Portal de Periódicos da CAPES

It’s About Time: A Call for More Longitudinal Consumer Research Insights

2020; University of Chicago Press; Volume: 5; Issue: 3 Linguagem: Inglês

10.1086/709858

2378-1823

Pradeep K. Chintagunta, Aparna A. Labroo,

Economics of Agriculture and Food Markets

Previous articleNext article FreeLongitudinal Effects and ConsumptionIt’s About Time: A Call for More Longitudinal Consumer Research InsightsPradeep Chintagunta and Aparna A. LabrooPradeep Chintagunta Search for more articles by this author and Aparna A. Labroo Search for more articles by this author PDFPDF PLUSFull Text Add to favoritesDownload CitationTrack CitationsPermissionsReprints Share onFacebookTwitterLinked InRedditEmailQR Code SectionsMoreThe most consequential of influences on, and influences of, consumer behaviors are likely to occur frequently over time, be pervasive over time, or have impact over extended durations of time. Such influences are likely to be among the most consequential because if they persist over time they are likely to be more observable and have bigger effect sizes. They also may be more complex—codetermined by a richer set of factors when they arise and impacting multiple outcomes over the extended periods when they persist. Longitudinal studies therefore hold the potential to provide some of the most invaluable insights into consumption behaviors. The prevalence of such studies in consumer research, however, remains few and far between because such studies can be costly to conduct and difficult to publish when they overclaim causality. Thus, longitudinal analysis has been both lacking among consumer research scholars, and when it is done, it is not always as strong as it could be.Our vision for this special issue on longitudinal effects and consumption, inspired by the journal’s first editor-in-chief, Joel Huber, is to encourage consumer researchers to engage in more research that provides longitudinal insights. To achieve this objective, we distinguish between a “longitudinal study” and a “longitudinal insight.” While a “longitudinal study” conventionally involves tracking cohorts or panels of consumers for changes in behaviors over an extended period of time, or correlating multiple variables for co-occurrence in historical data sets, we propose that a “longitudinal insight” can arise even without engaging in such classic longitudinal studies. By focusing on longitudinal insights, in this special issue, we thus broaden the definition of the types of questions consumer researchers might ask, and we include a consideration of more accessible methodologies that could be utilized to provide such insights. By doing so, we hope to popularize research providing such longitudinal insights. Second, to showcase the range of questions that might be asked within a topic area that provide longitudinal insights, and to facilitate a greater integration of insights across methodologies, we focus on three main topic areas. These topic areas include: sequential effects on financial decisions (by consumers and by firms), consumer learning and persistence, and relationship dynamics and consumer choices.In what follows, we first briefly describe how longitudinal studies are typically construed and why that approach might limit their popularity among consumer researchers. We next present our definition of what constitutes a longitudinal insight and illustrate different questions articles in our special issue ask that adhere to this broadened definition and utilize a broadened set of methodological approaches to provide longitudinal consumer insights. Finally, to demonstrate why such results matter, we turn to discussing each topic area and describe how all articles within each topic cumulatively provide rich longitudinal consumer insights.Traditional Approaches to Longitudinal ResearchLongitudinal research is a type of correlational research that involves looking at correspondences among variables collected over an extended period of time—usually days, weeks, months, years, or in some cases, several decades. One example of this type of longitudinal research involves, first, selecting a cohort of respondents that share some common personal characteristic, and second, tracking the cohort and collecting repeated observations on an outcome variable of interest at different points in time. For instance, to test whether childhood intelligence is correlated with positive life outcomes, in what has now become the longest still-running longitudinal study, psychologist and eugenicist Lewis Terman (1922) recruited 1,500 high-IQ children between ages 8 and 12 in the 1920s and recorded their emotional and social outcomes over decades. By correlating the original characteristic of interest with the repeated outcome observations, researchers can understand the trajectory of changes the cohort experiences over time.Rather than engage only one cohort, a second example of longitudinal research can sometimes recruit a panel instead. Panel studies involve collecting repeated responses from a cross-section of respondents, some of whom share a common characteristic and others who do not share that characteristic. Panel studies ensure generalizability across different populations; they also allow researchers to contrast across different populations the cumulative effects of repeated exposures to environmental contexts. This methodology allows researchers to ask several important questions about how effects endure over time—for instance, do consumers, over time, remain entrenched within the social class they were born into, or the experiences they gleaned in childhood, and whether firms’ outcomes, decades later, remain entrenched in early decisions they made. In particular, to test whether children born to a social class remain entrenched in that class throughout their lives, Michael Apted, a documentary producer, selected a cohort of children each from rich, poor, and middle-class families in the 1960s. He proceeded to interview them on film every seven years. Today, his documentation provides unique insight into the development of British consumption and culture since the 1960s (see https://en.wikipedia.org/wiki/Up_(film_series)). In an award-winning article, consumer researchers Connell, Brucks, and Nielsen (2014) tested enduring effects of childhood information exposures into adulthood. They found enduring effects of ad exposures in early childhood on product preferences and choices consumers made later in life. Bronnenberg, Dhar, and Dube (2007) tested the enduring effects of early location decisions firms make on their growth over time. They found that national brands continue to dominate over time in geographies where they originated, raising questions regarding what it means to be a national brand. All three of these studies provide invaluable longitudinal insights about consumer behavior or firm outcomes.A third type of longitudinal study is retrospective. Retrospective studies code historical information to uncover how different variables of interest repeatedly co-occur over a period of time. For instance, Baker, Bloom, and Davis (2016) coded 12,000 newspaper articles spanning between 1985 and 2010 in 10 outlets for frequency of coverage of events. The study showed that consumers are likely to seek more information in times of growing economic uncertainty, and therefore the extent consumers seek information at any point in time could be used to predict future economic uncertainty. These examples demonstrate critical consumer, policy, and firm behavior insights that longitudinal investigations can provide.Despite these invaluable insights, the prevalence of such longitudinal studies in consumer research remains limited. One reason for the paucity of longitudinal consumer research studies is that such studies typically rely on observational data and provide largely correlational conclusions. To address this issue, recent longitudinal research has employed difference-in-difference measures, also referred to as before-after designs. These studies typically examine the effects of a policy intervention and compare outcomes of the intervention group before the intervention against that group’s postintervention outcomes. Such studies also typically employ a control group that is similar to the intervention group on important dimensions but that did not experience the intervention. The control group data are used to control for naturalistic temporal changes and make inferences about the incremental effects of the intervention with greater confidence. An important point to note here is that the research insight from such studies typically pertains to the effect of the policy intervention controlling for any longitudinal effects; thus the researcher interest is in understanding the effect of presence-absence of the intervention rather than changes over time. While the study may be longitudinal by virtue of employing pre/post measures and methodology, the insight therefore, more often, is not longitudinal.For instance, in a powerful demonstration, Lin, Liu, and Qian (2014) investigated whether public policies that ban sex-selective abortion as a way to curb sex imbalance can have an unintended consequence on reducing survival rates of baby girls because parents reduce postnatal care of baby girls that they were unable to abort. They found support for their premise by comparing births and deaths pre- to postlegalization of sex-selective abortions in Taiwan. Similarly, Seiler, Song, and Wang (2017) investigated the effects of temporary banning of a microblogging site on viewership of TV shows. Neither of these studies provides longitudinal insight, even though both involve before/after comparisons—the key insight from both studies pertains to the presence versus absence of the intervention. Instead, Thaler and Benartzi (2004) proposed consumers fail to save enough for retirement because they increase expenses when they receive raises. They found encouraging people to commit to save now for tomorrow increased employee retirement savings from 3.5% to 13.6% of income after 40 months. This study provides longitudinal insight—it measures impact of an intervention on an outcome over a period of time.A second, major factor that might dissuade researchers from asking questions that hold potential to provide longitudinal insights could be the concerns over the difficulty, expense, and riskiness of running such studies. To address these researcher concerns, we broaden the definition of what we consider to be a longitudinal insight that is beyond what is typically considered to be a longitudinal study. We also exclude from our definition studies that may employ longitudinal designs but utilize a research question or main insight that is not longitudinal.Asking Questions That Deliver Longitudinal Consumer InsightsWe define research with a longitudinal insight to be any research that provides a richer understanding of the dynamics of consumer behavior. Longitudinal insights thus might arise from a consideration of the effects of a single event that occurs once, a single event that occurs repeatedly, or multiple events that occur in parallel or in sequence prior to having an impact on a subsequent temporally separated consumption decision. They also might arise from the consideration of a single or multiple events that set off a chain of subsequent outcomes. These effects may arise over varying time durations: an extended period of time, as is typical of longitudinal studies, or over a reduced period of time. They might also arise from the investigation of multiple events that co-occur over an extended period of time, as is typical in longitudinal studies. They might also arise from a consideration of a sequence of events that occur in close proximity to each other where their ordering might impact consumer or firm outcomes. They may arise from differences in information processing that employs sequential compared to simultaneous exposures. At the broadest level, they also might incorporate considerations of expected future outcomes on current choices, or of current choices on future expectations of outcomes. These latter questions all provide longitudinal insights but need not employ “longitudinal” studies.Our definition of longitudinal insight thus includes traditional longitudinal designs that employ cohort or panel data, and/or historical data. Indeed, our special issue showcases a set of studies that employ these traditional approaches and uncover important longitudinal consumer insights. As an example, Moore, Fitzsimons, and Fitzsimons (2020) present a quasi-cohort analysis in which they correlate whether making decisions on behalf of a relationship partner can evoke negative emotions among consumers with a more interdependent relationship style because making such choices cues anxiety of not being able to live up to the partner’s expectations. These negative emotions, over time, hurt relationship satisfaction. Moore et al. track a group of heterosexual couples over a four-month period. To begin, they measure the dominant relationship style of each respondent in their sample. After a period of three months they solicit recollections of the amount of choices each respondent made on behalf of their partner and emotions toward the relationship partner. One month later, they measure relationship satisfaction. While a cohort analysis might have measured all three variables—relationship satisfaction, actual choices, and negative emotions—multiple times, this quasi-cohort analysis still provides highly novel insights regarding relationship dynamics and how making choices for close others might hurt a relationship over time when consumers, ironically, view their relationships as more interdependent (see fig. 1).Figure 1. A typology of different pathways to longitudinal insight.View Large ImageDownload PowerPointRaghunathan, Yang, and Chandrasekaran (2020) instead present longitudinal consumer insights derived from a panel data analysis. They employ panel data from the National Longitudinal Survey of Children and Youth to investigate the correspondence between parental love that respondents received in childhood and their financial discipline observed in adulthood. Covering a sample of 1,428 respondents, their research finds that children who received higher levels of parental love during childhood demonstrate a higher emotional quotient, and they also tend to be more financially responsible adults. The methodology employs panel data, retrospective recall data, and cross-sectional data to provide important insights regarding how a consumer’s early life experiences might impact later financial outcomes.Also adhering to the traditional definition of a longitudinal study, Wood and Bollinger (2020) utilize historical data to provide important consumer insights with respect to how the extent of information in the environment, rather than the salience of any single event, prompts consumer health-related action. In this article, the authors investigate the co-occurrences of the amount of news coverage about a celebrity medical diagnosis, the number of celebrity medical diagnoses over time, and rates of breast exams and mammograms among consumers over a 19-year period. They find that the amount of news coverage about any diagnosis, but not the celebrity diagnoses per se, corresponds positively with breast exam and mammogram rates of consumers. Wood and Bollinger (2020) and Raghunathan et al. (2020) thus provide longitudinal insights employing traditional longitudinal studies in two different ways—by showing the effects of temporally distant variables (parental love during childhood) on current consumer outcomes (adulthood financial discipline), and by showing co-occurrence of two factors (extent of news about a celebrity medical diagnosis and breast exam/mammogram rates among consumers) as these arise over an extended 19-year period. To support their key theorizing and better establish causality, Raghunathan et al. (2020) and Moore et al. (2020) present additional experimental data, and Wood and Bollinger (2020) conduct several robustness checks that include suitable control comparisons.Several articles in this special issue instead employ nontraditional approaches to uncovering longitudinal insights pertaining to the dynamics of consumer behavior. A first of these nontraditional approaches to uncovering insights pertaining to the dynamics of consumer behavior includes randomized study designs where researchers assign respondents at random to a particular intervention and then measure the effect of this assignment on an outcome variable after a suitable lapse of time. For instance, Mehr et al. (2020) investigate the effects of a single intervention on consumer outcomes after a suitable lapse in time, while Brucks and Huang (2020) randomly assign respondents to different interventions and measure repeated outcomes over an extended period of time (Brucks and Huang 2020). Keller, Hesselton, and Volpp (2020) instead investigate the effects of a single intervention on prompting consumers to react sooner or later over an extended subsequent period of time. Finally, Cannon and Rucker (2020) investigate how chronic orientations of consumers might make them differentially responsive to naturally occurring environmental cues over an extended period of time.More specifically, Mehr et al. (2020) investigate downstream effects of asking consumers to source different strategies that might facilitate their goal persistence. Consumers who first report their motivation to persist toward a goal and then are directed to actively search their social networks for goal-pursuit strategies are more motivated after a 10-day period to persist on their goals compared to consumers who did not search for strategies but were merely provided them. Brucks and Huang (2020) similarly temporally separate the independent and dependent variable in a randomized longitudinal study design, but they additionally examine the effects of repeatedly practicing different creativity tasks over a two-week intervening period on daily performance changes. Respondents show a positive effect of repeated practice on performance over time on convergent creativity tasks, but a negative effect of divergent creativity practice on performance. They suggest that practice results in two opposing forces: it hinders accessibility to less routinized cognitive pathways but facilitates accessibility to routinized cognitive pathways. Interestingly, respondents failed to predict that practice might hurt some types of performance.Also employing a longitudinal randomized study design that temporally separates the independent and dependent variable, Keller, Hesselton, and Volpp (2020) additionally manipulate the perceived time that consumers believe they have to take action and their incentives to act sooner. They find that financial incentives increase enrollment in a health program, but counterintuitively this effect is especially likely to arise when the incentive reduces over time rather than when it remains constant. They suggest that declining incentives can cue regret of missing out on the incentive and prompt consumers to act sooner. Finally, Cannon and Rucker (2020) employ longitudinal designs to investigate how chronic orientations of consumers influence their sensitivity to naturally arising environmental cues. They find that consumers with a chronic nurturance orientation are more sensitive than consumers with an affiliation orientation to environmental nurturance cues, and vice versa, and environmental cues that match a consumer’s chronic orientation prompt compatible action.As a set, these four studies not only temporally separate the independent and dependent variables, but they all provide important longitudinal insights regarding how a single task or multiple intervening tasks a consumer performs might impact a single outcome or repeated consumer outcomes (Brucks and Huang 2020; Mehr et al. 2020), how different intervening incentives might prompt sooner action (Keller et al. 2020), and how some chronic orientations might direct differential attention to intervening environmental stimuli (Cannon and Rucker 2020). Notably, as we indicated previously, the mere comparison of before/after outcomes arising in response to a specific environmental change or policy decision does not imply from our perspective that a longitudinal insight has been gleaned. All four studies in our special issue that engage randomized longitudinal study designs provide important longitudinal insights in novel ways because their basic research questions pertain to how changes arise over time or how time might impact presence of outcomes. As a set, these articles ask a range of important longitudinal questions by considering persistence of effects of one or multiple independent variables on temporally separated dependent variables, by designing temporal incentives to respond, or by measuring the effects of naturally arising temporal factors on chronic orientations.The second of the nontraditional approaches to uncovering insights pertaining to the dynamics of consumer behavior involve brief temporal separations of the independent and dependent variables. Such studies can provide important longitudinal consumer insights into affective and memory processes when considering information that is presented sequentially rather than simultaneously, and of order effects when information is presented sequentially, on consumer outcomes. Several studies on order effects or information format exist in consumer research, but their longitudinal insights are often understated or not key to the question asked.In this respect, Khan and Pond (2020) propose that whether consumers first gauge their affect versus their cognitions in response to products marketed by firms engaged in socially responsible actions can differentially impact not only evaluation of the brand and the firm by consumers but also the persistence of consumer attitudes over time. The authors provide a novel longitudinal insight that consumers who first assess their affect toward a socially responsible brand compared to a control not only evaluate the brand more favorably, but this affect spills over to their product evaluations and positive perceptions of product attributes. However, when consumers first assess their product-related cognitions, there is no positive halo effect on subsequent brand evaluations. Thus, the nature of evaluative processes consumers first engage in when they encounter socially responsible brands can impact the extent and spread of positivity to subsequent firm-relevant evaluations they engage in. The article provides important longitudinal insights into the interplay between affective and cognitive processes on formation of consumer attitudes.Huang et al. (2020) also consider firm-relevant outcomes and how consumers learn information and develop associations in memory over time. They present novel theorizing around how repeated experiences with novel payment mechanisms in different contexts might impact their adoption among consumers. They posit and find that because many novel payment mechanisms are associated with social rather than professional purchases, consumers view firms using such mechanisms as warmer but less competent, regardless of actual service quality. Furthermore, consumers adjust their evaluations of such service providers over time only after several unambiguous positive experiences that highlight business competence. The article provides important longitudinal insights into how associations develop and endure in consumer memory over time.Two final articles in our special issue consider the effects of sequential exposures to information on consumer-relevant outcomes. Echelbarger, Maimaran, and Gelman (2020) demonstrate that for the most part, young children make choices that are similar to those made by adults, and they choose more variety when options are presented to them simultaneously rather than sequentially. Their findings suggest that factors that impact choices in adulthood are observed even in early childhood. Finally, Xu, Cheng, and White (2020) find that while consumers perceive a single negative feedback as opportunity to self-improve and prove themselves, repeated negative feedback serves as a self-threat and can demotivate performance. These latter two articles provide important longitudinal insights based on the effects of information format on evaluations and the effects of repeated feedback on emotional reaction, self-perceptions, and motivation. As a set, these four articles provide longitudinal insights by demonstrating an interplay of cognitive, affective, and motivational processes on consumer outcomes.Why Longitudinal Insights Matter: Three Topic Areas as ExamplesThe previous section highlighted research questions that provide longitudinal insights while leveraging methodological diversity. We now turn to discussing why longitudinal insights matter. We focus on key insights gleaned from each article within each topic area and then across all articles as a set within each topic area.The first topic area focuses on sequential effects on financial decisions of consumers and firms. The first two articles in this section provide important longitudinal and practical insights into the temporal effects associated with information consumers are exposed to on their financial decision making. The latter two articles provide important longitudinal and practical insights into temporal effects associated with consumers’ information exposure on their judgments that have a potential to influence a firm’s financial outcomes.First, Raghunathan et al. (2020) investigate temporal effects of experiences in childhood on financial decisions respondents make as adults. The article provides important longitudinal insight that children who experience more parental love are likely to develop higher emotional quotients that can facilitate better financial decisions in adulthood. It is important for showing the enduring effects of the quantity and quality of nurturance that a child receives in promoting prudent future financial decisions and well-being in adulthood. The article is important from a practical angle—it suggests that parents can improve the future financial well-being of their children by providing them adequate levels of parental love. The second article in this topic area by Keller et al. (2020) also considers how consumers make financial decisions and how the financial incentives they are offered might prompt them to make more prudent choices. But rather than investigate enduring childhood experiences into adulthood, this article considers temporal aspects of the design of different kinds of financial incentives and how these elements can cue regret of missing out on the incentive among consumers and prompt prudent action. The authors demonstrate that a “declining financial incentive” to enroll in a health management program (where the incentive declines over time) can cue regret of losing out on the incentive among consumers. This article is important for its longitudinal insights—it shows that temporal elements in the design of a financial incentive can impact the emotional response the incentive evokes among consumers and, consequently, on the success of the scheme. This article is also important for practical reasons because it shows that designing such financial incentives can benefit enrollment programs—declining incentives do just as well as fixed incentive schemes on enrollment, but costs are lower. Together, these two articles on financial decisions show temporal influences of childhood factors into adulthood financial choices and temporal influences of financial choices in adulthood.The third article in this section is by Huang et al. (2020). It shifts focus to consider financial outcomes of firms based on the associations consumers develop over time to the payment methods the firms employ. This article is important for its longitudinal insights because it shows that early associations consumers develop—that firms that employ novel peer-to-peer payment mechanisms are more social and therefore warmer but less competent—are enduring. Consumer attitudes persist once formed, and repeated exposures to unambiguous information may be needed to change such attitudes. This article therefore also is important for practical reasons. It suggests that firms should consider how employing payment mechanisms might impact their perceived competence and overall positioning as a more social but less competent organization. Firms using peer-to-peer payment mechanisms may wish to consider in advance the heightened importance for them of ensuring unambiguously positive outcomes with the firm to consumers. The final article in this section also investigates financial outcomes of firms based on the associations consumers develop over time but based on the kinds of information they first consider about the firm.In this article, Khan and Pond (2020) find that firms benefit from a halo effect of engaging in socially responsible corporate actions but only if consumers first consider evaluations that evoke an affective rather than a cognitive response. Firms that engage in socially responsible actions evoke a higher level of positive feelings compared to firms that do not engage in such actions. This article is important for its longitudinal insights into financial outcomes of firms—it shows that the order in which consumers make evaluations can spontaneously increase a reliance on affect versus cognition among them. This reliance can have different financial consequences for firms. The order in which consumers make evaluations can impact whether halo effects on other evaluations are observed, as well as the persistence of those attitudes over time. This article is also important for practical reasons because it shows that firms that engage in socially responsible actions may benefit by allowing consumers to rely on their affect rather than cognitions when evaluating firm-relevant outcomes. F