Revisiting the Age Construct: Implications for Service Research

Abstract

People in the older consumer segment spend more money on services than those in other segments, making them a desired target for service providers. This universal trend has led researchers to start discussing this trend’s implications for service research and marketing practice. These discussions’ results are ambiguous because service researchers and managers face the problem of having to choose between three main age constructs: chronological age, cognitive age, and future time perspective (FTP). Unfortunately, current age-related research lacks an understanding of their real value, as only a few studies have combined them to discuss their specific impact. Recognizing this gap in the literature, this article compares the three age constructs in behavioral and perceptional settings. We highlight each age construct’s merits and weaknesses as well as exploring which construct delivers the best results in which service context. Bayesian analyses of our data reveal that chronological age has its merits as a control variable but does not sufficiently discriminate between age groups’ behavior and perceptions. Cognitive age is useful if customers’ own age perceptions are included but only identifies age differences in specific service settings. FTP consistently detects age-related differences in customers’ perceptions and in their behavior in all contexts.

Keywords

service research older consumers segmentation market research age construct chronological age cognitive age future time perspective Bayesian interference

Seniors are an important target group for service providers. Not only do seniors tend to be more loyal, but service marketers also need to address the challenges arising from our aging population (Moschis, Bellenger, and Curasi 2003). One in four persons living in Europe and Northern America in 2050 will probably be aged 65 or older, although—for the first time in human history—people in that age bracket already outnumbered all the world’s children under 5 years in 2018 (United Nations 2019). Given that older consumers spend extensively on services, such as financial services, health, well-being, tourism, and hospitality (e.g., Nasco and Hale 2009), insights into the senior consumer segment are particularly important for service organizations. This is, however, not true of all senior populations across the globe. Many older people experience poor well-being due to inadequate financial resources and poor health (Mittal and Griskevicius 2016).

Our study focuses on age trends that primarily affect the United States, Western Europe, and Japan. We focus, however, on U.S. data to discuss the implications for service research and practice. Many service providers fail to adapt their offers to seniors, therefore also failing to engage with it (Yoon et al. 2009). Researchers highlight that this might, on the one hand, be due to stereotypes of this customer segment’s physical and psychological states, as well as their spending power, which influence companies (Szmigin and Carrigan 2001). On the other hand, older people tend to question their own abilities, especially in technology-driven service contexts (Marquié, Jourdan-Boddaert, and Huet 2002). Although opportunities for services aimed at this segment have increased (Kohlbacher and Chéron 2012), companies often miss opportunities to create value through their relationships with older consumers. Our study aims to provide research addressing the existing challenges that the role of age plays in service research and practice.

Traditionally, age is frequently used as a segmentation variable in service settings (Polo and Sese 2013). Service researchers often use the age construct to explain various behaviors (e.g., Tarasi et al. 2013) or use it as a control variable (Bock, Folse, and Black 2016; Sweeney, Danaher, and McColl-Kennedy 2015).

Service researchers acknowledge that the role of age goes beyond segmentation and control variables by, for example, positing age as a personal characteristic relevant in the context of consumer decision-making processes and purchasing as well as a determinant of loyalty behavior (e.g., Evanschitzky and Wunderlich 2006). In many service contexts, age is found to trigger significant differences in terms of value perceptions and satisfaction (Nysveen, Pedersen, and Thorbjørnsen 2005), and as a crucial success factor for pure services, such as mobile services (Kumar and Lim 2008). In other technology-related services, such as self-service settings, age is viewed as a determinant of customers’ attitude toward and usage of services (Weijters et al. 2007). In a review that covers 45 years of older people’s consumption-related decision making, Hettich, Hattula, and Bornemann (2017) point out that despite these advancements, the research is fragmented and the findings are mixed. They call for a more holistic view that, for example, investigates emotions’ crucial and often controversial role in the service context. These scholars suggest that the application of socioemotional selective theory, for example, by measuring consumers’ future time perspective (FTP), could be beneficial to advance our knowledge about emotions’ role in older consumers in a service setting.

Although the service research domain has recently started promising discussions (e.g., Grougiou and Pettigrew 2011), the current age-related research lacks an understanding of the three major time perspective constructs’ role in predicting behaviors or perceptions (Brothers et al. 2016). Moreover, the vast majority of studies only investigates one of the three concepts (Wei, Donthu, and Bernhardt 2013).

Furthermore, service research findings are ambiguous, regardless of the age concepts (see Table 1 for exemplary studies in service contexts). For example, some studies focusing on chronological age identify its impact on outcomes such as satisfaction and loyalty (e.g., Lambert-Pandraud, Laurent, and Lapersonne 2005; Mittal and Kamakura 2001), while others do not find that it has any significant influence on the same outcome constructs (e.g., Cooil et al. 2007). Some authors confirm that there is a difference between chronological age and cognitive age, while others contradict these findings (e.g., Guido, Amatulli, and Peluso 2014; Gwinner and Stephens 2001). When researchers use a single time perspective construct, the results remain heterogeneous or contradictive, therefore questioning a specific concept’s suitability in a given context (Collins and Tisdell 2002).

Table 1.

Exemplary Literature Age Versus Services.

Reference	Context	Age Construct	Exemplary Findings
Amatulli, Guido, and Nataraajan (2015)	Luxury	Cognitive age	Older consumers with a lower cognitive age rely more on brands than specific products.
Amatulli et al. (2018)	Various services	Felt age/cognitive age	Older consumers’ tendency to feel younger varies as a function of whether they are exposed to young or old (same-aged) social cues.
Cameron, Richardson, and Siameja (2016)	Various services	Chronological age	Dissatisfied older consumers are more than 12 times more likely to report an intention to switch providers than satisfied consumers.
Chaouali and Souiden (2018)	Banking	Cognitive age	Differences in usage, value, and risk perceptions.
Choi (2005)	Health care	Chronological age	No difference in the age groups.
Evanschitzky and Wunderlich (2006)	DIY retail customers	Chronological age	Loyalty is stronger in an older customer group.
Kumar and Lim (2008)	Mobile service	Chronological age	Differences in specific value perceptions.
Kuppelwieser and Sarstedt (2014)	Restaurant and grocery	Future time perspective (FTP)	FTP influences the satisfaction-loyalty relationship by driving customer loyalty and moderating the relationship between these two constructs.
Sinha and Wang (2013)	Various services	FTP	Convergent results of impulsive behaviors, both in terms of engagement in unplanned behaviors and in terms of preference for short-term gratification across self-reported intentions as well as actual behaviors.
Thakor, Suri, and Saleh (2008)	Various services	Chronological age	Presence of older consumers affects young adults’ attitude to a service and their patronage intentions.
Tarasi et al. (2013)	Banking	Chronological age	Older customers generate stable cash flows.
Wei, Donthu, and Bernhardt (2013)	Various services	Cognitive age	Time manipulation drives and predicts consumers’ product evaluations. Cognitive age interacts with time manipulation as a meaningful moderator/predictor, while a dispositional time view interacts with time manipulation to influence consumer evaluations.

In summary, researchers posit that all age constructs measure the influence that a set of different individual perceptions has on a specific respondent (Kuppelwieser 2016). If this is indeed the case, we need to explore which construct delivers the best insights and in which kind of service context. This article challenges the use of chronological age as an age measurement for service and consumer behavior research (Bytheway 2005). Since chronological age is a fundamental requirement for scale measuring and adoption, this research agrees with Liao and Carstensen’s (2018, p. 163) statement that it is a key unit of measurement and that its limitations were acknowledged in the past, but that “it is curious that developmental science…has attended so little to the [influence of] subjective sense of time on individual functioning.”

This is therefore the first study to clarify disparate prior research by comparing the three main age concepts used in research and practice, and which, as suggested by Barak and Gould (1985), provides quantitative data to identify the construct that works best. By highlighting the differences between the constructs, we provide quantitative evidence of previous age-related (non)findings in scholarly papers and of clear recommendations regarding the age construct that should be chosen in a specific context, thus advancing age-related service research (Grougiou and Pettigrew 2011). By focusing on different service settings, the study’s findings reveal that, as a measurement construct, chronological age does not provide insights that allow researchers to explore the drivers of consumer behavior in services. In addition, the study shows that cognitive age provides limited explanatory power. FTP explains and predicts consumer behavior (brand switching, word-of-mouth (WoM), and complaint behavior) and value perceptions (utilitarian, hedonic, and social) more precisely than the other two age constructs. We discuss the implications of this finding for service theory and practice.

In the following section, we discuss the three major age concepts, chronological age, cognitive age, and FTP.

Chronological Age

Traditionally, researchers use chronological age to compare a certain age cohort’s behavior with that of another age cohort. Researchers aim to identify aging’s influences to interpret individuals’ behavior or to extract variance from a variable or model. In line with previous works (e.g., Guo et al. 2013; Mittal and Kamakura 2001), our study defines chronological age as a linear count between the time a person is born and the current date.

Studies exploring the effects of age and age dispersion are found in service research (e.g., Grougiou and Pettigrew 2011), market research, organizational behavior research, psychology, marketing, and information systems. Managers use chronological age-based market research data in their segmentation strategies, contrasting older and younger customers’ attitudes, perceptions, and behaviors.

One of the most controversial notions about chronological age suggests that there is a (chronological) age-related decline in extrinsic and intrinsic motivation and in consumers’ subsequent behaviors. Although chronological aging is only a subprocess of the more general aging process, previous research has attempted to explain all possible changes that occur in people’s psychological, social, and even their societal functioning (Settersten and Mayer 1997). Recent research has found that individuals differ in their biological age, even if they are of the same chronological age, which impacts their physical and cognitive abilities, brain aging, and health. In summary, these findings highlight that chronological age cannot deliver reliable research data.

Cognitive Age

With researchers challenging chronological age’s usefulness, their focus turned toward cognitive age. In our study, cognitive age reflects individuals’ perceptions of how old they feel (Barak and Schiffman 1981; Sudbury and Simcock 2009b). Cognitive age elicits societal expectations and, subsequently, influences individuals’ behavior (Cleaver and Muller 2002). Cognitive age scales often use marker ages—that is, they segment people as being in their 20s, 30s, 40s, and so on—to measure the age people themselves feel, look, act, and share interests (Barak and Schiffman 1981). These age markers inherently assume that the role individuals ascribe to that age grouping is true, thus evoking these individuals’ expectations, beliefs, and assumptions about that specific age bracket (Kornadt, Meissner, and Rothermund 2016). Despite these challenges, researchers maintain that cognitive age provides a deeper insight into customer behavior than chronological age allows (Cleaver and Muller 2002; Stephens 1991). Researchers therefore use customers’ own understanding of their age and compare the different age brackets, motivations, and behavior.

In sum, cognitive age has a strong association with chronological age, either actively through individuals’ perceptions or passively through the social role society ascribes to an individual. While a more advanced concept than chronological age, cognitive age is considered insufficient to explore the way consumers perceive age and its influence on their behavior. This leads researchers to investigate other age constructs that also allow them to explore and measure consumers’ motivations.

Socioemotional Selectivity Theory (SST)—FTP

SST provides a theoretical explanation of individuals’ motivations across their lifetime (Carstensen 2006). As a life-span theory of motivation, SST explains a broader range of cognitive, emotional, and motivational variables than chronological age does (Carstensen 2006). SST is based on individuals’ ability to monitor time since they adjust their time horizons according to this perception and ultimately accept that time runs out. Consequently, changes in individuals’ motivations and behaviors are primarily due to changes in their time perspectives. In our study, we apply this perspective to explore age-related implications and identify different customer groups on the basis of their FTP.

Their FTP determines individuals’ perspective of time, which in turn explains changes in their social motives, their motivations, and their behavior. Consequently, people with limited time horizons focus on emotional rather than on functional aspects in their decision making (e.g., Carstensen 2006; Pour Mohammad and Drolet 2019). An FTP focuses on subjective time experiences and refers to how much time individuals believe they have left. An FTP is centered on the guiding question of “how long do you expect to live,” rather than emphasizing the time lived, articulated in “how old are you?” By inverting the perspective, an FTP can explain motivations, emotions, and behavior. Psychology (e.g., Carstensen 2006; Holman et al. 2016) and marketing studies (e.g., McKay-Nesbitt et al. 2011; Pyone and Isen 2011; Wei, Donthu, and Bernhardt 2013) suggest that younger and older consumers’ FTP describes differences in their behavior more accurately than their chronological age does.

Method

Study Overview and Sample

We carried out an empirical study to test for a recommended and applicable age measurement scale. Building on previous findings when contrasting chronological age and cognitive age (e.g., Schau, Gilly, and Wolfinbarger 2009), chronological age and an FTP (e.g., Grühn, Sharifian, and Chu 2016), and cognitive age and an FTP (e.g., Wei, Donthu, and Bernhardt 2013), the study focuses on individuals’ age perception across these three constructs. In the first step, the methodological routines focus on the discriminant effect that the two perceived and distinct age groups have on each scale. The second step emphasizes these differing age perceptions and identifies the recommended age construct in keeping with the research context.

The study therefore included the age constructs discussed above as well as the behavioral and perceptional variables. For this study, we collected responses from 1,096 respondents of whom 52.1% were female. The respondents stem from a U.S. panel and answered the survey online in November 2018.

To ensure the generalizability, our data covered each of Silvestro et al.’s (1992) major service types. The data set thus stems from professional service contexts (n = 398), the respondents’ main bank (n = 261), and from their main grocery retailer (n = 437). We also ensured that the sample was representative of the U.S. population in terms of gender (52.1% female), chronological age (median 40–44 years), and state.

The study relies on Bayesian inference because this approach has several advantages over the standard framework of frequentist significance testing. In respect of our study, this approach specifically allows for quantifying uncertainty about effect sizes (Wagenmakers et al. 2018), obtaining evidence in favor of statistically insignificant differences (Dienes 2014). The approach also enables a comparison of the predictive adequacy of competing (statistical) models (Dienes and McIatchie 2018).

Given the diverging results of age-related scales’ previous applications, this article compares the three mostly used age constructs. In doing so, this study contrasts the scales and reveals the need for each scale’s careful application (e.g., Kuppelwieser 2016). Previous research mainly focused on the scale’s specific domain and tried to identify the best measurement model within this domain (e.g., Kozik et al. 2019). This paper expands and extends our knowledge of age-related measurements and their applicability. Since previous studies defined age-related scales’ aging dimensions inconsistently, we tested the adopted scales’ dimensionality and checked their formal distributional assumptions in our preliminary analyses. Step 1 focuses on the difference in the scales in terms of the age groups. This step compares the focal constructs’ age groups and tests for the presence or absence of an effect determining their size. Step 2 assesses which age construct explains the data better than the alternative constructs. We compare the three age constructs in six contexts and examine the scales’ suitability. Our example demonstrates the impact of choosing an age construct unthinkingly. This step reminds practitioners and researchers to select an appropriate age construct carefully. Figure 1 shows the methodological steps and their results.

Figure 1.

Methodological steps.

Measurement

We adapted all of this study’s items from established measures and adjusted them to fit this study’s context. At the beginning of the questionnaire, we asked the participants to indicate their year of birth and asked them to fill in the age construct scales. This study relies on Lang and Carstensen’s (2002) scale, which many previous studies also use, to measure an FTP. Cognitive age’s measurement relied on Barak and Schiffman’s (1981) and Barak’s (1987) scale to measure subjective age. In keeping with Barak (1987), this study refers to youth age as the concept; that is, the difference between chronological and cognitive age. Appendix Table B1 presents the measurement scales for each age construct.

Thereafter, the participants indicated their last service experience with either their professional service, such as an attorney or a consultant, their main bank, or their main grocery store. If they had not experienced any of these services within the last 3 months, they were excluded from the sample. Subsequently, the participants indicated the name of their specific service provider, which made them remember their recent experience.

The participants then completed the self-reported behavioral and perceptional constructs, which also stem from established scales and were adjusted to fit the study’s context. To measure brand switching behavior, this study adopted Romani, Grappi, and Dalli’s (2012) four-item subset of Bougie, Pieters, and Zeelenberg’s (2003) measurement scale. The measurement of word-of-mouth behavior relied on Harrison-Walker’s (2001) six-item scale, while Bougie, Pieters, and Zeelenberg’s (2003) four-item scale measured complaint behavior. Utilitarian value perception was gauged by adapting relevant items from Sweeney and Soutar’s (2001), Jones, Reynolds, and Arnold’s (2006), and Anderson and Srinivasan’s (2003) scales. Hedonic value perception was measured by using Sweeney and Soutar’s (2001), Soutar and Sweeney’s (2003), and Jones, Reynolds, and Arnold’s (2006) scales. Social value perception was assessed by using Sweeney and Soutar’s (2001) and Rintamäki et al.’s (2006) items. Value perceptions are a key construct in service research (e.g., Ostrom et al. 2015) and are regarded as a crucial outcome variable to explain service experiences (e.g., De Keyser et al. 2020). Appendix Table A1 provides an overview of the items used in this study.

Preliminary Methodological Routines

We started our methodological considerations with a parallel analysis (PA; Horn 1965) of each of the scales individually. A PA formally tests the probability that a factor is due to chance and identifies the number of factors prior to performing an exploratory factor analysis. Based on a Monte Carlo simulation on randomly generated data that match the sample size and the number of scale items in the original data set, PA generates a 95 percentile cutoff (Glorfeld 1995). This result allows researchers to confidently specify the number of factors to extract in a subsequent analysis (Timmerman and Lorenzo-Seva 2011). Our PA created 5,000 permutations of the raw data by using a common factor analysis approach. Based on a Monte Carlo simulation, this routine generates eigenvalues from the raw data, mean eigenvalues, and eigenvalues representing the 95th percentile. In a comparison of the original eigenvalues with the 95th percentile eigenvalues (Glorfeld 1995), the PA suggested a one-component solution for all the scales.

Follow-up factor analyses with maximum likelihood extraction and an oblique rotation (promax, κ = 4) specified the solutions and indicated no item loading below the common threshold of .6. Appendix Table A1 shows the items and their loadings.

Before executing the analysis, it is important to confirm that the intended analyses are appropriate, and the models are not incorrectly specified for the data. We plotted the data of each scale to examine the model assumptions’ validity. Figure 2 includes exemplary histograms and density estimates of WoM, which are split into chronological age and the FTP conditions.

Figure 2.

Exemplary histograms of WOM.

The histograms suggest that the data are not approximately normally distributed. Additional Shapiro-Wilk tests of the samples indicated that the data deviate significantly (p < .001) from a normal distribution in respect of all the dependent variables. We thus assess the result’s robustness by conducting the Mann-Whitney test as an equivalent of the t test (Mann and Whitney 1947).

Step 1: Do the Age Constructs Behave Differently?

The analysis follows two main steps. In the first step, we aim to ascertain the presence or absence of the effect that the specific age construct has on behavior or perception (hypothesis testing). If the effect is present, we determine its size in a second step (parameter estimation). The analyses were conducted in JASP (JASP Team 2020), which allows for Bayesian testing procedures (e.g., Marsman and Wagenmakers 2017).

Next, we specify the statistical model (Etz et al. 2018). The study’s focus is on behavior or perception’s difference between two between subjects’ conditions, suggesting a two-sample t test. As in similar studies, we focus on the differences between two perceived age groups on each scale. As in previous studies, we split the chronological age data based on the median (38 years, see U.S. Census Bureau 2018) of the whole population (e.g., Lambert-Pandraud and Laurent 2010), the cognitive age data on the mean of the scale, indicating cognitively older and younger customers (e.g., Chaouali and Souiden 2018), and the FTP data on the rescaled construct scores (T scores; see, e.g., Lang and Carstensen 2002) because the differences in their age-related goals are more pronounced (e.g., Allemand 2008). Specifically, because our data were already deemed not normally distributed and we only have a little knowledge of the prior distribution, the standardized effect size was assigned a Cauchy prior distribution with spread $r = \frac{1}{\sqrt{2}} ≍ .707$ . Since the credible intervals treat their bounds as fixed and the estimated parameters as random, the uninformative Cauchy prior distribution avoids the influence of situation-specific assumptions. This prior distribution therefore provides answers that are only inspired by the likelihood function’s shape instead of being induced by a random opinion about the unknown parameters’ range. In other words, because the literature does not provide a clear indication, we applied equal weights to both models.

For hypothesis testing (whether an effect is present or absent), we focus on the Bayes factor because it quantifies the relative predictive performance of two rival models (Wagenmakers et al. 2018). Figure 3 presents an example of the two-sided t tests’ WoM results, Appendix C1 provides a complete overview.

Figure 3.

Exemplary Bayesian two-sample t tests for the parameter δ. Note. Dependent variable is WoM behavior, Mann-Whitney tests with 1,000 samples, H1:δ∼Cauchy represents the prior distribution indicated by a dashed line. The solid line indicates the posterior distribution. The graphical output also shows the posterior median, the central 95% credible interval, and the Bayes factors BF10 and BF01.

We also assigned a range of stretched prior widths and tested their effect on the Bayes factor values (Quintana and Williams 2018). This test clearly suggested that the prior width did not influence the age-related differences in the FTP. In respect of chronological age and cognitive age, the groups’ differences interacted with the dependent construct. For example, the chronological age example’s calculations (Figure 3) suggest that the two age groups’ WoM differs. Overall, the results indicate a clear preference for the FTP construct. The methodological routines suggest that there is very strong evidence of all the outcome variables differing significantly (min BF₁₀ = 6.76). The data are minimally 49.6 times more likely to differ between the FTP groups (brand switching behavior, the remaining variables are higher). An additional robustness check assesses the sensitivity of the prior distribution, with an ultrawide prior yielding a min BF₁₀ = 26.728.

In contrast, in respect of chronological age groups’ brand switching behavior, the Bayes factor only provides anecdotal evidence of the significant difference between the age groups (BF₁₀ = 2.28) with a posterior median of .18 and a 95% credible interval¹ range of .04 to .31. The data are therefore approximately 2.3 times more likely to differ in the old age-group than in the young one. The additional robustness check with an ultrawide prior distribution yields a BF₁₀ = 1.979. Although the results indicate a difference between the chronological age groups, substantial uncertainty about their size remains, with values close to 0 having the same posterior density as those close to 1.

Step 2: Which age Construct Should I Choose?

This second analysis focuses on the model selection criteria. In this part, we assess which age construct explains the data better than the alternative models do. Specifically, we compare chronological age to youth age and to FTP. Since these models are nonnested, they cannot be readily compared with each other by means of commonly used fit indices. However, the models share the same prior probabilities (Cauchy distribution with spread $r = \frac{1}{\sqrt{2}} ≍ .707$ ); we therefore compare the logarithm of each model’s Bayes factor. The highest model logarithm is the best model of these pairwise comparisons.

To identify the best fitting model, we report the logarithm of each model in Table 2.

Table 2.

Logarithmic Bayes Factor (BF) for Each Age Model.^a

Dependent Variable	Chronological Age		Cognitive Age		Future Time Perspective
Dependent Variable	BF₁₀	ln(BF₁₀)	BF₁₀	ln(BF₁₀)	BF₁₀	ln(BF₁₀)
Brand switching behavior	2.279	0.824	298.545	5.699	6.758	1.911
WoM behavior	0.334	−1.097	0.105	−2.254	1.976e⁶	14.497
Complaint behavior	17.459	2.860	0.083	−2.489	985.029	6.893
Utilitarian value	6.677	1.899	0.101	−2.293	1,006.322	6.914
Hedonic value	19.140	2.952	0.081	−2.513	1.248e⁵	11.735
Social value	0.131	−2.033	1.030	0.030	4.589e⁴	10.734

Note. Best models are given in bold.

^a All t tests are Mann-Whitney tests with 1,000 samples.

Since the analyses alter the age construct, the results allow for identifying a pattern. With the exception of the switching behavior model, the calculations consistently identify an FTP as the best performing age construct. More specifically, an FTP construct exhibits a substantially better logarithmic value than the chronological age, Δ ln(BF₁₀) = (1.087; 15.593), and the cognitive age models, Δ ln(BF₁₀) = (3.789; 16.750). In respect of the switching behavior model, the comparison favors cognitive age, ln(BF₁₀) = 5.699. More importantly, with the exception of switching behavior, these differences in the logarithmic Bayes factor constantly exceed 4 in respect of all age constructs, which provides very strong evidence of an FTP construct’s superior fit.

The cognitive age construct also outmatches chronological age in two of our six models. Cognitive age only outperforms switching behavior, Δ ln(BF₁₀) = 4.875, and social value, Δ ln(BF₁₀) = 2.062.

An Example

The results have important implications for research and practice. Choosing the age construct rashly could cause researchers and practitioners to misinterpret their results’ true value (Kuppelwieser 2016; Kuppelwieser et al. 2014). A convenient example demonstrates this impact and highlights its consequences.

In this example, we are interested in age groups’ different WoM behavior. Comparing the data of our two chronological age groups in the sample along this variable clearly suggests that the younger group (n = 366) and the older group (n = 515) behave similarly (Figure 3, BF₁₀ = .334). Detailing the groups’ FTP reveals that the older group and the younger group include respondents with limited and open-ended FTPs. The younger group consists of 142 respondents with a limited FTP and 224 respondents with an open-ended FTP, with the data suggesting distinct WoM behavior (BF₁₀ = 18,417). The older group includes 269 respondents with a limited FTP and 246 respondents with an open-ended FTP, with the Bayesian calculations also suggesting distinct behavior (BF₁₀ = 193,279). Our previous calculations (Figure 3) confirm the two FTP groups’ different behavior (BF₁₀ = 1.98e⁶).

Although the chronological age groups’ WoM behavior was similar, that of the FTP groups differed. Only 224 of the 366 (61.2%) younger customers are in the open-ended FTP group, and only 269 of the 515 (52.23%) older customers are in the limited FTP group. Consequently, if their chronological age is applied, 388 of 881 (44.04%) customers are misspecified. Table 3 provides details of this example.

Table 3.

Sample Composition.^a

Sample Group	Chronological Age	Limited FTP		Open-Ended FTP		BF_{10, U}	Misspecified
Sample Group	Chronological Age		%		%	BF_{10, U}		%
Young	366	142	38.8	224	61.2	18,417	142	38.8
Old	515	269	52.2	246	47.8	193,279	246	47.8
Total	881	411	46.7	470	53.3	1.98e⁶	388	44.0

Note. FTP = Future time perspective; BF = Bayes factor.

^a Individual comparisons are based on the default t test with an uninformative Cauchy (0, r = 1/sqrt(2)) prior. The “U” in the Bayes factor denotes that it is uncorrected.

Discussion

This article does not seek to challenge the core notions in age-related research but explores which age concepts best explain age-related perceptions and behaviors in which specific service context. This exploration is necessary because although current literature has used different concepts of individuals’ time perception, relatively little scholarly inquiry has been aimed at connecting the dots and gaining a complete understanding of recent findings that could be useful to explore the best way to apply different age concepts in market research (cf. Kuppelwieser 2016). In the following paragraphs, this article not only provides a new perspective on these recent findings but also offers guidance on how to interpret them in a service context. Service researchers traditionally use chronological age as a control variable. Our study highlights that an FTP would be a more meaningful control variable to use.

Recognizing the deficiency of one-dimensional perspectives on aging, the literature proposes several multidimensional models (see, e.g., Settersten and Mayer [1997] for an overview). These multidimensional models either explain behavior at a given stage in life by means of earlier life experiences—offering only a few insights into studying them—or by means of the dynamic interplay between individuals and their environment, which is constantly changing.

Compared to customers’ behavior according to their chronological age and due to various fields’ recent analyses and emerging new models, our current understanding of age is far more developed than before. Conversely, researchers often disregard what exactly age indicates, often not even specifying the mechanisms through which age plays a role in their models. In addition, researchers should think more critically about whether the impacts they hope to index by using age could be measured more directly. They should, to begin with, also critically reflect on the conditions under which age is a relevant dimension (Kuppelwieser and Sarstedt 2014).

This study reveals the specific time perspective construct that market researchers and practitioners can use to obtain meaningful results. Using chronological age as a starting point, this article finds that researchers and practitioners should use it as a covariate when they assume that age explains a significant portion of the variance in their dependent variable(s). Cognitive age, on the other hand, explains some behaviors. Although cognitive age does not express individuals’ motivation, it can be used to build behavioral clusters, for example, when defining customer segments. An FTP builds on the foundation of individuals’ motivation. An FTP can therefore be recommended when a researcher’s project or managerial application aims to investigate and disclose age-related motivational and behavioral differences (see Table 4).

Table 4.

Recommendations and Central References.

Aims of the Study	Purpose of the Age Dispersion	Recommendation	References
Examine a possible age dispersion in the model	To identify the influences of age, general comments about different age groups’ behavior, and to extract variance from a variable or model	Chronological age as a control variable	Guo et al. (2013) Lambert-Pandraud and Laurent (2010) V. Mittal and Kamakura (2001)
Study focuses on cognition and affect	To compare customers’ different age brackets and to use customers’ own understanding of their age	Cognitive age	Barak and Schiffman (1981) Cleaver and Muller (2002) Sudbury and Simcock (2009b)
Study focuses on behavior	To identify different customer groups and to find age-related implications	SST/FTP	Carstensen (1992, 2006) Carstensen et al. (1999) Kuppelwieser and Sarstedt (2014) Lang and Carstensen (2002) Liao and Carstensen (2018) This article

Note. SST = Socioemotional selectivity theory; FTP = future time perspective.

Theoretical Implications

The aim of this study was to identify the best context-fitting age construct. Our analyses’ results indicate an FTP’s superior value regarding predicting age differences in customer behavior.

Our study emphasizes that the traditional way of measuring the age construct often does not mirror how older consumers perceive themselves. Today, more than ever before in history, older consumers redefine their identity. This indicates that older consumers are more likely to reevaluate their brand and service preferences, realigning themselves with services that more consistent with their sense of self. They may prioritize service experiences’ quality more and reassess their value perceptions. These shifts need to be detected and implemented in a service provision to ensure competitiveness.

Importantly, the first results demonstrate that chronological age is not discriminant in all contexts. Recent research emphasizes that individuals differ in their biological age, even if they are of the same chronological age. The aging process consists of several dimensions, of which chronological aging is only one (e.g., Carstensen et al. 1999; Drolet et al. 2019). The age concept is therefore an individual perception and loses its stability assumption (Grougiou and Pettigrew 2011).

When should a specific construct be applied?

Our findings broadly underline the mentioned theoretical propositions. Since chronological age is not discriminant, service researchers are misled to compare their studies’ results to the participants’ age. For example, while some studies confirm age’s impact (e.g., Lambert-Pandraud, Laurent, and Lapersonne 2005; V. Mittal and Kamakura 2001), others do not find any significant age influences (e.g., Cooil et al. 2007). Consequently, previous findings regarding chronological age’s role in customers’ behavior are ambiguous. While our findings confirm the differences in the chronological age groups in general, the size of the effect is often marginal.

Some scholars identify chronological age as a proxy for age-related processes. Our analyses suggest that chronological age might identify value perceptions but loses its predictive power in behavioral contexts. For example, while chronological age compares the groups in a hedonic value setting reasonably well (BF₁₀ = 19.14, Figure 3), it completely loses its predictive power in WoM behavior (BF₁₀ = 0.33, Figure 3). In other words, in a hedonic setting, the discriminating hypothesis (old vs. young) is approximately 19 times more likely to occur than the nondiscriminating null hypothesis (no difference). In the behavioral setting, chronological age does not discriminate; a statistically significant difference is only .33 times more likely to occur than an insignificant difference.

The study’s second results underline the previous results. Chronological age consistently seems to be the inferior or the least stable age construct. Our results build a clear indication of a need to “stop accepting chronological age as a factor” (Griffiths 1997, p. 208). Indeed, chronologically older individuals outperform younger ones at times and vice versa (Drolet et al. 2019)—a finding that our results support. More specifically, older adults use their deliberative capacity selectively and show, for example, an increased preference for satisficing or “good enough” choices. However, their accumulated experience might be a substitute for their lack of sufficient deliberative capacity due to their age-related decline (Drolet et al. 2019).

Why is one age construct superior to another?

As our findings and the example clearly support, researchers might not detect customers’ underlying motivations when only applying chronological age. Consequently, there is solid support for transferring service marketing to promise management (Grönroos 2020). If promises need to be made and successfully kept, promise-making, promise-keeping, and promise-enabling all need to adapt to the way customers change in an FTP. For example, promises need to become more functional for customers with an extended FTP. Likewise, an emotional address and response might be the key in promise-keeping to engage limited FTP individuals and to support their cocreational behavior. Promise-enabling should allow dyadic value cocreating parties to emphasize their orientation either functionally or emotionally. From a promise theory perspective, our results question previous age-related works that identified or neglected age differences as well as their application of age constructs. Keeping a promise is therefore not the automated process that traditional service marketing models assume it is; instead, it is a customer-focused experience that needs to be adapted to customers’ expectations in order to satisfy and engage them.

Cognitive age stresses the individual’s perception of how old they feel. Sudbury and Simcock (2009a) posit that these feelings induce behavior and value perceptions, but our results challenge this notion. We only found a strong effect on brand switching behavior, which suggests that thinking of oneself as being younger (older) converges with preferring new (or established) brands. Gwinner and Stephens (2001) maintain that cognitive age captures older people’s lifestyles and predicts their choices more effectively than other variables. From an SST perspective, people with a limited time horizon emphasize feelings and emotions more; consequently, their interest in new information will subsequently decline. Such people prioritize close and well-known contacts rather than new, informative ones (Isaacowitz et al. 2000). Their cognitive decline limits their consideration set to known or previously owned brands. Since they are more committed to preserving the status quo, they do not embrace new and exciting experiences because they do not want to compromise their positive state (Kuppelwieser and Sarstedt 2014). Novelties such as new brands could potentially harm their self-concept, creating negative emotions. People with a limited time perspective need solid evidence to either support or refute their blind judgment (McKay-Nesbitt et al. 2011). Our findings underline this chain of effects and contribute, for example, to the open questions regarding emotions’ influence on customer (dis)satisfaction and (dis)loyalty (Furrer et al. 2020).

Our results indicate that cognitive age is a suitable measurement for social contexts. Cognitive age is, however, a poor predictor of utilitarian and hedonic value perceptions. The lack of model fit stems from cognitive age’s relation to chronological age. Researchers and practitioners should thus refrain from applying this concept in customers’ hedonic and utilitarian value perception measurement. In social value, cognitive age performs better than chronological age.

Our findings show that the FTP performs well in terms of predicting age-related consumer behavior. An FTP is discriminant in all focused contexts except brand switching, therefore allowing pronounced research effects to be identified. Differences between younger and older consumers’ behavior can be more accurately described by their FTP than by their cognitive or chronological age (Kuppelwieser 2016). That said, discussions on individuals’ behavior based on their chronological or cognitive age seem spurious. For example, reflections on younger generations (Generation X, Y, or Z) and their specific behavior seem obsolete when comparing them with other age groups. These generation clusters build on the individuals’ birthdate, a chronological age measure that, considering our results, allows only limited predictions of their behavior.

In addition, the FTP scale’s strength is that it is embedded in the SST and is associated with a number of individual traits (e.g., Jiang et al. 2016; Kozik, Hoppmann, and Gerstorf 2015) across different samples and settings.

The discrepancy between certain older consumers’ biological age and their perceived age is associated with the opportunity to create a new life narrative (Schau, Gilly, and Wolfinbarger 2009). Their subsequent changing lifestyles, renewed energy, and sense of self will translate into an increasing demand for a range of services not normally associated with the older consumer segment but those more closely mirroring their desire and their social context. Using the correct age measurement will allow service providers to cater better to these needs and demands than merely relying on demographic insights that might not be accurate. An FTP underlines an emotional or functional consumer focus’s dominance. A limited time perspective challenges the prevalent age-alignment perspective in services, matching consumer and service staff based upon their chronological age. Applying an FTP instead indicates that staff and consumers should be matched according to their emotional or functional perspective, not their chronological age.

Likewise, service innovation depends strongly on customers’ perception and adoption. Our findings underline the need to develop new or adopt existing services along the SST dimensions. The different age groups’ distinct behavior might partly explain their resistance to innovative services as well as their adoption and drivers of customer innovation. The development of service robots might benefit from our findings. While it appears very obvious that health care robots need to provide emotional support, learning robots’ functional component should also be emphasized. In the light of our findings, we refrain from discussing older customers’ motivation for adopting e-services or how they perceive new smart services. It is not age but individuals’ FTP that strongly influences their perceptions of, engagement with, and adoption of service robots.

From a more theoretical perspective, our findings also emphasize the need to identify customers’ time orientation when focusing on cocreation processes. Despite cocreation processes’ underlying intention to increase well-being, customers who lack motivations might result in such processes having negative outcomes. Value will subsequently either be created on a low level, or will not be created at all, a situation service-dominant organizations need to avoid. If customers are regarded as resources, proper resource integration—confirming their functional or emotional orientations—will influence value creation, customers’ emotional needs, and their satisfaction.

In conclusion, aging societies are becoming a reality in many countries. Research does not keep up with societies’ and companies’ pressing questions on how to deal with aging consumers when marketing their services (e.g., Drolet et al. 2019). Measuring age, contrasting age groups’ behavior, and focusing on a specific age-group are not solely about measuring a construct of interest. While our study highlights an FTP’s performance, researchers try to measure the impact that some of the respondents’ individual perceptions, decisions, and behavior by means of a simple demographic variable, namely chronological age. Our study’s results clearly indicate that being less rigid about the age measurement concept will enhance marketing theory.

Managerial Implications

Researchers suggest that it is important that service marketers not only understand older consumers’ cognitive or subjective age but also how their interaction with service firms may influence their age perception. Service managers need to create awareness of misconceptions and age-based stereotypes that may be culturally entrenched in staff as well as need to influence how staff engages with older consumers. Failure to do so, for example, by relying on practices based on chronological age, could, as our data and example indicate, lead to differences in spreading WoM.

Our analysis reveals that WoM behavior is becoming increasingly important in respect of the older consumer segment, as this segment is increasingly participating in social media activities, and therefore of importance for service organizations. If a service organization instead chooses to measure, for example, chronological age to determine its communication effect, it might be misled. This could in turn lead to poor resource allocation and, ultimately, to customers spreading positive or negative WoM.

As with any measurement, including age, a key component of market research is that managers need to make the right choice regarding how age should be measured in order to support their efforts in the best possible way. If service managers and market researchers choose a measurement unwisely, the information provided could decrease the results’ reliability and validity significantly and, ultimately, their management efforts’ chances of success. The analyses of this study offer service researchers and sales managers the following general guidelines for using age as a measurement.

First, one needs to pay close attention to the purpose of age in respect of the insights that one could gain. Even though chronological age is easy to measure and to collect, managers should not be careless and use it instead of age diversion. Since chronological age is an adequate proxy for other age concepts, it might already moderate the results one is interested in. Chronological age calculations are easy because the data are readily available. However, the interpretation of the results is, in most cases, ambiguous, which leads to (theoretically) questionable conclusions. Therefore, if a service provider wants to cluster its customers roughly, chronological age might be an applicable construct. If the same store wants to specialize in a specific customer segment, another age concept might be more advisable in order to maximize the marketing and business purposes’ output. However, the reasons for choosing and applying any age concept need to be carefully articulated.

Second, careful thought should be given to the respondents’ age characteristics. If researchers want to focus on age-related models and constructs, a good knowledge is required of the sample’s age distribution. Earlier research already tried to link consumer responses and aging, which resulted in ambiguous results. Consequently, contemporary models need to explicitly measure what they intend to measure and choose an adequate age model. Depending on the sample’s characteristics, different approaches to age could influence the results and lead to different outcomes.

Third, greater use should be made of the available age concepts, especially when assessing the differences between certain age groups, although all of them have their merits and downsides. While managers are used to running surveys on their customers’ satisfaction or loyalty, adapted age measurement is far less known. Following our results, we recommend asking customers about their FTP to gain additional insights into them and to allow managers to adapt their tools.

Fourth, managers should pay particular attention to which outcome they wish to measure. Our findings indicate that it is important to choose the age measurement according to the outcome that managers wish to explore. For example, our study highlights that while one age measurement might be suitable for investigating how customers perceive value, the same measurement should not be used to measure the behavior managers want to stimulate. However, sophisticated concepts inherently provide scope for developing theoretical and methodological approaches. In addition, the three major age concepts presented in this article provide a good opportunity to refine and strengthen theory and its implications for management.

Limitations and Future Research

Although this research comprises several different contexts, it has limitations. First, the research context is an important consideration when selecting the age construct. Even our FTP construct is prone to contextual effects and relies on data selection (Kozik et al. 2019). Since our data stems from a representative U.S. sample, different locations, cultural backgrounds, and focused variables could influence the results. Our results should therefore be broadly replicated in different contexts.

Second, this study only focuses on the current three major age constructs used in past research. Several other measurement scales, such as Zimbardo and Boyd’s (1999) time perspective inventory,² were developed in the past but were only applied to a limited degree in marketing. Future research should thus investigate these scales, testing and contrasting their usefulness for measuring consumers’ age-related behavior. In addition, most of the previously developed age constructs only focused on limited aging aspects, for example, cognitive age as dominantly affective. An FTP, on the other hand, comprises cognitive, emotional, and motivational aspects (e.g., Lang and Carstensen 2002). As a complex process, aging is inherently multidimensional. Individuals as biological, social, psychological, and spiritual beings age in several dimensions (e.g., Moschis 1994). Our article highlights only three alternative ways of considering age in research and practice, although it is still unclear how the relevant constructs relate to life’s changes during the aging process. We encourage service researchers to explore this dynamic in more detail.

Third, just like prior studies, our study examined the age constructs as a single measure with multiple items indicating different dimensions. Although previous publications adopted this approach, recent research using the FTP scale reports multiple distinct dimensions (Kozik et al. 2019). Further research should therefore extract these dimensions and individually test them to explain age-related perceptions, decisions, and behavior. Such an endeavor would also allow for contrasting these dimensions and identifying their specific importance.

Similarly, although our study indicates that in most of the contexts on which we focused FTP outperforms the other two age constructs, the scale might not cover the whole aging process (e.g., Drolet et al. 2019). Combining previously suggested age scales, or including additional variables, would provide a broader understanding of aging and behaviors.

Fourth, the study’s focus is on consumers’ age perceptions, and we acknowledge that future research should explore service personnel’s perceptions to test the findings’ generalizability. We encourage researchers to explore the service contexts in which service interactions should be designed and aligned with the FTP, instead of relying on cognitive and chronological age.

Fifth, our results show that each age-group defines, experiences, and creates value differently. Adopting recent research on value cocreation, our study could not connect cocreational efforts with value perceptions. Further research should thus investigate these relationships and emphasize the different age groups’ different needs.

Finally, to avoid situation-specific assumptions, we assumed an uninformative Cauchy prior distribution in our Bayesian calculations. Prior distributional assumptions depend on the knowledge of the relationship under examination. In other words, assuming that the age construct and/or the dependent variable will behave in a certain way would change the prior distribution and, subsequently, alter the results. Although this seems unlikely in our case due to the data’s nonnormal distribution, further research should examine the prior distributions’ different assumptions and test the results’ stability.

In sum, there seem to be limitless research avenues for identifying differences in aging; Table 5 provides just a few examples of when specific age constructs can alter underlying assumptions, theoretical models, and managerial applications.

Table 5.

Exemplary Research Questions.

Domain	Research Need
General	Do established relationships between constructs still hold when considering specific age constructs? Do our service models still work when focusing on specific age groups and age constructs? What happens to our models if we assume that consumers become older? How does the changing motivation during a life span impact our models? Do hypotheses work for all consumers or do they favor specific age groups and motivational backgrounds? How does aging relate to well-being?
Service provision	Does the servicescape represent customers’ expectations when we take age-related changes into account? If not, what needs to change and how does it need to change? How does declining functional motivation influence the servicescape? How does service provision need to change when customers become older? How do service design and delivery decisions advantage or disadvantage older (younger) customers?
Service experience	How does service experience change during a life span? What and where do we need to adapt? How do customers engage with brands and services when they become older? Which dimensions of the customer experience change during different life stages? How do they change? What influence does aging have on requested cocreational efforts? Given the FTP, do customers experience the same value? Do we intend to provide the same value to each customer, regardless of their age? If not, which value do we want to provide to customers of various ages? What are the differences between these values?
Methodological	Apply informative priors, for example, adopted from a meta-analysis Reexamine well-known relationships in the light of SST Replicate this study’s findings in different contexts

Note. SST = Socioemotional selectivity theory; FTP = future time perspective.

Now that we have established the FTP as a highly precise construct when considering age in theory and practice, additional and finer granulated research is warranted to refine our findings. The main theories highlighted in this article reflect the aging process’s main dimensions. However, this article also clarifies that examining age-related behavior and perceptions differs with regard to these main dimensions and, subsequently, to the age construct. Researchers therefore need to have a clear understanding of which dimension is suitable for their research context. Further, research needs to identify and carve out the age constructs’ underlying dimensions with regard to different customer decisions. Once identified, the dimensions need a thorough extension of their impact on customer behavior and how it changes over time. Furthermore, a decreasing FTP implies an individual’s greater emotional focus. The customer’s emotional need must be reflected in the service provision and in service employees’ behavior. Further research should therefore investigate these relationships and identify different needs and typologies regarding time perspectives and customers’ emotional needs. While the challenges of taking the elderly into account in research and practice are huge, they also provide opportunities.

Supplemental Material

Supplemental Material, 200921_Web_Appendix - Revisiting the Age Construct: Implications for Service Research

Supplemental Material, 200921_Web_Appendix for Revisiting the Age Construct: Implications for Service Research by Volker G. Kuppelwieser and Philipp Klaus in Journal of Service Research

Supplemental Material

Supplemental Material, 201001a_Executive_Summary_Revisting_the_Age_Construct_Implications_for_Service_Research_FINAL - Revisiting the Age Construct: Implications for Service Research

Supplemental Material, 201001a_Executive_Summary_Revisting_the_Age_Construct_Implications_for_Service_Research_FINAL for Revisiting the Age Construct: Implications for Service Research by Volker G. Kuppelwieser and Philipp Klaus in Journal of Service Research

Footnotes

Appendix A

Table A1.

Measurement Items and Loadings.

Item	Standardized Loading
Brand switching behavior (α = .90)
I have acquired the services of COMPANY less frequently than before.	.847
I have switched to a competitor of COMPANY.	.880
I will not acquire services of COMPANY anymore in the future.	.862
I intend to switch to a competitor of COMPANY in the future.	.862
Word-of-mouth behavior (α = .92)
I mention COMPANY to others quite frequently.	.838
I’ve told more people about COMPANY than I’ve told about most other organizations.	.849
I seldom miss an opportunity to tell others about COMPANY.	.824
When I tell others about COMPANY, I tend to talk about the organization in great detail.	.845
I have only good things to say about COMPANY.	.703
I am proud to tell others that I use COMPANY.	.778
Complaint behavior (α = .86)
Complain to the service provider about the service quality.	.890
Ask the service provider to take care of the problem.	.656
Complain to the service provider about the way I was treated.	.871
Discuss the problem with the service provider.	.665
Hedonic value (α = .95)
Compared to other (service providers), the time spent at COMPANY is enjoyable.	.842
I chose COMPANY not because I had to but because I wanted to.	.752
I feel that I made a smart decision by choosing COMPANY.	.869
I am happy with COMPANY as my CATEGORY.	.878
I feel relaxed when dealing with COMPANY.	.883
The personnel at COMPANY give me positive feelings.	.810
In general, I feel at ease with COMPANY.	.885
At COMPANY I feel really appreciated.	.812
Utilitarian value (α = .93)
COMPANY provides the service I am looking for.	.847
I feel COMPANY is convenient.	.710
I feel COMPANY is efficient.	.851
With COMPANY I accomplish my goal(s).	.802
COMPANY provides the value I am looking for.	.838
By engaging with COMPANY in the process I accomplish my goals.	.797
I feel I really can count on COMPANY when it matters most.	.862
Social value (α = .95)
COMPANY is very well considered at a social level.	.795
The fact that I am a customer of COMPANY looks good to the people I know.	.879
Being a customer of COMPANY makes me feel accepted.	.826
Being a customer of COMPANY improves the way I am perceived.	.904
Being a customer at COMPANY makes a good impression on other people.	.882
Being a customer at COMPANY gives me social approval.	.902

Note. N = 1,096.

Appendix B

Table B1.

Age Measurement Scales.

Chronological age
Scale	What is your year of birth?
Scoring	Respondents type in their birthyear. Chronological age is computed through simple subtracting birth year from measurement year.
Cognitive age ^a
Instructions	We all have a chronological age, but this age does not necessarily reflect the way and age we feel, look, behave, and so forth. Please specify which of these age decades you FEEL you really belong to: preteens, teens, 20s, 30s, 40s, 50s, 60s, 70s, or 80s.
Scale	I FEEL as though I am in my ________
	I LOOK as though I am in my ________
	I DO most things as though I were in my________
	My INTERESTS are mostly those of a person in their ________
Scoring	Feel/age, look/age, do/age, and interest/age are independently scored through the midpoints of each age decade (preteens, teens, 20s, 30s, etc.) a respondent identifies with. To illustrate, an individual who responds to the feel/age statement “I FEEL as though I am in my…” with 30s, receives a score of 35 years for feel/age. The cognitive age score is computed through a simple average of the four age dimension scores: Cognitive age = (feel/age + look/age + do/age + interest/age)/4. A nonresponse to any of the four age dimensions eliminates a respondent from the data analysis.
Future time perspective ^b
Instructions	Read each item and, as honestly as you can, answer the question: “How true is this of you?”
Scale	Many opportunities await me in the future.
	I expect that I will set many new goals in the future.
	My future is filled with possibilities.
	Most of my life lies ahead of me.
	My future seems infinite to me.
	I could do anything I want in the future.
	There is plenty of time left in my life to make new plans.
	I have the sense that time is running out.
	There are only limited possibilities in my future.
	As I get older, I begin to experience time as limited.
Scoring	7-point Likert scale—1 means the statement is very untrue, and 7 means the statement is very true.

^a See, for example, Stephens (1991) and Szmigin and Carrigan (2001).

^b Scale from Lang and Carstensen (2002), also see http://psych.stanford.edu/∼lifespan/links.htm

Appendix C

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Volker G. Kuppelwieser

Philipp Klaus

Supplemental Material

The supplemental material for this article is available online.

Notes

References

Allemand

(2008), “Age Differences in Forgivingness: The Role of Future Time Perspective,” Journal of Research in Personality, 42 (5), 1137–1147

Amatulli

Peluso

A. M.

Guido

Yoon

(2018), “When Feeling Younger Depends on Others: The Effects of Social Cues on Older Consumers,” Journal of Consumer Research, 45 (4), 691–709.

Amatulli

Guido

Nataraajan

(2015), “Luxury Purchasing among Older Consumers: Exploring Inferences about Cognitive Age, Status, and Style Motivations,” Journal of Business Research, 68 (9), 1945–1952.

Anderson

R. E.

Srinivasan

S. S.

(2003), “E-Satisfaction and e-Loyalty: A Contingency Framework,” Psychology & Marketing, 20 (2), 123–138.

Barak

(1987), “Cognitive Age: A New Multidimensional Approach to Measuring Age Identity,” International Journal of Aging and Human Development, 25 (2), 109–128

Barak

Gould

(1985), “Alternative Age Measures: A Research Agenda,” Advances in Consumer Research, 12 (1), 53–58.

Barak

Schiffman

L. G.

(1981), “Cognitive Age: A Nonchronological Age Variable,” in Advances in Consumer Research, Vol. 8, Monroe

K. B.

, ed. Ann Arbor, MI: Association for Consumer Research, 602–606.

Bock

D. E.

Folse

J. A. G.

Black

W. C.

(2016), “When Frontline Employee Behavior Backfires: Distinguishing between Customer Gratitude and Indebtedness and Their Impact on Relational Behaviors,” Journal of Service Research, 19 (3), 322–336.

Bougie

Pieters

Zeelenberg

(2003), “Angry Customers Don’t Come Back, They Get Back: The Experience and Behavioral Implications of Anger and Dissatisfaction in Services,” Journal of the Academy of Marketing Science, 31 (4), 377–393.

10.

Brothers

Gabrian

Wahl

H.-W.

Diehl

(2016), “Future Time Perspective and Awareness of Age-Related Change: Examining Their Role in Predicting Psychological Well-Being,” Psychology and Aging, 31 (6), 605–617.

11.

Bytheway

(2005), “Ageism and Age Categorization,” Journal of Social Issues, 61 (2), 361–374.

12.

Cameron

M. P.

Richardson

Siameja

(2016), “Customer Dissatisfaction among Older Consumers: A Mixed-Methods Approach,” Ageing and Society, 36 (2), 420–441.

13.

Carstensen

L. L.

(2006), “The Influence of a Sense of Time on Human Development,” Science, 312 (5782), 1913–1915.

14.

Carstensen

L. L.

Isaacowitz

D. M.

Charles

S. T.

(1999), “Taking Time Seriously: A Theory of Socioemotional Selectivity,” American Psychologist, 54 (3), 165–181.

15.

Carstensen

L. L.

(1992), “Social and Emotional Patterns in Adulthood: Support for Socioemotional Selectivity Theory,” Psychology and Aging, 7 (3), 331–38.

16.

Chaouali

Souiden

(2018), “The Role of Cognitive Age in Explaining Mobile Banking Resistance among Elderly People,” Journal of Retailing and Consumer Services, 50, 342–50.

17.

Choi

K. S.

(2005), “The Service Quality Dimensions and Patient Satisfaction Relationships in South Korea: Comparisons across Gender, Age and Types of Service,” Journal of Services Marketing, 19 (3), 140–149.

18.

Cleaver

Muller

(2002), “I Want to Pretend I’m Eleven Years Younger: Subjective Age and Senior’s Motives for Vacation Travel,” Social Indicators Research, 60 (1-3), 227–241.

19.

Collins

Tisdell

(2002), “Age-Related Lifecycles: Purpose Variations,” Annals of Tourism Research, 29 (3), 801–818.

20.

Cooil

Keiningham

T. L.

Aksoy

Hsu

(2007), “A Longitudinal Analysis of Customer Satisfaction and Share of Wallet: Investigating the Moderating Effect of Customer Characteristics,” Journal of Marketing, 71 (1), 67–83.

21.

De Keyser

Verleye

Lemon

K. N.

Keiningham

T. L.

Klaus

(2020). “Moving the Customer Experience Field Forward: Introducing the Touchpoints, Context, Qualities (TCQ) Nomenclature,” Journal of Service Research, 23 (4), 433–455.

22.

Dienes

McIatchie

(2018), “Four Reasons to Prefer Bayesian Analyses over Significance Testing,” Psychonomic Bulletin & Review, 25 (1), 207–218.

23.

Dienes

(2014), “Using Bayes to Get the Most out of Non-Significant Results,” Frontiers in Psychology, 5 (781), 1–17.

24.

Drolet

Jiang

Pour Mohammad

Davis

(2019), “The Influence of Aging on Consumer Decision-Making,” Consumer Psychology Review, 2 (1), 3–16.

25.

Etz

Haaf

J. M.

Rouder

J. N.

Vandekerckhove

(2018), “Bayesian Inference and Testing Any Hypothesis You Can Specify,” Advances in Methods and Practices in Psychological Science, 1 (2), 281–295.

26.

Evanschitzky

Wunderlich

(2006), “An Examination of Moderator Effects in the Four-Stage Loyalty Model,” Journal of Service Research, 8 (4), 330–345.

27.

Furrer

Kerguignas

J. Yu

Delcourt

Gremler Dwayne

(2020), “Twenty-Seven Years of Service Research: A Literature Review and Research Agenda,” Journal of Services Marketing, 34 (3), 299–316.

28.

Glorfeld

L. W.

(1995), “An Improvement on Horn’s Parallel Analysis Methodology for Selecting the Correct Number of Factors to Retain,” Educational and Psychological Measurement, 55 (3), 377–393.

29.

Griffiths

(1997), “Ageing, Health and Productivity: A Challenge for the New Millennium,” Work & Stress, 11 (3), 197–214.

30.

Grönroos

(2020), “Viewpoint: Service Marketing Research Priorities,” Journal of Services Marketing, 34 (3), 291–298.

31.

Grougiou

Pettigrew

(2011), “Senior Customers’ Service Encounter Preferences,” Journal of Service Research, 14 (4), 475–488.

32.

Grühn

Sharifian

Chu

(2016), “The Limits of a Limited Future Time Perspective in Explaining Age Differences in Emotional Functioning,” Psychology and Aging, 31 (6), 583–593.

33.

Guido

Amatulli

Peluso

A. M.

(2014), “Context Effects on Older Consumers’ Cognitive Age: The Role of Hedonic versus Utilitarian Goals,” Psychology & Marketing, 31 (2), 103–114.

34.

Guo

Arnould

E. J.

Gruen

T. W.

Tang

(2013), “Socializing to Co-Produce: Pathways to Consumers’ Financial Well-Being,” Journal of Service Research, 16 (4), 549–563.

35.

Gwinner

K. P.

Stephens

(2001), “Testing the Implied Mediational Role of Cognitive Age,” Psychology and Marketing, 18 (10), 1031–1048.

36.

Harrison-Walker

L. J.

(2001), “The Measurement of Word-of-Mouth Communication and an Investigation of Service Quality and Customer Commitment as Potential Antecedents,” Journal of Service Research, 4 (1), 60–75.

37.

Hettich

Hattula

Bornemann

(2017), “Consumer Decision-Making of Older People: A 45-Year Review,” The Gerontologist, 58 (6), e349–e368.

38.

Holman

E. A.

Silver

R. C.

Mogle

J. A.

Scott

S. B.

(2016), “Adversity, Time, and Well-Being: A Longitudinal Analysis of Time Perspective in Adulthood,” Psychology and Aging, 31 (6), 640–651.

39.

Horn

J. L.

(1965), “A Rationale and Test for the Number of Factors in Factor Analysis,” Psychometrika, 32 (2), 179–185.

40.

Isaacowitz

D. M.

Charles

S. T.

Carstensen

L. L.

(2000), “Emotion and Cognition,” in Handbook of Aging and Cognition, Craik

Salthouse

, eds. Mahwah, NJ: Lawrence Erlbaum, 593–631.

41.

JASP Team (2020), “JASP (Version 0.12.2).” https://jasp-stats.org/

42.

Jiang

Fung

H. H.

Sims

Tsai

J. L.

Zhang

(2016), “Limited Time Perspective Increases the Value of Calm,” Emotion, 16 (1), 52–62.

43.

Jones

M. A.

Reynolds

K. E.

Arnold

M. J.

(2006), “Hedonic and Utilitarian Shopping Value: Investigating Differential Effects on Retail Outcomes,” Journal of Business Research, 59 (9), 974–981.

44.

Kohlbacher

Chéron

(2012), “Understanding “Silver” Consumers through Cognitive Age, Health Condition, Financial Status, and Personal Values: Empirical Evidence from the World’s Most Mature Market Japan,” Journal of Consumer Behaviour, 11 (3), 179–188.

45.

Kooij

D. T. A. M.

Kanfer

Betts

Rudolph

C. W.

(2018), “Future Time Perspective: A Systematic Review and Meta-Analysis,” Journal of Applied Psychology, 103 (8), 867–893.

46.

Kornadt

A. E.

Meissner

Rothermund

(2016), “Implicit and Explicit Age Stereotypes for Specific Life Domains across the Life Span: Distinct Patterns and Age Group Differences,” Experimental Aging Research, 42 (2), 195–211.

47.

Kozik

Drewelies

Düzel

Demuth

Steinhagen-Thiessen

Gerstorf

Hoppmann

C. A.

(2019), “Future Time Perspective: Dimensions of Opportunities, Life, and Time Are Differentially Associated with Physical Health, Cognitive Functioning, and Well-Being in Old Age,” Aging & Mental Health, 24 (9), 1487–1495.

48.

Kozik

Hoppmann

C. A.

Gerstorf

(2015), “Future Time Perspective: Opportunities and Limitations Are Differentially Associated with Subjective Well-Being and Hair Cortisol Concentration,” Gerontology, 61 (2), 166–174

49.

Kumar

Lim

(2008), “Age Differences in Mobile Service Perceptions: Comparison of Generation Y and Baby Boomers,” Journal of Services Marketing, 22 (7), 568–577.

50.

Kuppelwieser

V. G.

(2016), “Towards the Use of Chronological Age in Research—A Cautionary Comment,” Journal of Retailing and Consumer Services, 33 (11), 17–22.

51.

Kuppelwieser

V. G.

Sarstedt

(2014), “Exploring the Influence of Customers’ Time Horizon Perspectives on the Satisfaction-Loyalty Link,” Journal of Business Research, 67 (12), 2620–2627.

52.

Kuppelwieser

V. G.

Sarstedt

Tuzovic

(2014), “The Role of Context and Motivation Variables in Mobile Commerce Usage—A Further Perspective on Chong (2013),” Technological Forecasting and Social Change, 88, 156–161.

53.

Lambert-Pandraud

Laurent

(2010), “Why Do Older Consumers Buy Older Brands? The Role of Attachment and Declining Innovativeness,” Journal of Marketing, 74 (5), 104–121.

54.

Lambert-Pandraud

Laurent

Lapersonne

(2005), “Repeat Purchasing of New Automobiles by Older Consumers: Empirical Evidence and Interpretations,” Journal of Marketing, 69 (2), 97–113.

55.

Lang

F. R.

Carstensen

L. L.

(2002), “Time Counts: Future Time Perspective, Goals, and Social Relationships,” Psychology and Aging, 17 (1), 125–139.

56.

Liao

H.-W.

Carstensen

L. L.

(2018), “Future Time Perspective,” GeroPsych, 31 (3), 163–167.

57.

Mann

H. B.

Whitney

D. R.

(1947), “On a Test of Whether One of Two Random Variables Is Stochastically Larger than the Other,” The Annals of Mathematical Statistics, 18 (1), 50–60.

58.

Marquié

J. C.

Jourdan-Boddaert

Huet

(2002), “Do Older Adults Underestimate Their Actual Computer Knowledge?” Behaviour & Information Technology, 21 (4), 273–280.

59.

Marsman

Wagenmakers

E.-J.

(2017), “Bayesian Benefits with JASP,” European Journal of Developmental Psychology, 14 (5), 545–555.

60.

McKay-Nesbitt

Manchanda

R. V.

Smith

M. C.

Huhmann

B. A.

(2011), “Effects of Age, Need for Cognition, and Affective Intensity on Advertising Effectiveness,” Journal of Business Research, 64 (1), 12–17.

61.

Mittal

Griskevicius

(2016), “Silver Spoons and Platinum Plans: How Childhood Environment Affects Adult Health Care Decisions,” Journal of Consumer Research, 43 (4), 636–656.

62.

Mittal

Kamakura

W. A.

(2001), “Satisfaction, Repurchase Intent, and Repurchase Behavior: Investigating the Moderating Effect of Customer Characteristics,” Journal of Marketing Research, 38 (1), 131–142.

63.

Morey

R. D.

Hoekstra

Rouder

J. N.

Lee

M. D.

Wagenmakers

E.-J.

(2016), “The Fallacy of Placing Confidence in Confidence Intervals,” Psychonomic Bulletin & Review, 23 (1), 103–123.

64.

Moschis

Bellenger

Curasi

C. F.

(2003), “Financial Service Preferences and Patronage Motives of Older Consumers,” Journal of Financial Services Marketing, 7 (4), 331–340.

65.

Moschis

G. P.

(1994), “Consumer Behavior in Later Life: Multidisciplinary Contributions and Implications for Research,” Journal of the Academy of Marketing Science, 22 (3), 195–204.

66.

Nasco

S. A.

Hale

(2009), “Information Search for Home, Medical, and Financial Services by Mature Consumers,” Journal of Services Marketing, 23 (4), 226–235.

67.

Nysveen

Pedersen

P. E.

Thorbjørnsen

(2005), “Intentions to Use Mobile Services: Antecedents and Cross-Service Comparisons,” Journal of the Academy of Marketing Science, 33 (3), 330.

68.

Ostrom

A. L.

Parasuraman

Bowen

D. E.

Patrício

Voss

C. A.

(2015). “Service Research Priorities in a Rapidly Changing Context,” Journal of Service Research, 18 (2), 127–159.

69.

Polo

Sese

F. J.

(2013), “Strengthening Customer Relationships: What Factors Influence Customers to Migrate to Contracts?” Journal of Service Research, 16 (2), 138–154.

70.

Pour Mohammad

Drolet

(2019), “The Influence of Age and Time Horizon Perspective on Consumer Behavior,” Current Opinion in Psychology, 26, 94–97.

71.

Pyone

J. S.

Isen

A. M.

(2011), “Positive Affect, Intertemporal Choice, and Levels of Thinking: Increasing Consumers’ Willingness to Wait,” Journal of Marketing Research, 48 (3), 532–543.

72.

Quintana

D. S.

Williams

D. R.

(2018), “Bayesian Alternatives for Common Null-Hypothesis Significance Tests in Psychiatry: A Non-Technical Guide Using JASP,” BMC Psychiatry, 18 (1), 178.

73.

Rintamäki

Spence

M. T.

Kuusela

Kanto

(2006), “Decomposing the Value of Department Store Shopping into Utilitarian, Hedonic and Social Dimensions: Evidence from Finland,” International Journal of Retail & Distribution Management, 34 (1), 6–24.

74.

Romani

Grappi

Dalli

(2012), “Emotions that Drive Consumers Away from Brands: Measuring Negative Emotions toward Brands and Their Behavioral Effects,” International Journal of Research in Marketing, 29 (1), 55–67.

75.

Schau

H. J.

Gilly

M. C.

Wolfinbarger

(2009), “Consumer Identity Renaissance: The Resurgence of Identity Inspired Consumption in Retirement,” Journal of Consumer Research, 36 (2), 255–276.

76.

Settersten

R. A.

Mayer

K. U.

(1997), “The Measurement of Age, Age Structuring, and the Life Course,” Annual Review of Sociology, 23 (1), 233–261.

77.

Silvestro

Fitzgerald

Johnston

Voss

(1992). “Towards a classification of service processes,” International Journal of Service Industry Management, 3 (3), 62–75.

78.

Sinha

Wang

(2013), “How Time Horizon Perceptions and Relationship Deficits Affect Impulsive Consumption,” Journal of Marketing Research, 50 (5), 590–605.

79.

Soutar

G. N.

Sweeney

J. C.

(2003), “Are There Cognitive Dissonance Segments?” Australian Journal of Management, 28 (3), 227–249.

80.

Stephens

(1991), “Cognitive Age: A Useful Concept for Advertising?” Journal of Advertising, 20 (4), 37–48.

81.

Sudbury

Simcock

(2009a), “A Multivariate Segmentation Model of Senior Consumers,” Journal of Consumer Marketing, 26 (4), 251–262.

82.

Sudbury

Simcock

(2009b), “Understanding Older Consumers through Cognitive Age and the List of Values: A U.K.-Based Perspective,” Psychology and Marketing, 26 (1), 22–38.

83.

Sweeney

J. C.

Danaher

T. S.

McColl-Kennedy

J. R.

(2015), “Customer Effort in Value Cocreation Activities: Improving Quality of Life and Behavioral Intentions of Health Care Customers,” Journal of Service Research, 18 (3), 318–35.

84.

Sweeney

J. C.

Soutar

G. N.

(2001), “Consumer Perceived Value: The Development of a Multiple Item Scale,” Journal of Retailing, 77 (2), 203–220.

85.

Szmigin

Carrigan

(2001), “Time, Consumption, and the Older Consumer: An Interpretive Study of the Cognitively Young,” Psychology and Marketing, 18 (10), 1091–1116.

86.

Tarasi

C. O.

Bolton

R. N.

Gustafsson

Walker

B. A.

(2013), “Relationship Characteristics and Cash Flow Variability: Implications for Satisfaction, Loyalty, and Customer Portfolio Management,” Journal of Service Research, 16 (2), 121–137.

87.

Thakor

M. V.

Suri

Saleh

(2008), “Effects of Service Setting and Other Consumers’ Age on the Service Perceptions of Young Consumers,” Journal of Retailing, 84 (2), 137–149.

88.

Timmerman

M. E.

Lorenzo-Seva

(2011), “Dimensionality Assessment of Ordered Polytomous Items with Parallel Analysis,” Psychological Methods, 16 (2), 209–220.

89.

United Nations (2019), “World Population Prospects: The 2019 Revision,” (accessed September 28, 2020), [available at https://population.un.org/wpp/].

90.

U.S. Census Bureau (2018), “Midwest Home to Most of the Counties with Decreases in Median Age,” (accessed September 28, 2020), [available at https://www.census.gov/newsroom/press-releases/2018/popest-characteristics.html].

91.

Wagenmakers

E.-J.

Marsman

Jamil

Verhagen

Love

Selker

Gronau

Q. F

Šmíra

Epskamp

Matzke

Rouder

J. N.

Morey

R. D.

(2018), “Bayesian Inference for Psychology. Part I: Theoretical Advantages and Practical Ramifications,” Psychonomic Bulletin & Review, 25 (1), 35–57.

92.

Wei

Donthu

Bernhardt

K. L.

(2013), “Effects of Cognitive age, Dispositional time Perceptions, and Time View Manipulations on Product Attribute Evaluations,” Journal of Business Research, 66 (11), 2171–2177.

93.

Weijters

Rangarajan

Falk

Schillewaert

(2007), “Determinants and Outcomes of Customers’ Use of Self-Service Technology in a Retail Setting,” Journal of Service Research, 10 (1), 3–21.

94.

Yoon

Cole

C. A.

Lee

M. P.

(2009), “Consumer Decision Making and Aging: Current Knowledge and Future Directions,” Journal of Consumer Psychology, 19 (1), 2–16.

95.

Zimbardo

P. G.

Boyd

J. N.

(1999), “Putting Time in Perspective: A Valid, Reliable Individual-Differences Metric,” Journal of Personality and Social Psychology, 77 (6), 1271–1288.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.79 MB

0.01 MB