Internet-Enabled Knowledge Acquisition and Power in Family Decisions

ABSTRACT - The relationship between knowledge and power appears assumptive in marketing, the study reported here is designed to justify this assumption. The study is set in a family decision context. Families are divided according to the domain-specific Internet use of sons, then family decision-power patterns are inspected for a range of products. The relationship between the Internet enabled, domain-specific acquisition of knowledge and power in group decisions is supported.

Citation:

Roger Marshall, Peter Alan Reday, and Na WoonBong (2005) ,"Internet-Enabled Knowledge Acquisition and Power in Family Decisions", in AP - Asia Pacific Advances in Consumer Research Volume 6, eds. Yong-Uon Ha and Youjae Yi, Duluth, MN : Association for Consumer Research, Pages: 70-73.

Asia Pacific Advances in Consumer Research Volume 6, 2005      Pages 70-73

INTERNET-ENABLED KNOWLEDGE ACQUISITION AND POWER IN FAMILY DECISIONS

Roger Marshall, Nanyang Technological University, Singapore

Peter Alan Reday, Ashland University, U.S.A.

Na WoonBong, Kyunghee University, South Korea

ABSTRACT -

The relationship between knowledge and power appears assumptive in marketing, the study reported here is designed to justify this assumption. The study is set in a family decision context. Families are divided according to the domain-specific Internet use of sons, then family decision-power patterns are inspected for a range of products. The relationship between the Internet enabled, domain-specific acquisition of knowledge and power in group decisions is supported.

DEVELOPING THE RESEARCH ISSUE

The relationship between knowledge and power has been a matter of more than just academic interest for centuries. Sun Tzu discussed the power of information within a military context some 300 years before Christ. In the late Sixteenth Century, when Francis Bacon wrote of knowledge being power, he was a herald for an industrial age when knowledge was no longer to be hoarded by an elite few, but was to become far more widely shared and applied. The application of the rapidly-growing body of knowledge, mostly as technology, led to the rapid growth of physical power, productivity, profits and standards of living for those who harnessed it. The link between organized, useful information (hereafter called knowledge) and power in a scientific, technical sense is well-entrenched, and a scan of the technical innovation literature shows that the relationship is still deemed to be relevant (Carter and Scarbrough 2001; Hislop, Newell, Scarbrough, and Swan 2000).

Closer to marketing, negotiators speak of knowledge as giving power, too. It is widely recognized that information about the objects of negotiation and opponents in a negotiation situation can lead to a clear advantage to a negotiator (e.g. Weingart, Hyder, and Prietula 1996). The Internet and computer technology have lead to yet another tightening of the relationship between knowledge and power, as companies have moved to knowledge-management and data-based marketing in order to optimize the use of the flood of information that has become so readily available to them. The relationship is clearly recognized in both the popular literature (e.g. Teresko 1996) and in the academic literature (e.g. Grimshaw 2001).

The objective of the research reported here is to find out whether or not an assumption can be made by marketers that organized information, or knowledge, leads to power in group purchasing decisions. Logic certainly supports such a view. If a group decision can be viewed as a negotiation (and it often is, if the marketing literature is taken at face value) then the acquisition of pertinent information not held by others in the group should lead to the information-gatherer gaining power in the decision. Again, the whole rationale for information gatekeepers and opinion leaders is based upon the idea that relevant information assists persuasion through either adding to the persuader’s perceived credibility or by controlling the flow of information (Allen 1984; Childers 1986).

The topic takes on a new relevance and importance when the possible impact of the Internet on group purchase decisions is considered. Patterns of influence are fairly well understood, in general, in both industrial and family group decisions. ButBif the knowledge-power link is indeed realBthen the emergence of the Internet has brought a whole new perspective to bear on group decision power structures, as Internet access is not controlled by marketing people in the same way that direct mail (electronic or not) and interpersonal communication channels are. Even the reach of information contained in advertising materials is generally understood by marketers, but there is no way, at this time, that the Internet habits of specific role-players in decision groups can be predicted satisfactorily. If knowledge acquisition is enhanced through Internet use and knowledge does indeed lead to power, then the familiar patterns of power in group purchase decision centers may well be disrupted in hard-to-forecast ways. A search of the literature (not just the marketing literature, but several related literatures as well) reveals that the issue of knowledge and power in group decisions has remained assumptive; rigorous attempts to verify the relationship have never been reported. This gap in the literature provides an motivation and a focus for this research.

RESEARCH METHOD

Hypothesis development

The basic contention under consideration is that domain-specific information acquisition will lead to an increase of decision-power in purchase situations involving the specific domain. To test this idea data is gathered from families which are carefully matched on a number of relevant dimensions, in order to minimize contamination from alternative explanations. First, the Internet usage of adolescent sons with respect to their parents is ascertained and, second, the family buying power patterns exhibited for thirteen products are identified. There is, once again, only a single, simple research hypothesis:

H1: The acquisition of domain-specific knowledge by an individual will enhance the power of that person, within the relevant domain, in a group-decision situation.

Sample

As the drive of this research is to build theory and not simply to provide empirical information, representation of the sample was sacrificed in favor of control. The literature on family decision making is well established, and a number of factors have emerged which are understood to exert an influence on the power of the family decision role-players. Control of these factors is discussed below.

Adolescent boys were selected as the focus for the research for several reasons. In order to give the effect under consideration a good chance to show, adolescents were chosen as they are more likely to have input in family decisions than their younger siblings (Foxman, Tansuhaj, and Ekstrom 1989; Lee and Beatty 2002). This is particularly pertinent as the knowledge under consideration is being acquired via the Internet; adolescent children are more likely to be Internet users than their younger siblings, especially with respect to the use of the Internet for purchase-related search. Boys were selected, rather than girls, as a control measure. Given that the focus falls on families whose eldest child is a 13 to 15 year old son, the family life cycle (which has been shown to have a strong impact on family decision patterns) is also effectively controlled.

Ethnicity obviously does have an influence on family decision patterns, as much empirical research has demonstrated (e.g. (Lee and Marshall 1998; Na, Marshall, and Son 2003). An attempt was made to optimize control by limiting even sub-cultural ethnic participation, so that whatever effect emerges can be reasonably ascribed to the information-acquisition manipulation and not to a contaminant. Consequently, responses from minority ethnic groups were removed, so all 255 nuclear families providing data are from the same ethnic group.

Social class is strongly related to culture, and there is an established literature that shows that lower- and upper-class families tend to exhibit a stronger, more traditional sex-role orientation than the more liberal, middle and upper-middle classes, resulting in a more democratic process in family decisions (Lee and Marshall 1998). As social class is such a difficult construct to operationalize an employment proxy was used here, and all the respondent families were drawn from the white-collar group, who are more likely to have Internet access than blue-collared.

Data Preparation

    Independent variable, Comparative Internet usage

The extent of Internet usage, for the purposes of acquiring purchase-related information, is quite hard to assess objectively. The first item collected was a straightforward question asking respondents to state the average number of hours per month which were spent seeking product/service-specific information on the Internet. Although this measure has strong face validity, there are a number of problems with it regarding reliability, as not only is it difficult to accurately recall the "average hours" spent doing almost anything, but also it is very hard to distinguish between directed search and casual surfing. A single item is, in any case, intrinsically unreliable. Consequently, a seven-item scale was also constructed, asking each son, mother and father about their preferences and willingness to search for information prior to purchase.

After deleting one (reverse-scored) item the scale has an Alpha of .82 for sons, and .92 for both mothers’ and fathers’ usage. In order to test the validity of the usage scales, they were compared to the relevant single-item statements concerning the number of hours each respondent claimed to spend searching for data in an average month. In each case there is a low, but positive and significant correlation between them (Sons’ r=.24, p=<.001; Fathers’ r=.34, p=<.001 Mothers’ r=.48, p<.001). Even the fact that sons’ correlation is lower adds to the validity, as it seems likely that sons would have greater difficulty than their parents in sorting out in their minds the difference between searching for domain-specific information and Internet cruising.

Furthermore, there is a need to compare comparative, rather than absolute usage of the Internet for information search purposes, of the adolescents. That is, if one son has an absolute high usage rate but his parents have an even higher exposure, then his real, comparative Internet usage is less than another son with an absolutely lower usage but whose parents barely use the medium. Thus sons’ Internet usage was re-expressed as a proportion of the total family Internet use.

An inspection of this scale for Sons shows the distribution to be somewhat leptokurtic (kurtosis=1.59), and slightly skewed (skewness=.76). This centralizing tendency was expected both because of the homogeneity of the respondent families and the large sample size. There do not seem to be any natural breaks in the distribution of the variable, however, so the data set was divided into three equal groups and only the two extreme groups were used in consequent analysis. Thus, the final sample is composed of two homogenous groups of nuclear families, each fulfilling the sample requirements concerning ethnicity, family life cycle and social class. Group 1 consists of 83 families (mean Sons’ Comparative Internet Usage=30.1, SD=4.5); while Group 2 consists of 85 families (mean Sons’ Comparative Internet Usage=49.2, SD=7.6). This represents a significant difference in comparative Internet usage (t=19.7, p<.001).

    Dependent variable, Decision Power

Measurement of decision power follows documented prior research (Lee and Beatty 2002; Mangleburg 1990). Data were collected from each of the three family members, on a 100-point, constant sum scale, regarding the perception of each others’ (and their own) decision power for 13 products, over three decision stages (Initiation, negotiation and final decision). Means were then calculated across each decision stage to yield a single number representing the mean power of each family member for each product.

ANALYSIS

The analytical focus

First, it should be noted that the decision topics selected were deliberately chosen, from topics used previously, to offer a range of products with regard to the expected change of influence of an adolescent son (Mangleburg 1990), particularly when he becomes more Internet enabled. Thus it was thought that sons would try to exert influence in the selection of their own sports clothing, music CDs, computer hardware/software and computer games. It was further considered that these are not "bookish" young people, so books might not be important to them and neither might family activities such as eating out, going to the movies or going on family holidays. Although family decisionsBsuch as the purchase of TV and audio equipmentBare less easy to hypothesize about, there seems at least a possibility that these products might come under the influence of a maturing, Interne-savvy young son. Certainly it was not expected that their parents’ clothing or banking activities would hold the remotest interest for these teenagers!

Factor analysis, purchasing power

Factor analysis of the purchasing power of sons for the 13 products was undertaken using SPSS. This is a Q-type analysis, where there is an assumption about an underlying change of behavior of the respondents in relationship to the type of product being considered, rather than an R-type where the relationship between the factors is of primary importance. The Kaiser-Meyer-Olkin measure of sampling adequacy is .77, and Bartlett’s test of sphericity is both high (806.4) and highly significant p<.001). Furthermore, the number of observations is 168 and the number of variables (products) is only 13; giving a satisfactory observations-to-item ratio (see, for instance, Hair et al (1995)).

General Least Squares analysis was used to extract the factors, as there are only 13 items in the variable list and several fail to reach the .6 communality level suggested by Gorsuch (1983). An orthogonal rotation was used, because uncorrelated factors are a preferable output of the analysis which is intended for use in further analysis by ANOVA. A VARIMAX rotation was selected, (although a test showed that neither EQUIMAX, PROMAX nor QUARTIMAX yield significantly different factor solutions).

Three factors emerged from the analysis, with Eigenvalues of 3.74, 2.79 and 1.45 respectively; the total variance explained is 62%. The Rotated Factor Matrix is shown in the Table. At a significance level of 0.05 and a power level of 80%, the sample size of 168 would conservatively suggest that loadings exceeding .45 should be considered as significant.

There is little ambiguity in the data displayed in the Table; the factors seem quite clear and are easy to name. Factor 1, "Youth-centric", consists of the few product decisions that it was considered adolescent males would care to exert an influence on and where their new-found knowledge might give them an edge, although books do feature on the list and were not expected to. The second factor, "Parent-dominated", consists of items where a youth would normally be excluded from participation (later analysis shows a mean comparative purchase influence for sons of only 12%). The final factor, "Syncratic", represents the joint decisions that are made in the light of fairly extensive product knowledge of all the participants, whether or not they access the Internet (again, later analysis shows a fairly equal decision influence here).

TABLE

FACTOR LOADINGS

The inclusion of books in Factor 1 surprised. Hence, an investigation was undertaken, by t-test, to gauge the extent to which Internet empowerment effects the influence children have in the selection of books, and it was found that there is no change whatsoever. This could, perhaps, be a function of their already-strong influence in the area (in excess of 50% of the family’s decision weight). It was consequently determined to remove this item and form a new "Youth-centric" factor, consisting of CDs, computer games, sports products and computer hardware/software.

Testing the hypothesis; Change in Influence

Three separate ANOVA models are run to test for differences in power-structure of sons between the two groups. There is no significant difference for either the Parent dominant or the Syncratic factors, but the mean difference for the Youth-centric factor between the high-Internet group and low is significant (M_{(High usage)}=57, M_{(Low usage)}=48; p<.001, F=13.66). Hay’s c² reveals an effect size of .07 for the Youth-centric factor; Cohen describes this as "medium-sized".

Decision stages

The decision influence data were collected over three stages; Initiation, Negotiation and Final decision. The effect of Internet-enabled knowledge is significant at each stage (Initiation_(High)=68, Initiation_(Low)=59, F=12.3, p=.001; Negotiation_(High)=59, Negotiation_(Low)=51, F=8.7, p=.004; Final_(High)=43, Final_(Low)=35, F=9.0, p=.003).

DISCUSSION

The study reported above offers strong evidence that the acquisition of information specific to a purchase domain, which increases an individual’s knowledge, will add influence to that individual in a group decision situation. The precise mechanics involvedBthat lead from information acquisition to power in group decision-makingBare not clear for the research. It seems probable that new knowledge merely aids it’s owner to negotiate his/her point of view more persuasively, but it is also possible that the knowledge provides more credibility to the role-player in the eyes of his/her fellow decision-group members. Nevertheless, the fact that there does seem to be a significant change in the balance of power in group (family, at least) decisions based upon the information-acquisition strategies of the decision protagonists is of interest and importance in itself.

Implications for marketing practice

As the main research reported here is fundamentally of a theory-development rather than empirical nature, the sample design focused on control rather than representation and thereby sacrificed much empirical relevance. The fact that the research design used worked so well, though, holds real promise for any company marketing products or services which might be purchased in a situation where group influence comes into play. It is not complicated research to replicate for a particular product/service situation and inspection of the results of a similar survey with an appropriate sample and decision-objects would reveal very valuable strategic information about not only the influence structures at play but also the effect of providing information to particular role-players.

Implications for marketing theory

Information acquisition does lead to power-enhancement in group decisions. Although this important fact has hitherto simply been assumed, it seems as if it is a reality. Moreover, the relationship does not always hold, as was evidenced above by the failure of sons’ knowledge enhancement to affect their decision power for products in which his parents were equally, or more, powerful and interested. This effect requires yet more rigorous researchBpossibly using more sophisticated modeling techniquesBto fully unravel precisely what variables do intervene between knowledge and power and moderate the knowledge-power relationship.

Another aspect of interest is the emergence of Internet as a knowledge-enhancement vehicle. The interest stems from the difficulties surrounding assessment of individual’s exposure to Internet. Thus it seems as if further research could also be fruitfully conducted to ascertain the relative importance of this new vehicle vis-a-vis other, more conventional media.

REFERENCES

Allen, Thomas J. (1984). Managing the Flow of Technology. Cambridge: MIT Press.

Carter, Charles and Harry Scarbrough (2001). Regimes of knowledge, stories of power: A treatise on knowledge management. Creativity an Innovation Management, 10 (3), 34-45.

Childers, T. (1986). Assessment of the psychometric properties of an opinion leadership scale. Journal of marketing research, 23 (May).

Foxman, Ellen R., Patriya S. Tansuhaj, and Karin M. Ekstrom (1989). Family members’ perception of adolescents’ influence in family decision making. Journal of Consumer Research, 15 (March), 482-91.

Gorsuch, Richard L. (1983). Factor Analysis. Hillsdale, NJ: Erlbaum.

Grimshaw, Davis J. (2001). Harnessing the power of geographic knowledge; The potential for data integration in an SME. International Journal of Information Management, 21 (3), 14-27.

Hair, Joseph F., Rolph E. Anderson and Ronald L. Tatham (1995). Multivariate data analysis: with readings (4th ed.). Englewood Cliffs. NJ: Prentice Hall.

Hislop, David, Simon Newell, Harry Scarbrough, and John Swan (2000). Networks, Knowledge and Power: Decision making, politics and the process of innovation. Technology Analysis & Strategic Management, 3 (12), 21-29.

Lee, Christina Kwai Choi and Sharon E. Beatty (2002). Family structure and Influence in family decision making. Journal of Consumer Marketing, 19 (1), 24-41.

Lee, Christina Kwai Choi and Roger Marshall (1998). Measuring influence in the family decision making process using an observational approach. Qualitative Market Research, 1 (2), 88-98.

Mangleburg, Tamara (1990). Childrens= influence in purchasing decisions: A review and critique. Advances in Consumer Research, 17, 813-25.

Na, WoonBong, Roger Marshall, and Youngseok Son (2003). Purchase role structure in Korean families: Revisited. Psychology & Marketing, 20 (1), 47-57.

Teresko, John (1996). Data warehouses; build them for decision-making power. in Industry Week.

Weingart, Laurie R., Elaine B. Hyder, and Michael J. Prietula (1996). Knowledge matters: The effect of tactical descriptions on negotiation behavior and outcome. Journal of Personality and Social Psychology, 70 (6), 1205-17.

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