Cybersenior Mobility: Why Some Older Consumers May Be Adopting the Internet

Charles A. McMellon, Hofstra University
Leon G. Schiffman, The City University of New York
ABSTRACT - This study explores the relationship between limited out-of-home mobility and older consumers’ Internet behavior. Limited mobility can be the result of physical and social deterioration in older individuals as they proceed through the normal process of aging. We use principal component and canonical correlation analysis to examine a variety of mobility and Internet activity variables. We suggest continuity theory as a possible explanation why some individuals go on-line. This theory proposes that as older individuals age, they make adaptive choices to maintain the internal and external structures of their physical, psychological, and social environments. Within this context, our study examines how older consumers are using the Internet as an adaptive tool to maintain some structures limited by the effects of aging. Specifically, various aspects of personal communication, necessities, financial matters, and searching for entertainment and information are identified as Internet activities that are associated with the characteristics of those who are limited in mobility outside-of-the-home. These findings have specific managerial and public policy implications.
[ to cite ]:
Charles A. McMellon and Leon G. Schiffman (2000) ,"Cybersenior Mobility: Why Some Older Consumers May Be Adopting the Internet", in NA - Advances in Consumer Research Volume 27, eds. Stephen J. Hoch and Robert J. Meyer, Provo, UT : Association for Consumer Research, Pages: 139-144.

Advances in Consumer Research Volume 27, 2000      Pages 139-144

CYBERSENIOR MOBILITY: WHY SOME OLDER CONSUMERS MAY BE ADOPTING THE INTERNET

Charles A. McMellon, Hofstra University

Leon G. Schiffman, The City University of New York

ABSTRACT -

This study explores the relationship between limited out-of-home mobility and older consumers’ Internet behavior. Limited mobility can be the result of physical and social deterioration in older individuals as they proceed through the normal process of aging. We use principal component and canonical correlation analysis to examine a variety of mobility and Internet activity variables. We suggest continuity theory as a possible explanation why some individuals go on-line. This theory proposes that as older individuals age, they make adaptive choices to maintain the internal and external structures of their physical, psychological, and social environments. Within this context, our study examines how older consumers are using the Internet as an adaptive tool to maintain some structures limited by the effects of aging. Specifically, various aspects of personal communication, necessities, financial matters, and searching for entertainment and information are identified as Internet activities that are associated with the characteristics of those who are limited in mobility outside-of-the-home. These findings have specific managerial and public policy implications.

INTRODUCTION

The Internet is one of the fastest growing technologies ever, with approximately 60 to 70 million Americans of all age groups now on-line. This growth, interestingly, is driven by the "young" (i.e., those less than 20) and the "old" (i.e., those more than 50), who are more likely than any other age group to be new users of the Internet (GVU 1998). The current study focuses on the older age group because of their importance as an economic group and their growth in size.

Older adults appear to use the Internet for a variety of reasons including searching for information, interacting socially, passing time, finding entertainment, and seeking advice (Dixon 1993; Kaye 1998). One possible explanation why some older individuals use the Internet in the ways that they do is in response to the effects of the normal aging process. In normal aging (i.e., without catastrophic life events which limit the individual’s activity either through illness or care-giving), individuals may experience a slow physical and social deterioration as they grow older (Birren and Fisher 1991). For example, thinking slows down and socializing diminishes. One effect of this physical and social deterioration is a lessening of an individual’s out-of-home activities (Brail and Chapin 1973; Hawes 1977). Limited out-of-home mobility may influence Internet behavior because it offers the user opportunities to substitute activities that may normally be diminished due to aging. For example, if friends move to retirement communities in other cities, there will be reduced opportunities to go out with them to socialize. In addition, the aging individual may not have the money or the energy to make a trip to visit with friends. Thus, individuals may turn to e-mailing or chatting on the Internet as an alternate means to communicate with friends or like-minded others in an effort to maintain their former levels of socializing.

The purpose of the current study is to explore the potentially complex relationships between limitations on out-of-home mobility in older adults and the amount of time they allocate to activities on the Internet. We recognize that mobility may be one of many influences on older adults in their allocation of time on the Internet. We focus on mobility in this study because of its possible link to quality-of-life when the individual stays home more than he or she might otherwise do. We also focus on limited mobility because of the potential marketing and public policy implications that may emerge from a better understanding of this phenomenon.

We begin with a review of the consumer behavior literature on mobility. Next, we discuss our method and present our results. We then discuss the possible relationship between mobility and time allocated to on-line activities. We conclude with a discussion of the limitations, the marketing and public policy implications, and the possible future research directions that emerged from this exploratory study.

MOBILITY AND OLDER ADULTS

Mobility is defined as an individual’s ability to participate in outside-of-the-home activity (Rahtz, Sirgy, and Meadow, 1989). Mobility may be an important issue for older adults because many are limited in their out-of-home activities due to the normal effects of aging such as a diminished social structure and a slow deterioration of physical and cognitive ability (Birren and Fisher 1991). Importantly, mobility may influence the individual’s quality-of-life (Lumpkin and Hunt 1989). Yet, there is little empirical research on the consumer related issues of the influence of mobility among older adults and nothing could be found examining its influence on Internet behavior. Two studies in the literature were found that examined mobility among older adults; restricted life space (Lumpkin and Hunt 1989) and limied activity (Rahtz, Sirgy, and Meadows 1989).

Lumpkin and Hunt (1989) examined older adults’ shopping behavior, information search, and psychological profiles in relation to their our-of-home activities. They conceptualized mobility as a public and private transportation problem and segmented their sample into self-reliant and dependent older adults. Of special interest to our study are their results that suggest individuals with restricted life space (i.e., limited mobility) develop their own personal sources of information (e.g., neighbors) rather than the information sources provided by the marketer (e.g., advertisements). Other findings suggest that individual’s with limited mobility appear to be less active in the community, less self-confident, less healthy, do not cope as well, and score lower on internal locus of control (Rotter 1966) than those individuals’ who have more mobility. These findings suggest a lower quality-of-life than those who score higher on these items.

Rahtz, Sirgy, and Meadow (1989) examined a variety of demographic and psychological correlates for television orientation (i.e., a disposition to use television for entertainment and information). Their results show that education and income are negatively correlated with an individual’s television orientation, while morale, personal concern, financial concern, and limited activity are positively correlated with an individual’s television orientation. Of special interest to our study is the positive correlation between limited activity and television orientation which suggests that those older adults who stay home more may use television for entertainment and information more than those who go out of their homes more.

These two studies, Lumpkin and Hunt (1989) and Rahtz, Sirgy, and Meadow (1989), suggest that older adults, when limited in their out-of-home mobility, will find compensating sources of information and entertainment to meet their continuing needs.

RESEARCH QUESTION

We suspect that going on the Internet, for some older computer users, may be a substitute for out-of-home activity. At any time in "virtual reality," on-line users can travel to any place in the world they chose and talk with a variety of other like-minded individuals. The Internet offers many avenues for social interaction (e.g., chat rooms, e-mailing to friends, reading and posting messages in news groups and bulletin boards) and the satisfaction of personal interest needs (e.g., hobbies, genealogy, financial). Research also suggests that older adults with limited out-of-home mobility may use different sources for information and entertainment than those older adults with fewer limitations on their mobility (Lumpkin and Hunt 1989; Rahtz, Sirgy, and Meadow 1989). For example, older adults who use books as sources of information may turn to television or the Internet because they can no longer travel to the library or a bookstore as easily as in the past.

How an older individual’s limited out-of-home mobility relates to the amount of time and specific activity behavior on the Internet appears to be a complex phenomenon. Mobility may be influenced by a diminished social life (e.g., loss of friends), lack of physical energy, or a combination of these and other factors. In addition, their behavior on the Internet may also be complex because of the myriad choices the individual faces when going on-line. This complexity of issues suggests a need for a better understanding of limited mobility and Internet behavior, the relative strengths of the variables involved, and how they interact with each other.

METHOD

A convenience sample consisting of older adults (i.e., 55+) who participated in on-line activities (e.g., emailing to friends, surfing) was used in the analysis. The respondents were on-line members of SeniorNet, a group of older adults on America Online, and from two groups accessible through CompuServe (i.e., Retirement + and AARP). The respondents were solicited via e-mail to participate in the study. Then a 12-page, 131 question survey was sent by regular mail to those who had agreed to participate in the study. The questions gathered data on the dependent variables representing the amount of time spent on various on-line activities (e.g., e- mailing to children and grandchildren, searching for information) and the mobility related independent variables. The survey also contained other questions which are not discussed in this paper.

A total of 3,918 individuals was solicited by e-mail. The first e-mail solicitation letter included a number of elements known to aid in increasing response rates; (a) naming the institutional source, (b) egoistic, altruistic, and help-the-sponsor appeals, (c) announcement of a lottery for participation, and (d) assurances of confidentiality (Church 1993; Faria and Dickinson 1992; Fox, Crask, and Kim 1988). The e-mail solicitation letter also asked the potential respondents for their home address. This request proved problematic because of the potential respondent’s concerns about the legitimacy of the survey. For example, some respondents thought our e-mail was part of a confidence game to bilk them of money. The number of older adults agreeing to participate by sending us their home address, then completing the survey and returning it by regular mail was 369 (9%). Follow-up e-mails were sent to those who did not respond to the first solicitation in an effort to increase response rates. This strategy had to be abandoned because a few non-survey responding adults sent irate messages to the researchers to stop e-mailing them.

The response level for completed surveys from the mailing to those who agreed to participate in the survey was 284 (78%) of which 243 (65.8 %) were 55 years of age and older. Although the overall solicitation e-mail response rate (i.e., 9%) appears low compared to regular mail survey research, the response rates did not seem low for on-line solicitation. This may be especially salient given our requirement that asked potential respondents to send researchers their home address. A review of the response rate literature suggests there is no accepted range for the response rates of an e-mail solicitation or questionnaire. Small scale e-mail studies using academicians as respondents produced rates of more than 50% (e.g., Mehta and Sivadas 1995), while larger non-academic respondent studies similar to ours produced rates of about 6% (e.g., Tse, et al. 1995). Commercial electronic direct marketers guarantee a rate of about 3% on targeted markets (e.g., Quantum Communications 1997). This approach limits the generalization of the study because of sample and respondent bias.

Dependent Variables

The dependent variables used in this study are based on 12 specific on-line activities that the authors uncovered during their literature review and earlier exploratory research. They all measure the amount of time per week the individual estimated they spent on each activity. These activities were: (a) exploring or surfing (SURF); (b) entertainment, playing games, and viewing graphics (ENT); (c) e-mailing to friends or people with similar interests (FMAIL); (d) e-mailing to children or grandchildren (CMAIL); (e) reading message boards (MESS); (f) participating in chat groups (CHAT); (g) searching for non-product related information like weather or travel (INFO); (h) educational activities like seeking information for school reports (EDUC), (h) paying bills (BILL); (i) window shopping, looking at advertisements, and product information searchs (SHOP); (j) financial activities such as checking stock prices or company information (FIN); and work activities (WORK).

Independent Variables

The two existing mobility studies wer used as the source for questions used in our survey. Each question used a seven-point Likert-type response scheme (i.e., from "strongly agree" to "strongly disagree") for examining the influence of mobility. The questions for restricted life space (Lumpkin and Hunt 1989) were: (a) "I’m not as socially active as I used to be (SOCIAL);" (b) "I’m not as active as I used to be because I have no transportation (TRANSPO);" and (c) "I don’t have as many friends or acquaintances as I used to have (FRIENDS)." The questions for limited activity (Rahtz, Sirgy, and Meadow 1989) were: (a) "I stay home most of the time (HOME)," and (b) "I would rather stay home than go out with friends (STAYHOME)."

In addition two questions were developed by the authors to examine health and perceived guilt: " I really don’t have any physical problems (PROBLEMS)," and "I get out of the house as much as I should (IGETOUT)."

RESULTS

Our research objective was to better understand the relationships, (a) among the 7 mobility independent variables, (b) among the 12 dependent Internet allocation of time variables and, (c) between an older individual’s limited mobility and the amount of time older adults devote to various activities on the Internet.

Our first step was to examine the 7 independent and 12 dependent variables for any underlying structural factors that might condense the number of variables into meaningful groupings before further analysis. Principal component analysis is recommended as an exploratory tool to examine this type of data (Hair, Jr., Anderson, Tatham, and Black 1992; Tabachnick and Fidell 1989). Principal components analysis was performed on the seven independent mobility variables. No factors emerged from the analysis. Principal components analysis was then performed on the 12 Internet behaviors (e.g., surfing, information search) which did factor suggesting that there were underlying structural factors that might condense the list of variables. The procedure extracted four factors from the 12 Internet behaviors with an Eigenvalue of greater than one. The factors accounted for 58% of the variance. The factor meanings, variable names, and factor loadings are shown in Table 1.

TABLE 1

PRINCIPAL COMPONENT ANALYSIS LOADINGS

All four factors appear to have meaning. The first factor suggests a variety of entertainment seeking and information searching behaviors related to the personal interests of the user. These personal interests include scanning message boards and visiting consumer product and other Internet sites for relevant information and entertainment. The second factor suggests necessary behaviors. E-mailing to children can be seen as a parental duty while work can also be seen as a necessary duty. The third factor suggests personal communication behaviors of the respondents with like-minded others. This personal communication factor contains the elements of e-mailing and participating in chat groups (i.e., real-time discussions on the Internet) with friends and like-minded others. The fourth factor was financial in nature. This factor contains the elements of paying bills on-line and financial matters (e.g., checking stock prices). Both of these elements are reflective of the personal monetary areas of an individual’s life.

Our second step was to analyze the possible relationships and strengths of the variables; (a) among the four factors representing the amount of time spent on specific Internet activities (i.e, dependent variables), (b) among the seven mobility variables (i.e, independent variables), and (c) between the dependent and independent variables. One approach to understanding these types of complex relationships is canonical correlation analysis because it examines both the within structure and the functional relationships between the dependent and independent variables (Westbrook and Fornell 1979). Canonical correlation analysis constructs a dummy variable (i.e., canonical variate) for ech side of the equation finding their maximum correlation. The statistical results from this analysis, which will aid in understanding these relationships, are the standardized canonical weights and the canonical loadings. Standardized canonical weights are the equivalent of beta weights, although they are somewhat unstable if multicollinearity exists. Canonical loadings, which may offer a better explanation of the relationship, reflect the degree to which a variable is represented by the canonical variate (Fornell 1978).

Canonical correlation analysis was performed on the seven mobility independent variables and the four factors representing the dependent time variables. One canonical variate of interest was constructed with a canonical correlation of .31 (10 % of the variance). The Wilk’s test of significance was approximate F=1.387, p=.089. The remaining three variates correlated at .22, .13, and .09. The first pair of canonical variates, therefore, accounted for the significant relationships between the sets of variables. Table 2 depicts both weights and loadings.

Redundancy for the dependent variables indicates little shared variance. Redundancy for the independent variables indicates a moderate amount of shared variance. Examination of the dependent variate suggests there is an inverse relationship between necessities and the other three variables. In addition, the magnitude of the standardized weights and loadings suggest that the most influential variable in the dependent set is information and entertainment with personal communication also playing a role. Necessities and financial matters appear to play a lesser role. Examination of the independent variables suggests a more complex relationship. Among the independent variables, IGETOUT appears to play the strongest role with HOME, in an inverse relationship, also of some importance. PROBLEMS, SOCIAL, TRANSPO, FRIENDS, and STAYHOME appear to play a much lesser role.

In addition, there were differences between some of the weights and loadings among the independent variables indicating the possibility of instability. This may be caused by multicollinearity. Consequently, a split sample canonical correlation analysis was performed to examine instability. The two split samples had canonical correlations of 45.2% and 44.4% respectively. As shown in Table 3, the split analysis indicated that the variate maintained the directional relationship of the dominate variables in all loadings, although weights and loadings changed in some cases.

Examination of the dependent variable split sample analysis indicates a similar inverse relationship between necessities and the other factors. In addition, the loadings are in a similar relationship with the information/entertainment and personal communication variables retaining their influence. This suggests some stability in the dependent variables. Examination of the independent variable split sample analysis indicates a similar inverse relationship between IGETOUT and HOME. The loadings for these variables also maintain their relationships. There is also some apparent instability in that a few of the lesser influential independent variables changed signs and values.

TABLE 2

RELATIONSHIP ANALYSIS FOR THE CANONICAL VARIATE

TABLE 3

SPLIT SAMPLE ANALYSIS FOR THE FIRST CANONICAL VARIATE

DISCUSSION

The principal component analysis of the 12 Internet behaviors suggested a possible underlying structure for the allocation of time on the Internet. All four factors could be explained. The first factor contained a wide range of entertainment and information elements that suggested that individuals who participated in these behaviors were pursuing personal interests on the Internet. Earlier exploratory research by the authors with similar respondents suggested a wide range of personal interests ranging from searching genealogical records to discussing unidentified flying objects. The Internet behaviors that make up this factor were educational, reading messages, window shopping, searching fr information, entertainment, and exploring. As shown in the canonical analysis, this factor appears to be significant in the relationship between mobility and amount of time spent on Internet activities. The second factor, which contained the obligatory elements of work time and e-mailing to children and grandchildren, suggested necessary behavioral tasks. Parents should write to their children and grandchildren. In addition, many individuals have to work. The third factor contained the personal communication elements of e-mailing and chatting with friends and like-minded individuals. Our respondents allocated more time to e-mailing than any other Internet behavior. The fourth factor contained the monetary elements of paying bills and checking finances on-line. These findings suggest that Internet behaviors can be categorized in meaningful groups and might be reflective of basic Internet behaviors.

The canonical analysis suggested an inverse relationship between necessary time and leisure time. That is the more an individual has to do necessary tasks on the Internet, the less time they will allocate to more leisurely activities. The canonical analysis also indicated that "...not getting out as much as they should" was the most influential variable in the independent variate and that "entertainment/information" behavior was the most influential dependent variate. The other variables played a less influential role. The canonical variate analysis also indicates a relationship between the independent and dependent variates. That is, that older Internet users’ with limited out-of-home mobility may spend more time on the Internet than those individuals with less limited out-of-home mobility.

One explanation for these findings is that as individuals age, they begin to spend more time at home due to the deterioration of their physical and social lives (Birren and Fisher 1991). If they can no longer get out of the house the way they have in the past, they may be using the Internet as a tool to compensate for their deteriorating physical and social conditions. For example, exploring on the Internet may be a substitute for their inability to go outside-of-the-home as often as they have in the past. On the Internet, an individual can go anywhere they want to go. Thus, travel on the Internet becomes the substitute for travel outside-of-the-house. Atchley (1989) and Kaufman (1987) suggest a continuity approach for successful aging that proposes that some older adults develop adaptive strategies that help maintain their internal and external structures.

Considering the Atchley-Kaufman continuity approach, we suggest that the continuing increase in the numbers of older consumers who are learning how to operate computers and are going on the Internet may be an indication that some individuals in this segment are using the Internet as an adaptation strategy to maintain past aspects of their life. Moreover, it is our position that such older adults are making an adaptive choice of participating in "out-of-home cyber-experiences" that maintain their internal and external structures without ever leaving their homes. Specifically, we identify searching for information and entertainment as an important on-line activity that may be related to the effects of mobility in older adults. In addition, the canonical analysis suggests personal communication with friends and like-minded others, and dealing with financial matters on-line are also related in some manner to out-of-home mobility.

This study supports and extends the earlier research of Lumpkin and Hunt (1989) and Rahtz, Sirgy, and Meadow (1989) that found relationships between limited mobility and the older adult seeking new sources of information and entertainment. In this study, we suggest that the Internet may be a new source of information and entertainment for those older adult with limited mobility. In addition, it may also be a tool for attending to financial matters and for personal communication with friends and other like-minded individuals for those older adult with limited mobility.

CONCLUSIONS

This study suggests that a relationship exits between older computer users who perceive some level of limited out-of-home mobility and the amount of time spent on the Internet for specific activities. We suggest that one explanation for this relationship is Atchley’s (1989) and Kaufman’s (1987) contention that, to be successful in aging, individuals should develop adaptive strategies that maintain their internal and external structures. In this context, our research indicates that older computer users with limitations on out-of-home mobility may be attempting to maintain their past social and cognitive structures through use of the Internet by e-mailing or chatting with friends or like-minded others, searching for information or entertainment and, tending to financial matters more than those individuals with fewer limitations on their out-of-home mobility.

The findings of our study are restricted because the study used a convenience sample consisting of older adults who were members of specific on-line groups (i.e., SeniorNet, Retirement+, and AARP). Thus, the results shown here may not be representative of all older computer users who go on-line. Although we felt the response rate was acceptable given the request for a home address, the respondents who participated in our survey may be different from those who did not answer. For example, they may be willing to accept more risk because they were willing to send us their home addresses. Thus, generalizations from the study are not recommended.

This study also suggests a number of managerial and public policy implications. Promotions and products designed for the Internet may be improved if managers know that older computer users are using Cyberspace as a substitute for going out-of-the house because they were limited in their out-of-home mobility. For example, product web sites might be designed to look a little more like out-of-home places and a little less like Cyberspace. In addition, promotional strategies might emphasize the socializing and informational aspects of the Internet, and actual advertising executions might depict older adults in situations more reflective of the findings presented here.

There are also public policy implications to our findings. We have suggested that the quality-of-life of older adults may be influenced by their strategic choice to go on-line. Public policy initiatives could direct more older adults to the benefits of going on-line. It appears that many of the well educated and financially well off are already going on-line, but there is a large segment of the older population who, because of money limitations or other concerns, may not be going on-line. Public policy can help correct this imbalance.

We have examined mobility within the context of the Internet and have presented results that suggest that the amount of time some older adults spend on line with specific activities may be related to their out-of-home mobility. Not all the factors that influence mobility have been examined for their relationship with the allocation of time on-line (i.e., fear of crime, financial problems, care giving). In addition, some older individuals who use the Internet may have made the decision to stay home to use the Internet for various activities such as communicating with like-minded individuals instead of communicating face-to-face. Thus, they would also report that they do not go out as much as they should. More research is needed to better understand the dynamics of mobility and its relationship with the amount of time spent on the Internet.

REFERENCES

Atchley, Robert C. "A Continuity Theory of Normal Aging," The Gerontolgist 29, 2 (1989): 183-190.

Birren, James E. and Laurel Fisher. "Aging and Slowing Behavior: Consequences for Cognition and Survival," in Psychology and Aging. Eds. J. J. Berman, and T. B. Sonderegger. Nebraska: Nebraska Symposium on Motivation 1991, (1991): 1-37.

Brail, Richard K. and F. Stuart Chapin, Jr. "Activity Patterns of Urban Residents," Environment and Behavior 5, 2 (1973): 163-190.

Church, Allan H. "Estimating the effect of incentives on mail survey response rates: A meta- analysis." Public Opinion Quarterly, 57, 1 (1993): 62-79.

Dixon, Janet Marie "Uses and gratifications theory to predict senior’s use of an electronic community." Dissertation Abstracts International, 55-01, 4686A. (1993): (University Microfilms no. 9413198)

Faria, A. J., and John R. Dickinson. "Mail survey response, speed, and cost." Industrial Marketing Management, 21, 1(1992): 51-60.

Fox, Richard J., Melvin R. Crask, and Jonghoon Kim. "Mail survey response rate. A meta- analysis of selected techniques for inducing response." Public Opinion Quarterly, 52, 4 (1988): 467- 491.

Fornell, Claes. "Three approaches to canonical analysis." Market Research Society Journal, 20 3 (1978): 166-181.

GVU (Graphics, Visuals, and Usability Center). "April, 1998 GVU’s 9th WWW User Survey." http://www.cc.gatech.edu/user_surveys/survey-1998-04/. Accessed August 29, 1998.

GVU (Graphics, Visuals, and Usability Center). AOctober, 1995 GVU’s 4th WWW User Survey." http://www.cc.gatech.edu/user_surveys/survey-10-1995/. Accessed March 6, 1996.

Hair, Joseph. F., Jr., Rolph E. Anderson, Ronald L. Tatham, and William C. Black. Multivariate Data Analysis. (3rd ed.). New York: Macmillan Publishing Company, 1992.

Hawes, Douglass K. "Time budgets and consumer leisure-time behavior." Advances in Consumer Research 4, (1977): 221-229.

Kaye, Barbara, K. "Uses and gratifications of the World Wide Web: From Couch Potato to Web Potato." The New Jersey Journal of Communication, 6, Spring (1998): 21-40.

Kaufman, Sharon R. The Ageless Self: Sources of Meaning in Late Life. Madison, WI: University of Wisconsin Press 1987.

Lumpkin, James R. and James B. Hunt. "Mobility as an Influence on Retail Patronage Behavior of the Elderly: Testing Conventional Wisdom, Journal of the Academy of Marketing Sciences 17, Winter (1989): 1-12.

Maddox, George L. "Themes and issues in sociological theories of human aging," Human Development 13, (1970): 17-27.

McMellon, Charles A. "Mature Consumers: Their Allocation and Consumption of Time On-line," Dissertation Abstracts International, 76, 044-338-01, 1997.

Mehta, Raj and Eugene Sivadas. "Comparing Response Rates and Response Content in Mail Versus Electronic Mail Surveys." Journal of the Market Research Society 37, 4 (1995): 429-439.

Quantum Communications. "Customers for You," Quantum Communications <sales@quantumcom.net, e-mail message sent to authors April 5, 1997. A copy is available upon request from the authors.

Rahtz, Don. R., Joseph M. Sirgy and H. Lee Meadow" The Elderly Audience: Correlates of Television Orientation," Journal of Advertising 18, 3, (1989): 9-20.

Rotter, Julian B. "Generalized Expectancies for Internal Versus External Control of Reinforcement," Psychological Monographs: General and Applied 80 (1, Whole No. 609),1966.

Sirgy, M. Joseph, John T. Mentzer, Don Rahtz and H. Lee Meadow. "Satisfaction with Health Care Services Consumption and Life Satisfaction Among the Elderly," Journal of Macromarketing 11, Spring (1991): 24-39.

Tabachnick, Barbara G. and Linda S. Fidell. Using Multivariate Statistics. New York: Harper & Row, 1991

Tse, Alan C. B., Ka Chun Tse, Chow Hoi Yin, Choy Boon Ting, Ko Wai Yi, Kwan Pui Yee, and Wing Chi Hong. "Comparing Two Methods of Sending our Questionnaires: E-mail Versus Mail." Journal of the Market Research Society 37, 4 (1995): 441-446.

Westbrook, Robert A. and Claes Fornell. "Patterns of Information Source usage Among Durable Goods Buyers." Journal of Marketing Research 17, 3 (1995): 303-312.

----------------------------------------