Values and Decision Strategies For Non-Comparable Choices

ABSTRACT - This paper explores the area of consumer choice between products from unrelated categories. The methodology examines a series of case studies using means-end chain analysis. The results showed some general support for previous finding regarding consumers using an abstraction process to allow for within attribute processing. However, they also indicate that some consumers do not attempt any direct comparison of the alternatives but evaluate each separately on the basis of prominent values that may differ between the alternatives. Choice is made on the basis of the most important value at the time of consideration.


Annalisa Boyd and Rob Lawson (2001) ,"Values and Decision Strategies For Non-Comparable Choices", in AP - Asia Pacific Advances in Consumer Research Volume 4, eds. Paula M. Tidwell and Thomas E. Muller, Provo, UT : Association for Consumer Research, Pages: 46-52.

Asia Pacific Advances in Consumer Research Volume 4, 2001      Pages 46-52


Annalisa Boyd, University of Otago, New Zealand

Rob Lawson, University of Otago, New Zealand


This paper explores the area of consumer choice between products from unrelated categories. The methodology examines a series of case studies using means-end chain analysis. The results showed some general support for previous finding regarding consumers using an abstraction process to allow for within attribute processing. However, they also indicate that some consumers do not attempt any direct comparison of the alternatives but evaluate each separately on the basis of prominent values that may differ between the alternatives. Choice is made on the basis of the most important value at the time of consideration.


Understanding how consumers choose between products and services has been a major part of the study of the discipline of consumer behaviour for the last forty years. The actualidea of a choice is seen to comprise of the following component parts: alternatives, attributes of value and uncertainties (Bettman, Johnson and Payne 1991). This paper is concerned with the characteristics of alternatives and the attributes of value when the choice involves non-comparable items. Non-comparable alternatives are defined as those alternatives within a choice set which are dissimilar products or services from divergent product categories. In such cases it is not possible to compare the alternatives on any concrete attributes other than price. Within the vast array of research that has been performed on consumer decision making relatively little has directly addressed the processes involved in this area and most previous investigations have taken a quantitative experimental approach. This paper documents findings from a series of in depth interviews that were used to investigate some of these decisions with respect to the underlying values for the choice and the type of decision strategy employed.


The actual level of prevalence of these decisions is not really understood but they are presumed to be fairly common events. Johnson (1983) identified four different circumstances have been identified in which consumers might be faced with decisions between non-comparable alternatives:

$after initial independent processing within categories;

$when supply constraints are present within categories;

$when product elimination is initially required because of processing constraints;

$and when there are substantial differences within categories.

In the first situation consumers may have previously considered alternatives within two or more product categories. After having made a selection from each of these alternatives, they may find themselves having to choose between the alternatives, for example because of budget constraints. The second scenario involves supply constraints within product categories. For example, the only options available to a consumer for an evening’s entertainment in a small country town may be totally different items such as the local movie theatre or the local bar. Processing constraints result in non-comparable alternatives being considered in situations such as when a consumer has to learn about many new product categories. One example might be a person setting up home for the first time and looking for many new items. Because of the volume of things to be considered they will end up only comparing the best items across categories. Again, this scenario implies some prior processing within categories and the existence of a budget constraint that is determining the existence of alternatives. The final circumstance is when substantial differences exist within product categories such as motor vehicles or sound systems.

Michael Johnson has been the most prolific author on the topic (1983, 1984, 1986, 1988 and 1989). Johnson proposes that, like other decision models, strategies for evaluating non-comparable alternatives fall into two essential groupsBacross attribute strategies and within attribute strategies. These two strategies are not exclusive but may be combined into what is termed as a "phased strategy". Across attribute strategies involve grouping separate attribute values into one overall value which can then be directly compared to another overall value. However, Johnson proposes that consumers are more partial to within attribute strategies that allow direct comparison and that attributes that are non-comparable at "concrete" levels are made comparable by moving to a higher level of abstraction. In this context abstraction implies a concentration of information and is best understood by picturing attributes lying on a continuum from the absolute concrete ascending through to the highly abstract. Concrete attributes can also be explained as feature representations while abstraction is concerned with dimensions. In the case of a chocolate bar, such as a "Mars" or "Moro", a consumer may describe just one aspect of it in several ways such as the presence or absence of caramel, sugar content or overall sweetness. These descriptions represent different levels of abstraction that concentrates information from individual features into representative dimensions. Clearly such a process may allow consumers to overcome the problem of non-comparability.

Following Johnson there have been several studies that have tried to clarify the nature of the abstraction process and understand the basis for choice in non-comparable situations (Corfman 1991, Corfman, Lehmann and Narayaman 1991, and Park and Smith 1989). In particular Corfman’s work, which was aimed at clarifying the levels of abstraction and the location of the decision, introduced the notion of values as a relevant concept in understanding the determinants of choice for non-comparable situations. Corfman (1991) suggests that five basic levels of comparability exist, varying from specific features through micro and macro functions to basic values and the overall utility of the choice options. Since values are a foundation and standard for much behaviour, including consumption, they may be used as actual choice criteria. The research addressed in this study attempts to build upon these ideas by directly investigating the link the link between values and behaviour in some non-comparable decisions using means-ends chains analysis.

Means-end chain theory may be described as a model of consumers’ own cognitive compositions and describes the way in which the concrete characteristics of a product are related to the self-relevant end result (Grunert and Grunert 1995). Means-end chain analysis focuses on the key defining elements between the product and the particular role it plays in the life of the consumer (Gutman 1990). It shows how the product characteristics (concrete or abstract) are linked to the consequences (functional or psychological) of consumption, which in turn may be linked to the attainment of life values (instrumental or terminal). Examining non-comparable decisions using means-end techniques should enable the process of abstraction to be assessed using a different method. Comparisons between the ladders for the alternatives under consideration should help identify the level of abstraction at which a comparison might be made. This method provides a more direct description of the abstraction process and an alternative methodological approach to the problem from the experimental form of most previous studies. The results aim to reveal any prevalent values or utilities on which the decisions between non-comparable alternatives are based and to show the kind of decision strategy employed by respondents. The latter is assessed both by direct probing and by comparison of the ladders for the two alternatives considered by the respondent.




Data was collected in two ways. In the first case it was facilitated by the co-operation of a local electrical retailer who gave access to his list of recent buyers. Customers who had spent over $300 on a single item were phoned and screened to identify those that had considered non-comparable alternatives. A target of 20 interviews was aimed for in order to give some assurance that there would be some variation within the responses. In effect it proved difficult to recruit sufficient respondents and a halt was called when the purchase had taken place over four months previously. At that point, although there had been no obvious indiations, there was some thought that the purchase event might be becoming too far back for meaningful recall. Also it seemed more and more difficult to interest potential respondents as the purchase became more remote. At this stage 200 customers had been approached (see Table 1).

The large number of non-qualifying customers may be seen as a comment both upon the prevalence of both non-comparable decisions and possibly on the overall extent of choices in consumer decision making. In order to extend the database to the original target a further ten interviews were conducted using a convenience sample of people who were presented with a set choice scenario between two predetermined non-comparable alternativesBa home entertainment centre and a holiday both valued at $2000. At the time of interview it was sensed that there were some differences between those based on their own decisions an those based on the set choice. In particular the set choice interviews seemed to produce longer, fuller interviews revealing more emotive and personal sets of feelings. Several factors could have contributed to this. Firstly, some of the richness of description from the own choice interviews may have been lost because of the time delay between purchase and interview, though respondents did not appear to have any problems with recall as the interviews had been conducted. Secondly, the holiday alternative in the set choice was quite different to any of the alternatives in the own decision set (see table two below). The elements of escape and fantasy that are often associated with tourist motivation may have had some impact on the nature of the information offered. Finally, and perhaps most importantly in the opinion of the interviewer, the constructed nature and hypothetical situation detailed in the set choice removed all personally threatening aspects. They felt more comfortable expressing themselves in a role playing format than the interviewees who were being questioned about their own decisions with fears about revealing potentially irrationalities or sub-optimal aspects. For these reasons the results from the two sets of interviews are presented separately in the following discussion.

The interviews were transcribed verbatim as they were completed and the software package NUD*IST was used to aid analysis. NUD*IST is one of the few programmes developed especially for qualitative data analysis. NUD*IST enables the coding and search of data to identify themes presented in text. The principles related by Thompson (1997) were followed in order to avoid projecting any set of predetermined meanings on the data. These involve identifying key patterns in the data, comparing across the cases and then making broad interpretations. In this case the key patterns are essentially in the ladders depicting the attributes, consequences and values. Use of NUD*IST helped greatly in this process because it was evident that respondents showed little order in their descriptions of the purchase process. Frequently they moved from one level to another and back again and did not provide information for the ladders in a neatly ordered sequence.




1. Categorisation of Comparability.

Johnson’s (1983) levels of comparability were used to appraise the differences in the product classes considered in the ten interviews involving consumers’ own decisions (Table 2). One comparison (video versus speaker drivers) was considered as moderately non-comparable (i=1), while all the remaining comparisons were classified as more non-comparable (i=2). At a level two categorisation it is likely that most will be evaluated at more abstract levels of comparison as defined by Corfman (1991).

It can be seen that no two respondents actually chose between the same pair of alternatives. This range of comparisons is far more extensive than in any other revious investigation of this subject.

2. Values identified with Decisions

All the interviews were summarised into means/end chains as illustrated in Figure 1 from own choice interview number ten. Both ladders clearly show how the attributes are translated to consequences and subsequently values and in this case both result in the common value of self-fulfilment even though the underlying attributes and consequences are quite different.

Tables 3 and 4 summarise the values identified in the ladders. Initially it was decided to base the classification of values on LOV (Kahle 1983) but there were several values and other objectives that could not be incorporated into this scheme. This may be a reflection of either the comprehensiveness of LOV in representing values relevant to consumer behaviour or the specification of the ideas by respondents. Certainly, there are themes in table four that are identifiable in more detailed classifications such as Rokeach (1973) or Schwartz and Bilsky (1980). Other items, such as "laziness" are difficult to attribute to any values scheme and furthermore it is not unusual in means-end analysis for respondents not to be able to articulate the whole of a chain but instead to draw a conclusion at a lower level. However, in the "laziness" case the respondent was quite definite that there was no other term to describe what was guiding his choice. He admitted to being lazy and achieving this was an end in itself different from possibly related ideas such as convenience or relaxation. The frequencies in the tables show values or other ends associated with both chosen and rejected alternatives. The figures in parenthesis in the tables indicate those values that could be identified as the basis for the decision.

There are at least two issues of note in the two tables. Firstly, if one takes into account the different nature of the holiday in the set choice compared to all the items in the own decisions, there seems to be little difference between the two sets. Fun and enjoyment, general well being and relaxation are all more likely to be the result of the holiday option and longevity is related to the transient nature of the experience as opposed to the durability of the alternative. Interestingly one respondent judged this characteristic the other way round and spoke of holiday memories lasting a lifetime. Overall fun and enjoyment is the most common value expressed which is quite different to Johnson whose most common abstraction in his different studies related to use and necessity.

As in the illustrated examples, the ladders that underlay this summary tend to be fairly complex. Ladders frequently depicted both different attributes leading to the same values and one attribute leading to different values. The first situation was the anticipated form where, for example, both an improved sound system and a large screen provided the consequence of increased realism that was subsequently linked to fun and enjoyment. In the second situation the example of a large screen increased the clarity of the picture and the entertainment value, allowing relaxation and finally leading to "well being". In another case however, the large screen enabled the respondent to see the picture more clearly, aiding comprehension and contributing to overall knowledge. This allows the respondent to stay ahead in their career and ultimately provides a sense of accomplishment. A further complexity evident in the ladders was that the values identified were often linked to other values. As an example, fun and enjoyment was in two cases seen as a subsidiary value to self-fulfilment. This pattern is one that fits with more complex approaches to considering values structures such as Schwartz and Bilsky as opposed to LOV or Rokeach (op cit).

3. Decision Strategies Used by Respondents

Perhaps the first point to note is that there appear t be no differences between the strategies used between the respondents in the two data sets. To a large degree the results seem to agree with the ideas proposed by Johnson who maintained that, in following the observed desire of consumers to make direct within attribute comparisons between comparable alternatives, consumers also endeavour to find "comparable representations for non-comparable alternatives allowing within-attribute comparison". Consequently, it was suggested that consumers attempt to make attributes comparable by moving through an abstraction process until a point is reached upon which very dissimilar alternatives may be compared. This process was observed with almost one-third of the respondents and a further third could be seen to use a phased strategy that also incorporates an element of within attribute processing. In this case the respondents use one strategy on a subset of applicable attributes and then switch to the alternative strategy. In this case the most common example was the judgement of performance quality in the set choice home entertainment centres or their own stereo and video purchases. Several attributes relating to sound and vision performance were assessed using an across attribute strategy to assess overall performance quality. The comparison was then effected employing a within attributes strategy using this single dimension previously generated from several attributes. The two cases labelled as "other" were essentially too confused in the description to be classified in any of the other four categories.



In many ways the most interesting cases are those respondents who apparently made no direct comparison between the alternatives. On the face of it this seems illogical when the respondents either said they had chosen between two options or were asked to choose between two in the set choice. In these cases the alternatives cannot even be compared on the values, or ends, that are being served and our interpretation is that the final choice was made on the value that held most significance at that particular point in time. For example, in interview five of the own choices, there are no common themes identified anywhere through the ladders for the breadmaker or the portable phone. In choosing the breadmaker, the respondent clearly identifies in her own words that the most important aspect is the sense of accomplishment arising from making her own bread and that, at that particular time, this was more important than the extra convenience of the portable phone. This is a very simple idea that fits with the idea of some overall utility as the basis for choiceBCorfman’s highest level of abstraction. It recognises that consumers have competing values, or goals, driving their decisions and limited budgets to meet these. However, what is really fascinating is the apparent simplicity of the process by which this overall utility is assessed. This is not by any complex across attribute strategy but perhaps more akin to the process involved in a simple disjunctive decision heuristic where high performance on single attributes or dimensions is regarded as significant. Key attributes and dimensions were identified by respondents as serving particular values and these become the focus of the decision. Providing these are regarded as being able to serve the value in question, the decision revolves around the value with the highest priority at that point in time. Such an interpretation supports both the abstraction process and the proposed role of values as a basis for comparison but it acknowledges that different values carry different importance weightings and that these may vary over time.










This paper has attempted to build upon the limited amount of research that has specifically considered consumer choices between non-comparable alternatives. Using a contrasting methodology it has tried to obtain some further description of the abstraction process and explicitly lik values into the decision process using means-end chain analysis. Across a wide range of alternatives the study identified self-fulfilment, fun and enjoyment and sense of accomplishment as the most frequently sought values. Though one example of across attribute processing was observed as well as some phased decisions, the decision strategies identified tend to emphasise within attribute processing. In particular six of the twenty cases seemed to make no direct comparison at all between the alternatives and based on any macrofunction or common basic value, but the decision seemed to be based on the single value viewed as most important at the time. This goes beyond a comparison of different means to the same end and reflects a higher order of overall utility between competing ends. It is still built upon a relatively simple within attribute strategy where the consumer judges the alternatives on one or two attributes that lead to the different competing ends. This type of decision strategy would seem to fit well with at least the first and third type of situation where Johnson suggested non-comparable choices might be made. In these situations where budgeting and processing constraints have constrained the options available it is easy to visualise how no direct comparison might eventuate. This is in contrast to the second or fourth option where the nature of the supply determines the nature of the choice and one might expect to find the alternatives being judged on how well they meet the particular end required.


Bettman, J.R., Johnson E.J. and J.W. Payne (1991), "Consumer Decision Making", in Robertson, T.S. and H.H. Kassarjian (eds), Handbook of Consumer Behavior, New Jersey, Prentice Hall Inc.

Corfman, K.P. (1991), "Comparability and Comparison Levels Used in Choices Between Consumer Products" Journal of Marketing Research, vol XXVIII (Aug), 368-394.

Corfman, K.P., Lehmann D.R. and S. Narayaman (1991), "Values, Utility and Ownership: Modelling the Relationships for Consumer Durables" Journal of Retailing, vol 67, No. 2, 104-124.

Grunert K.G. and S.C. Grunert (1995) "Measuring Subjective Meaning Structures by the Laddering Method: Theoretical Considerations and Methodological Problems", International Journal of Research in Marketing, vol 12, No. 3, 209-225.

Gutman J. (1982), "A Means-End Chain Model Based on Consumer Categorisation Processes", Journal of Marketing, vol 46, Spring, 60-72.

Johnson M.D. (1983) "Decision Processing and Product Comparability: A Theory of Strategy Selection", A dissertation submitted to the faculty of the Graduate School of Business in candidacy for the degree of Doctor of Philosophy, Chicago, Illinois.

Johnson M.D. (1984) "Consumer Choice Strategies for Comparing Non-Comparable Alternatives", Journal of Consumer Research, vol 11, December, 741-753.

Johnson M.D. (1986) "Modelling Choice Strategies for Non-Comparable Alternatives", Marketing Science, vol 5, Winter, 37-54.

Johnson M.D. (1988) "Comparability and Hierarchical Processing in Multialternative Choice", Journal of Consumer Research, vol 15, December, 303-314.

Johnson M.D. (1989) "The Differential Processing of Product Category and Non-Comparable Choice Alternatives", Journal of Consumer Research, vol 16, December, 300-309.

Kahle L. (1983) Social Values and Social Change: Adaptation to Life in America, New York, Praeger.

Park C.W. and D.C. Smith (1989) "Product-Level Choice: A Top-Down or Bottom-Up Process", Journal of Consumer Research, vol 16, December, 289-299.

Rokeach M. (1973) The Nature of Human Values, The Free Press, New York.

Schwartz S.H. and W. Bilsky (1990) "Towards a Psychosocial Structure of Human Values", Journal of Personality and Social Psychology, vol 53, 550-562.

Thompson C.J. (1997) "Interpreting Consumers: A Hermeneutical Framework for Deriving Marketing Insights from the Texts of Consumers’ Consumption Stories", Journal of Marketing Research, vol XXXIV (Nov), 438-455.



Annalisa Boyd, University of Otago, New Zealand
Rob Lawson, University of Otago, New Zealand


AP - Asia Pacific Advances in Consumer Research Volume 4 | 2001

Share Proceeding

Featured papers

See More


P8. Understanding Financial Literacy: a Meta-Analysis of the Antecedents, Consequents and Moderators

Frederike Monika Budiner Mette, ESPM, Brazil
Mateus Canniatti Ponchio, ESPM, Brazil
Wagner Junior Ladeira, Unisinos

Read More


Proximity Bias: Interactive Effects of Event Valence and Event Nearness on Probability Estimates

Jennifer Seok Hwa Hong, New York University, USA
Chiara Longoni, Boston University, USA
Vicki G. Morwitz, New York University, USA

Read More


J4. A Large Pack of Toilet Paper is Bad for Me: Self-control and Consumers’ Responses to Product Quantity

(Joyce) Jingshi Liu, Hong Kong University of Science and Technology
Keith Wilcox, Columbia University, USA
Amy Dalton, Hong Kong University of Science and Technology

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.