Influence of Problem Recognition on Search and Other Decision Process Variables: a Framework For Analysis

Girish Punj, University of Connecticut
Narasimhan Srinivasan, University of Connecticut
ABSTRACT - Though Problem Recognition (PR) is a critical phase of the consumer decision making process, it has been relatively under-researched. Since PR frames the problem-solving situation, the remaining stages in the purchase decision are dependent on it. An analytic framework of problem recognition, consisting of four segments -- "new need," "product depletion," "expected satisfaction" and "current dissatisfaction" -- is proposed. A sample of 1056 new car buyers is used for empirical validation. The "new need" and "product depletion" groups differ significantly from the others in terms of several pre-search, search, alternative evaluation and satisfaction variables.
[ to cite ]:
Girish Punj and Narasimhan Srinivasan (1992) ,"Influence of Problem Recognition on Search and Other Decision Process Variables: a Framework For Analysis", in NA - Advances in Consumer Research Volume 19, eds. John F. Sherry, Jr. and Brian Sternthal, Provo, UT : Association for Consumer Research, Pages: 491-497.

Advances in Consumer Research Volume 19, 1992      Pages 491-497

INFLUENCE OF PROBLEM RECOGNITION ON SEARCH AND OTHER DECISION PROCESS VARIABLES: A FRAMEWORK FOR ANALYSIS

Girish Punj, University of Connecticut

Narasimhan Srinivasan, University of Connecticut

ABSTRACT -

Though Problem Recognition (PR) is a critical phase of the consumer decision making process, it has been relatively under-researched. Since PR frames the problem-solving situation, the remaining stages in the purchase decision are dependent on it. An analytic framework of problem recognition, consisting of four segments -- "new need," "product depletion," "expected satisfaction" and "current dissatisfaction" -- is proposed. A sample of 1056 new car buyers is used for empirical validation. The "new need" and "product depletion" groups differ significantly from the others in terms of several pre-search, search, alternative evaluation and satisfaction variables.

INTRODUCTION

The problem recognition stage in the consumer decision making process has been generally regarded as the event or "trigger" that initiates a purchase decision. It is the precursor of all consumer-initiated activities, such as pre-purchase information search, evaluation and choice processes, preceding any transaction. The idea that initial consumer actions affect future ones is both simple and intuitively appealing. First, the problem recognition phase frames the problem-solving situation. Second, other stages of the decision making process sequentially linked to the problem recognition stage have a dependent relationship with it. Hence we can expect this stage to have a crucial influence on all subsequent decision process activities. Surprisingly, despite the potential significance of the problem recognition phase, it continues to be an area of limited research (Bruner 1985; Bruner and Pomazal 1988), though its importance is recognized in most models of consumer behavior (Howard 1989; Engel, Blackwell and Miniard 1986; Wilkie 1990).

Some of the difficulties associated with problem recognition (PR) research appear to be the consequence of an inadequate theory of problem recognition and hardly any empirical specification of the construct. In this paper, we present an exploratory study which attempts to address both these issues to some degree. Given the preliminary nature of this investigation, we propose and test a theoretical framework of PR. We validate the proposed framework rather than test specific hypotheses.

First, we specify a theoretical framework of the problem recognition process, drawing upon previous PR research (Dewey 1910; Bruner 1986, 1985; Bruner and Pomazal 1988; Sirgy 1983) and existing descriptive concepts in consumer decision making models (Howard 1989; Engel, Blackwell and Miniard 1986; Hawkins, Best and Coney 1989; Wilkie 1990). Second, we attempt to validate our framework of the PR process using data from a large sample of new automobile buyers. The results of the empirical test are then used to modify our framework of problem recognition.

PREVIOUS RESEARCH

Previous research on PR tends to view the construct in a variety of ways. For instance, Engel, Kollat and Blackwell (1968) describe PR as being caused by a significant difference between a desired state and an actual state, with respect to a particular want or need. While such a conceptualization is useful in understanding PR as the event or "trigger" initiating the consumer decision process, the operational aspects of the precedents to problem recognition, such as the definitions of the 'actual' and 'desired' states and their relative strengths are left unspecified. Similar descriptions of the PR construct are offered by Wilkie (1990) and Hawkins, Best and Coney (1989).

An abstract conceptualization of PR is provided by Sirgy (1983) who regards the construct as representing a homeostatic process disrupted whenever there is a significant incongruity between a "perceptual stimulus" and an "evoked referent." Also, a distinction is offered between problem and opportunity recognition. Essentially, a consumer is perceived to be vigilant and aware of the 'actual,' 'desired' and 'available' states. Bruner (1985, 1986) makes a contribution to our understanding by regarding the fulfillment of the discrepancy between the "desired" and "actual" need states as a matter of consumer style, rather than the manifestation of particular situations i.e. problem recognition is a person-specific construct, rather than a situation-specific construct. Bruner and Pomazal (1988) attempt a comprehensive model offering a process view of the PR construct, using the earlier conceptualizations. However, in their attempt at providing a detailed cognitive understanding of the PR process, they offer a model that cannot be easily specified and tested.

TOWARD A FRAMEWORK OF PR

A primary consideration in developing a framework for the PR process is the recognition that the framework be empirically testable to encourage further theoretical developments. A starting point would be a parsimonious conceptualization that retains the essence of the elaborate cognitive descriptions of the construct.

A second criterion deals with the importance of incorporating both the internal states of the consumer and the overt actions in depicting the PR phenomenon. Taking both aspects assists in developing an actionable framework of PR useful in developing marketing strategy.

A final consideration relates to the importance of linking PR to other aspects of the decision process. Ideally we would expect our framework to be both descriptive and predictive of the subsequent constructs and processes. However, the degree to which this might occur could vary across different purchase contexts.

FIGURE

A FRAMEWORK OF PROBLEM RECOGNITION

Based on these considerations, we assume that PR may be regarded as a cognitive event that is simultaneously manifest as an overt action representing the start of the consumer decision process. We acknowledge that several perceptual and motivational sub-processes are likely to lead up to the PR event but these are appropriately viewed separately as they are quite different in character. Such a conceptualization allows us to consider PR as a construct rather than a process, thus providing an useful anchor for formulating an analytic framework.

A FRAMEWORK OF PR

A framework of PR is shown in the Figure. The main theoretical proposition in our framework is that the nature of the occurrence of the PR event influences the cognitive and physical resources utilized for the subsequent stages of the consumer decision processes. The cognitive and physical resources are measurable in terms of typical decision process activities such as the nature of presearch decisions, the amount of information search, decision time, usage of a purchase pal, etc. Thus corresponding to "routine" PR events such as replenishment, we can expect "routine" decision activities with a low degree of search, and in new purchase contexts, the expenditure of cognitive/physical effort would be akin to "extended problem solving" i.e. a greater degree of effort and decision time.

The second theoretical proposition in our framework is that the perceptual and motivational sub-processes which precede the PR event are predictive of it, thus providing an understanding of how and why various PR events differ from one another.

A third proposition in our framework refers to the role of long term memory in determining the decision process activities triggered in response to various manifestations of the PR event. We suggest that the latter serve as important retrieval cues in facilitating the remaining steps in the purchase decision.

Finally, we posit that the nature of the PR event influences the criteria consumers use in assessing satisfaction with the purchase decision. For some contexts these criteria are likely to be of a satisficing nature, but not so for others.

Consistent with the literature, the present framework of problem recognition proposed in this paper assumes that the PR event occurs in response to a discrepancy between a desired state and an actual state with respect to a consumer need (generic PR) or want (selective PR). We may not be able to empirically distinguish between needs and wants because this difference merely influences the scope of the decision, and not the nature of it. Further, we assume that the desired state is typically influenced by external factors (i.e. culture, social class, reference group, lifestyle, household and marketer influence) while the actual state is generally influenced by factors internal to the decision maker (i.e. motivation, learning, memory, personality, etc.) The above partitioning is obviously a simplification, but it is consistent with those frequently used in understanding the multitude of influences on consumer decision processes (e.g. Hawkins, Best and Coney 1989). Furthermore, it parallels the distinction often made between the controllable and uncontrollable determinants of PR, from the viewpoint of a marketer.

In understanding the influence of the actual state with respect to a need/want on the PR event, we can see that there are basically two sets of internal circumstances that are likely to trigger the decision process. The first has to do with the depletion of the current product corresponding to a existing need/ want, whereas the second relates to dissatisfaction with the current state or product performance. Turning to the influence of the desired state corresponding to a need/want on the PR event, we similarly find two major external conditions initiating the consumer purchase decision. The first condition relates to acquisition of a product representing a completely new need/want, while the second corresponds to the possession of products promising higher satisfaction than the current one.

It is important to recognize that the time element (Bloch, Sherrell and Ridgway 1986) corresponding to both the depletion of current product and the acquisition of a product corresponding to new need is not explicitly modeled in our framework. For instance, in the former case the depletion can be sudden (like in a product failure/ breakdown) or more routine (like in planned replacement). In the latter instance, the acquisition could also be sudden (like in an impulse purchase) or more contemplative in nature. The incorporation of this time effect in our framework awaits additional research (see Bayus 1991 for current work in this area). Additionally, it is possible that the time effect is also related to the magnitude of the discrepancy between the desired and actual states, thus leading to the classification of certain consumer purchase problems as "active" or "inactive" (Hawkins, Best and Coney 1989). However, this is again in the realm of future research.

Hence our framework of PR proposes that discrepancies between desired and actual states is driven by completely new or routine product purchases at the extremes and by existing dissatisfactions (with current products) or potential satisfactions (with new products) at intermediate levels. It is important to note that the above four part categorization of the PR event captures the essence of the broad range of causes of PR currently considered in the literature (Wilkie 1990; Hawkins, Best and Coney 1989; Bruner and Pomazal 1988). Yet, it provides a parsimonious and empirically verifiable formulation of the first stage of the consumer decision process, as shown in the Figure.

DATA

The data used in this research were collected as part of a study of new car buyers in three geographically separate metropolitan markets. The sample consisted of households that had recently purchased a new automobile for personal use. A stratified sample (by make of car) was drawn from all new car registrations in the markets during a particular time period, and was obtained through R.L.Polk & Co. Potential respondents were contacted by telephone to solicit their participation in the study --and also to ensure that they belonged in the sampling frame. Approximately 2400 questionnaires were mailed to households that could be contacted via phone and 1056 usable responses were received, representing a 44% cooperation rate.

Prior to their use in the analyses, the data were subjected to a series of extensive consistency tests to identify biases that would require correction during the analyses phase. In almost all instances the data held up well to the verification checks. (For more details of the data, please refer to Punj and Staelin 1983).

TESTING THE FRAMEWORK OF PR

The main variable in this study is PR. Four groups corresponding to the Figure were formed. Details of the formation of these groups and operational specifications of the other variables used in this study are given below.

Problem Recognition (PR): We recoded nine specific categories (and the open ended responses) which best described the reason that led people to think about buying their new car, into the following four PR groups, corresponding to the framework of PR presented earlier. Presenting respondents with the major possible PR scenarios and asking them to identify with the most applicable in their case appears to be a reasonable method of assessment. The task is simple and recall did not appear to be a problem.

Group I: HIGHER EXPECTED SATISFACTION

* Had a car but wanted one more.

* Old car ran fairly well, but the new models had better styling.

* Old car ran fairly well, but could get better gas mileage with a new car.

Group II: CURRENT DISSATISFACTION

* Old car needed repairs too often and was not reliable.

* Old car ran fairly well, but if it broke down, it would not be worth fixing.

Group III: PRODUCT DEPLETION

* Old car stopped running and had to be replaced.

* Old car ran fairly well, but it is best to trade every two or three years.

Group IV: NEW NEED

* Old car ran fairly well, but wanted a car for a different purpose -- recreation, hauling things, carrying more people (or fewer people).

* Did not have a car and wanted to get one.

The sizes of the segments were appreciable, as seen from the Table. The four PR groups (labeled as Expected Satisfaction, Current Dissatisfaction, Product Depletion and New Need groups hereafter) account for 24%, 43%, 19% and 14% respectively. A fifth of the market (PRODUCT DEPLETION segment) is seen to exhibit regularity in product purchase. Two out of three buyers (combining the EXPECTED SATISFACTION segment and the CURRENT DISSATISFACTION segment) are susceptible to product promotions/new product introductions at any time. Apparently, a lot more buyers express dissatisfaction with the state of the currently owned car, compared to the buyers who can be attracted into the market with promise of new potential benefits, such as improved styling, new product features, design, etc.

The remaining variables examined in our study were split into four categories corresponding to the well known stages in the consumer decision making process:

(a) Presearch Stage: This included the number of makes which consumers were willing to consider seriously and the degree of predecisions about manufacturer/dealer at the start of the information search process.

(b) Information Search Stage: This consisted of the length of time from serious consideration of purchase to the actual purchase; the count of several search activities; the pattern of search, such as dealer/non-dealer; and use of others to assist in the decision process.

(c) Alternatives Evaluation Stage: This dealt with the number of models examined and the number of retail visits made, and finally

(d) Postpurchase Stage: We examined the self rating of buying performance and the satisfaction with the new car purchased.

TABLE

RESULTS OF ONEWAY ANALYSIS ON PR GROUPS

Details of the operational specifications of the above variables are described below:

Number of Makes willing to consider for purchase (NMAKES): Respondents were queried as to the makes they were willing to consider buying, before visiting any dealer. The multiple choice format considered divisional level makes, such as Buick, Cadillac, Chevrolet, Dodge, Ford, Lincoln, Mercury, Oldsmobile, Plymouth, Pontiac, Honda, Toyota, etc. It is assumed that the NEW NEED group would be willing to consider more makes compared to those who engage in replacement purchases. Also, the EXPECTED SATISFACTION group may be expected to have a smaller latitude of acceptance compared to the CURRENT DISSATISFACTION group, since the latter would probably eliminate only their current make and show greater acceptance of the other alternatives in the market.

Degree of Pre-Decisions made (PREDEC): Responses to queries about whether manufacturer/model/dealer were known at the beginning of the search process were categorized into the following four groups: (a) Nothing was predecided, (b) Manufacturer pre-decided (like GM, Ford, etc.), (c) Specific make or dealer pre-decided, and (d) both the specific make and dealer were predecided. It may be expected that the PRODUCT DEPLETION group would exhibit the highest degree of pre-decisions compared to the other groups, whereas the NEW NEED group would show the least amount of pre-decision making.

Purchase Decision Time (DEC.TIME): The elapsed time between first considering a car purchase seriously and the actual purchase was recoded into the following four categories: (a) Less than a month (b) Between 1 and 3 months (c) Between 3 and 6 months, and (d) More than 6 months. Once again, we expected that the NEW NEED group would take the longest time for the purchase decision since it is a new problem-solving situation. The PRODUCT DEPLETION group which may be expected to have engaged in a lot of prior planning would take the least time for making the purchase decision.

Search Activities (ACTVTS): Count of the participation in the following activities: (a) Talking to friends/relatives about new cars or dealers (b) Reading books and magazine articles (c) Reading advertisements in newspapers and magazines (d) Reading about car ratings in magazines (e) Reading automobile manufacturer brochures and pamphlets (f) Driving to and from dealerships (g) Looking around the showrooms (h) Talking to salespersons and (i) Test driving cars.

The NEW NEED group may be expected to engage in the greatest degree of search and the PRODUCT DEPLETION group the lowest, due to the greater degree of prior deliberation in the latter case. It is difficult to make a distinction between the EXPECTED SATISFACTION group and the CURRENT DISSATISFACTION group since the source and degree of the dissatisfaction is unknown.

Non-store search (NON.STOR) is the difference of the total hours spent in information search and the time spent at dealerships. For the major part, it may be taken to reflect the effort spent in gathering of general or preliminary information about models considered suitable for purchase. Generally, the decision about the make/model has to be made by a consumer before entering a dealership. Of course, the gathering of non-store information does continue concurrently while going through the process of dealer visits. Non-store information sources includes all non-dealer sources, such as friends/relatives, advertisements, brochures and auto ratings in magazines. Time is assumed to be a common denominator for measuring the effort expended across a variety of sources which may require differential cognitive resources.

Once again, we expect the NEW NEED group to spend the greatest effort in non-store search, since past experience may not be available to play a compensatory role, while the PRODUCT DEPLETION group would spend the least time due to better preparedness. The EXPECTED SATISFACTION group may conduct less non-store search compared to the CURRENTLY DISSATISFIED group, but as explained earlier, this depends on the nature and scope of dissatisfaction.

Dealer Search (DLR.SEAR) is the aggregate time, in hours, spent in visiting dealers. This represents focused search since dealer visits occur during the later part of the car buying process, usually after one has narrowed down the choice to some specific model(s). While some dealer visits may be made for information gathering or assessing the suitability of any model considered, some other visits might be accounted for by price shopping behavior. Non-Store search and Dealer Search constitute the pattern of information search considered in this paper and represent a dimension different from the degree of information search.

The NEW NEED group is expected to have the highest number of dealer visits due to the extensiveness of the decision process necessary whereas the PRODUCT DEPLETION group would have the lowest. The EXPECTED SATISFACTION group may be expected to be lower than the CURRENT DISSATISFACTION group if particular and desired feature(s) are not readily available.

Use of a purchase pal in the decision process (PUPL.USG): This is a dimension not commonly studied due to the assumption that the decision making is done by individuals, for the most part. However, it may be a joint decision to some degree, involving significant others such as friends/relatives/mechanics outside the immediate family. Respondents were queried whether such a purchase pal helped in deciding the new car to buy. It is a binary measure and does not capture the degree to which there was reliance on the purchase pal. It is possible to conceive of the PRODUCT DEPLETION group to have the least necessity of depending on someone to assist in the purchase decision, whereas the NEW NEED group would probably seek outside assistance. As before, the other two groups may not be distinguishable.

The number of models shopped for (NBRNDS): The respondents were queried on the number of models seriously considered for purchase during each dealer visit. NBRNDS is the aggregate of all the models shopped for during all the dealer visits. The NEW NEED group is expected to have the highest number of models shopped for, because of less prior knowledge. The PRODUCT DEPLETION group is expected to have the fewest NBRNDS since the search process was initiated much earlier, compared to the other groups.

Number of retail visits (RTL.VSTS): This captures the store dimension found in studies investigating search across stores and brands. Consumers might be shopping for the same model across dealers or shopping for different models at the same dealer. Since the former is more likely, this measure can be expected to reflect inter-store shopping. The PR groups are expected to show a similar pattern, as discussed for NBRNDS.

Certainty about the purchase (FNL.CER): Buyers were asked to assess how good a buy they thought they had gotten when they first rode in the new car after the purchase. The response was gathered on a 7 point scale (1 = Worst buy, 7 = Best buy). We did not have any expectations on how this variable would differ across the PR groups.

The overall satisfaction with the decision made (SATISFAC) was measured on a 7 point scale (1 = Totally dissatisfied, 7 = Totally satisfied). We expected the NEW NEED group to have the lowest satisfaction due to cognitive dissonance, while the PRODUCT DEPLETION is expected to show the highest satisfaction, given prior purchasing experience. Relative to the CURRENT DISSATISFACTION group, we expected the EXPECTED SATISFACTION to have a greater score since this is the motivation driving the purchase for this PR group.

ANALYSIS AND DISCUSSION

Essentially, we were interested in determining whether there were any differences among the four PR categories, in terms of the patterns of relationships identified in the previous section. Hence, several one way analysis of variance were conducted to identify significant differences across the PR groups. Also, the mean and standard deviation for each variable across groups are shown in the Table.

The empirical results provide encouraging support for the PR framework postulated in this paper. At the extremes, the NEW NEED group and the PRODUCT DEPLETION group stand out distinctively. However, the EXPECTED SATISFACTION and the CURRENT DISSATISFACTION groups do not show any statistical significance (p<0.05) between them. The nature and scope of dissatisfaction could be a reason for this finding.

To illustrate the results, consider the pre-search stage variables (NMAKES and PREDEC). As expected, the PRODUCT DEPLETION group shows the smallest value for NMAKES (m=2.61, s.d.=1.8). The other three groups differ significantly (p<0.05) from this group, but not among themselves. The degree of pre-search decision making (PREDEC) follows an identical pattern, demonstrating consistency.

In terms of the alternatives evaluation stage, the results are stronger. The PRODUCT DEPLETION group differs significantly from the other three groups (for NBRNDS, m = 3.30, s.d. = 2.4). So does the NEW NEED group (for NBRNDS, m = 4.8, s.d. = 4.0). The directional expectation of the means is most evident in this stage.

The most surprising finding in our study is that all the four PR groups do not differ significantly in terms of the satisfaction (SATISFAC) with he new car they bought. All's well that ends well? As long as customers engaged in whatever they deemed necessary and sufficient under the circumstances, which they appear to be controlling for, then they are satisfied with the new purchase made.

The findings indicate that the EXPECTED SATISFACTION and the CURRENT DISSATISFACTION group do not differ on any of the variables examined in this study. Yet, it is useful to conceptually distinguish between the two groups, since availability of an improved product causing PR is different from the case when it is caused by dissatisfaction with the currently owned product. However, the availability of better products may increase dissatisfaction with the currently owned brand and this might be the reason for the lack of statistical difference between the two groups.

We repeated the one way analysis of variance after collapsing the EXPECTED SATISFACTION and CURRENT DISSATISFACTION groups. As could be expected, the results showed the three groups being statistically different from one other for the following variables: PREDEC, DLR.SEAR, NBRNDS and RTL.VSTS. The PRODUCT DEPLETION group had significantly lower NMAKES, DEC.TIME, ACTVTS and higher FNL.CER, compared to the remaining two groups. However, in terms of satisfaction with the new car purchased, the three groups remained indistinguishable.

SUMMARY AND CONCLUSIONS

Problem Recognition, though recognized as being a critical part of the consumer decision process, remains relatively under researched. In this paper, we present a conceptual framework of Problem Recognition. The framework is tested using a sample of 1056 new car buyers. Results indicate that two segments -- a new need segment and a product depletion segment -- are clearly distinguishable in terms of the subsequent decision process variables. Two other groups, though conceptually distinct, do not reveal any statistical uniqueness. Additional research is needed to gain a richer and more precise understanding of how this important construct might actually frame and guide consumer decision making.

REFERENCES

Bayus, Barry L. (1991), "The Consumer Durable Replacement Buyer," Journal of Marketing, 55 (1) (January), 42-51.

Bloch, Peter H., Sherrell, Daniel L. and Ridgway, Nancy, M. (1986), "Consumer Search: An Extented Framework," Journal of Consumer Research, 13 (June), 119-126.

Bruner II, Gordon C. (1985), "Recent Contributions to the Theory of Problem Recognition," American Marketing Association Educators' Proceedings, 11-15.

Bruner II, Gordon C. (1986), "Problem Recognition Styles and Search Patterns: An Empirical Investigation," Journal of Retailing, 62 (Fall), 281-297.

Bruner II, Gordon C. and Pomazal, Richard J. (1988), "Problem Recognition: The Crucial First Stage of the Consumer Decision Process, Journal of Consumer Marketing, 5 (1), Winter, 53-63.

Dewey, John (1910), How We Think, Heath & Co., Boston.

Engel, James R., Kollat, David T. and Miniard, Paul W. (1986), Consumer Behavior, Fifth Edition, Dryden Press. Chicago.

Hawkins, Del I., Best, Roger J. and Coney, Kenneth A. (1989), Consumer Behavior: Implications for Marketing Strategy, Fourth Edition, BPI/Irwin Publications, Homewood, IL.

Howard, John A. (1989), Consumer Behavior in Marketing Strategy, Prentice-Hall, Englewood Cliffs: NJ.

Punj, Girish and Staelin, Richard (1983), "A Model of Consumer Information Search for New Automobiles," Journal of Consumer Research," 9 (March), 366-380.

Sirgy, Joseph M. (1983), Social Cognition and Consumer Behavior, Praeger Publishers, NY.

Wilkie, William L. (1990), Consumer Behavior, Second Edition, John Wiley and Sons, NY.

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