An Investigation of Self Reported Search Stage Versus Point of Purchase Information Search Behavior

Donald R. Lehmann, Columbia University
William L. Moore, Columbia University
ABSTRACT - This paper relates actual and stated search behavior over time. Specifically, subjects acquired information on, made choices of, and consumed their choices of five health breads once a week for six weeks. Search behavior was monitored using a behavioral process technology. Subjects also rated their search behavior each week by indicating which of five scenarios best described their behavior. These two measures were found to be significantly related.
[ to cite ]:
Donald R. Lehmann and William L. Moore (1980) ,"An Investigation of Self Reported Search Stage Versus Point of Purchase Information Search Behavior", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 733-736.

Advances in Consumer Research Volume 7, 1980     Pages 733-736

AN INVESTIGATION OF SELF REPORTED SEARCH STAGE VERSUS POINT OF PURCHASE INFORMATION SEARCH BEHAVIOR

Donald R. Lehmann, Columbia University

William L. Moore, Columbia University

ABSTRACT -

This paper relates actual and stated search behavior over time. Specifically, subjects acquired information on, made choices of, and consumed their choices of five health breads once a week for six weeks. Search behavior was monitored using a behavioral process technology. Subjects also rated their search behavior each week by indicating which of five scenarios best described their behavior. These two measures were found to be significantly related.

INTRODUCTION

Both the problem solving approach to consumer choice (Howard, 1963, 1977, and Howard and Sheth, 1969) and the related information processing approach (Bettman, 1979; Chestnut and Jacoby, 1977; Lehmann, 1978; and Newell and Simon, 1972) consider the entire choice process rather than just the purchasing act. A key part of this process is external search. Marketers have long recognized the importance of external search and have conducted numerous studies on the determinants of external search (for reviews see Bettman, 1979; Engel, Kollatt, and Blackwell, 1973, pp. 376-83; and Newman, 1977). Still a recent review states, "the empirical evidence on [external) search and its determinants is very limited." (Newman, 1977, p. 93) One of the reasons for this lack of findings is the inability of researchers to measure search. This paper extends previous work on this topic by examining the link between actual and stated search behavior in a longitudinal laboratory study.

BACKGROUND

Howard has suggested that prepurchase behavior can be classified into three categories: Extensive Problem Solving, Limited Problem Solving, and Routinized Response Behavior. The stage that a person is in depends primarily on familiarity with the product class and the available alternatives. Bettman (1979) proposes that consumers search for information in pursuit of particular goals and this search may be internal (memory) or external (packages, Consumer Reports, advertisements, etc.) Furthermore, he suggests that an internal search is usually performed initially and it is followed by external search if there is insufficient information in memory to make a decision. Thus, both authors argue that the amount of external search would decrease as a consumer learns more about the choice alternatives.

Most studies of external search have used self-reported search behavior, usually taken weeks or months after the purchase, as the measure of external search behavior. However, the validity of these findings might be questioned as Newman and Lockeman (1974) found "little or no correlation between observation-based and survey-based scores of in-store information seeking." This may be due to forgetting in some instances and may occur because individuals have great difficulty in accurately recalling the cognitive processes they use (Nisbett and Wilson, 1977; Smith and Miller, 1978).

In order to overcome problems with the after-the-fact, self-reported measures, marketers have recently focused on four methodologies to measure external search behavior: verbal protocols (Bettman and Zins, 1977; Payne and Ragsdale, 1978), eye movements (Russo and Rosen, 1975; Russo, 1978), chronometric analysis (Gardner, Mitchell, and Russo, 1978), and behavioral process techniques (Chestnut and Jacoby, 1977; Green, Mitchell, and Staelin, 1978; Holbrook and Maier, 1978; Jacoby, Chestnut, Weigl, and Fisher, 1976; Jacoby and Chestnut, 1977; Jacoby, Chestnut, and Fisher, 1978; and Jacoby, Szybillo, and Busato-Schach, 1977). This study employs the number of items accessed in a behavioral process methodology as the operationalization of actual search. Chestnut (1975) asked subjects to state how many bits of information they had collected in an immediate recall situation and found correlations between stated and actual of about 0.4. Jacoby and Chestnut (1977) found a median Spearman correlation of 0.22 between the number of items actually chosen and self-rated typical behavior across 30 attributes of breakfast cereals, .26 across 23 margarine attributes, and .32 across 13 headache remedy attributes. Similar results are reported by Jacoby et al. (1978) and VanRaiij (1977).

This paper extends the work in this area in two ways. First, stated search is measured by asking subjects to choose which of five search scenarios best described their behavior. Second, the relation between stated and actual search is studied longitudinally. In this study five new brands in an existing product class have been used. A longitudinal design was used to examine the changes in information acquisition because considerable learning should take place during the first six purchases.

Hypothesis

H1:  The self reported stage of search will be significantly related to actual search.

This is essentially a test for convergent validity between stated and actual search. It also tests the construct validity in terms of Howard's categorization.

H2:  The amount of search associated with stated search stage will decrease overtime.

This hypothesis recognizes that the amount of search should decrease over time. Moreover, since respondents will see a lower level of search as appropriate or necessary, the amount of search required for them to describe it as extensive will also decrease.

METHOD

Subjects

Approximately 150 students and staff members of the Columbia University Graduate School of Business were recruited during the summer of 1978 to participate in a bread purchasing study. They were told they would receive one loaf of bread each week and an amount of money at the end of the study that would be equal to the difference between the price of the brands they had chosen and the price of the most expensive brand in the study. (This was the least expensive and easiest to manage way to add a price dimension to the study) One hundred twenty subjects completed the entire experiment.

Test Product

Five types of bread from a small health food bakery were used in the experiment. The types were whole wheat, pumpernickel, date bran, cinnamon raisin, and five grain bread. While most of the subjects have heard of some of these "brands", the taste of bread varies considerably between bakeries so these brands should have been somewhat unfamiliar to the subjects.

Experiment

Subject tasks.  During the week prior to the experiment subjects were required to fill out a questionnaire covering areas such as familiarity with types of bread, normal consumption, importance of attributes when choosing bread, similarities of different types of bread, attitudes toward bread, and demographics.

Each week the subjects were required to choose one leaf of bread. Prior to their choice, they were able to acquire as much or as little information as desired about the brands on nine attributes. These attributes consisted of all the information found on the packages, the physical product (loaf of bread in its clear package), and estimated protein and calories. When the subjects asked for a piece of information other than physical product, they were given a card with that piece of information on it. When they asked to see the physical product, they were handed a loaf of bread enclosed in a clear wrapper (the normal packaging for these breads). After making their choice, subjects were given a short questionnaire concerning their satisfaction with last week's choice, breads consumed during the week, time pressure and other situational variables, and a self-reported description of their information search behavior.

At the end of the experiment, the subjects were required to fill out another questionnaire. The final questionnaire contained items on importance of attributes, similarity of types of bread, recall of the information they could have acquired during the experiment, and perceived complexity and realism of the task. At this time, the subjects were given an amount of money equal to the difference between the price of their six choices and the price of the most expensive alternative they could have chosen. A more detailed description of the study and additional results may be found in (Lehmann and Moore, 1979a, b).

VARIABLE OPERATIONALIZATION

Search

Stated.  Stated search was measured by asking the respondents to indicate which of five scenarios best described their behavior immediately after choosing a loaf of bread. The responses were:

1.  Searched extensively for information in order to make an informed choice.

2.  Gathered a few pieces of information to fill out gaps in my knowledge.

3.  Gathered a few pieces of information out of curiosity.

4.  Chose a bread in order to try it.

5.  Chose my favorite bread without gathering much information.

As all subjects were familiar with the general product category, they should have been engaging in either limited problem solving or routinized response behavior. Scenario 1 represents an attempt to capture limited problem solving. Scenarios 4 and 5 represent routinized response behavior. The other two scenarios attempt to capture a transition stage.

Actual.  Actual search was measured by number of pieces of information accessed in each week during the experiment. It did not include any information acquired by experience or word of mouth communications.

RESULTS

Table 1 gives the amount of actual and stated search for each week of the experiment. The total amount of information acquired dropped from 17.2 bits in week one to 4.5 bits in week six. The decrease between each pair of weeks was significant at the .05 level. Both the number of brands looked at and the number of attributes accessed decreased significantly over time, in accordance with H2.

Similarly the amount of search as measured by self reports changed significantly over time. During the first two weeks, "gathered information to fill out gaps in my knowledge" was the most common response. In the middle two weeks the biggest group of people said they "chose a bread in order to try it", and during the final two weeks the largest group indicated that they "chose my favorite brand without gathering much information". The percent of respondents who indicated they searched extensively decreased from 25% to 2.5% between weeks one and six. Hence both the stated and actual search measures follow the expected pattern.

In order to assess the correspondence between stated and actual search, the results were compared on a week-by-week basis. This was done by means of a dummy variable regression analysis which used the first four stated search categories as the independent variables and actual search as the dependent variables. The average number of bits of information chosen are portrayed in Table 2 and contain four major findings:

1.  The amount of information taken decreases for all the stated search categories over time except for the extensive category in accordance with H2.

2.  Week 3 is very different from the others. The fit is worse (R2 = .03) and the average number of pieces of information selected for each category is noticeably and significantly different. (This may have been caused by a vacation week between the second and third purchase occasions.

3.  The amount of information taken varies substantially across the stated search categories. "Searched extensively'' had by far the most actual search associated with it. People who answered "gathered a few pieces" out of either "curiosity" or "to fill out knowledge" or "chose a brand to try it" all used about the same average number of bits of information. Those who said they chose without gathering much information selected the least information, as expected. Hence H1 is not rejected.

4.  The subjects' ability to accurately classify their search strategy increased over the course of the experiment (R went from .22 in week 1 to .60 in week 6). This may simply indicate that the subjects learned how to participate in the experiment. On the other hand, some of the increase may be due to the subjects becoming more expert at recalling what they did, which suggests that self reports of experienced consumers may be more valid then those of less experienced consumers.

In summary, stated search is significantly related to actual search. While the degree of association is less than we would like (possibly due to the somewhat ad-hoc nature of the self-rated scale used), the level of relation between stated search measured by scenarios and actual search compares favorably with either direct numerical assessment (Chestnut, 1975) or self-rated typical behavior (Jacoby and Chestnut, 1977). Thus self-rated scenarios alter promise as measures of information search.

SUMMARY

This paper has related self-rated search measures with actual search. The results indicate that a scenario-based rating scale used immediately after a purchase has promise as a measure of stated search.

Further work in this area should center on the development and testing of a set of scenarios that could be used with a wide variety of product classes. In retrospect a primary problem with our scenarios was a lack of clarity and a confounding of choice and search. For example if a person search extensively and then chose one of the Brands to try it, he might not know whether to check response 1 or 4. Thus future work should focus on scenarios that describe either search only, or the list of scenarios should be expanded so several choice descriptions could go with each search scenario (e.g. searched extensively - chose regular brand, searched extensively - chose one based on package information).

TABLE 1

STATED AND ACTUAL SEARCH OVER SIX WEEKS

TABLE 2

ACTUAL VS. STATED SEARCH OVER SIX WEEKS

REFERENCES

Bettman, James R., (1979) An Information Processing Theory of Consumer Choice. Reading, Mass., Addison-Wesley Publishing Co.

Bettman, James R. and Michel A. Zins, (1977) "Constructive Processes in Consumer Choice," Journal of Consumer Research, Vol. 4, September, pp. 75-85.

Chestnut, R. W., (1975) "The Expenditure of Time in the Acquisition of Package Information," unpublished M.S. Thesis, Purdue University.

Chestnut, R. W. and J. Jacoby, (1977) "Consumer Information Processing; Emerging Theory and Findings," in A. Woodside, P. D. Bennett, and J. N. Sheth, eds., Foundations of Consumer and Industrial Buying Behavior, New York, American Elsevier.

Engel, James F., David T. Kollat, and Roger D. Blackwell, (1973) Consumer Behavior. Second Edition, New York, Holt, Rinehart, and Winston.

Gardner, M. P. A. A. Mitchell, and J. E. Russo, (1978) "Chronometric Analysis: An Introduction and an Application to Low Involvement Perception and Advertisements,'' in H. Keith Hunt, ed., Advances in Consumer Research.

Green, Richard, Andrew Mitchell, and Richard Staelin, (1977) "Longitudinal Decision Studies Using a Process Approach: Some Results from a Preliminary Experiment," in B. A Greenberg and D. A. N. Bellinger, eds., Contemporary Marketing Thought, American Marketing Association.

Holbrook, M. B. and K. A. Maier, (1978) "A Study of the Interface Between Attitude Structure and Information Acquisition Using a Questionnaire-Based Information-Display Sheet," in H. Keith Hunt, ed., Advances in Consumer Research, Vol. 5, Association for Consumer Research.

Howard, John A., (1963) Marketing Management: Analysis and Planning, Homewood, Ill., Richard D. Irwin, Inc.

Howard, John A., (1977) Consumer Behavior: Application of Theory, New York, McGraw-Hill Book Company.

Howard, John A. and Jagdish N. Sheth, (1969) The Theory of Buyer Behavior, New York, John Wiley & Sons.

Jacoby, Jacob and Robert Chestnut, (1977) "Amount, Type, and Order of Package Information Acquisition in Purchasing Decisions,: Final Report to National Science Foundation (GI - 43687), June.

Jacoby, Jacob, Robert W. Chestnut, and William A. Fisher (1978) "A Behavioral Process Approach to Information Acquisition in Nondurable Purchasing," Journal of Marketing Research, Vol. 15, November, pp. 532-544.

Jacoby, Jacob, Robert W. Chestnut, Wayne D. Hoyer, David A. Sheluga, and Michael J. Donahue, (1978) "Psychometric Characteristics of Behavioral Process Data: Preliminary Findings on Validity and Reliability,'' in J. Keith Hunt, ed., Advances in Consumer Research, Vol. 3, Association for Consumer Research, pp. 306-314.

Jacoby, Jacob, George J. Szybillo, and Jacqueline Busato-Schach, (1977) "Information Acquisition Behavior in Brand Choice Situations," Journal of Consumer Research, Vol. 3, March, pp. 209-216.

Lehman, Donald R., (1978) "Determinants of Choice Approach: A research Paradigm," working paper, Columbia University, Graduate School of Business.

Lehman, Donald R. and William L. Moore, (1979a) "An Empirical Study of Information Acquisition Over Time," working paper, Columbia University, Graduate School of Business.

Lehman, Donald R. and William R. More (1979b) "Individual Differences in Search Behavior and Recall of Package Information on a Nondurable," working paper, Columbia University, Graduate School of Business.

Myers, James H. and Mark I. Alpert, (1968) "Determinant Buying Attitudes: Meaning and Measurement," Journal of Marketing, Vol. 32, October, pp. 13-20.

Newell, Allen and Herbert A. Simon, (1972) Human Problem Solving, Englewood Cliffs, NJ, Prentice-Hall.

Newman, Joseph W., (1977) "Consumer External Search: Amount and Determinants," in Arch G. Woodside, Jagdish N. Sheth, and Peter D. Bennett, eds., Consumer and Industrial Buying Behavior, Amsterdam, Holland, North Holland Publishing Company, pp. 79-94.

Newman, Joseph W. and Bradley D. Lockeman, (1975) "Measuring Prepurchase Information Seeking," Journal of Consumer Research, Vol. 2, December, pp. 216-222.

Nisbett, Richard E. and Timothy DeCamp Wilson, (1977) "Telling More Than We Can Know: Verbal Reports on Mental Processes," Psychological Review, Vol. 84, pp. 231-253.

Payne, John W. and E. K. Easton Ragsdale, (1978) "Verbal Protocols and Direct Observation of Supermarket Shopping Behavior: Some Findings and a Discussion of Methods," in H. Keith Hunt, ed., Advances in Consumer Research, Vol. 5, Association for Consumer Research, pp. 571-577.

Russo, J. Edward, (1978) "Eye Fixations Can Save the World: A Critical Evaluation and a Comparison Between Eye Fixations and Other Information Processing Methodologies," in H. Keith Hunt, ed., Advances in Consumer Research, Vol. 5, Association for Consumer Research, pp. 451-570.

Russo, J. E. and L. D. Rosen, (1975) "An Eye Fixation Analysis of Multialternative Choice," Memory and Cognition, Vol. 3, May, pp. 267-276.

Smith, Eliot R. and Frederick D. Miller, (1978) "Limits on Perception of Cognitive Processes: A Reply to Nisbett and Wilson," Psychological Review, Vol. 85, pp. 355-362.

VanRaiij, W. Fred, (1977) "Consumer Information Processing for Different Information Structures and Formats," in W. D. Perrault, ed., Advances in Consumer Research, Vol. 4, Association for Consumer Research, pp. 176-184.

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