INTERACTIVE VERSUS ONE WAY PROMOTION:

THE EFFECT OF INFORMATION, TRANSFORMATION AND INVOLVEMENT

Dr Nigel Pope (author for contact re copy editing)

Faculty of Commerce and Administration

Griffith University

Nathan

Queensland 4111

Australia

Tel: +61 7 3875 7443

Fax: +61 73875 7126

Email: n.pope@mailbox.gu.edu.au

University of Alaska, Anchorage

University of Sydney

INTERACTIVE VERSUS ONE WAY PROMOTION:

THE EFFECT OF INFORMATION, TRANSFORMATION AND INVOLVEMENT

This paper examines viewer responses to promotional material in both paper and electronic form. The instrument used was an adaptation of the Puto and Wells Informational versus Transformational advertising questionnaire and the Mittal Purchase Decision Involvement Inventory. It was proposed that respondents to the electronic copy would have similar responses across involvement levels. By contrast, it was suggested that respondents to the paper copy would display differences in their responses based on their levels of product involvement. The proposition was confirmed in that, while the group exposed to the paper version of the material showed significant differences in brand confidence according to their product involvement level, such a difference did not occur in the group exposed to an electronic version of the same material. This suggests that the possession of a web presence by companies - and the inclusion of promotional material in those sites - will enhance the overall confidence in their brands for both high and low involvement customers.

Running Headline: INTERACTIVE ADVERTISING

Keywords: advertising; world wide web; involvement

In consumer behaviour literature, consumers have been typically viewed as cognitive thinkers who engage in a step by step mental process for making purchase decisions and evaluating promotional material (Biehal, Stephens and Curlo, 1992; Fishbein and Ajzen, 1975). Such a view would suggest that an individual would undergo the same step by step mental approach in "surfing" through a web site (Donthu and Garcia, 1999; Korgaonkar and Wolin, 1999). Another view is that of the affective school. It adds a different element of thinking to the cognitive perspective by paying attention to the emotions or affective responses, which motivate purchase decisions and evaluation of promotional material. This view would suggest that, rather than following a linear thought process in viewing a web site, an individual absorbs affective cues through an experiential process of surfing. This would further imply that the propensity to engage in surfing a particular site is based on an individual’s involvement with the subject matter of the web site.

Research into the Internet shopper has focussed on demographic characteristics of the individual. For example, this person has been found to be older and of higher income than the non-shopper (Donthu and Garcia 1999; Korgaonkar and Wolin 1999), slightly more likely to be a male (Korgaonkar and Wolin 1999), and less risk averse (Donthu and Garcia 1999; Korgaonkar and Wolin 1999). There is also evidence that the Internet shopper is convenience-oriented (Donthu and Garcia 1999; Korgaonkar and Wolin 1999), innovative and variety-seeking (Donthu and Garcia 1999). He or she does not appear to be brand- or price-sensitive (Donthu and Garcia 1999). Importantly, no research to date has examined the nature of the web-site involved in communicating with the viewer and the particular trigger intended to be used on the potential consumer.

In this paper, we present the results of an examination of viewer responses to a promotional web site. We modified the Puto and Wells’ (1984) Informational and Transformational Scale for use with web sites rather than advertisements. That particular scale is one that has been successfully used in a range of applications and not just advertising (see, for example, Erevelles, 1998; Meenaghan and Shipley, 1999). Our modification suited the web site analysed, http://www.levi.com/, as, at the time of our research, it was a promotional and informational site with no transaction facility available. We begin our discussion with an examination of the constructs involved: cognition versus affect, and informational versus transformational advertising.

Research on cognition and affect originated in the discipline of psychology, and is dominated by the works of Fishbein and Ajzen (1975), Lazarus (1981), and Zajonc and Markus (1982). Their research spread to the marketing discipline and has significant implications in understanding how consumers go through purchase decision-making processes and develop attitudes toward brands. There is, however, some debate as to which comes first, cognition or affect. On the one hand, Zajonc and Markus (1982) argue that affective factors play the more important role in development and maintenance of preferences. On the other hand, Lazarus (1981) argues that to experience any emotions, cognitive activity must be undertaken first.

Whichever does come first, it is accepted theory that thinking and mental processing represent cognition, while emotions and feelings come under the realm of affect. These are mirrored in the concepts of informational (cognitive) and transformational (affective) promotion content. Characteristics of informational promotion content are:

factual, relevant information about the brand

information immediately and obviously important to the potential consumer

data which the consumer accepts as being verifiable

(Erevelles, 1998; Puto and Wells, 1984).

On the other hand, transformational promotion works by ‘…transforming the experience of using the advertised brand to reflect the feelings evoked during the exposure to the ad’ (Puto and Wells, 1984, p.638). Thus, transformational advertising is characterized by the following:

arousing or appealing to the senses and emotions

makes the experience of using the product richer, warmer, more exciting, and/or more enjoyable, than that obtained solely from an objective description of the advertised brand

creates feelings, images, meanings, and beliefs about the product or service that may be activated when consumers use it and thus ‘transforms’ their interpretation of the usage experience

(Erevelles, 1998; Puto and Wells, 1984).

While Puto and Wells (1984) suggest that ads and other forms of promotion can be assessed in terms of either transformational or informational content, they do not address the issue of how the two forms interact in the one form of promotion. This is understandable. With conventional advertisements, the opportunity for the viewer to explore is limited by both the size and the content of the message. On the other hand, a web site allows a viewer the opportunity to explore and thereby offers the researcher the chance to examine differences between transformation and information response to the one promotion message source.

The way that an individual moves through a web site has been described as one of flow (Hoffman and Novak, 1996), and is related to attention and involvement. Again, this has been discussed in terms of cognition and affect. Hoffman and Novak (1996) suggest that flow is a measurable construct and has resulted in the merging of informational and transformational messages into the one site. It is our contention that, logically, the propensity to enter a web site and remain or leave will be governed by the attention and involvement at immediate entry as well as throughout the course of the stay at the site. In that sense, involvement would be an antecedent to flow, as well as a part of it.

If this is the case, we find it reasonable to assume that if one has a particular involvement with the product one is more likely to visit the site in the first place, and to be more responsive to promotional communications which are presented, be they affective or not. As a result of this, we are interested to know whether, through the use of interactive promotional communications, it is possible to reduce the difference in response among respondents who have different involvement levels with the product under examination. This leads to our research proposition:

P1: Differences in response to marketing communications that result from different involvement levels will be reduced in interactive communication contexts as opposed to one way communication contexts.

We are also interested to know in which areas of response, cognitive (informational) or affective (transformational), this may occur. Our reason for asking this is as follows. It is possible that the possibility of interactivity may of itself have an impact emotionally. That is to say that the flow effect identified by Hoffman and Novak (1996) may work on the affective level more than the informational. Equally, it may be that the ability to check information and verify it through flow will result in a stronger informational rather than transformational effect.

SAMPLING FRAME AND CONTEXT

In order to test our research proposition and also to examine the applicability of the Puto and Wells (1984) scale to the Internet, we needed a brand and product context that was (1) promoted via the Internet and, (2) would be familiar and likely to be used by our sample. We selected the category of Jeans-ware and the brand Levi’s as being able to fulfill these criteria. The decision was a subjective one on our part and was made on the basis of the brand’s well-established presence in the market place and the Internet.

Our sample frame was a class of 149 undergraduate marketing students from an Australian east-coast university. The group was surveyed initially to establish age, gender and cultural background. From this survey, 86 subjects were paired on the basis of age and gender and randomly assigned to two groups.

MEASURES AND SURVEY INSTRUMENT

Our dependent variables in this study related to responses to the web-site. The independent variable was purchase decision involvement measured through Mittal’s (1995) PDI. We also allowed for stimulus presentation (either print or electronic form). We used Puto and Wells’ (1984) Informational and Transformational Ad Content Scale to measure web-site response. This scale allows for two constructs to be measured (informational and transformational) by summation of certain item responses. We anticipated some difficulties in application of this instrument to a new medium.

The first difficulty we addressed was that of nomenclature in some of the questions which were originally designed for television commercials. This required deletion of some items. Items for deletion were selected by an expert panel of four academicians chosen from colleagues at an east coast Australian university. Five items were deleted (statements 2, 12, 16, 17 and 18 of the original scale). Elsewhere in the scale, the word commercial was replaced with the words "web-site". The original scale is presented at Appendix A.

The next problem we anticipated was a result of the first: that changes to the original scale, especially when tested in a new medium, would adversely affect internal reliability. To overcome this we decided in advance that we would test for item reliability within the informational and transformational contexts and further conduct factor analyses of the item responses to identify any additional measurable constructs. Any identified constructs apart from informational and transformational would be treated as dependent variables.

Four commonly used scales of product involvement have been tested for content validity and unidimensionality by Mittal (1995). These are Zaichkowsky’s Personal Involvement Inventory (hereafter, PII); Laurent and Kapferer’s Consumer Involvement Profile (hereafter, LandK); the Foote, Cone and Belding scale (hereafter, FCB); and Mittal’s Purchase-Decision Involvement scale (hereafter, PDI) (Laurent and Kapferer, 1985; Zaichkowsky, 1985; Ratchford, 1987; Mittal, 1989). The scales were tested across six parameters of unidimensionality, convergent validity, nomological validity, parsimony, simplicity and minimum response set bias.

Of the four scales examined, the FCB and PDI, in that order, performed best. Each is related directly to the construct of purchase decision involvement. The FCB is a three item scale, and the PDI four item. The difference lies in the addition in the PDI of perceived brand difference. The remaining three items in each scale relate to thinking/feeling dimensions. We elected to use the PDI scale as its use would not detract from the overall validity and offered an extra item for examination.

The resulting independent variable of involvement we treated as categorical, either high or low. We designated respondents who scored higher than the mean as high involvement and those who scored below as low involvement. That is to say, we merged the values into single bins, thereby reducing value spaces in accordance with procedures suggested by Weiss and Indurkhya (1998). We allowed for stimulus presentation by treating the groups discretely.

ADMINISTRATION AND ANALYSIS

One group of subjects was presented with a color printout of each page of the corporate web-site http://www.levi.com/, the other was presented with the site electronically in a computer laboratory. Each group was given twenty minutes perusal time, after which they were asked to complete our survey. All eighty-six responses were useable.

Dependent variable data were subjected to item reliability and factor analyses. Each group’s responses were then separated and a discriminant analysis of the data conducted with the grouping variable being involvement (high or low), and the dependent variables being the extracted factors from the Puto and Wells scale.

As we had anticipated, our revised Puto and Wells scale lacked reliability in some aspects. While the total scale showed a Cronbach µ of .70, when divided into its two constructs it showed µ ’s of .33 (Informational) and .70 (Transformational) respectively. We therefore conducted a varimax normalized factor analysis of the items contributing to the Informational construct. This produced three factors with items possessing loadings greater than 0.6. We named these factors educational (two items), brand differentiation (two items) and confidence (one item) as these were the concepts they represented. Together these factors explain fifty-six percent of the total variance. Results of this factor analysis appear at Table 1.

<<insert Table 1 about here>>

Levene’s test for homogeneity of variance indicated an abnormal distribution for the dependent variable Brand Differences (F=7.96, df=1,84, p<.05). This variable was therefore removed from the analysis. All other variables showed normal distributions under this test (p>.05).

Our measure of Purchase Decision Involvement showed strong internal reliability (µ = .89). The mean was 20.27 (sd=4.51) with a maximum possible score of 28. When divided into bins of either high or low involvement in accordance with Weiss and Indurkhya (1998), the print group had 23 high involvement subjects and 20 low. The electronic group had 22 high and 21 low.

TEST OF RESEARCH PROPOSITION

We suggested in our research proposition that because of the nature of interactive communication, differences in communication due to involvement level with a brand or product would decrease when an individual was stimulated by web media as opposed to print copies of the same material. To test this proposition, we conducted discriminant analyses of responses from our two groups, using the same dependent and grouping variables.

The predicting power of the predictor variables was determined by calculation of Wilks’ Lambda (.70). In the case of the group who viewed the material in print, the effectiveness of the discrimination was significant along only one vector, that of confidence in the brand (F=16.70 df=1,39 p<.001), in regard to which those with high involvement showed a higher score than those with low involvement (4.3 as opposed to 3.5 on a scale of 1 to 5). No other variable was significant, and Confidence in the Brand accounted for 72% of the predictable group variation. Results of this analysis appear at Table 2.

<<Insert Table 2 about here>>

In the case of the group who viewed the material electronically, no significant differences were found. On this data therefore, we accept our research proposition that involvement based differences in responses to marketing communications are ameliorated by electronic delivery via web sites.

DISCUSSION

Earlier literature has suggested that humans are rational creatures who enjoy logical, linear thought processes which can be moderated by affective states and responses (see, for example, Fishbein and Ajzen, 1975). Whether or not cognition precedes affect or vice versa is still debated (Lazarus, 1981; Zajonc and Markus, 1982). Irrespective of that debate, it is accepted that advertisements can be either informational (cognitive) or transformational (affective) to varying degrees (Puto and Wells, 1984), and Hoffman and Novak (1996) suggest that informational and transformational messages can be merged in interactive promotional web-sites.

In our research proposition we suggested that individuals with high product involvement would differ from individuals with low product involvement in their responses to paper versions of promotional material. It was our contention that these differences would be reduced when individuals were exposed to the material in interactive, electronic form. We also sought to establish if this occurs more in the informational or transformational areas.

We found differences between high and low involvement respondents in only one area, that of confidence in the brand, a sub-set of the informational category. Confidence in the brand was higher among those respondents with high involvement. This was only found in the group exposed to the print version of the material, and not among those exposed to the electronic medium. We consider two possibilities might account for this.

First, it is possible that those who are highly involved with the product category, in this case jeans, already have considerable knowledge of the brands within the category and the brand name itself (Levi’s) has connotations of quality and confidence already embedded in the respondents’ perceptions. Such a condition has support from the literature (for example, Biehal and Chakravarti, 1982; Nedungadi, 1990). A possible counter to this argument is that the brand may be so well known that even those with low involvement could also hold such perceptions. We note that this possibility also suggests that the promotional material is, to an extent, redundant.

A second possibility is that those with higher involvement were more inclined to peruse the material and pay greater attention to the material provided. This would suggest that the promotional effort in this form is redundant when applied to low involvement consumers if one is seeking to increase their confidence in the brand. It is also possible that a combination of these two possibilities is at work.

Respondents from the electronic group did not show a significant difference in any area. Of the possibilities discussed above, regarding the print group, only the second can be applied to account for the lack of significance. That is to say, the reduction in differences between low and high involvement respondents as we expose them to one way or interactive material may well be accounted for by the inclination of respondents to peruse the promotion. We would suggest that interactive, electronic material is easier to examine and "flick" through than its print counterpart, allowing for greater absorption of information resulting in brand confidence.

No significant differences were found between high and low involvement respondents in the area of transformational response for either print or electronic groups. This surprised us. We had expected that this would be more likely in the print area than the informational response. On consideration however, we came to the conclusion that this may well be a function of the site itself, which was designed to provide information about the company rather than to create a mood. Overall, we are unwilling to make a commitment as to the reasons for this.

LIMITATIONS AND FURTHER RESEARCH

Before drawing conclusions regarding this research, it is necessary to raise several limitations. First, we only examined one site. This research was intended as an exploration only, so replication is necessary and we would welcome it. We believe that a valid study would involve the examination of several sites from several product categories and intended for different purposes, either transformational or informational. Again, a print comparison would be necessary.

Second, our sample, although paired, was a student sample and a multi-product, multi-site, multi-purpose investigation would need to select respondents from a broad cross-section of the community. There may have been some effect of students knowing they were in an experiment but they were not informed beforehand of the nature of the experiment or of the activity assigned to the other group. So, while we must allow for that, we do not believe it had any major effect. Third, the scale we used needs further work to adapt it for web site analysis. This is particularly so in the informational area.

CONCLUSION

In this research we sought to examine whether interactive, electronic marketing communication reduced differences between high and low involvement respondents in their responses when measured in both informational and transformational contexts. We found that this occurred only in informational contexts and then only in the case of confidence in the brand. We attribute this to easier perusing of material in electronic format, which in turn allows those of low involvement to browse and absorb information more easily than in print format. We can offer no explanation for why this does not apply to other informational constructs in our instrument.

An important implication of our research relates to management attitudes to the use of the web for promotional purposes. It appears that customers or potential customers – regardless of product involvement - will find increased brand confidence in electronically delivered material than in the same material delivered in paper form. It is also possible that mere presence on the web has the effect of inspiring confidence. If that is the case, companies that do not have web sites, or that have non-promotional sites, may be doing themselves a great disservice with regard to generating overall confidence in their brands.

The non-impact of electronic presentation in the transformational context we attribute to the nature of the site we examined. We believe that we have established the existence of an effective difference in responses due to electronic delivery of printable promotional material, but that further research, in the context of the limitations we discussed above, is necessary to further explore this phenomenon.

REFERENCES

Biehal, G. and Chakravarti, D. (1982) ‘Information-Presentation Format and Learning Goals as Determinants of Consumers’ Memory Retrieval and Choice Processes’, Journal of Consumer Research 8 (March) 431-441.

Biehal, G., Stephens D. and Curlo, E. (1992), "Attitude Toward the Ad and Brand Choice," Journal of Advertising, 21 (3), 19-36.

Donthu, N. and Garcia, A. (1999) "The Internet Shopper," Journal of Advertising Research, May/June, 52-58.

Erevelles, S. (1998) "The Role of Affect in Marketing," Journal of Business Research, 42 (3) 199-215.

Fishbein, M. and Ajzen, I. (1975) Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research Addison-Wesley Reading, MA.

Hoffman, D. L. and Novak, T. P. (1996) ‘Marketing in Hypermedia Computer-Mediated Environments: Conceptual Foundations’, Journal of Marketing 60 (3) 50-68.

Kargaonkar, P. K. and Wolin, L. D. (1999) "A Multivariate Analysis of Web Usage," Journal of Advertising Research, March/April, pp. 53-68.

Laurent, G. and Kapferer, J. N. (1985) ‘Measuring Consumer Involvement Profiles’, Journal of Marketing Research 22 41-53.

Lazarus, R. S. (1981) ‘A Cognitive Scientist’s Reply to Zajonc on Emotion and Cognition’, American Psychologist 36 222-223.

Meenaghan, T. and Shipley, D. (1999) "Media effect in Commercial Sponsorship," European Journal of Marketing, 33 (3/4) 328-348.

Mittal, B. (1995) ‘A Comparative Analysis of Four Scales of Consumer Involvement’, Psychology and Marketing 12 (7) 663-682.

Nedungadi, P. (1990) ‘Recall and Consumer Consideration Sets: Influencing Choice without Altering Brand Evaluations’, Journal of Consumer Research 17 (December) 263-276.

Puto, C. P. and Wells, W. D. (1984) ‘Informational and transformational Advertising: the Differential Effects of Time’, Advances in Consumer Research 11 638-643.

Ratchford, B. R. (1987) ‘New Insights about the FCB Grid’, Journal of Advertising Research, August/September 24-38.

Weiss, S. M. and Indurkhya, N. (1998) Predictive Data Mining: A practical guide, San Francisco, CA: Morgan Kaufmann.

Zaichkowsky, J. L. (1985) ‘Measuring the Involvement Construct’, Journal of Consumer Research 12 (December) 341-352.

Zajonc, R.B. and Markus, H. (1982) ‘Affective and Cognitive Factors in Preferences’, Journal of Consumer Research 9 (September) 123-131.

APPENDIX A

INFORMATIONAL AND TRANSFORMATIONAL AD CONTENT

(Puto and Wells 1984)

I learned something from this commercial that I didn’t know before about (this brand).

I would like to have an expertise like the one shown in the commercial.

The commercial did not seem to be speaking directly to me.

There is nothing special about (this brand) that makes it different from the others.

While I watched this commercial, I thought how (this brand) might be useful to me.

The commercial did not teach me what to look for when buying (this product).

This commercial was meaningful to me.

This commercial was very uninformative.

(This brand) fits my lifestyle very well.

I could really relate to this commercial.

Using (this brand) makes me feel good about myself.

If they had to, the company could provide evidence to support the claims made in this commercial.

It’s hard to give a specific reason, but somehow (this brand) is not really for me.

This commercial did not really hold my attention.

This commercial reminded me of some important facts about (this brand) which I already knew.

If I could change my lifestyle, I would make it less like the people who use (this brand).

When I think of (this brand), I think of this commercial.

I felt as though I were right there in the commercial, experiencing the same thing.

I can now accurately compare (this brand) with other competing brands on matters that are important to me.

20. This commercial did not remind me of any experiences or feelings I’ve had in my own life.

I would have less confidence in using (this brand) now than before I saw this commercial.

It is the kind of commercial that keeps running through your head after you’ve seen it.

It’s hard to put into words, but this commercial leaves me with a good feeling about using (this brand).

Table 1: Factor loadings of informational items in Puto and Wells’ (1984) scale

Note items 4, 8 and 21 are reverse scored

Table 2: Results of discriminant analysis for Print Subjects with grouping variable High or Low Involvement by Confidence, Education and Transformation

N=43