Printer Friendly

Group-reference effect and the recall of consumer brands.

Mantonakis, Whittlesea, and Yoon (2008) noted that consumer decision making can be based upon three processes: stimulus based processes (i.e., consumers have all the information they need right in front of them), memory based processes, or a mixture of stimulus and memory based processes. Of these three options, Mantonakis et al. argued that strictly stimulus based decision making is unusual because consumers simply would not have all the needed information. Therefore, memory is a critical component of consumer information processing that can be expected to interact "with many other components of choice processes such as perception, information acquisition, and decision processes" (Bettman, 1986, p. 261).

There are two interconnected ways that memory can influence consumer behavior. First, researchers have examined how certain variables can affect memories of consumers (e.g., Burke & Srull, 1988; Herrmann, Walliser, & Kacha, 2011). For example, Burke and Srull investigated how the amount of advertising clutter impacted upon whether consumers could recall commercials. Second, researchers have examined how consumer memories impact upon consumer decisions, judgments, and choices (e.g., Kardes, Posavac, & Cronley, 2004; Puccinelli et al., 2009). For example, Puccinelli et al. outlined how basic memory processes (e.g., encoding, retrieval) can impact upon both how consumers search for information and how they evaluate information. Although these two ways of studying consumer memory are linked, the current study is primarily interested in exploring the former approach by examining how the group-reference effect (GRE) can impact the recall of brand information.

Group-Reference Effect

Craik and Lockhart's (1972) depth of processing theory contends that information processed at a deeper level should be remembered better than information processed in more superficial ways. Rogers, Kuiper, and Kirker (1977), building upon the depth of processing theory, found that trait adjectives processed in relation to oneself were better recalled than information processed for either its meaning (i.e., semantic content) or for some structural aspect of the stimulus word (e.g., the word was printed in big letters). This finding eventually became known as the selfreference effect (SRE); furthermore, a subsequent meta-analysis (Symons & Johnson, 1997) showed that this effect has been a consistent finding.

The SRE only looks at information processing as it relates to an individual's sense of personal self. However, Johnson et al. (2002) noted that individuals also incorporate the groups to which they belong into their self-representations. Therefore, Johnson et al. hypothesized that a group-reference effect (GRE) could occur: information processed in reference to important social groups would be better recalled than information processed in another way. These investigators found that when people processed personality traits in reference to relevant social groups (i.e., their university or their family), they recalled the personality traits better than when they processed these traits for their meaning. Johnson et al. also found that participants in the GRE condition had similar levels of recall as compared with participants who engaged in self-referential processing of personality traits. The GRE has been replicated in several other studies (Bennett, Allan, Anderson & Asker, 2010; Bennett & Sani, 2008; Stewart, Stewart, & Walden, 2007).

Johnson et al. (2002) speculated that the two explanations offered for the SRE may also explain the GRE. In the elaborative processing explanation, the argument is that people process the items on the to-beremembered list item-by-item by elaborating whether each word is descriptive of themselves (Bellezza, 1984; Rogers et al., 1977). For example, if the word was "conscientious" then people might think about instances in their lives that show them being conscientious, such as being detail oriented at work or never being late for meetings.

In the organizational view, people are not processing item-by-item but instead are making connections between items on the list (Klein & Kihlstrom, 1986). This type of processing could assist recall in several ways. First, individuals could organize the to-be-remembered list into categories of items that are self-referent versus items that are not selfreferent (Klein & Kihlstom, 1986). Additionally, in the organizational view, words may be connected with each other because these stimuli are related to some higher-order category. Therefore, the words "conscientious" and "diligent" may be coded both under the "self-referent" category and also connected to each other because they both relate to the higher order "good employee" category (Klein & Kihlstrom, 1986; Klein & Loftus, 1988).

GRE and Consumer Brands

While the SRE is firmly established in psychology, the GRE research is still in its nascent stages and therefore, further research is needed. One area that could use more attention involves the stimulus domains used in the experiments. Klein (2012) noted that the SRE has been explored in numerous stimulus domains, but the GRE has only considered personality traits (Bennett et al., 2010; Bennett & Sani, 2008; Johnson et al., 2002; Stewart et al., 2007; Yang, Liao, & Huang, 2008). Therefore, one goal of the present research was to explore whether the GRE could increase recall in another stimulus domain, namely consumer brands. One reason that consumer brands represent an interesting extension of the GRE is because its SRE predecessor has tended to be weaker when nouns (specifically concrete nouns) are used as stimulus materials (Maki & McCaul, 1985; Symons & Johnson, 1997).

Despite the difficulty in establishing the SRE with concrete nouns, we propose that a GRE will occur if the concrete nouns take the form of consumer brands. While brands are clearly concrete nouns, it is also true that marketing professionals try to imbue brands with personality characteristics (Aaker, 1997) that make it easier for consumers to relate to and identify with the brand. This aspect of the brand may enable consumers to elaborate on and to organize these brands more easily relative to other types of concrete nouns, thus making brand names operate somewhat differently than other nouns when it comes to self-referential or group referential processing.

If our reasoning is supported then results should show that the SRE can occur when concrete nouns are represented by consumer brands, thus increasing our confidence that the GRE could also occur for consumer brands. We only know of two studies that have directly examined the SRE and consumer brands (D'Ydewalle, Delhaye & Goessens, 1985; Stewart et al., 2015). While the findings of these two studies are mixed, both provide tentative evidence that the SRE occurs with consumer brands.

D'Ydewalle et al. (1985) used pictorial advertisements and had participants evaluate them from a SRE perspective, a semantic perspective, and a structural perspective. While these authors did find evidence that the SRE increased the recall of consumer brands over a structural condition on a cued recall test, on the more critical comparison they found no difference between recall in the SRE and semantic condition. However, there are two features to this study that may have made it less likely for them to find the SRE with consumer brands. First, the participants' ability to elaborate and to identify with the brand may have been lessened because the orientating questions were about the product (e.g., toothpaste) and not the brand itself (e.g., Crest). Second, they used pictorial advertisements of the brands. Research has found that pictures tend to increase recall (e.g., Stewart, Stewart, Tyson, Vinci & Fioti, 2004), so the use of pictures may have overwhelmed the differences noted between the SRE and semantic conditions.

Stewart et al. (2015, Exp. 1), which is similar to the current studies in both procedures and stimulus materials, found more direct support for the SRE occurring for consumer brands. Specifically, these investigators found that the participants were more likely to recall consumer brands in the SRE condition than either the semantic or structural conditions. Therefore, based on the limited findings that the SRE occurs with consumer brands, we predicted that the GRE also would extend to consumer brands. More formally, we hypothesize that GRE processed brands should receive the highest level of recall followed by semantically processed brands, with the structurally processed brands resulting in the lowest level of recall (Hypothesis 1).



Participants. There were 46 University of Minnesota, Morris (UMM) undergraduate participants run in this study, but only 43 (14 males, 29 females, [M.sub.age] = 19.93, SD = 1.53) were retained for use in this study. At the conclusion of the study, because of the surprise recall task, we asked if anyone would like to withdraw their data by circling a statement at the bottom of the debriefing form. One participant elected to do so and therefore, that participant was dropped from all analyses. Two other participants were eliminated because they did not complete many of the orientating questions. Participants received either extra credit in a course or a $10 Amazon gift card for participation in the study. Since we were dealing largely with brands from the United States, we asked when recruiting for the study that people participate only if they had lived in the US for at least the last two years.

Design. Consistent with other studies on the GRE (Bennett et al., 2010; Johnson et al., 2002), this was an entirely within subjects design. The design was created by asking participants to answer "yes" or "no" to three orientating questions. The GRE question asked participants "Is X a brand generally used by UMM students" (e.g., Is Crest a brand generally used by UMM students). The semantic question asked participants "Is the X brand a type of Y product" (e.g., Is the Cottonnelle brand a type of toilet paper). The structural question asked participants "Is the X brand in a larger print" (e.g., Is the brand Ragu in a larger print). For the "yes" responses, the font of the brand was twice as large as the rest of the question and for the "no" responses the brand was in the same size font as the rest of the question. The semantic and structural items were counterbalanced so that half the items would be "yes" and the other half of the items would be "no." This could not be done for the GRE items because we could not determine how many items a participant would say "yes" or "no" to prior to the study.

Stimulus Materials and Counterbalancing. We used 66 brand names from Stewart et al. (2015). The brand names used in this study came from a variety of product categories. Examples of some of the product categories used were: beverages (e.g., Minute Maid), electronics (e.g., Sony), hotels (e.g., Marriott), cars (e.g., Chevrolet), fast food restaurants (e.g., Taco Bell), toys (e.g., Nerf), and laundry supplies (e.g., Tide). The brand names also did not include any items that were obviously more likely to be used by either men or women.

The 66 brand names were divided into three brand lists of 22 items each. Stewart et al. (2015) had matched the brand lists on the following dimensions: how much people liked the brand, how much people used the brand, how easily the brand could be identified with the correct product category, and word length. The reason for creating the brand lists was so that each orientating question could be associated with a brand list that was matched on the various item dimensions. While we believe that the brand lists were well-matched on the previously described dimensions, it is possible that they still differed on some other dimension (e.g., cost of the brand). Therefore, we also included two other counterbalancing methods to ensure that any differences observed between the conditions in the study were unlikely due to any differences in the brand lists. First, the presentation order of the orientating questions was completely counterbalanced. Second, within any presentation order of the orientating questions, the three brand lists were set up in a Latin Squares design such that every brand list would show up with each orientating question. For example, in the GRE- Structural Semantic presentation order, each one of the three brand lists were assigned to each one of the orientating questions.

Procedure. The procedure was a modified version of the one used by Bennett et al. (2010) and Stewart et al. (2015). The experiment was run in small groups of 14 or fewer participants. Participants were given a packet that gave instructions about the task and asked them to fill in their sex and age. The preliminary instructions explained the nature of the orientating questions that participants would answer and that these questions could occur in any order. The participants were also instructed to work as quickly and accurately as possible and if they were uncertain of an answer that they should give their best guess. The participants then completed three pages (22 items per page) of the orientating questions. All 22 items on a page were of the same orientating task and participants were given approximately 85 seconds to complete a page before moving onto the next page. Following the orientating questions, the participants completed a distracter task in which they had 2 minutes to circle as many "4s" as they could from an array of numbers. Participants then were given a surprise free recall task in which they had 6 minutes to recall as many brand names as they could from the previous list. They were instructed that they did not have to recall the items in any particular order and that they should just write down a brand name when it came to mind.

After the recall task was finished, participants were debriefed about the experiment. As part of this debriefing, we explained our reasoning for using a surprise recall task and that, because of this surprise recall task, we asked participates to circle a statement either giving us permission to use their data or a statement asking withdrawal of their data.

Results and Discussion

Although there were 22 items each for the GRE, semantic, and structural conditions, the first and last items on each page were not analyzed in order to control for potential primacy and recency effects in recall. All analyses were done on the proportion of items recalled (i.e., item recall in an orientation question category/20). (1) Preliminary analyses for participant's sex, compensation, and presentation order of the orientating task showed neither main effects of these variables nor interactions with orientating questions at the p <.05 level, so these variables will be considered no further. (2) A one-way within measure ANOVA was conducted, which indicated an orientating question main effect, F(2, 84) = 89.39, p < .001, [[eta].sup.2.sub.p(partial)] = .68. Hypothesis 1 was supported by the finding that the GRE had the highest level of recall followed by the semantic and structural conditions (see top of Table 1). Using Bonferroni tests (p < .05), we found that all three conditions differed significantly from each other.

We noted that one logical direction to move with GRE research was to explore the boundary conditions in which the GRE might not apply. One such area was whether the GRE applied to concrete nouns because this was an area in which even the fairly robust SRE had trouble replicating its effects. While consumer brands may differ in a number of ways from traditional concrete nouns used in other research studies (Symons & Johnson, 1997), the results of Experiment 1 provide preliminary support that the GRE can be extended to nouns in the form of brand names. Second, the finding that the GRE works with consumer brands suggests a number of applied uses for the GRE, such as embedding a GRE message into an advertisement. These issues will be elaborated upon in the general discussion section.


In Experiment 1, we used a research design to study the GRE that closely paralleled a prior design used to study the SRE (Symons & Johnson, 1997). However, in Experiment 2, we used a design that closely mirrored how the GRE has been studied by other researchers in the GRE area (e.g., Bennett et al., 2010, Johnson et al., 2002; Stewart et al., 2007); this involved comparing the GRE, SRE, and semantic conditions. The typical finding is that both the SRE and GRE increase the recall of personality traits over the semantic condition, and that the SRE and GRE do not significantly differ from each other. We hypothesized that this was the expected pattern of results we would observe for the recall of brands in our second study (Hypothesis 2).

Another reason that we omitted the SRE condition in Experiment 1 and included it in Experiment 2 had to with our wondering whether the presence of the SRE condition might suppress the strength of the GRE. Across three GRE experiments, a pattern has occurred for individual recall which showed that the GRE condition has had slightly lower recall (i.e., non-significantly lower) than the SRE condition (Bennett et al., 2010, Johnson et al., 2002, Exp 1; Stewart et al., 2007). One possible explanation for why the GRE might be smaller than the SRE is provided by a meta-analytic review by Gaertner, Sedikides, Vevea, and Iuzzini (2002). These investigators concluded that people are more likely to rely on the personal self than the collective self when these identities conflict with each other. Stewart et al. (2007) suggested this reasoning might explain why they found a marginal tendency for the GRE to be smaller than the SRE for individual recall. While Stewart et al. offered this explanation, their study did not test directly for it. Therefore, by comparing the results of Experiment 1 and Experiment 2, we provide a preliminary test of the hypothesis that GRE recall is greater when the self-referent questions are absent than when self-referent questions are present (Hypothesis 3).


Participants. There were 64 UMM undergraduate participants run in this study, but only 61 (29 males, 32 females, [M.sub.age] = 19.72, SD = 1.45) were retained for use in this study. One participant was dropped because the person lived in the US less than the last two years. One other participant was eliminated because the person did not complete many of the orientating questions. One participant misunderstood the instructions on the orientating task. Participants received either course extra credit or a $10 Amazon gift card for participation in the study.

Design, Materials and Procedure. Experiment 2 was largely completed in the same manner as Experiment 1 with a few exceptions. First, we replaced the brands of Amazon and Dell with Perkins and Hershey, respectively. Amazon was replaced because we thought it might be problematic that Amazon was used as a brand when one of the ways we compensated participants was with an Amazon gift card. (3) Dell appeared on the same brand list with Apple, which we thought might make them too similar. Second, the SRE condition was added and it asked participants "Is X a brand you use" (e.g., Is Levi's a brand you use?). The structural condition was eliminated in this experiment. Third, in Experiment 1, when we recruited participants we asked that they had lived in the United States for the last two years or more. However, we did not ask them about it when they came to the experiment. In Experiment 2, we asked an additional question on the day of the experiment regarding how long they had lived in the United States.


Hypothesis 2. The same procedures used to analyze the data in the first experiment were used again for the second experiment. All analyses were completed on the proportion of items recalled (i.e., items recalled in an orientation question category/20). Preliminary analyses for participant's sex, compensation, and presentation order of the orientating task showed neither main effects of these variables or interactions with orientating questions at the p <.05 level, so these variables will be considered no further. (4) We performed a one-way within measure ANOVA and found an orientating question main effect, F(2, 120) = 7.33, p = .001, [[eta].sup.2.sub.p(partial)] = .11. Bonferroni tests (p < .05) showed that Hypothesis 2 was supported because both the GRE and the SRE conditions had greater recall of brand names than the semantic condition (see bottom of Table 1). The GRE and SRE conditions were not significantly different from each other.

Hypothesis 3. This hypothesis stated that the GRE should be larger when the SRE was absent as a comparison task than when it was present as a comparison task. In order to examine this hypothesis, we compared the GRE mean from Experiment 1 (SRE absent; M = .392) to Experiment 2 (SRE present; M =.362). While our data trended in the direction predicted by Hypothesis 3, this difference was non-significant, f(102) = 1.24, p = .218.


Hypothesis 1 and 2

Our main objective in this research was to demonstrate whether the GRE would work with the more applied stimulus domain of consumer brands. A secondary interest was to show that the GRE worked with concrete nouns. Support for Hypothesis 1 and Hypothesis 2 is consistent with these two main goals. In Experiment 1, we found that the GRE increased the recall of consumer brands compared to the semantic and structural conditions. In Experiment 2, we again found that the GRE increased recall compared to the semantic condition. Also, the GRE produced the same amount of recall as the SRE condition.

These results add a valuable contribution to the GRE literature by being the first set of studies to demonstrate that the GRE can be applied to stimulus materials other than personality traits. Moreover, given that the brand lists had been equated on several dimensions (e.g., liking, usage), and that extensive counterbalancing in our list creation procedures (i.e., presentation order, Latin Squares design) was implemented, it is likely that these findings are due to differences between the processing conditions and not due to any idiosyncratic differences in the brand lists. However, it is premature to conclude that the GRE will apply equally well to all types of concrete nouns. As suggested in the introduction, it may be that the GRE works particularly well with consumer brands because thinking about whether a brand is used by a salient group increases the likelihood of more elaborative and/or organizational processing strategies.

Hypothesis 3

Most previous GRE studies have used the SRE as a comparison condition (Bennett & Sani, 2008; Bennett et al., 2010; Johnson et al., 2002; Stewart et al., 2007). We wondered whether this actually suppressed the strength of the GRE since previous research has suggested that when the personal self and collective self are in conflict, participants are more inclined to follow the personal self (Gaertner et al., 2002). While the means observed in the current studies were in the predicted direction of Hypothesis 3 in that the GRE was larger when the SRE was absent (Experiment 1) than when it was present (Experiment 2), the difference was non-significant; thus, Hypothesis 3 was not supported.

While the results did not confirm Hypothesis 3, we think that it may be worth further study for several reasons. First, it seems likely that if the GRE is being suppressed due to the presence of the SRE that this is going to be a relatively small effect. Therefore, it may take a much larger sample to be able to observe this effect. Second, our approach to examining this effect was a bit awkward because the GRE condition was being compared to the absence or presence of the SRE across two studies instead of within a single study. A more direct approach to examining whether the SRE comparison task suppresses the GRE would be to create the appropriate conditions within a single study.


The literature review noted two interconnected ways that memory research and consumer behavior are linked. The first way involved how various factors can increase the availability of items in the memories of consumers. The second way was how the memories of consumers impacted upon consumer decisions, judgments, and choices. While the current results are more connected to the first approach, any factor that influences the availability of brand information in memory can have an impact on consumer judgments and decisions. For example, since there are usually far too many brands available for consumers to consider when making a purchase, consumers frequently only consider a subset of brands called a consideration set (Shapiro, MacInnis, & Heckler, 1997). Often these consideration sets are completely or partially memory based such as when deciding on a restaurant for dinner (Shapiro et al., 1997). Our research suggests that reminding consumers that a product is used by other group members may increase the likelihood that it becomes part of the consideration set.

Admittedly the depth of processing (DOP) paradigm (Craik & Lockhart, 1972) used in our studies does not correspond directly to the way consumers would naturally encounter brand name products. However, we feel that there are a couple of ways in which the DOP paradigm is a useful way to approach investigating the GRE and recall of consumer brands. First, we think a logical step in the progression of the GRE literature is to establish first that the GRE occurs for consumer brands using the most common research methodology (i.e., DOP paradigm) for studying it prior to trying to expand the GRE into different methodological approaches. Second, previous research has found that on any given day a consumer is exposed to a large number of advertisements about brand name products. Britt, Adams, and Miller (1972) estimated that people are exposed to between 300 and 600 ads per day. Moreover, the number of advertisements grows even larger when newer strategies for delivering brand name information are considered, such as online advertisements (Goldstein, Suri, McAfee, Ekstrand-Abueg, & Diaz, 2014) and product placements (Cain, 2011). Therefore, the DOP paradigm may represent a good way to simulate advertisement exposure since participants have to examine a relatively large number of brands in a short period of time.

Limitations and Future Research

Embedding Advertisements. While the DOP paradigm approach to studying the GRE was useful for both practical and applied reasons in our studies, future GRE research should occur in a more naturalistic context such as embedding GRE messages into advertisements. When exploring the impact of embedded GRE messages in advertisements, we think that two main dependent variables should be explored. First, we think that further attention should be paid to recall measures. For example, researchers could examine if GRE messages increase memory not only for the brand name, but also for attributes of the product. Second, researchers could also look at whether the GRE occurs on both immediate and delayed recall measures.

Elaborative vs. Organizational Processing. Future research should examine whether organizational and elaborative processing differs for the GRE versus the SRE. For organizational processing, the GRE may provide a greater number of categories in terms of dividing the list up into the "me" and "not me" as compared with the SRE. Johnson et al. (2002) found that for the GRE participants did not necessarily describe a single family member, but often listed several family members. Therefore, the organizational structure could involve multiple categories when it comes to list item organization such as "family/not family", "sibling/not sibling", and so forth. For elaborative processing, the SRE may involve a consideration of whether a person likes a brand. However, for the GRE, participants might elaborate about several different people before coming to a definitive conclusion about whether the group likes the brand.

Participants. We largely used participants from the millennial generation, so we believe that future research may wish to expand into other age groups for several reasons. First, older consumers have greater prior knowledge (Bettman, 1986) and expertise (Alba & Hutchinson, 1987), which are both variables found to impact consumer memory. While it is difficult to predict how prior knowledge and expertise might interact with the GRE, we think it is an interesting area of future study. Second, Lambert-Pandraud and Laurent (2010) have found that older consumers are less likely to change brands than younger consumers. This greater brand loyalty of older consumers suggests that the GRE might be even more robust for older consumers than younger consumers.

Given that our study was conducted at a Minnesota school, many participants were probably from an individualistic culture. It would be useful to investigate how the GRE works in more collectivistic cultures. Given the greater weight that collectivistic culture places on group membership (Triandis, 1985), we would hypothesize that such cultures might show an increased GRE compared to more individualistic ones. (5)


Aaker, J. (1997). Dimensions of brand personality. Journal of Marketing Research, 34(3), 347-356.

Alba, J. W., & Hutchinson, J. W. (1987). Dimensions of consumer expertise. Journal of Consumer Research, 13(4), 411-454. 1086/ 209080

Bellezza, F. S. (1984). The self as a mnemonic device: The role of internal cues. Journal of Personality and Social Psychology, 47(3), 506-516. doi:http:/ /

Bennett, M., Allan, S., Anderson, J., & Asker, N. (2010). On the robustness of the group reference effect. European Journal of Social Psychology, 40(2), 349-354.

Bennett, M., & Sani, F. (2008). Children's subjective identification with social groups: A group-reference effect approach. British Journal of Developmental Psychology, 26(3), 381-387. doi: 6268

Bettman, J. R. (1986). Consumer psychology. Annual Review of Psychology, 37, 257-289. doi: 10.1146/

Britt, S. H., Adams, S. C., & Miller, A. S. (1972). How many advertising exposures per day? Journal of Advertising Research, 12, 1--9.

Burke, R. R., & Srull, T. K. (1988). Competitive interference and consumer memory for advertising. Journal of Consumer Research, 75(1), 55-68. http ://

Cain, R. M. (2011). Embedded advertising on television: Disclosure, deception, and free speech rights. Journal of Public Policy & Marketing, 30(2), 226-238. doi:

Craik, F. I., & Lockhart, R. S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning & Verbal Behavior, 11(6), 671-684. (72)80001-x

D'Ydewalle, G., Delhaye, P., & Goessens, L. (1985). Structural, semantic, and self-reference processing of pictorial advertisements. Human Learning: Journal of Practical Research & Applications, 4(1), 29-38.

Gaertner, L., Sedikides, C., Vevea, J. L., & Iuzzini, J. (2002). The "I," the "we," and the "when": A meta-analysis of motivational primacy in self-definition. Journal of Personality and Social Psychology, 83(3), 574-591. doi:http:/ /

Goldstein, D. G., Suri, S., McAfee, R. P., Ekstrand-Abueg, M., & Diaz, F. (2014). The economic and cognitive costs of annoying display advertisements. Journal of Marketing Research, 51(6), 742-752. doi:

Herrmann, J., Walliser, B., & Kacha, M. (2011). Consumer consideration of sponsor brands they do not remember: Taking a wider look at the memorisation effect of sponsorship. International Journal of Advertising, 30, 259--281, doi: 10.2501/IJA-30-2-259-281.

Johnson, C., Gadon, O., Carlson, D., Southwick, S., Faith, M., & Chalfin, J. (2002). Self-reference and group membership: Evidence for a group-reference effect. European Journal of Social Psychology, 32(2), 261-274. doi:

Kardes, F. R., Posavac, S. S., & Cronley, M. L. (2004). Consumer inference: A review of processes, bases, and judgment contexts. Journal of Consumer Psychology,14(3),230-56. doi:

Klein, S. B. (2012). Self, memory, and the self-reference effect: An examination of conceptual and methodological issues. Personality and Social Psychology Review, 16(3), 283-300. doi:

Klein, S. B., & Kihlstrom, J. F. (1986). Elaboration, organization, and the selfreference effect in memory. Journal of Experimental Psychology: General, 115(1), 26-38. doi:

Klein, S. B., & Loftus, J. (1988). The nature of self-referent encoding: The contributions of elaborative and organizational processes. Journal of Personality and Social Psychology, 55(1), 5-11. /0022-3514.55.1.5

Lambert-Pandraud, R., & Laurent, G. (2010). Why do older consumers buy older brands? The role of attachment and declining innovativeness. Journal of Marketing, 74(5), 104-121.

Maki, R. H., & McCaul, K. D. (1985). The effects of self-reference versus other reference on the recall of traits and nouns. Bulletin of the Psychonomic Society, 23(3), 169-172. doi:

Mantonakis, A., Whittlesea, B. W. A., & Yoon, C. (2008). Consumer memory, fluency, and familiarity. In C. P. Haugvedt, P. M. Herr & F. R. Kardes (Eds.) Handbook of consumer psychology. (pp. 77 -102). Taylor & Francis Group/Lawrence Erlbaum Associates, New York, NY. 10.4324/9780203809570.ch3

Puccinelli, N. M., Goodstein, R. C., Grewal, D., Price, R., Raghubir, P., & Stewart, D. (2009). Customer experience management in retailing: Understanding the buying process. Journal of Retailing, 85(1), 15-30. doi:http :// .jretai.2008.11.003

Rogers, T. B., Kuiper, N. A., & Kirker, W. S. (1977). Self-reference and the encoding of personal information. Journal of Personality and Social Psychology,35(9), 677-688. doi:

Shapiro, S., MacInnis, D. J., & Heckler, S. E. (1997). The effects of incidental ad exposure on the formation of consideration sets. Journal of Consumer Research, 24(1), 94-104. doi: 10.1086/209496

Stewart, D. D., Stewart, C. B., Moret, T., Chaplin, A., Gelbmann, A., & Harder, R. (2015). Self-reference effect and consumer brands. North American Journal of Psychology, 17(1), 89 -104.

Stewart, D. D., Stewart, C. B., Tyson, C., Vinci, G., & Fioti, T. (2004). Serial position effects and the picture-superiority effect in the group recall of unshared information. Group Dynamics:Theory, Research, and Practice, 8(3), 166-181. doi:

Stewart, D. D., Stewart, C. B., & Walden, J. (2007). Self-reference effect and the group-reference effect in the recall of shared and unshared information in nominal groups and interacting groups. Group Processes & Intergroup Relations, 10(3), 323-339. doi:

Symons, C. S., & Johnson, B. T. (1997). The self-reference effect in memory: A meta-analysis. Psychological Bulletin, 121(3), 371-394. doi:http:/ /

Triandis, H. C. (1989). The self and social behavior in differing cultural contexts. Psychological Review, 96(3), 506-520. doi:

Yang, H., Liao, Q., & Huang, X. (2008). Minorities remember more: The effect of social identity salience on group-referent memory. Memory, 16(8), 910917. doi:


(1.) Two coders went through all the recall forms in Experiment 1 and all but one in Experiment 2. Most of the coding was very objective and disagreement between coders had to do with things like one coder omitting an item or putting an item in the wrong category. These items were corrected by the first author. The only subjective aspect of coding was that misspelled brands were counted as correct if they could be recognized as a correct brand name. Additionally, we accepted abbreviations that were obvious (e.g., Coke for Coke-Cola). The instances of disagreements were few and they were resolved by the first author in Experiment 1 and the second author in Experiment 2.

(2.) The Latin Squares portion of the design was not analyzed in either Experiment 1 or Experiment 2 because the number of participants in each condition would have been too small to be analyzed meaningfully.

(3.) The number of times the brand Amazon was mentioned was quite small (6 times) and unlikely to have changed the results in Experiment 1.

(4.) We were missing data on whether one participant was compensated with extra credit versus gift card, so that person was dropped from this analysis.

(5.) The authors wish to thank the anonymous reviewers for their comments that were very helpful in the revision of this article.

Note: This project was funded by a grant to Dennis D. Stewart from the Grant-inAid program, the Office of the Vice President for Research, University of Minnesota. The first experiment was a poster presentation at the 2015 meeting of the Midwestern Psychological Association.

Dennis D. Stewart, Cheryl B. Stewart, Amanda Wiener, & Alyssa C. Chaplin

University of Minnesota, Morris

Author info: Correspondence should be sent to: Dr. Dennis D. Stewart, University of Minnesota, 600 East Fourth Street, Morris, MN 56267 Email:

TABLE 1 Means and Standard Deviations for
Experiments 1 & 2

Experiment 1     GRE      SEMANTIC   STRUCTURAL

  Mean         .392 (a)   .279 (b)    .086 (c)
  SD            (.119)     (.102)      (.10)

Experiment 2     GRE      SEMANTIC      SRE
  Mean         .362 (a)   .287 (b)    .348 (a)
  SD            (.126)     (.125)      (.125)

For each experiment, the means within the same row with
different subscripts differ significantly at p Bonferroni
<.05. The standard deviations are in parentheses.
COPYRIGHT 2015 North American Journal of Psychology
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Stewart, Dennis D.; Stewart, Cheryl B.; Wiener, Amanda; Chaplin, Alyssa C.
Publication:North American Journal of Psychology
Article Type:Report
Geographic Code:1USA
Date:Dec 1, 2015
Previous Article:Individual differences in everyday helping situations: attractiveness, cost, and gender.
Next Article:The impact of religion on the development of young children's factuality judgments.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters