Editor’s note: David J. Pagnucco is a senior account manager with Maritz Marketing Research Inc.'s Detroit office. He is a graduate of the University of Michigan with a B.S. in Computer Science and Psychology. For more information about Natural Group Interviewing, contact him at (313) 540-2200. Robert P. Quinn, Ph.D. is a senior research scientist at the University of Michigan's Survey Research Center. He is a specialist in organizational behavior and survey research methods.

Since motivational research has its roots in psychology, it isn't surprising that it concentrates on individual perceptions, preferences and motives. Accordingly, the ways we assess consumer preferences, such as in questionnaires and one-on-one interviews focus exclusively on individual purchasing situations. But what is most common is not necessarily universal. Under certain circumstances, the appropriate unit of analysis may be a group of individuals.

The term "focus group" suggests superficially that we already have a methodology for dealing with group purchasing decisions. A focus group is a group in the sense that it develops a structure, goals and norms. In spite of this, it remains ephemeral and artificial. A focus group lacks two properties of a "natural" group: a continuing history and influence over the lives of its members.

Natural groups

Regarding consumer behavior, the most obvious natural group is the family. While most household purchases end up in the hands of an individual, the family may become more salient where purchases involve large amounts of money or major changes in lifestyles. For example, almost half (49%) of car purchase decisions are made with at least one other person. and in 25% of the purchases, the joint decision-maker is the spouse (1987 Newsweek/Maritz Marketing Research Study).

Even a corporate purchaser doesn't necessarily operate independently. Although a single manager may be charged with concluding a contract for a fleet of trucks, the actual decision to buy is likely to be made by a group, each member of which has a stake in the purchase. Such corporate groups are usually ad hoc, their membership changing depending on the nature of the purchase.

Therefore, whenever the appropriate unit of analysis is a group rather than an individual, traditional data collection methods must be supplemented. It is risky to infer a group's preference from the isolated preferences of its members. Simple statistical averaging of individual preferences fails to consider power differentials among group members. It also fails to account for the negotiations, compromises and tradeoffs that take place when group members initially disagree.

Sometimes, a decision that appears to be wholly an individual one really isn't. Not every decision affecting group members is made jointly because it would be too time consuming. Instead, the final decision is delegated to a specific individual, and other group members participate only in the sense that they make their general preferences known. Family members, for example, rarely discuss, the details of a grocery shopping list. Rather, the person actually doing the shopping is trusted to take into account the likes and dislikes of all family members.

Identifying and investigating

The occurrence of natural groups is certainly widespread, but difficult to investigate empirically. The preferences of group members are likely to be communicated to the "delegate" in subtle ways, often over an extended period of time. They may even be internally inconsistent and unstable. More approachable empirically is the less frequent situation wherein all group members participate actively in the final purchasing decision.

The first step in investigating joint decision-making is to identify the situations in which is occurs. The most likely candidates are those having a major impact on the home lives and jobs of group members. For example, decisions which involve considerable cost (financial or otherwise) will probably be made jointly since they entail serious personal consequences if made incorrectly. With the range of decision situations thus narrowed, intuition and keen observation can restrict the field still further. Occasionally, one will have harder data to work with, such as that cited in the above study. Focus groups can also provide clues. Finally, the identification of 'true" natural group decision situations can be attacked frontally. Individuals being screened for participation in surveys can be asked directly about the extent to which others will take an active part in the final purchasing decision.

Since the impact of natural group decision? making on major buying decisions continues to be an empirical desert, we are currently considering studying it in a three?pronged methodological investigation. The purposes of this investigation are:

1. To assess how joint preferences and evaluations of natural groups differ from “average” preferences and evaluations of separate group members;

2. To investigate the interpersonal dynamics through which individual preferences become consolidated into group buying preferences, and ultimately, group behavior; and,

3. To fine tune methods of collecting data from natural groups, rather than from individuals or focus groups.

Three studies

The one-on-one study: This study concentrates on people who intend to buy cars. As part of our customary screening procedures, a subsample of couples who indicate their car buying decisions are largely joint ones is selected. Both members of each are asked to come to a central interviewing facility at the same time. Each one is isolated from the other and is asked to evaluate various features of a prototype vehicle. A separate evaluation questionnaire is then given to each member of the couple. The couple is never asked to make a joint evaluation.

Additional qualitative information is obtained by a semi-structured, one-on-one interview conducted with each member. The projected joint evaluations of the couple are estimated simply by averaging their separately collected evaluations.

The one-on-two study: This study differs from the one-on-one study in several ways. Rather than being given separate questionnaires to fill out independently, the couple is shown the same prototype at the same time, and is asked to discuss its attributes between themselves. They record their final "group" evaluation on their joint questionnaire. It is left completely up to the couple to resolve any differences that may emerge. The couple's deliberations are observed directly and videotaped. Additional qualitative information is obtained by a semi-structured, on-on-two interview conducted with both members of the couple at the same time. The joint evaluations of the couple are obtained directly from the joint questionnaire they complete.

The final joint evaluations, as inferred in the one-on-one study and obtained directly in the one-on-two study, are then compared. This indicates 1) to what extent and 2) in what ways inferred group evaluations (i.e. statistically manufactured) differ from actual group evaluations. But, even if such differences are evident, these two proposed studies tell us little about how such differences arise.

The combined study: To understand how individual preferences combine to generate group preferences and decisions, a more complicated study design is needed. This study begins like the first one-on-one study in which the evaluations of both members are obtained separately. Following this, each party is presented with both his or her own evaluations and those of his or her partner. The couple is then required to consolidate their initial evaluations into a single, final joint evaluation. Their deliberations in doing so are observed and videotaped. A semi-structured one-on-two interview is then used to discover how the couple resolved disagreements.

Natural group interviewing cannot be substituted for either one-on-one interviewing or focus groups, nor is it intended to do so. Its application extends to a limited number of, but nevertheless real, decision-making situations.