William D. Wells, Needham, Harper & Steers Advertising, Inc.
[ to cite ]:
William D. Wells (1980) ,"Discussion", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 473-474.

Advances in Consumer Research Volume 7, 1980     Pages 473-474


William D. Wells, Needham, Harper & Steers Advertising, Inc.

These three quite different papers are united by the fact that they all make some use of Life Style or psychographic data. This discussion will focus on the "So What?" question: assuming that the results can be accepted at face value, what more do we know about Life Style than we knew before the studies were done?

The paper by Liisa Uusitalo presents a very brief overview of an extensive investigation which attempted to derive abstract consumption patterns by analyzing relationships among consumption activities and other activities of households. A factor analysis of household expenditures produced three major dimensions, "modernity'', "mobility" and "variosity." These dimensions were then combined to produce eight "consumption style" segments.

One value of attempts to organize masses of data is that the organization, once achieved, may produce an interesting and appealing new way to think about a field. Riesman's division of society into "other-directed" vs. "inner-directed" captured a great many diverse observations in an intuitively appealing way. Maslow's organization of motives into a hierarchy running from safety to self-actualization stimulated a great deal of controversy, analysis, and further research. The question is, will this organization of consumer behavior into a new set of typologies based upon three major variables capture the imagination of people in our field in a similar way? Maybe. Maybe not. We shall see.

The author proposes that this classification should also be quite helpful in thinking about specific marketing policy or public policy problems. For instance, she suggests that the recreation and home decoration markets might find their potential customers in certain segments, and she believes that certain other segments might be especially subject to health care problems arising from "one-sided" consumption of certain kinds of foods.

The typology would probably be very difficult to use in this way. If one wants to make a serious investigation of the recreation or home decoration markets, it would be absolutely essential to study the specifics of these markets in great detail, rather than approach them from the perspective of such a highly abstract typology. Similarly, if one were concerned about the "one-sided" food consumption of certain consumers, it would first be necessary to find out exactly which consumers showed the "one-sided" pattern. Even though some of the segments tend to show this pattern, some individuals in the segment would not be "one-sided," and some individuals not in the segments would be "one-sided." Thus any conclusions drawn about the segments which tend to show the "one-sided" pattern might not apply at all well to the individuals who do show it.

The second paper, on consumer mobility, examines mobility as a dimension by which to describe consumers. The answer to the "So What?" question depends upon whether this dimension is thought of as a dependent variable or an independent variable. As an independent variable, mobility would join many other dimensions used to describe consumers: age, sex, income, stage in the life cycle, innovativeness, brand loyalty, and so on.

This use of the mobility dimension seems to have limited value. Age is an extremely valuable dimension because when we know a person's age we know a great deal more about him or her. When we know a person's position on the mobility scale we do not know a great deal more about that person without intimate knowledge of all of the correlates of mobility. Most of us don't have that intimate knowledge, and few of us will ever acquire it.

As a dependent variable, mobility would be of great interest to anyone interested in locating or communicating with consumers who are likely to make use of the services of real estate companies, or moving vans, or any of the goods or services directly associated with relocation. It would be of great value to anyone wishing to contact movers before the fact, and it would be of great value to scholars wishing to study moving as a phenomenon.

It is interesting to note that both the reliability and the meaning of the consumer mobility scale probably change dramatically with the age of the respondent. The scale is probably considerably more reliable for older respondents than for younger respondents because, for older respondents, it is based upon a much longer record. The meaning of the score probably changes with age because "life opportunities" influence so many of the moves of young people, while "subjective willingness to move," as an enduring trait, has had relatively more opportunity to influence the number of moves the older individual has made.

The final paper, on "nomological" validity, attempts to determine whether some expected relationships among Life Style types do in fact appear. Specifically, it attempts to determine whether persons of similar life styles tend to marry each other, and whether "masculinity'' ratings of Life Style segments correlate with masculinity as measured by a personality scale.

The answer to the "So What?" question in this case is quite complex. If the expected relationships emerge we gain some confidence that Life Style descriptions are sufficiently accurate to be correlated to choice of spouse, or with a score on a masculinity-femininity scale. Such findings would perhaps increase our confidence in Life Style measurements a little bit.

But what if the expected relationships are not found? The theory might be wrong. The analysis might be inappropriate. The data or the measurement might be contaminated with method variance and sampling error.

In the present case the finding was that the expected relationships did appear more or less. But, knowing that, we really don't know much more than we knew before. Were the relationships strong enough to confirm our faith in Life Style research? No. If the relationships had been a lot weaker, would Life Style research have lost credibility? Not at all. Weak relationships could have occurred in too many ways.

One final comment about these three papers. All make use of powerful multivariate techniques which produce abstract dimensions that get pretty far away from the real answers of real consumers to real questions. Thus all three, in varying degrees, eliminate the richness of detail which is one of the most useful aspects of Life Style descriptions. All three -- again in varying degrees -- ask us to grasp the essence, expressed as factor scores or beta weights, of whatever it is that produces correlations among large numbers of variables. This essence may be useful in helping us understand consumer behavior, but when it gets too far from the data, its usefulness becomes very elusive indeed. Abstraction is necessary. We can't get along without it. But we are too often (not always!) better off when we are on Cloud Two or Cloud Three than when we are on Cloud Nine.