How Strong is the Pull of the Past? Measuring Personal Nostalgia Evoked by Advertising ALTAF MERCHANT Marketers frequently evoke personal nostalgia in their advertising. To date, scaies University of Washington, have been developed to measure the propensity to get nostalgic but not the actual Tacoma aitafnn@u.washington.edu dimensions of personal nostalgia. Results from four studies show that advertising- evoked personal nostalgia comprises four correlated but distinct dimensions: past KATHRYN LATOUR imagery, positive emotions, negative emotions, and physiological reactions. This Corneii University kai276@comeii.edu multidimensional scale showed a high level of validity and reliability. Moreover, due to careful choice of sampling frames, the study demonstrates a high level of JOHN B. FORD Old Dominion University external generalizability. Evaluating nostalgia-based advertising using the study's jbford@odu.edu multidimensional scale may provide marketers with strategic insights for developing MICHAEL S. LATOUR and fine-tuning advertising aimed at inducing nostalgia among consumers. Corneii University msi275@corneii.edu INTRODUCTION There has been an increasing use of personal nos- During times of crises—whether financial or talgia in the advertising for consumer goods and political—many marketers' main communication services such as colas, cereals, beer, insurance, and response has been to reassure consumers by mak- banking (Sullivan, 2009). In fact, personal nostalgia ing them feel safe and secure. More often than not, has been found to influence preferences for certain the technique they chose to create such comfort products and services (Loveland, Smeesters, and was through nostalgic advertising (Boyle, 2009; Mandel, 2010). A content analysis of 1,000 U.S. tele- Elliott, 2009b). As a result, many advertising execu- vision advertisements found that nostalgia was tives believe that, when it comes to boosting brand used via theme, copy, or music in 10 percent of the sales in tough times, nostalgia is the new "new" advertising (Unger, McConocha, and Faiere, 1991). (Foley, 2009). The use of nostalgia in advertising has not been Over time, marketers have developed differ- just an American phenomenon. For example, in ent types of techniques to evoke nostalgia, vary- the United Kingdom, brands such as Richmond ing from commercials that directly ask consumers Sausages, Cadbury's, and Walkers Crisps have to remember their past (i.e., Disney's "Remember evoked nostalgia through their advertising (Foley, the Magic" campaign) to vignettes lifting brand 2009). Marketers in Russia and India as well often moments from different eras (i.e., Pepsi's "Genera- use nostalgic themes in their advertising (Razdan, tion" campaign featuring Britney Spears singing 2004; Holak et al, 2007). music from different decades) to reviving old com- Though nostalgia-driven techniques have mercial jingles (i.e.. Bumble Bee reviving its 1970s evolved in their type and usage, surprisingly their "Yum, Yum, Bumble Bee, Bumble Bee Tuna" jingle measures of effectiveness have not. Although there in 2009). More generally, the use of nostalgic adver- are a number of measures for attitudes toward tising is seen as a means to reconnect the consumer advertising and identifying individual differences to the brand (Sujan, Bettman, and Baumgartner, in reaction to nostalgic messages (i.e., nostalgia 1993), with a hope to connect favorably with the proclivity), the authors believe that no measure consumer (Foley, 2009). captures the complexities of the nostalgia evoked 1 5 0 JDÜRRflL DE RDÖERTISIRG RESERRCH June 2 0 1 3 DOi: 10.2501/JAR-53-2-150-165 PERSONAL NOSTALGIA IN ADVERTISEMENTS by advertisements. In fact, recognizing the memory (Goulding, 2002). The focus of 1977). Building from that point, another practitioner need for a scale to measure the the current inquiry is centered on personal piece of research concluded that "one can complex nature of advertising evoked nos- nostalgia. remember without being nostalgic but one talgia, one study recently called for more A variety of definitions of personal nos- cannot be nostalgic without remembering" academic research in this area (Ford and talgia can be found in the literature, among (Batcho, 2007, p. 362). Merchant, 2010). them "A preference toward objects that were The scale the authors have used in the The authors' research responds to this more common when one was younger" (Hol- current paper was developed based on the call and fills this gap in the literature. brook and Schindler, 1991, p. 332). Among conceptualization that personal nostalgia Based on the standard-scale develop- the interpretations that portray nostalgia is a multidimensional experience with ment process (Churchill, 1979; DeVellis, as a positive emotion is a "positively toned cognitive and affective components. And, 2003), the authors initiated four studies evocation of a lived past" (Davis, 1979, p. 18; in this instance, they have deñned "per- to develop a measure of personal nostal- see also Batcho, 1995; Pascal, Sprott, and sonal nostalgia" as "a reflection on the past, gia evoked by advertising. This tested and Muehling, 2002; Wildschut, Sedikides, comprising a mix of memories and multiple validated scale in theory would enable Arndt, and Routledge, 2006). emotions." advertisers to consider multidimensional By contrast, several other researchers Personal nostalgia has been demon- responses to various nostalgia-based pro- have described nostalgia as a negative strated to infiuence the consumer's pref- motional stimuli. emotion, including "A wistful mood that erences for a variety of products and Further, the research demonstrated may he prompted by an object, a scene, a smell services. Extant research shows that it that nostalgia elicited by advertising was or a strain of music" (Belk, 1990, p. 670; see influences the consumer's purchase of so engaging that it influenced A^̂ , bond- also Best and Nelson, 1985; Peters, 1985). automobiles (Brown, Kozinets, and Sherry, ing with brand and brand choice. Thus, it Still other theorists have defined nostal- 2003; Braun-LaTour, LaTour, and Zinkhan, sought to reaffirm the practice of employ- gia as a basket of positive and negative 2007); foods and cosmetics (Loveland et al., ing nostalgic advertising as it nurtured emotions; one study described nostalgia 2010); perfumes (Lambert-Pandraud and brand-consumer relationships. The find- as a positive emotion with tones of loss Laurent, 2010); cigarettes and tea (Holak ings also indicated that, among less loyal (Johnson-Laird and Oatley, 1989); another et al, 2007); songs (Batcho, 2007); arts and consumers, nostalgia-based advertising is stated that nostalgia is a "wistful pleasure, entertainment (Holbrook and Schindler, likely to work better than non-nostalgia- a joy tingled with sadness" (Werman, 1977, 2003); and movies (Holbrook, 1993). based communication. Thereby, advertis- p. 393). ers may be advised to use nostalgic themes Offering a more comprehensive defini- EXISTING MEASURES to engage even a less loyal consumer tion of nostalgia, encompassing cogni- A review of the advertising literature segment. !• , tive and affective elements, one study revealed a variety of well-established The balance of the current paper offers described nostalgia as a "positively valenced measures that gauged how consumers complex feeling, emotion or mood produced by reacted to and processed advertising. • a review of the current literature; reflection on things (objects, persons, ideas) Current scales measure attitude toward • a discussion on the general methodology; associated with the past" (Holak and Hav- advertising (e.g.. Spears and Singh, 2004); • four sequential studies (and a detailed lena, 1998, p. 218). Other research shared advertising effectiveness (Moreau, Mark- discussion of related findings from this description and further proposed that man, and Lehmann, 2001); persuasiveness each); and autobiographical memories are affectively (Reichert, Heckler, and Jackson, 2001); cre- • managerial implications along with a charged and that the affect associated with ativity (Kim, Han, and Yoon, 2010); trust series of suggestions for future research. these memories can either be positive, in advertising (Soh, Reid, and Whitehill negative, or both (Baumgartner, Sujan, King, 2009); advertising-evoked pleasure CONCEPTUAL DEFINITION and Bettman, 1992). and arousal (Poels and Dewitte, 2008); Past research characterized two types of Other work distinguished nostalgia empathy during message processing (Liji- nostalgia: personal and vicarious. Personal from reminiscence by defining reminis- ang, 2010); advertising-evoked cognitive nostalgia deals with the actual "lived" cence as the act of remembering the past processing (McQuarrie and Mick, 1999); past, whereas vicarious nostalgia evokes and nostalgia as the bittersweet affect that advertising-evoked emotions and feel- a period outside of the individuals living accompanies certain memories (Werman, ings (Heath and Nairn, 2005; Aaker and June 2 0 1 3 JDURIlflL OF flDUEflTISIflG flESEflflCH1 5 1 PERSONAL NOSTALGIA IN ADVERTISEMENTS Williams, 1998); advertising-evoked mood from your past?" and provided a host of indicated that personal nostalgia evoked (Ellen and Bone, 1998); and advertising items including toys, television shows, by advertising is a four-dimensional influencing attitudes toward the brand friends, and the like. construct. The authors also generated a (Okazaki, Katsukura, and Nishiyama, There are several older scales (e.g., the list of 65 candidate items for their nos- 2007; Brown, Homer, and Inman, 1998). antiquarianism scale [McKechnie, 1977] talgia scale. None of these measures, however, actu- and the experience scale [Taylor and Kon- ally captured the complexity of the nostal- rad, 1980]) that also measured personal • Study 2: Through two rounds of data gic reaction to advertising. dispositions toward the past. These con- collection, the authors first conducted struct scales did measure the propensity exploratory analysis and then confirma- Existing Nostaigia Scaies to get nostalgic, but they did not meas- tory factory analysis using the items A 2002 study offered a 10-item single- ure the actual dimensions of the nostal- generated in Study 1. The factor analy- dimension scale to measure ad-evoked gic experience as evoked by marketing ses resulted in a final list of 34 items nostalgia (Pascal et al., 2002). This scale communications. loading on the four factors: included items such as "the ad reminds The current research seeks to fill this gap me of the past"; "makes me nostalgic"; in the literature. - past imagery, and "evokes fond memories." It should - physiological reactions, be noted that the scale did not tap into the FOUR STUDiES: GENERAL - positive emotions, and various cognitive and emotional dimen- iVIETHODOLOGY - negative emotions. sions of the nostalgic experience. In this research, the authors followed the The scale offered in the current study scale-development guidelines (Churchill, The correlated four-factor model was builds on the 2002 work and com- 1979; DeVellis, 2003) that reconunended found to be superior to various alternate prehensively deconstructs the per- that any scale development should com- models, and the subscales had accept- sonal nostalgia as evoked by marketing mence with the conceptual definition. able reliability estimates. communications. Once that definition has been determined, Some individuals show higher propen- a list of potential scale items needs to be • Study 3: The authors established sities for and proneness to nostalgia than generated through a review of the litera- criterion-related validity for the scale by others. One study defined the proclivity to ture along with qualitative research. With demonstrating that advertising-evoked "nostalgia" as "a facet of individual char- the list in place, the individual items need personal nostalgia (measured by the acter—a psychographic variable, aspect to be refined and shortlisted through fac- authors' scale) predicted higher levels of life-style, or general customer charac- tor analysis, and the underlying dimen- of A ĵ,  and behavioral intentions. teristic—that may vary among consum- sions need to be confirmed. Reliability ers" (Holbrook 1993, p. 246). That same coefficients and alternate models are also • Study 4: In the final study, the authors research offered a nostalgia-proneness examined at this stage. Finally, the valid- established nomological validity by scale, which has become widely used, that ity of the scale needs to be established: linking ad-evoked personal nostalgia Litilizes 20-items (e.g., "products are get- criterion-related validity (i.e., can the (measured by their scale) to its various ting shoddier and shoddier"; "the truly construct—measured by the scale—help antecedents and consequences. great sports heroes are long dead and predict some outcomes), and nomological gone") to measure nostalgia. Again, how- validity (i.e., does the construct—meas- STIiVIULUS DEVELOPiVIENT ever, it does not measure the cognitive ured by the scale—link to its theoretical Three print advertisements (Disney Parks, and emotional elements of the nostalgia antecedents and consequences). Public Broadcasting Services [PBS], and evoked by advertisements. In this paper, the authors have employed "Homemade" brand cookies) were devel- Two years later, yet another piece of these guidelines through four studies oped to be used as stimuli in the subse- research tested the nostalgia inventory (Table 1): quent studies. The authors deemed the but also examined nostalgia proneness product categories of the offerings desira- as a personality trait (Batcho, 1995). This • Study 1: The authors reviewed exist- ble because they often are associated in the 20-item survey asked respondents, "How ing literature and conducted qualita- literature with an outcome linked to nos- much do you miss each of the following tive research. The findings of this study talgia (Baumgartner, 1992; Batcho, 1995). 1 5 2 JDÜRReL OF ROOERTISinO RESERRCH June 2 0 1 3 PERSONAL NOSTALGIA IN ADVERTISEMENTS TABLE 1 , Summary of Studies Undertaken study Nature of Study Sample Objectives Findings/Results 1 Literature Identify items from current literature 36 items identified from extant literature review Qualitative 13 focus groups: Explore dimensions and generate list of 4-dimensional construct, 71 items generated 58 consumers items (non-student) Quantitative 5 expert judges Content validity: how well each item 65 items retained out of a pool of 107 represented its respective dimension and if there were any overlaps between the items ' - 2 Pilot: 143 Exploratory Factor Analysis: Scale 4 factors, 34 items retained Quantitative (student) refinement Main study: 200 Confirmatory Factor Analysis: Reliability and Acceptable reliabilities of sub-scales, and fit Quantitative (non-student) Dimensionaiity with 4-factor model 3 Quantitative 145 Criterion-related validity: We expected that Path co-efficients were statistically significant (non-student) higher levels of advertising-evoked nostalgia would predict higher levels of  ,̂ A ,̂ and Bl 4 Quantitative 262 Nomological validity: We proposed a series Nostalgia proneness, loneliness, and brand (non-student) of relationships between advertising-evoked loyalty influenced nostalgia evoked by personal nostalgia and its antecedents and advertising. Ad-evoked nostalgia impacted consequences perceptions of social support, enhanced brand bonds, and had a bearing on consumer choice. Nostalgic advertisements worked better among less loyal consumers as compared to non-nostalgic advertisements For each product/service, the advertise- advertisements one at a time and were evoked by advertisements. A list of 65 can- ment evoked nostalgia and used words asked whether the advertising evoked didate items measuring the four dimen- such as "relive" or "remember the past" nostalgia or did not. Feedback was sions also was generated. and cued nostalgia through a series of pic- recorded, and changes were made to the tures and graphics. This technique aligns various stimuli. Introduction with past research on advertising-evoked The revised stimuli were presented to In line with the recommendations of com- nostalgia (e.g., Braun-LaTour, LaTour, four more focus groups comprising 21 parable previous studies (Churchill, 1979; Pickrell, and Loftus, 2004). participants in all with an average age of DeVellis, 2003), a detailed review of the The advertisements were developed 44 years. nostalgia literature was undertaken, gen- using an iterative process. In the first The focus group findings indicated that erating an initial pool of 36 items. The con- round, the stimulus was presented in the advertisements for each of the prod- ceptualization was advanced at this point three focus groups. Each focus group was ucts evoked personal nostalgia (Appendi- through a qualitative study composed of composed of six or seven participants and ces A-C). 13 separate focus groups. lasted for approximately 30 to 40 minutes. The objectives of the qualitative study There were 20 respondents in all, with an STUDY 1 : ITEM GENERATION were average age of 41 years. Key Finding After a couple of introductory ques- In this study, the authors found that there • to enhance the understanding tions, the subjects were shown the were four dimensions to the nostalgia of personal nostalgia evoked by June 2 0 1 3 JDOROflL DF BDÖERTISinG RESEfiRCH 1 5 3 PERSONAL NOSTALGIA IN ADVERTISEMENTS In this study, the authors key statements to an analysis work- nostalgia research had deliberated ("Peace- sheet, and ful and warm" [Holak and Havlena, 1998]) found that there were • the assessors met to discuss and achieve along with some additional emotions that consensus on the results at the summary had not been discussed such as "relaxed"; four dimensions to level. "cahn"; "pleasant"; "connected"; "spir- itual"; and "secure." the nostaigia evoked Findings The results of the focus groups revealed Negative Emotions. The third factor dealt by advertisements. that there were four dimensions to the with negafive emodons. Once again, there personal nostalgia evoked by advertise- were some items mentioned that past ments. There were 71 manifestations of research had discovered ("sadness and nostalgia that were identified through the regret" [Baumgartner et al., 1992; Batcho, advertisements, and to identify any focus groups: 2007]). There also were 18 items that were dimensions not captured by the current novel to the nostalgia literature, among literature, and • past-imagery factor, them were such considerations as "anxiety"; • to generate an exhaustive list of items • positive emotions, "tensed"; "guilty"; "depressed"; and "grief." for the nostalgia scale. • negative emotions, and • physiological reactions. Physiological Reactions. In addition to Procedure articulating various emotions in response Each focus group consisted of four to six Past-Imagery Factor. The first set was to the series of advertisements, some consumers and lasted between 2 and 3 composed of statements related to the respondents were moved so intensely by hours. In total, participants included 33 images of the past that came to the con- the nostalgic experience that they pro- females and 25 males. To achieve a com- sumers' mind—what the authors termed duced a number of physiological reactions. prehensive perspective, respondents were "past imagery factor." Unlike any past research on nostalgia, selected from a variety of different ages, There were 18 manifestations for this the authors of the current study identi- incomes, and educational backgrounds. factor. Among these, there also were some fied a fourth factor to the personal nostal- The youngest respondent was 19; the old- manifestations that past nostalgia research gia experience that was grounded in such est was 60. The discussions were mod- had referenced (e.g., "I relived the event physiological reactions. There were 17 erated by two researchers and were from my past" [Baumgartner, 1992]; "I was such nranifestations. audio-recorded. transported to the past" [Baumgartner Examples included "My breathing To trigger nostalgia, the respondents et al., 1992]; "I could see many images" became steady/slow"; "I could taste/ were presented with three nostalgic print [Braun-LaTour et al., 2007]). smell/hear things from my past"; "I could advertisements (Disney, PBS, and "Home- The authors also found several other feel shivers/trembling"; "I had goose made" brand cookies). They then were manifestations that were new to the litera- bumps"; "I was sweating." asked to talk about the thoughts, memo- ture. Examples included "I remembered a ries, and feelings evoked by the advertise- specific event"; "It was like a flashback"; Content Validity ments. The same process was followed for "There was a montage of images"; "It was Five marketing faculty members served each advertisement, one at a time, and the a dreamlike experience"; "Images were as expert judges and rated how well each order of the advertisements was rotated like flashing pictures"; "The images were item represented its respective dimen- across the groups. impressionistic." sion and whether there were any overlaps Two assessors completed in-depth ana- between the items. All judges had earned lysis of the transcripts in two stages: Positive Emotions. The second dimen- doctoral degrees and regularly conducted sion was composed of a variety of positive behaviorally oriented research. • Each assessor conducted an inde- emotions that the advertisements evoked. A total of 107 items were generated pendent review of the transcripts, There were 18 manifestations for this through the combined process of literature highlighting the transcripts based on dimension. In this grouping, the authors review and the focus groups (literature common themes and then transferred also found a set of emotions that previous review, 36 items; focus groups, 71 items). 1 5 4 JOÜRURL OF RDÜERTISIRG RESEHRCR June 2 0 1 3 PERSONAL NOSTALGiA iN ADVERTiSEiVIENTS Only those items that were classified • item-to-total correlations above 0.50; All factor loadings were significant at as representative or highly representative • an average inter-item correlation above p < 0.001 and were above 0.50. The Cron- were retained (Zaichkowsky, 1985). In all, 0.30; and bach's alphas, average variance extracted, 65 items were retained of the full set of 107. • a factor loading above 0.50 (see Spector, and composite reliability coefficients for 1992). each of the dimensions were consistent with STUDY 2: ITEM REFINEMENT, recommended ranges (Fomell and Larcker, RELIABILITY, AND DIMENSIONALITY The factor analysis resulted in 34 items 1981; Clark and Watson, 1995; Table 2). Key Findings loading on four factors that the authors Factor analyses resulted in a final list of identified as past imagery (14 items); Dimensionality 34 items loading on the four factors (past physiological reactions (9 items); positive Several alternative measurement models imagery, physiological reactions, positive emotions (5 items); and negative emotions were examined (Anderson and Gerbing, emotions, and negative emotions). This (6 items). 1988): correlated four-factor model was found to The four factors were selected on the be superior to various alternate models, bases of scree plot and interpretability • Model 1 is the base model and corre- and the subscales had acceptable reliabil- (explained 61 percent of the variance). lated four factors; ity estimates. • Model 2 is a second-order factor model; Main Study • Model 3 is a one-factor model; and Pilot Study The main study was conducted with data • Model 4 has the four factors Data were collected from 143 undergradu- collected from 200 consumers using an uncorrelated. ate students in a large American univer- online consumer panel. sity. The respondents were made up of 36 At the time of the study, there was a In Model 5, the correlation between posi- percent male, with an average age of 26 unique opportunity to test the authors' tive emotions and negative emotions is set years. scale with a 2009 Pepsi Super Bowl tel- to 1. In Model 6, the correlation between Each respondent was exposed to one of evision commercial. Pepsi had launched positive emotions and physiological reac- the three nostalgic print advertisements a nostalgic "Refresh Anthem" commercial tions is set to 1; whereas in Model 7, the (Disney, PBS, and "Homemade" brand (http://tinyurl.com/4krbrat) that was correlation between negative emotions cookies) and responded to a list of 65 pos- based on the classic song "Forever Young". and physiological reactions is set to 1. sible reactions to those advertisements. Aimed to evoke nostalgia among consum- Last, in Model 8, the correlations between Specific instructions were "Listed below ers (Elliott, 2009a), a song in the advertise- positive emotions, negative emotions, and are statements that describe the thoughts/ ment was sung by its original lyricist Bob physiological reactions are all set to 1. feelings that come to your mind (or how Dylan and rapped by The Black Eyed Peas. As per the fit indices and difference of you feel) right now after reading the Nuances of older films were built into chi-square test, all the alternate models advertisement." the new commercial, and the advertising were significantly worse fit as compared For items related to past imagery, presented a visual collage of good times, to Model 1 (the four-factor model in the respondents were asked to indicate how celebrating generations past and present. current study; Table 3). ' . ' likely they were to agree or disagree with The commercial lasted for 60 seconds. This implies that the four-dimensional each of the statements (scale of 1 to 5, After seeing the advertising, the construct structure proposed by the 1 being "strongly disagree" and 5 being respondents completed our 34-item nos- authors is the most robust. "strongly agree"). For all the other items, talgia scale. Half the sample was com- respondents were asked to indicate to posed of men, and the mean age of the STUDY 3: CRITERION-RELATED VALIDITY what extent they felt each of the following respondents was 49 years. On the bases of Key Findings (scale of 1 to 5,1 being "very slightly or not the previous findings, a confirmatory fac- In this study, the authors established at all" and 5 being "extremely"). tor analysis was conducted with a corre- criterion-related validity—more specifi- Exploratory factor analysis was run lated four-factor model. cally, whether the scale helped predict using the 65 items, and the factors were The CFA model showed good fit some outcomes. In these instances, the rotated using varimax rotation. Statistical (f (488) ^ 1223, CFI = 0.95, IFI = 0.95, current study revealed that advertising- criteria for item retention were TLI = 0.94, GFI = 0.85, RMSEA = 0.058). evoked personal nostalgia—measured by June 2 0 1 3 JDÜRflflL DF flDÜERTISIflGR ESEBflCH 1 5 5 PERSONAL NOSTALGIA IN ADVERTISEMENTS TABLE 2 In this study, the authors Personal Nostalgia Scale Items and Factor Loadings '^^^ CPA*** established criterion- Past imagery 0.93% 0.51^ 0.93"= . . . m-^-M. r.^^}.\^^.^^.'y. ^ r related validity—more I relived the eyerit from rny past 0.86 .'..^.^.?..Í.';^.".?P.?.!'.t.^.íí..'í9..t.^.?..p.^.?.'í .° ' .?.^ specifically, whether It was like a flashback 0.85 The images were distinct 0.81 the scale helped predict It was a dreamlike experience 0.80 I rernerribered a specific event 0.77 SOme OUtCOmeS. The memories were in bits and pieces 0.74 The images were impressioriistic 0.71 .9.".?..'.'!^.?i?..'.?.'?..î.°..?'̂ °^!?.?r. 9.:^.1 the four-factor scale—predicted higher 1 could see rnany images 0.62 levels of A^̂ , which, in turn, had a positive The image/s were vivid 0.58 effect on A^and resulted in higher levels of The image/s were sharp 0.55 behavioral intentions. The image/s w^enkeflashirigpic^^^^ 0.54 Keeping in mind past research on nos- '"" ^ t • ^ n c/1 talgic advertising (Pascal et al., 2002) and There was a montage of images 0.54 & & 7 T I _ __h „ „_,. research on adverfising-generated affect Physiological reactions 0.93% 0.72% 0.95' * ^ and its impact on advertising and brand atd- was pouriding 0,95 ^ ^ ^ ^^^^^^^ ^^^ ^.^^^^ ^^^^^^ ^^^ ^^^^^^^ feel shivers/trembling 0.94 ^^p^^^^^ ^^^^ ^^^^^ ^^^^j^ ^^ adverfising- .9;.?.? evoked personal nostalgia would predict 9.:?.9. higher levels of A^̂ , which would posi- I was sweating 0.86 dvely impact Aĵ and would result in higher My stomach was churning 0.85 levels of behavioral intentions. There were tears in my eyes 0.85 I could taste/smell/hear things frorri my past 0.66 Methodology _ „_ In this study, the authors used the three I laughed/smiled 0.63 ~ nostalgic print advertisements—for Dis- Positive emotions 0.92% 0.72% 0.92'= ^ ^ , , , , _ , ney Parks, PBS, and Homemade brand .• cookie—that had been developed earlier .°;.?Z intheproject. Pleasant 0.84 Data were collected from 145 consumers Relaxed .9;.?.? using an online consumer panel. The sam- Calm 0.79 pie contained a balanced age and gender Negative emotions .9-?.3% 9.-75'. 9-94°... "^i^- ̂ ^ * respondent was exposed to one Sadness 0 91 advertisement and then completed our ' '. „ oo nostalgia scale. In addition, the respond- Anxiety 0.88 » r ents answered questions on behavioral Tensed 0.88 ^ intentions (BI; Spears and Singh, 2004), •• •' A^j—attitude toward the advertising 0.83 (MacKenzie and Lutz, 1989), and A , - 0-82 attitude toward the brand (Cox and Locan- " Cronbacb's a estimates: '' Aivrnge variance extracted: ' Composite reliability: *** all loadings significant at p < 0.001 der, 1987; Park and Young, 1986). 1 5 6 JOÜRflllL DF flOOEflTISIflG RESEflflCH June 2 0 1 3 PERSONAL NOSTALGIA IN ADVERTISEMENTS TABLE 3 Dimensionality Description t df CFi NFI IFI TLI RMSEA A xVdi Model 1 Base model—4 factors correlated 1323 488 0.95 0.90 0.95 0.94 0.058 Model 2 Second order factor 1387 490 0.93 0.88 0.94 0.92 0.064 64/2*** Model 3 1 factor model 3426 494 0.76 0.73 0.76 0.73 0.127 2103/6*" Model 4 4 factors—uncorrelated 2141 494 0.86 0.83 0.86 0.85 0.095 818/6"' Model 5 Oorrelation between positive and negative emotions set to 1 1348 489 0.92 0.89 0.92 0.91 0.069 25/1"' Model 6 Correlation between positive emotions and physiological 1357 489 0.92 0.89 0.92 0.91 0.069 34/1*" reactions set to 1 Model 7 Correlation between negative emotions and physiological 1347 489 0.92 0.89 0.92 0.91 0.069 24/1*** reactions set to 1 Model 8 Correlation between positive, negative emotions and 1358 491 0.92 0.89 0.91 0.91 0.068 35/3*" physiological reactions set to 1 *** significantly worse fit than base model (p < 0.001) Findings of Study 3 In conclusion, the results of this iVIethodoiogy The results of the structural equations study illustrated that higher levels of For this study, the authors proposed a series modeling show that the model had adverfising-evoked personal nostalgia of relationships between advertising- acceptable fit (x" (df) = 83(9); CFI = 0.91, predicted higher levels of A^̂ , which, in evoked personal nostalgia and its anteced- GFI = 0.91, IFI = 0.90, NFI = 0.90). We turn, led advertisements to higher levels ents and consequences (Figure 1). found the coefficients of the path from of A^ and BI. three dimensions of advertising evoked This demonstrates criterion validity for • Antecedents personal nostalgia to the attitude toward the authors' nostalgia scale. - Nostalgia proneness: Some indi- the advertising to be positive and statis- viduals show higher propensifies (or tically significant (ßp,̂ , ¡̂ .ĝ ^̂ .̂j = 0.65, STUDY 4: A NOMOLOGiCAL NETWORK proneness) for nostalgia than others p < 0.001; ßp,„,„,„,,,_„„^,,, = 0.45,p < 0.001; FOR ADVERTiSING EVOKED PERSONAL (Holbrook, 1993). One study indicated ßp„s.,ve™.,o„..Aad = 0-66, P< 0.001). NOSTALGiA that a person highly prone to nostalgia The authors found a negative effect, Key Findings would have a better capacity for emo- however, of the negative emotions of nos- In this study, the authors linked fionality (Batcho, 1998) and, therefore, talgia on Â^̂ (ß = -0.44, p < 0.001). This advertising-evoked personal nostalgia— would be very happy when experienc- finding is in line with the recent work of measured by their four-factor scale—to ing happiness and very sad when expe- previous studies that concluded that nega- its various antecedents and consequences. riencing sadness. The subject's capacity tive emotions evoked by advertising were They found that nostalgia proneness, lone- to feel emofions more intensely would likely to have a negative impact on the liness, and brand loyalty influenced nos- increase the likelihood of the individual consumer's attitudes toward the adver- talgia evoked by advertising. They also to experience nostalgia. tisement (Hong and Lee, 2010; Lau-Gesk found that nostalgia affected perceptions The authors, therefore, argue that and Meyers-Levy, 2009). of social support, enhanced brand bonds, the consumer's nostalgia proneness is Further, the current study found that the and had a bearing on consumer choice. likely to positively influence the per- coefficients of the path  ^ to A,̂ (ß = 0.81, Further, they learned that nostalgic adver- sonal nostalgia evoked by advertising. p < 0.001) and from Â^ to behavioral inten- tisements worked better among less loyal fions (ß = 0.76, p < 0.001) to be statistically consumers as compared to non-nostalgic - Loneliness: Loneliness is an emotional significant. advertisements. state in which a person experiences a June 2 0 1 3 JDÜRIlfiL OF flDüEBTiSillG RESEflRCH 1 5 7 PERSONAL NOSTALGIA IN ADVERTISEMENTS Antecedents Focal Construct Consequences Emotions J o.4i(7.62)* 'Standardized path estimates: figures in parentheses are tvaiues: *fvalues significant at p< 0.05 Figure 1 Advertising-Evoked Personal Nostalgia and Relationships with Other Constructs powerful feeling of emptiness and iso- The attitudinal approach argues a positive effect on the nostalgia lation (Hawthorne, 2006). One recent that loyalty exists when there is a evoked by the advertised brand. The study also found that higher levels of favorable belief toward the brand authors, therefore, offer the following loneliness resulted in higher levels of (Agustin and Singh, 2005). propositions about the antecedents of the consumer's nostalgia (Wildschut Behavioral loyalty, conversely, sees nostalgia: et al, 2006). loyalty as an expressed behavior (i.e., In line with these arguments, the the consumer's propensity to buy with PI: Higher levels of the consumer's authors propose that the consumer's reference to the pattern of past pur- nostalgia proneness will gener- levels of loneliness positively impact chases [Russell and Kamakura, 1994]). ate higher levels of advertising- the personal nostalgia evoked by Some researchers have argued that evoked personal nostalgia. advertising. for low-risk, frequently purchased P2: Higher levels of the consumer's products, loyalty may be the joint loneliness will generate higher - Brand loyalty: A consumer's loyalty outcome of habit and attitude (Zhang, levels of advertising-evoked per- to a brand has been defined in the Dixit, and Friedman, 2010). sonal nostalgia. literature using both attitudinal and The authors propose that the con- P3: Higher levels of the consumer's behavioral approaches. sumer's loyalty to a brand will have loyalty toward the focal brand 1 5 8 JDDRRRL DF RDUERTISIRG RESERRCR June 2 0 1 3 PERSONAL NOSTALGIA IN ADVERTISEMENTS will generate higher levels of by advertising is likely to enhance brand (Pepsi). Perceived social support advertising-evoked personal the consumer's bonds with the focal was assessed using one measure (Sarason, nostalgia. brand. Levine, Basham, and Sarason, 1983); bond- The authors, therefore, offer the fol- ing with the focal brand was measured Consequences lowing propositions about the conse- using another (Aaker et al., 2004; Fournier, - Perceived social support: Being quences of nostalgia: 1998). socially connected—and feeling a sense of belonging—is a basic human P4: Higher levels of advertising- Findings of Study 4 . drive (Leary, Kelly, Cottrell, and evoked personal nostalgia will The nomological network was tested Schreindorf er, 2009). Recent work of result in higher levels of per- using structural equations modeling. The Loveland et al (2010) has shown that ceived social support for the model demonstrated a reasonably good consumers who have a goal to belong consumer. fit ix^idfi = 34(9); CFI = 0.95, GFI = 0.96, choose to consume nostalgic products P5: Higher levels of advertising- IFI = 0.96, NFI = 0.94). as a means to address this goal. evoked personal nostalgia will The estimated path coefficients (i val- Further, their inquiry also found enhance the consumer's bonds ues of path coefficients tested at p < 0.05) that consumption of nostalgic prod- with the focal brand. generally were supportive of the expected ucts successfully fulfills this need to relationships embodied in propositions 1 belong. These findings are consistent Methodology to 5, between the four factors of the per- with past work that also found that Data were collected from 166 consumers sonal nostalgia scale and the five add- nostalgia enhanced the consumer's using an online consumer panel. The sam- itional constructs (See Figure). perception of social support, and ple contained a balanced age and gender To elaborate on these results, the reduced loneliness (Zhou, Sedikides, mix. authors found that the consumer's nos- Wildschut, and Gao, 2008). Respondents initially answ^ered talgia proneness positively influenced the The authors, therefore, argue that questions related to the antecedents. past imagery, physiological reactions, and personal nostalgia evoked by adver- Specifically, positive emotions evoked by the nostalgic tising also would result in higher advertising (See Figure). levels of perceived social support for • loyalty for the focal brand (Pepsi) was The current study also foimd that the the consumer. determined by asking the respondents more a consumer feels lonely, the more "Out of the last ten times you would he or she is likely to experience past - Brand bonds: Brand bonds are an have had a soda, how many times did imagery, physiological reactions, and neg- important facet of the relationship the you drink Pepsi Cola?" ative emotions after watching a nostalgic consumer has with a brand (Aaker, • loneliness was measured using a 20-item advertisement. Fournier, and Brasel, 2004). In fact, loneliness scale (Russell, Peplau, and The authors found that higher levels of there may be nostalgic roots in the Cutrona, 1980); and loyalty with the focal brand led to signifi- relationships consumers share with • nostalgia proneness was measured cantly higher levels of past imagery, posi- brands (Fournier, 1998). In the context using a 20-item scale developed by tive emotions and physiological reactions, of cigarette and tea advertising, for Batcho (1995). and lower levels of negative emotions instance, one study found that even evoked by the nostalgic advertising. newer brands benefited by using The respondents then were exposed to On the consequences side of the nostalgic imagery, as the positivity the nostalgic Pepsi "Refresh Anthem" research, the authors found that past associated with the past depicted by 60-second television commercial. After imagery and positive emotions evoked the nostalgic advertisements rubbed the panelists had reviewed the television by the nostalgic advertising led to higher off on the focal brand and enhanced advertisement, they completed the 34-item levels of perceived social support for the the consumer's emotional bonds to nostalgia scale. consumer, whereas higher levels of nega- the brand (Holak et al., 2007). In light In turn, the respondents then answered tive emotions associated with personal of these findings, the authors pro- questions related to the two consequences: nostalgia led to lower levels of perceived pose that personal nostalgia evoked social support and bonds with the focal social support. June 2 0 1 3 JDÜlillllL DF BDÖEÜTISinG BESEflBCII 1 5 9 PERSONAL NOSTALGiA IN ADVERTISEMENTS Moreover, the inquiry showed that had acceptable levels of loading on the nostalgic Pepsi "Refresh Anthem" adver- past imagery, physiological reactions, and discriminant function (loading greater fising), producing a total sample of 262 positive emotions evoked by the nostalgic than 0.40). The Wilks X for the discrimi- consumers. advertising may enhance bonding with nant model was statistically significant The authors found that the nostalgic the focal brand, whereas higher levels of {X = 0.93, x~ (df) = 12(4), p < 0.05). Fur- advertising—when compared to the non- negative emotions associated with per- ther analysis also showed that the discri- nostalgic Pepsi advertising—scored sig- sonal nostalgia led to lower levels of bond- minant function predicted brand choice nificantly (p < 0.05) higher across the four ing with the focal brand. significantly better than chance (Press's dimensions of their nostalgia scale. Taking these results, along with the Q = 31.41, p< 0.01). The mean number of times the respond- Study 3 findings, the authors concluded Overall, the results of this study sup- ents consumed Pepsi (of the last 10 soda that even though overall nostalgia port the premise that the personal nos- consumption occasions) was 2.87 (median enhances the consumer's relationship with talgia (measured by its four factors—past = 2). For analysis purposes, those who the brand, the negative emotions of nostal- imagery, positive emotions, negative had consumed Pepsi two or fewer times gia could enervate these relationships. emotions, and physiological reactions—) were termed "less loyal" consumers These findings are new to the litera- evoked by advertising influences the (145 respondents), and the rest w êre ture and build on extant research on choice of brand made by the consumer. termed as "more loyal" consumers (117 advertising-evoked affect (Brown et al., respondents). 1998; Lee and Han, 2002). Effects on Types of Users (Less versus Comparisons between the two adver- This series of findings establishes nomo- More Loyal) tisements (nostalgic and non-nostalgic) logical validity for the authors' scale that Past advertising research shows that for more- and less-loyal consumers were satisfactorily relates with its theoreti- advertisements had elicited different reac- tested separately using one-way analy- cal antecedents and consequences. The tions among different types of consumers. sis of variance (ANOVA). The results of authors also compared their new scale to The key question the authors examined the one-way ANOVA among more loyal an alternate measure of personal nostal- in the current study: Do nostalgic adver- consumers showed that there was no sig- gia (Pascal et al, 2002) and found the new tisements (as compared to non-nostalgic nificant difference between the purchase index to be superior (Appendix D). advertisements) work better among more intentions across the two advertisements loyal consumers—as compared to less (^Nosta l ic ad ~ ^ • ' ^ ^ ' ' '^Non noslal ic ad ' ^•^^' Predicting Consumer Choice loyal—consumers? F = 1.09, n.s.). This implies that more Another test of the author's new scale was To investigate this issue, the authors loyal consumers are so engaged with the its ability to predict consumer choice of the collected additional data from 96 con- brand that they are predisposed to buy the advertised brand. At the end of the sur- sumers using the online consumer panel. product anyway, regardless of the type of vey, the respondents were told that if the The respondents first answered a ques- advertising. research agency were to consider offering tion related to the loyalty for the focal Conversely, the results of the one-way them $5 as a gift, which option would they brand (Pepsi) ("Out of the last ten times ANOVA among less loyal consumers choose: a gift coupon for Pepsi Cola for $5 you would have had a soda, how many showed that there was a significant dif- or a gift card for $5 (which could be used times did you drink Pepsi Cola?"). The ference in the levels of purchase inten- to buy any product). respondents then were exposed to the tion generated by the two advertisements. A multiple discriminant analysis was non-nostalgic "Justin Timberlake" Pepsi The nostalgic advertising generated run with choice as the dependent vari- advertisement (http: / /bit.ly/h4t47m). higher purchase intentions as com- able (1 = Choosing gift coupon for Pepsi, Consequently, they expressed their pur- pared to the non-nostalgic advertising 0 = Choosing non-Pepsi gift card). The chase intention—specifically, how likely (̂ Nostalgic ad = ^.50, M^^_, _^^ ĵ̂ ,̂ ._. ^̂ = 2.69, four dimensions of advertising-evoked are they to purchase the product shown f = 5.45, p< 0.05). personal nostalgia were included as the in the advertisement—and completed the These findings are new to the literature independent variables in the discriminant study's 34-item nostalgia scale. and suggest that even less loyal consum- model. This sample was integrated with the ers can be engaged using nostalgia-based The analysis revealed one discrimi- previous sample of 166 respondents advertising, resulting in positive outcomes nant function, and all the four variables (the group that had been exposed to the for the focal brand. 1 6 0 JDÜRflflL DE flflüERTISIRG RESEflRCfl June 2 0 1 3 PERSONAL NOSTALGIA IN ADVERTISEMENTS DISCUSSION AND IMPLICATIONS Academic relevance is shown as this nostalgic advertising as it nurtures Advertising research has moved away paper shifted the focus on nostalgia to a brand-consumer relationships. from looking at whether an advertisement more complex, robust, multidimensional "is liked" toward examining how well an theoretically grounded formulation mani- • It indicated that, among less loyal advertisement "engages" consumers. fested in response to marketing communi- consumers, nostalgia-based adver- One way to engage consumers is to cations. Physiological response based on tising was likely to work better than make advertising content personally rel- memories evoked by the advertising was non-nostalgia-based communication. evant to them by invoking situations (or identified as an important new factor in Thereby, advertisers may be advised to events) that they have experienced in the measuring personal nostalgia. use nostalgic themes to engage an even past (i.e., through personal nostalgia). The current study also found that per- less loyal consumer segment. The current research sought to develop sonal nostalgia involved invoking both a scale that could identify the degree to positive and negative emotions (rather • The research suggested that effective which an advertisement successfully than either/or) and that nostalgia resulted nostalgia-inducing advertisement must aroused personal memories of the past in higher levels of attitude toward the - evoke images from the consumer's and facilitated the positive transfer of not advertising and strengthened bonds with past, only emotion but relevance and meaning the focal brand. - conjure up positive emotions and toward the advertised brand. Indeed, the authors believe that the physiological reactions, and Toward that end, the authors executed effect of nostalgia is so potent that it plays - curtail the negative emotions associ- four studies employing recommended a role even in enhancing the consumer's ated with nostalgia. qualitative and quantitative analyses, perceptions of social support. Though item generation, and item reduction, and the nostalgic experience on the whole has This investigation explicates that, even successfully assessed various forms of positive ramifications, the current inquiry though the ramifications of nostalgia are reliability and validity across multiple showed that the negative emotions of nos- overall positive, the negative emotions data sets. The findings culminated with talgia had an adverse effect on A^̂ , and the have a negative effect on  ^ and bonds Study 4, which tested four dimensions consumer's relationship with the brand. with the advertised brand. of the study's focal construct preceded This research also has several implica- These insights into the consumer's nos- by theoretically grounded antecedents tions from the practitioner's perspective: talgic response are novel to the literature and outcome variables in the context of a and may be of strategic help to the adver- nomological network. Study 4 also found • It demonstrated that advertising-evoked tising executive for copy development the authors' four-dimensional formulation personal nostalgia is multidimensional: and testing. For example, using standard was superior to an alternative nostalgia past imagery, positive emotions, nega- A ĵ and likeability measures to assess a measure (See Appendix D), and the discri- tive emotions, and physiological reac- nostalgic advertising may only indicate minant analysis showed the new nostalgia tions. Advertising executives may find the consumer's overall impressions of the scale did predict brand choice. it useful to evaluate nostalgia-based advertising. Using the new nostalgia scale Overall, the evidence was strong that the advertising using the authors' multidi- diagnostically, however, would reveal authors' nostalgia scale made a significant mensional scale. A tested and validated how effectively the advertisement evokes contribution to the literature from the van- scale would enable advertisers to engage the various dimensions of nostalgia, tage points of both theory and application. in deep thinking about their target mar- prompting ability to fine-tune the copy The authors also sought to make this kets' multidimensional responses to accordingly. paper more than a "scale-development various nostalgia-based promotional piece" to be added to the extant litera- stimuli. FUTURE RESEARCH ture and previous measures. Of overrid- Future researchers could consider study- ing interest was the "bridge" built from • This research demonstrated that nos- ing the effects of the use of framing con- complex multidimensional theory sur- talgia elicited by advertising was so cepts (Levin, Schneider, and Gaeth, 1998) rounding what the authors call "personal engaging that it influences A^̂ , bonding in presenting nostalgic advertising to nostalgia" and how practitioners' market- with brand, and brand choice. Thus, it consumers. For instance: Would posi- ing techniques evoke this construct. reaffirmed the practice of employing tively framed messages be more effective June 2 0 1 3 JOÜfiOBL DF flDOERTISlOGR ESEflRCH 1 6 1 PERSONAL NOSTALGIA IN ADVERTISEMENTS as compared to negatively framed ones, advertising strategy issues, in particuiar, he has BATCHO, K. I. "Nostalgia and the Emotional as they are less likely to evoke negative specialized in viewer perceptions of sex-roie portrayals Tone and Content of Song Lyrics." American emotions? In both the basic and applied in internationai advertising. His v/orW has been Journal of Psychology 120,3 (2007): 361-381. context, what would be the research published in a wide variety of journais inciuding Journal opportunities to extend this new nostalgia of Advertising Research, Journal of Advertising, and BAUMGARTNER, H . "Remembrance of Things scale to a cross-cultural context? What are Journal of the Academy of Marketing Science. Past: Music, Autobiographical Memory and the possibilities of studying the manifes- Emotion." Advances in Consumer Research 19 tations of nostalgia in other cultures and MICHAEL S. LATOUR (PhD, University of Mississippi) is a (1992): 613-620. the potential for this new scale (and/or visiting professor of services mari