AN INVESTIGATION OF CHRONOLOGICAL VERSUS COGNITIVE AGE IMPACT IN THE KUWAIT COFFEE SHOP MARKET

Given the growing body of research dedicated to self-perceived age and the cognitive age construct, it is widely believed that one’s self-perceived age may actually be a be!er predictor of age-related psychological states or a!itudes than mere chronological age. Extending the research on cognitive age, the current study examines the impact of both cognitive age and traditional chronological age on the behaviors of co"ee shop users in Kuwait. #e study $nds that chronological age and cognitive age are highly correlated, both in age levels and in terms of consumer behavior. Nevertheless, a large portion of the sample perceived themselves to be younger than their chronological age. #is is especially true of consumers aged 55 and over. #e main $ndings that di"erentiate chronological age %om cognitive age are that as Kuwaiti consumers become chronologically older, co"ee drinks become more important to them. Also, as cognitive age increases, consumers are less likely to drink co"ee with %iends.


Introduction
With longer life expectancies and declining fertility rates, the world's population is slowly and inexorably moving towards an older average age (Guido, Amatulli & Peluso, 2014).In fact, it has been suggested that the ageing of the world's population is the most signi cant demographic change in the history of humankind (Kohlbacher, Sudbury & Hofmeister, 2011).In light of this trend, understanding how age interacts with consumption pa erns has become essential to the world's marketers.is is especially true in global markets where li le is known about the unique behavioral pa erns of the various age segments.In addition, researchers have begun to consider a variety of age-oriented constructs, including the concept of "cognitive age".Cognitive age can be thought of as the psychological or self-perceived age of a given consumer.First introduced in 1981 by Barak and Schi man, the concept of cognitive age has proven to be a valuable psychological construct not only in the eld of marketing, but in the elds of psychology and gerontology as well (Goldsmith & Heiens, 1992;Mathur & Moschis, 2005;Van Auken, Barry & Bagozzi, 2006).Moreover, recent research indicates that the tendency for consumers to possess and be inuenced by cognitive age self-perceptions is a global phenomenon, applying to consumers in markets as diverse as the U.S., Western Europe, Japan, and China (Eastman & Iyer, 2012;Gianlugi, Amatulli & Peluso, 2014;Teller, Gi enberger & Schnedlitz, 2013;Van Auken, Barry & Bagozzi, 2006;Wei & Talpade, 2009).Yet li le research has been conducted on the impact of either age or age-related psychological constructs on consumer behavior in the rapidly growing Middle Eastern market.
Over the last several decades, the Middle East has become more signi cant as a potential market.For example, the countries of the Gulf Cooperation Council (GCC), including Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates, in aggregate now represent one of the 20 largest economies in the world (Yates, 2013).In particular, the retail market in Kuwait is expected to grow at a steady annual rate of more than seven percent over the next several years (Gavin, 2013).Yet there is one essential characteristic of Middle Eastern society that di erentiates it from the majority of consumer markets throughout the world.is unique di erence is the restriction on the Arab lifestyle driven by the social and religious precepts of the Muslim faith, which include regulations that preclude the sale and use of alcohol.Consequently, social interactions in the Muslim world that might normally occur in venues as iconic as the English pub, German beer garden, Italian wine bar, or even the old American saloon, have been redirected to the Arab co ee house.In Kuwait, for instance, si ing rooms known as "Dewaniya" have served as gathering places where men would congregateto discuss politics or social issues while nibbling snacks, smoking, and enjoying the mildly stimulating e ect of Arab co ee.ese gathering places take the form of either the family "Dewan" or the public "Dewan".Hence, co ee has long been featured as prominent in the social life of the typical Arab consumer.Co ee, in fact, is one of the most widely consumed beverages throughout the entire world, and the retail co ee industry is currently poised to undergo a sustained period of growth (Co er, 2014).Consequently, the investigation into the relationship between age (including both chronological and cognitive age) and co ee consumption in the Kuwait co ee shop market, the focus of the current study, is especially germane to service marketers seeking to understand the demographic and psychological factors that may impact retail co ee sales in the potentially lucrative markets of the GCC and the Middle East as a whole.

Literature review
Age is a descriptor variable of tremendous signi cance and interest to virtually all of the social sciences.Its use by marketers as a basis for segmenting and targeting consumer populations is widespread, and a great many segmentation schemes and typologies employ age in some respect.is is largely because the monochromic time perspective, which views time as a linear and stable construct, dominates the Western mindset, and leads to the expectation that all populations experience the e ects of aging at a constant and unvarying rate (Bluedorn, Kaufman, & Lane, 1992).Chronological age, as we are all aware, is the actual age of a person as measured in years, months, and days from the date the person was born.Nevertheless, it has been proven in the physical sciences that time is actually a relative concept, the value of which depends on perspective, measurement, and context (Einstein, 1905).erefore, the in uence of time and age on consumers may be subject to a wider array of interpretations than the standard linear model may imply.Instead, a personal and subjective view of one's age could possibly provide greater insight into the behavior and lifestyle pa erns of consumer segments than chronological age alone.
According to Birren (1960), ageing is a multi-dimensional construct comprised of three independent yet inter-related mechanisms.ese include biological aging, psychological aging, and sociological aging.Each of these aging mechanisms is driven by subjective consumer self-perceptions.e theoretical basis for the notion that one's "age" may be subjectively determined is self-concept theory.According to selfconcept theory, an individual tends to possess a wide variety of thoughts and feelings regarding themselves as a distinct "a itude object" (Rosenberg, 1979).Consequently, chronological age is at best a proxy measure for the biological and psycho-social changes that occur throughout a person's life.Instead, it has been argued that when we consider the unique temperament and make-up of a person, people may actually "age" at di erent rates (Barak & Schi man, 1981).
e most in uential measure of self-perceived age in a marketing context has been the concept of "cognitive age", rst proposed by Barak and Schi man (1981).As originally developed by Barak and Schi man (1981), the cognitive age construct is comprised of four sub-dimensions, "feel-age", "look-age", "act/do-age", and "interestage".In terms of self-concept theory, it is the "age reality" that is perceived subjectively, rather than the "age reality" that can be measured objectively, that is the basis of the age-oriented dimension of one's self-concept (Barak & Gould, 1985).Over the years, as researchers began to accept the concept of self-perceived age, additional dimensions were added to the cognitive age construct.For instance, Clark, Long, and Schi man (1999) extended the construct by adding the dimensions of "health age" and "think age".Whether consisting of four or even six dimensions, the reliability and validity of the cognitive age construct has been well-established (Clark, Long & Schi man, 1999;Moschis & Mathur, 2006). is is true even when using a variety of di erent measurement techniques, including ratio scales, semantic di erential scales, Likert scales, and age-decade scales (Eastman & Iyer, 2012).
Additional evidence for the validity of the cognitive age construct comes from the fact that, although the magnitude of age discrepancies may vary across chronological age segments, the percentage of people who perceive themselves exactly at their actual chronological age is relatively low.In most cases, respondents tend to see themselves as perceptually younger than their chronological age. is is especially true of older consumers, whose cognitive ages are o en ten years or more below their chronological ages (Ca erall & Maclaran, 2001).Consistent with these previous studies, Chang (2008), in a study focusing speci cally on younger consumers, discovered that the majority of the respondents, despite an average chronological age of only 19.54, also reported feeling younger than their chronological age.Indeed for much a consumer's adult life, cognitive age tends to lag chronological age, at least until the decline of one's health or some other signi cant life-stage event (Van Auken, Barry & Bagozzi, 2006).
Given the growing body of research dedicated to self-perceived age and the cognitive age construct, it is widely believed that one's self-perceived age may actually be a be er predictor of age-related psychological states or a itudes than mere chronological age (Iyer & Reisenwitz, 2010;Teller, Gi enberger & Schnedlitz, 2013).It is also increasingly apparent that an individual's time perspective may do a be er job of predicting customer loyalty pa erns than one's chronological age (Kuppelweiser & Sarstedt, 2014).In summary, the growing body of research on cognitive age has challenged the complete reliance on chronological age as a proxy for life-stage oriented behavioral tendencies.As the explanatory and predictive powers of chronological age continue to be questioned, there have been calls for still additional research regarding the relative impact of chronological versus cognitive age on speci c pa erns of consumer behavior across a wider range of industries (Teller, Gi enberger & Schnedlitz, 2013).Co ee is a ubiquitous consumer product with global appeal.Consequently, in order to examine the impact on consumer behavior of cognitive age versus chronological age, the investigation of co ee shops as a retail service category allows for possible future replication of the study across multiple national or regional contexts.

Hypotheses
According to Guido, Amatulli, and Peluso (2014), it must be acknowledged that cognitive age is not an absolute concept, but is instead relative in nature.us, Guido, Amatulli, and Peluso (2014) suggest that cognitive age must always be compared to chronological age in order to make sense.Consequently, in order to determine whether or not di erences exist between the two constructs on a variety of dimensions related to the Kuwait retail co ee market, we o er a series of hypotheses related to both chronological age and cognitive age.To begin with, a great deal of evidence exists to suggest that di erent pa erns of a itudinal loyalty, behavioral loyalty, and customer expectations exist between chronologically older and chronologically younger consumers.For example, Berbel-Pineda et al. (2011) report that younger customers are more demanding of service based retailers, while older customers are more likely to build long-term relationships and exhibit greater degrees of a itudinal and behavioral loyalty.e work by Berbel-Pineda et al. (2011) suggests that older consumers are likely to have more stable satisfaction ratings with the retailers that they choose to patronize, giving rise to the rst set of research hypotheses.

H 1a : As chronological age increases, consumers exhibit greater satisfaction with their retail co ee shop choices.
H 1b : As cognitive age increases, consumers exhibit greater satisfaction with their retail co ee shop choices.
One of the most important motives that drives older shoppers is the need for social interaction.According to Myers and Lumbers (2008), older consumers tend to view shopping as both a means of socialization and a leisure pursuit.Moreover, in the Kuwait market, there is a long tradition of drinking co ee in social gathering places and si ing rooms known as "Dewaniya", particularly among tribal elders and leaders.Given the traditional role of the co ee shop in the Middle Eastern social structure, retail co ee shops are particularly well-positioned to ful ll an important lifestyle need for Kuwaiti consumers, leading to the second set of hypotheses.If age does have an impact on customer satisfaction, it is also likely to positively impact a consumer's degree of behavioral consistency as well.According to Evanschitzky and Woisetschläger ( 2008), age has a negative in uence on a consumer's willingness to gather information about new products and services.Moreover, their research indicates that age is positively related to brand loyalty.Similarly, Karani and Fraccastoro (2010) nd that older consumers resist switching brands once they have established a favorite.Lambert-Pandraud and Laurent (2010) suggest that older consumers may form a ective links and even emotional a achments over time with particular brands, leading to H 3a and H 3b .
H 3a : As chronological age increases, the percentage of visits to the most preferred retail co ee shop also increases.H 3b : As cognitive age increases, the percentage of visits to the most preferred retail co ee shop also increases.
According to Dick and Basu (1994), customer loyalty is a multidimensional construct which includes several di erent and distinct conceptualizations of loyalty based on a itude and behavior.Speci cally, "no loyalty" can be de ned as the absence of either positive a itudes or repeat purchase behavior."Spurious loyalty" exists when consumers exhibit positive behaviors towards a brand, but their a itudes are not fully supportive.Instead, their actions are driven by convenience or circumstances."Latent loyalty" exists when consumers possess positive a itudes toward a brand, yet are prevented from exhibiting repeat purchase behavior due to external factors.Finally, "true loyalty" only exists when repeat purchases are driven by positive consumer a itudes.Based on the Dick and Basu (1994) conceptualization of true loyalty, we o er the next set of hypotheses.
H 4a : As chronological age increases, the percentage of retail co ee shops toward which consumers exhibit "true loyalty" also increases.
H 4b : As cognitive age increases, the percentage of retail co ee shops toward which consumers exhibit 'true loyalty' also increases.
In a previous study of the Kuwait retail co ee market, Pleshko and Heiens (2015) demonstrated that satisfaction and loyalty are positively related.In other words, as satisfaction levels increase, levels of loyalty also increase.Also, according to Bogomolova and Grudinina (2011), as consumers gain knowledge and experience with a given product category, they are able to make more informed decisions.Consequently, it is likely that satisfaction levels with a particular co ee retailer that develop over time would also be correlated with pa erns of true loyalty, leading to H 5a and H 5b .
H 5a : As chronological age increases, the number of retail co ee shops toward which consumers are considered both "true loyal" and satis ed also increases.
H 5b : As cognitive age increases, the number of retail co ee shops toward which consumers are considered both 'true loyal' and satis ed also increases.
According to Richelieu and Korai (2014), the consumption of co ee at retail co ee shops is a ritualized experience associated with pleasure, companionship and extended relationships.Although there have not been any studies that have speci cally looked at the relationship between age, either chronological or psychological, and the consumer co ee experience per se, Richelieu and Korai (2014) note that co ee is a consumer product capable of becoming symbolically charged in the minds of consumers.Given the likelihood that the symbolic and ritualistic associations linked to the co ee shop experience may develop over time, we present the following two sets of hypotheses.Previous research suggests that people who feel relatively older tend to live less active lifestyles (Mathur & Moschis, 2005).It has also been shown that as consumers age, they generally become less mobile (Karani & Fraccastoro, 2010).Finally, it has been suggested that consumers with a higher cognitive age may visit fewer unfamiliar venues outside the home than do older consumers with a lower cognitive age (Guido, Amatulli, & Peluso, 2014).us, we propose H 8a and H 8b .
H 8a : As chronological age increases, the percentage of co ee drinks consumed at home increases.H 8b : As cognitive age increases, the percentage of co ee drinks consumed at home increases.
Finally, it has been demonstrated that older consumers a ach greater importance to familiar and emotionally close contacts (Carstensen, Charles & Fung, 2003).Also, as previously discussed, there is a long tradition in the Kuwaiti market of drinking co ee in a social se ing.erefore, rather than consuming co ee alone, we expect both chronologically and cognitively older consumers to be more likely to pursue the traditional practice of consuming co ee with friends, leading to H 9a and H 9b .
H 9a : As chronological age increases, the percentage of co ee drinks consumed with iends increases.
H 9b : As cognitive age increases, the percentage of co ee drinks consumed with iends increases.

Research methodology
e data for the current study were generated from a sample of Kuwaiti co ee shop patrons.At the time of this study, thirty-nine co ee shop retailers were operating in Kuwait.Most of these co ee shop retailers had multiple locations throughout Kuwait City.At the time of this study, these thirty-nine retailers were operating approximately two hundred and y co ee shops (Kuwait Chamber of Commerce, 2011).In Kuwait, there are no bars nor are there similar entertainment complexes that serve alcohol.erefore, co ee shops are places of gathering for many in the population wanting to meet with friends outside the home.us, the list of shops includes a wide variety of both local and international co ee shops.
Data for the project were collected using personal interviews to administer a standard questionnaire.Seventy interviewers were recruited for the study, and the interviewers were assigned ratios to guide the collection of data and to ensure representativeness of the sample with regard to the population as a whole.As detailed demographic, behavioral, or psychographic information is di cult to obtain in Kuwait, the authors selected two available descriptors, age and gender, to provide guidelines for selecting the sample.Secondary data sources provided the age and gender statistics which were used as guidelines for the percentages of adults to be included in each age and gender category (CIA World Fact Book, 2011;Kuwait Public Authority for Civil Information, 2011).A test of the expected versus sample frequencies revealed no signi cant di erences in age or gender between the sample and the population as a whole (X 2 = 2.03, p = 0.37).
As an initial screening procedure, only those consumers who had visited a co ee shop within the past three months were considered for inclusion.Considering the statistical tests to be performed, the number of companies being studied, and the desirability of obtaining reliable parameter estimates, a sample size of 600 was the goal for the project.A sample of 700 was actually taken with the knowledge that some of the surveys would likely be discarded as incomplete or on the basis of interviewer error.Ultimately, 82 surveys were discarded due to incomplete questionnaires, resulting in 618 usable surveys.
e study included a variety of constructs.Two general age indicators were employed: (i) chronological age and (ii) psychological age.A third age variable, Age di , is calculated by nding the di erence between Age chron and Age cog ,.Eight consumer variables were employed as well: (iii) average satisfaction, (iv) percent of brands to which the buyer is considered as satis ed, (v) percent of total visits to most preferred brand, (vi) true loyalty as a percent of brands currently in use, (vii) brands which the user is satis ed with and loyal to as a percent of brands currently used, (viii) experience as in the number of brands tried, (ix) number of brands currently used, and (x) number of purchase occasions/visits per year.Finally, four co ee-related variables were utilized: (xi) importance of co ee drinks, (xii) the number of co ee drinks per day (xiii), whether co ee is consumed at home or not, and (xiv) whether co ee is consumed with friends or not.

e Age indicators
Chronological age, Age chron , refers to the age of a respondent as measured in years.
Respondents were asked to write down their year of birth.en, Age chron was calculated for each respondent by subtracting the year of birth from the current year.Note that the study only included consumers eighteen years or above.e range of Age chron was from 18 to 70 with a mean of 34.69 years and a standard deviation of 0.540.
Cognitive age, Age cog , was also calculated in years.With cognitive age, however, the years refer to how old the respondents perceived themselves to be.To measure cognitive age, the authors utilized age-decade scales, which have the advantage of providing respondents with scale reference points (Van Auken, Barry & Bagozzi, 2006).Speci cally, respondents were asked to answer ve questions by circling an answer from the following scale for each question: 'teens' , '20s' , '30s' , '40s' , '50s' , '60s' , or '70s or older' .
e scale values were taken as the midpoint of each choice, except for the 'teens' and '70s and older' categories, resulting in the following scale values: 'teens'=17, '20s'=24.5,'30s'=34.5,'40s'=44.5,'50s'=54.5,'60s'=64.5,and '70s or older'=75.Consistent with Barak and Schi man (1981) and Clark, Long, and Schi man (1999), the ve questions were: (i) I feel like I am in my, (ii) I look like I am in my, (iii) I act like I am in my, (iv) my interests are like those of a person in his or her, and (v) I think like a person in his or her. e ve items were subjected to a principal components factor analysis with results explaining 83.24% of total variance in a single factor.Cronbach's alpha statistic was 0.949 for the ve items, which is indicative of a reliable indicator.en, as it was necessary to estimate an age in years, Age cog was calculated by summing the ve items and dividing by ve.Note that only whole numbers rounded down were used.
e process used to produce the nal scale is consistent with scaling research methods historically utilized in the transformation of ordinal data into higher-level rating or ratio scales ( urstone, 1927;Emory, 1980).e range of Age cog was from 17 to 68, with a mean of 33.53 years and a standard deviation of 0.434.

e consumer variables
e satisfaction variable, Sat avg , refers to the average satisfaction rating given by the respondents for each of the thirty-nine co ee shop retailers for which they were considered to be users.To be considered a user, the respondent had to have visited a given co ee shop retailer within the past three months.Respondents were asked to indicate their general experiences with those co ee shop retailers by writing an appropriate number on a scale ranging from one -not at all satis ed -to ten -extremely satis ed (Pleshko & Cronin, 1997;Dawes & Smith, 1985).e Sat avg variable was calculated for each respondent by summing the satisfaction responses and then dividing those satisfaction totals by the number of retail brands used in the past three months.Values for the variable Sat avg ranged from 3.29 to 10, with a mean of 7.22 and a standard deviation of 0.048.
e Sat % variable refers to the percentage of retail brands each respondent was more than marginally satis ed with.Consistent with Pleshko and Heiens (2015), a buyer was considered satis ed with a given retail brand if the satisfaction rating for the retailer was greater than ve on a scale of one to ten.Speci cally, for each respondent, the number of retail brands with which the customer was considered satis ed was totaled and then divided by the number of brands they were currently using.Values for the variable Sat % ranged from 10.5% to 100%, with a mean of 0.769 and a standard deviation of 0.009.e percent of visits to the most preferred retail brand is referred to as Visits %pr .Respondents were asked to indicate their top ve favorite co ee shop retail brands.
en, for each respondent, the total number of visits to retail co ee shops was found, allowing for the calculation of the percentage of total visits allocated to the number one most preferred retailer.Values for the variable Visits %pr ranged from 6.82% to 100%, with a mean of 0.400 and a standard deviation of 0.007.e true loyalty percent, Brands %loy , refers to the number of brands to which the respondent is considered to be truly loyal divided by the number of brands currently in use.Consistent with Dick and Basu (1994), a respondent was considered to be truly loyal if he or she had both high a itude and high behavior towards a given retail brand.A itude in this study was indicated by preference rankings (see Visits %pr above).Speci cally, a respondent was considered to have a high a itude towards a retailer if he or she ranked the given retailer in the top ve.Similarly, a respondent was considered to have a high behavior towards a retailer if he or she indicated the given retailer as one of their top ve most visited co ee shops.Combining the preferences and visits data together leads to an indication of whether a respondent was truly loyal or not.
e maximum number of truly loyal brands was ve for each respondent.Brands %loy was calculated by totaling the number of brands towards which the respondent was considered to have both high a itude and high behavior and then dividing this total by the number of retail brands currently in use.Values for the variable Brands %loy ranged from 0% to 100%, with a mean of 0.471 and a standard deviation of 0.010.e variable Brands %s+l , or the percentage of retail brands both satis ed and loyal, refers to the number of retail brands which the respondent was considered both truly loyal to (see Brands %loy above) and satis ed with (see Sat avg above) divided by the number of brands currently in use.Values for the variable Brands %s+l ranged from 0% to 100%, with a mean of 0.426 and a standard deviation of 0.010.

e co ee-related variables
e importance a respondent a ached to co ee drinks as a reason for visiting a retail co ee shop, Co imp , was calculated on a scale from one (not at all important) to ten (extremely important).Values for the Co imp variable ranged from 1 to 10 with a mean of 8.27 and a standard deviation of 0.100.e variable that indicated whether co ee was consumed at home or not, Co hom , was determined by asking the respondents to check a box on the questionnaire.A dichotomous question for Co hom was constructed with the response options being either 0/no or 1/yes.With a mean of 0.64 and a standard deviation of 0.019, approximately 64% of respondents indicated that they consumed co ee at home.
Whether co ee was consumed with friends or not, Co frn , was determined by asking the respondents to check a box next to the item that asked if he or she consumed co ee with friends while visiting co ee shops.e variable Co frn was measured via a dichotomous question with 0/no or 1/yes.With a mean of 0.81 and a standard deviation of 0.022, approximately 81% of respondents consumed co ee with friends on their visits to co ee shops.
e number of co ee drinks per day for each person, Co day , was found by asking the respondents to estimate the number of co ee drinks they typically consume per day in any location, including at home, in retail co ee shops, or at work.If a respondent indicated fewer than one co ee drink per day (for example, one per month), a fraction was calculated for the variable.Values for the variable Co day ranged from 0.008 to 10 with a mean of 2.07 and a standard deviation of 0.059.

Analyses and results
e hypotheses pertaining to the relationship between chronological and cognitive age on a variety of co ee-consumption outcome variables were tested using correlations, t-tests, and F-tests where appropriate.Prior to addressing the hypotheses, the authors investigated the two age indicators to determine their relationship.First, it should be noted that Age chron and Age cog exhibited a Pearson correlation of 'r'= +0.88 with a 'p'= 0.000.A test of mean di erences revealed that within the sample, Age chron (34.69 years) was slightly greater than Age cog (33.53 years) with a di erence of 1.16 years, 't'=4.18,and the 'p'=0.000. is suggests that while the two age indicators are highly correlated, there is a statistically signi cant di erence between the two age measures.Consistent with previous studies, the respondents in the sample perceived their age to be slightly younger than their actual chronological age. is is examined in greater detail.
Table 1 presents a cross-tabulation of the two age indicators a er combining the respondents into three groups (76.18% of the sample): age less than thirty, age between 30 and y-four, and age y-ve or greater.A Chi-square test reveals that the age groups are not distributed as would be expected if the respondents' cognitive ages were the same as their chronological ages (X 2 = 435.4,'p' = 0.000).Based on the raw numbers provided in the table, calculations show that while 76.18% of the sample perceived their cognitive age to be similar to their actual age, 16.31% of the sample actually perceived themselves to be younger than they actually are. is is especially true with older consumers.As mentioned previously, this nding is consistent with previous research in this area (Barak & Schi man, 1981;Ca erall & Maclaran, 2001;Teller, Gi enberger & Schnedlitz, 2013;Van Auken & Barry, 1995;Van Auken, Barry & Bagozzi, 2006;Wolfe, 1987).A smaller percentage of respondents considered themselves to be older than they actually are (46/613=7.50%).X 2 = 435.4d.f.= 4 'p' = 0.000 Table 2 shows the correlations of age with the consumer related and co ee related variables.Note that sample size varies between 613 and 616.e cut-o criterion for statistical signi cance is set at 'p'= .05.From Table 2, the statistical tests of correlations reveal that one or both age indicators are related to most of the consumer variables in the study.Speci cally, the tests reveal that as age (both chronological and cognitive) increases, respondents (i) are satis ed with more of the retailer brands which they are using, (ii) exhibit higher levels of loyalty, (iii) are more likely to be loyal if satis ed with the brands which they are using, (iv) drink more co ee, and (v) consume more co ee at home.Also from Table 2, the statistical tests of correlations reveal that one or both age indicators are related to most of the co ee variables in the study.Speci cally, the tests reveal that as Age chron increases, respondents (vi) perceive co ee drinks to be more important.Finally, as Age cog increases, respondents (vii) are less likely to consume co ee with friends.

Discussion and managerial implications
As we seek to build on our knowledge of consumer behavior, service marketers utilize a wide variety of psychological constructs that promise to facilitate the process of segmenting and targeting important consumer segments.One such construct is the cognitive age measure, a variable that continues to prove to be relevant and viable in an increasingly large number of international markets.e current study extends previous research on cognitive age to the Middle East as well.Adding to the unique contribution of the study is the investigation of the impact of both chronological age and cognitive age on consumer loyalty and consumption pa erns within the Kuwait retail co ee shop category.e rst nding of signi cance is the veri cation of the existence of the cognitive age construct in the Kuwait market.Although previous studies have validated the presence of distinctive psychological age pro les in international markets, no previous study has validated the use of the cognitive age construct in the Middle East.In addition, the pa erns of cognitive age self-designations found in Kuwait are consistent with those discovered in markets throughout the world.Speci cally, although there is a strong correlation between chronological age and cognitive age, many Kuwaiti consumers are inclined to view themselves as cognitively younger than their actual chronological age.
is tendency appears to be most pronounced in the older age range of 55 plus.
In addition, the study reveals several interesting ndings with regard to the impact of age on consumer loyalty and consumption pa erns in the retail co ee shop market.e rst signi cant result is that age tends to be positively related to several indicators of consumer loyalty.Speci cally, regardless of how age is measured, including both chronologically and cognitively, older consumers are satis ed with more of the retail brands which they are using.As a result, they also tend to exhibit higher levels of loyalty.Although previous research indicates that not all satis ed consumers necessarily become loyal (Pleshko & Heiens, 2015), there is greater consistency between satisfaction and loyalty in the case of both chronologically and cognitively older consumers, as these consumers are more likely to become loyal if satis ed with the brands which they are using.
Although up to this point the ndings are equally consistent for both chronological and cognitive age, some behavioral di erences were also observed between the two age constructs.First, it is only the chronologically older consumers who perceive co ee drinks to be more important.is may be at least partly due to the well documented health bene ts of co ee.For instance, it has been shown that co ee consumption tends to be associated with lower risks of coronary heart disease, lower risks of congestive heart failure, and lower risks of stroke (O'Keefe et al., 2013).In addition, recent evidence suggests that co ee consumption has a positive e ect on long-term memory (Mejia & Ramirez-Mares, 2014).erefore, it may not be surprising that the chronologically older consumers, for whom the health bene ts may be more essential, tend to view co ee as being more important than do the cognitively older consumers.
Nevertheless, as both chronological and cognitive ages increase, respondents consume more co ees per day. is means that co ee is a "mature" drink that appeals to both chronologically and cognitively older consumers, and that these consumers represent a potentially lucrative market for co ee retailers.Unfortunately, however, as both forms of age increase, these consumers tend to consume more co ee at home.Although the well-documented decrease in mobility associated with chronologically older consumers may help explain this pa ern, it is more di cult to explain the higher consumption of co ee at home among cognitively older consumers.Finally, as cognitive age increases, respondents are less likely to consume co ee with friends.Perhaps the relative sequestration of the cognitively older consumer is one factor that actually contributes to their self-perceived maturity.
Although the current study helps to shed light on the value of alternative age constructs in the Middle Eastern retail co ee shop market, future studies examining other industries and product and service categories are needed to fully appreciate the use of cognitive age as a segmentation variable in this market.Nevertheless, the implications for retail managers in the Kuwait market are clear.Age is an important segmentation variable, and speci c pa erns of consumer behavior might be expected depending on consumer age classi cation.In addition, age is not a uniform chronological construct, but the psychological or cognitive self-perceptions that consumers may hold regarding age can be equally e ective as a basis for consumer segmentation.Finally, although there are signi cant correlations between chronological and cognitive age categories and the consumer behaviors they may be associated with, there are several categories of behavior that the two constructs impact in di erent ways.erefore, both chronological age and cognitive age self-perceptions should be considered when seeking to fully and robustly pro le consumer segments.
H 2a : As chronological age increases, consumers are satis ed with more retail co ee shops overall.H 2b : As cognitive age increases, consumers are satis ed with more retail co ee shops overall.
H 6a : As chronological age increases, the importance a ached to co ee as a reason for visiting a retail co ee shop increases.H 6b : As cognitive age increases, the importance a ached to co ee as a reason for visiting a retail co ee shop increases.H 7a : As chronological age increases, the number of co ee drinks consumed per day increases.H 7b : As cognitive age increases, the number of co ee drinks consumed per day increases.

TABLE 1 .
Cross-Tabulation of Age chron and Age cog

TABLE 2 .
Correlations of Age with Consumer and Co ee Variables