Consumer Uses of Common Dimensions in the Appraisal of Services

J.E.G. Bateson, The London Business School
E. Langeard, I. A. E. University of Aix - Marseille
[ to cite ]:
J.E.G. Bateson and E. Langeard (1982) ,"Consumer Uses of Common Dimensions in the Appraisal of Services", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 173-176.

Advances in Consumer Research Volume 9, 1982      Pages 173-176


J.E.G. Bateson, The London Business School

E. Langeard, I. A. E. University of Aix - Marseille


This study is concerned with how consumers evaluate services. In particular it is concerned with whether evaluation of different services can be understood using the same criteria or dimensions. It is therefore an attempt to rise above the level of service specific attributes and instead look for commonality between services.

This paper first describes the extensive process of focus groups, in-depth interviews and pre-tests which were used to develop a set of dimensions. The validity of the dimensions is then discussed in the context of consumers' choice between more or less participative methods of receiving a number of different services, for example the choice between pumping their own gas and having an attendant do it or an automatic teller machine and a human teller. Finally some observations are offered on future research based on these dimensions.


The initial phases of the exploratory research consisted o two focus groups, twelve personal interviews and three 'mini-group' discussions. One focus group consisted of a cross-section of white collar workers, the other of blue collar. The professional group leader lead the white collar group into discussions of restaurants, hotels, and banks and the blue collar of restaurants, banks, and supermarkets.

The twelve personal interviews were conducted by a professional psychologist and took the form of semi-structured discussions of banks, fast food restaurants and hotels. The difficulty of getting individual respondents to talk and of getting sufficient depth in large groups lead to the idea of a 'mini-group'. Three groups of three respondents were assembled with a professional discussion leader.

This entire process was intended to surface the dimensions along which respondents evaluated all services. A number of such dimensions emerged and these are described below. It should be stressed that reactions to different services varied across respondents. Thus for example some respondents thought they could have more control of the situation using an automatic teller, others that they were more in control with a human teller. Consumers reactions to services also varied with the situation, for example who was i the party, but the dimensions did not vary. In total seven dimensions emerged and are listed below with representative respondent comments where appropriate:

Time. The respondents saw services as time consuming in a time budgeting sense.

Control. The respondents expressed the need to feel in control whilst receiving the service: "I much prefer to know what I am doing" "I don't like using the bell boy in hotels -I don't know what my purpose is walking along behind him. . . I don't like not knowing what I'm doing"

Effort. There was a general consensus that the more participative services required effort. This was especially obvious where the effort involved was physical e.g. in pumping gas but many respondents perceived more effort to be needed in the more intellectual tasks e.g. using an automatic teller machine

Dependence. There was quite a large reaction against the need to depend on other people to receive the service: "Machines make us more independent" "I like to depend on myself" "I don't need that other person..."

Efficiency. A large number of the respondents articulated their reaction to services in terms of how efficient it was for them.

Human Contact. Different respondents had different reactions to the need for human contact when receiving a service but all perceived differences between services along this dimension: "The people, courtesy, human contact...: enjoy that" "A bell hop is human, hotels are so big, a bell hop humanizes the hotel" "I don't have to get involved if I don't want to"

Risk. Many of the dimensions of risk originally suggested by Bauer (1960) and others surfaced doing the various discussions. There was little talk of financial risk but much of Psycho-social and performance risk.

Following these initial stages draft questionnaires were developed attempting to measure consumers reactions to services along these dimensions. To provide a preliminary test of these questionnaires two focus groups were conducted. Respondents were first asked to complete the questionnaire and then a professional discussion leader asked them to go through the questions and articulate what they thought they meant. This process produced a number of word changes but did not surface any new dimensions.


The problem of productivity is one that plagues the service sector. Across many industries within the sector, one approach being used to improve productivity is to increase the participation of the consumer in the production of the service. To encompass this idea Gartner and Riessman (1974) extended the traditional ideas of capital and labor intensity and suggested that in the service sector we should also be concerned with 'consumer intensity'.

The dimensions developed by the process described earlier were used to investigate how consumers evaluate the choice between the two alternative ways of receiving their service.


The exploratory research described earlier highlighted early in the study that consumers' choice between the different ways of receiving a service was highly influenced by situation specific variables (e.g. What is the occasion? What time is it? Who is in the party? etc.). To overcome this, situational scenarios were developed which were realistic from the standpoint of the respondents, controlled for a wide range of factors known to affect behavior (including monetary incentives) and offered respondents a choice between two alternative service delivery systems.

In total six scenarios were developed:

1. At a service station -- pump your own gas versus having an attendant do it for you.

2. At a bank -- using an automatic teller machine versus using the services of a human teller.

3. At a quick service restaurant -- getting your own food at the counter versus receiving table service from a waiter or waitress.

4. At an airport -- carry your own bags on to an aircraft with special storage facilities versus checking your bags.

5. At a hotel -- using a self-service food and drink dispenser versus obtaining the same food and drink from room service.

6. At a travel agent -- purchasing travellers checks from an automatic teller machine versus buying them from a clerk.

Appendix A shows an example of a scenario.

Based on each scenario respondents were asked to provide the following information:

a) An intention measure - on what percentage of occasions they would use the more participative alternative.

b) Their perceptions of the two alternative delivery systems along the seven dimensions developed earlier. The respondents were offered a five point scale for each dimension ranging from 'A Lot Less (1)' to 'A Lot More (5)', the indifference point was 'Same (3)'

c) How important they considered each dimension was in their choice between the alternatives on the following scale: Not at All Important (1), Not Very Important (2), Slightly Important (3), Very Important (4), Extremely Important (5).

The survey was conducted by mail and two different questionnaires were used. The first (Questionnaire A) included scenarios 1, 2 and 3; the second (Questionnaire B) scenarios 1, 4, 5 and 6. The questionnaires were pretested in a pilot mailing.

Questionnaire A was mailed to a stratified random sample of 1500 customers of three financial institutions located in three different regions of the United States. Five hundred of these customers were known to be users of a new self-service facility. Questionnaire s was mailed to a stratified random sample of 1000 customers of a nationwide financial institution equally divided between users and non-users of a new self-service facility offered by that institution

Respondents initially received an introductory letter which was followed one week later by the questionnaire, another cover letter and a 254 incentive. A reminder card was sent to all respondents, mailed to arrive one week after the questionnaire.

The overall response rate was 57.5% after deducting mail returned as undeliverable. The response rate for questionnaire (A) was 52.5% and for questionnaire (B) 67.5%. The range of response rates by sample and strata is shown in Figure 1.



We are inclined to believe that the overall differences in response rates between the two samples reflected subsample - related variables rather than questionnaire related variables. The results are consistent with less than up to date mailing lists and known poorer return procedures for undeliverable mail in certain cities.

As a general comment, a more than 50% response rate to a long questionnaire mailed to a nonspecialist audience and dealing with a general interest (as opposed to highly specialized) topic is a reasonably good result. Although some nonresponse bias may be introduced by the failure of many subjects to respond, it should be emphasized that the areas of research interest centered on observable differences between respondents rather than on identifying and generalizing the proportions within the population who might be expected to behave in a particular way.


The data was analyzed to see whether the dimensions that had been developed were capable of discriminating between respondents in terms of their intention to use the more participative alternatives. To do this, for each scenario, the respondents were broken into the following categories:


These categories were developed to reflect the tri-model nature of the frequency distributions of the intention measures.

For each scenario the intention group were analyzed in terms of both their perceptions along the dimensions and the importance they attached to the dimensions in making their choice. As an example, Figure 2 shows both the perception and importance data for the automatic teller (A.T.M.) versus human teller choice.



For this specific scenario the dimensions clearly discriminate between the intention groups in terms of their perceptions of the alternatives. All of the F-statistics are highly significant. The general pattern across dimensions is that those respondents who are less inclined to use the participative alternative perceive little difference between the alternative delivery systems. Those respondents who intend to use the more participative alternative on a greater percentage of occasions, see much bigger differences. They perceive the A.T.M. as taking less time, offering them more control of the situation; making them less dependent on others; being more efficient and having less human contact.

For this banking scenario, the importance attached to time control, efficiency, human contact and risk all discriminate between the groups. The more participative groups rate time and efficiency more highly and the less participative groups human contact and risk. Although there are differences along the human contact dimension the general level of importance attached to this dimension is low.

The analysis for the bank scenario has been presented in detail to illustrate the general process. Of more interest, in terms of the validity of the dimensions, is the pattern across all seven scenarios. Figure 3 tabulates only the F-statistics for all seven scenarios for both perception and importance. The findings tend to duplicate those found for the bank scenario alone. The perception of time, control, dependency, efficiency and risk discriminate between the participator groups. The importance attached to time and control stood out as key dimensions across all scenarios. These two dimensions must therefore be considered as particularly important since they discriminate on both 'perception of difference' and 'importance'.



Summarizing the detailed contents of the tables across scenarios is more difficult. Figure 4 tabulates the number of scenarios where a particular participator group perceived the alternatives as different. 'Different' is here arbitrarily defined as having a mean score differing by more than one from the indifference point of 3 in either direction on the scale. The table is therefore a crude indicator of where groups saw differences, but not of the direction of the differences



The bottom row of the table confirms the findings from the bank scenario Less participative groups perceive few differences between the alternatives This was not an altogether intuitively obvious finding since an alternative hypothesis would have been that the less participative groups would have perceived 'negative' differences between the alternatives The final column of the table confirms the importance of time, control, dependence and human contact as dimensions along which the different groups perceive differences between the alternatives offered to them across a number of scenarios

In a similar way, Figure 5 tabulates the number of scenarios where a particular participator group indicated a particular dimension was 'important' 'Important' is here arbitrarily defined as a mean score of 3 0 or above Once again the theoretical maximum in each cell is seven in all the scenarios



The results again mirror those found with the bank scenario Time, control, efficiency and risk stand out as dimensions perceived to be important in many scenarios The more participative groups across scenarios perceive time, control and efficiency as important whereas the less participative groups assign their importance to the risk dimension.


The dimensions that have been developed have little more than face validity They are however capable of discriminating between groups of consumers in terms of their intention to participate more in the production of their services Not only that but the pattern of responses by participator group is intuitively reasonable

The idea that consumers' evaluation of services could be understood using dimensions above service specific attributes is rooted in our belief that, from the consumers' perspective, services do have common characteristics. Elsewhere we have argued that those common characteristics result from the experiential nature of services and have suggested a theoretical model of services as an interactive process involving the consumer (Langeard, et. al.. 1981).

Of the dimensions that have been studied two are not new. The consumer behavior field is increasingly interested in the role that time plays in decision making. The idea of perceived risk was originally proposed by Bauer (1960) and produced a stream of work over the next fifteen years (see Ross 1974 for a review). The other five dimensions are, however, not only new but would be inappropriate if applied in the goods sector.

The five dimensions are all consistent with the idea of a service as an interactive process. Perhaps the most interesting and potentially the most worthy of further study is the dimension of control. This dimension was a valuable discriminator in terms of both perceived differences and importance rankings. The idea that consumers should wish to feel in control when receiving a service is also intuitively appealing. One obvious direction for future work would be to consider whether 'need for control' could be measured and whether it varied across individuals. It would be equally important to investigate the characteristics of a particular service which engendered in a respondent the sensation of being in control.

Understanding service interactions from the consumer point of view is a field that has received little attention. Czepiel (1980) in considering consumer satisfaction with services concluded that "little formal work has been done to investigate the elements of service encounters and concePtualize their content".

Potentially the work focusing on consumer involvement could provide a framework for investigating such interactions. Houston and Rothschild (1978) suggest three kinds of involvement: situational, enduring and response. All three could be helpful in understanding service interactions. However response involvement is probably the most interesting. Unfortunately, of the three measures this is the least standardized, dealing as it does with the various stages of the buying process. Since consumers are more and more being called upon to be the producer of their service it would also have to be extended to cover consumers participation as a producer.


The Bank Scenario

"It is 10.00 a.m. and you wish to withdraw $50.00 from your checking account. You have a card which would enable you to use an automatic teller machine or you could go to a human teller with your checkbook. So your choices are:

Either use the automatic teller machine; or use the human teller.

There are equally short lines of people waiting to use the machine and at the teller window."


Bauer, R. A. (1960), "Consumer Behavior as Risk Taking," in R. S. Hancock (ed.), Dynamic Marketing for a Changing World, Chicago: AMA, pp. 389-398.

Czepiel, John (1980), Managing Customer Satisfaction in Service Businesses, Boston, MA: Marketing Science Institute Report No. 80-109.

Gartner, A., and Reissman, F. (1974), The Service Society and the new Consumer Vanguard, new York: Harper and Row.

Houston, M. J., and Rothschild, M. L. "Conceptual and Methodological Perspectives on Involvement," in Subhash C. Jain (ed.), Research Frontiers in Marketing; Dialogues and Directions, Chicago: AMA.

Langeard, E., Bateson, J. E. G., Lovelock, C. H., and Eiglier P. (1981), Services Marketing: New Insights from Consumers and Managers, Boston M: Marketing Science Institute Report No. 81-104.

Ross, I. (1974), "Perceived Risk and Consumer Behavior: A Critical Review," in M. J. Schlinger (ed.), Advances in Consumer Research Vol. II, Proceedings of Association for Consumer Research, 1974 Meeting, Chicago: AMA.


The authors would like to thank the member companies of the Marketing Science Institute without whose support and co-operation this study would not have been undertaken. This paper is based on a much larger study and would not have been possible without the input of our fellow project team members Prof. Pierre Eiglier and Prof. Christopher Lovelock.