1,143 = 6.41, p < .01 for the contrast of L-SSPR to M-SSPR and M-SSPR to H-SSPR, respectively. The univariate F tests indicate that this multivariate effect is substantially due to the second contrast.
The third hypothesis we test is the multivariate ISPR-by-SSPR interaction. This effect is found to be significant at the .10 level (Wilks' A = .97, F2,142 = 2.39, p < .10). These results indicate that a person's level of ISPR has a differential effect on brand choice intentions which is dependent on the perceived level of SSPR. The univariate P statistics for contrasting the adjacent profiles are F1,143 = 3. 30, p < .10 and F1,143 = 2.72, p < .11 for the contrast of L-SSPR to M-SSPR and M-SSPR to H-SSPR, respectively. Here the univariate F tests indicate that the multivariate effect is due more tn the first contrast than the second contrast.
The results of this study provide evidence that subjects' brand choice decisions are influenced by situational and intra-personal factors. First, the results demonstrated that the subjects' level of ISPR influenced their brand usage intentions. As ISPR increased, the subjects tended to use more prestigious- fragrances-across situations. Second, subjects exhibited a positive shift towards more "prestige" fragrances as the level of SSPR increased across situations. This indicates, as expected, that the personal relevance perceived in situations affects consumers' brand choice decisions. Finally, the most important and interesting finding is the significant interaction between SSPR and ISPR. We find that the subjects exhibited a similar pattern of results when stating their brand choice intentions in the low and high-SSPR situations. However, in the moderate-SSPR condition, the subjects displayed a different pattern resulting in the interaction. When the subjects were presented with situations previously judged to be moderate, those with low levels of ISPR tended to respond to the moderate situations the same as they did to the low-SSPR situations. In contrast, subjects with high levels of ISPR tended to respond to the moderate situations the same as they did to the high-SSPR situations.
Why should the relationship between brand choice intentions and intra-personal relevance change with situations? Possible behavioral explanations are that the subjects' responses in the moderate-SSPR situations occurred as a function of (a) perceptual bias, (b) conscious choice based on the variance of possible outcomes in the moderate situations, or (c) a combination of "a" and "b".
In the case of perceptual bias, the high ISPR individual's world view, for example, may be such that end-states such as "looking good" or "maintaining one's self-image" are so highly valued that the individuals tend to perceive their world differently than low ISPR individuals. Individuals with high levels of ISPR may simply perceive a greater range of situations as highly relevant to these personal goals. In contrast, individuals with lower levels of ISPR may perceive most situations as having little relevance. Future research could examine this issue by having the subjects also rate the personal relevance of each situation.
In the case of the conscious choice explanation, consider that the SSPR in all situations is really an "expectation" or average of all possible outcomes that those situations will be personally relevant. If we view SSPR in this fashion, it is reasonable to assume that most people would view high and low-SSPR situations in a generally consistent manner. For instance, "going dancing" or "doing your homework" will be consistently viewed by most people as having high and low levels of SSPR, respectively, relative to the goals of "looking good". However, in moderate-SSPR situations a greater variance of potential outcomes are possible. For instance, "going shopping with a group of friends" could become highly relevant if one of these friends unexpectedly brings along another friend of the opposite sex. Thus the probability that a moderate situation might in fact turn out to be of "high" or "low" relevance is not clear. It is here that ISPR exerts its greatest influence on brand choice selection. For instance, high ISPR subjects may be simply less willing to risk the outcome that a moderate situation may turn out to be highly relevant. Here the variance of possible outcomes leads the high ISPR subject to treat the "expected" moderate situation as if it were a high relevance situation (e.g., better to be overdressed that underdressed). This model of decision making under risk is consistent with what Kahneman and Tversky (1979) termed prospect theory.
The same rationale may apply to the low ISPR subjects who respond to the moderate-SSPR situations as if they were of low relevance. This does not seem likely, however. It is doubtful that low ISPR individuals would expend the cognitive energies required to analyze the range of possible outcomes from moderate situations. It is just not that important to them. In this case, it seems more likely that low ISPR individuals simply categorize most situations as having low relevance. Therefore, option "c" may be the most plausible, that is, high ISPR subjects may tend to analyze moderate situations in a conscious, controlled manner because of their involvement, while low ISPR individuals may perceive little difference between low and moderate-SSPR situations.
Future Research Issues
The results of this study demonstrate the importance to consumer researchers to further understand the nature of situations. Future research could focus on and attempt to identify the specific cues and contingencies in situations that individuals perceive and respond to.- Such research would lead to a more complete spectrum of situations and to the discernment of perceived risk associated with specific situations. In addition, future research should attempt to understand further the relationship between personal factors such as ISPR and situational factors. Do individuals with differing levels of ISPR categorize situations differently? When might interactions be expected? How will they manifest themselves? Under what conditions can we expect different groups of consumers to respond differently to the "same" information? Are these interactions caused by perceptual biases or are they the result of controlled elaboration of the situational information?
We have extended previous research concerning the effects of ISPR and SSPR on consumers' behaviors to include brand choice intentions. SSPR and ISPR, individually as well as together, were found to affect individuals' brand choice intentions. The interaction highlighted the subtle way- that intra-personal factors can affect the way individuals respond to and possibly perceive situational information. These results emphasize the importance of examining both variables concurrently in consumer research.
Celsi, R. L. and J. Olson (1988), 'The Role of Involvement in Attention and Comprehension Processes," Journal of Consumer Research, 15, September, 210-224.
Houston, M. J. and M. L. Rothschild (1978), "Conceptual and Methodological Perspectives in Involvement," in Researcher Frontiers in Marketing: Dialogues and Directions, ed. S. Jain, Chicago: American Marketing Association.
Kahneman, D. and A. Tversky (1979), "Prospect Theory: An Analysis of Decision under Risk," Econometrica, 47 (March), 263-291.
Petty, R. E., and J. T. Cacioppo, and D. Schumann (1983), "Central and Peripheral Routes to Advertising Effectiveness: The Moderating Role of Involvement," Journal of Consumer Research, 10 (September), 135-144.
Richins, M. L. and P. H. Bloch (1986), "After the New Wears Off: The Temporal Context of Product Involvement," Journal of Consumer Research, 13, 2, 280-285.
Walker, B. A. (1988), "Integrating Self, Motivation, and Situations: A Cognitive Structure Model of Self/Product Relationships," unpublished doctoral dissertation, College of Business Administration at The Pennsylvania State University, University Park, PA 16802.
Wilkinson, L. (1984), SYSTAT: The System for Statistics, Evanston, IL.
Winer, B. J. (1971), Statistical Principles in Experimental Design, 2nd. ed. New York, NY: McGraw-Hill.
Zaichkowsky, J. L. (1985), "Measuring the Involvement Construct," Journal of Consumer Research, 12 (December), 341-352.