Consumer’S Acquisition of New Product Knowledge of Complex Service Products. Results of an Experiment on the Effect of Prior Product Knowledge

ABSTRACT - This paper describes an experiment to measure the ability of buyers to acquire new product information describinga complex service product from material written in the style of a sales brochure. The literature shows that measurements of product knowledge should be separated into two types of knowledge: declarative and procedural. The results of the research indicate that people with low levels of prior product knowledge acquired more new knowledge than people with higher levels of prior product knowledge, while people acquired an equal amount of new product declarative and procedural knowledge in percentage terms. A model is developed for the buyer’s acquisition of new product knowledge.

Citation:

Tony Ward (1996) ,"Consumer’S Acquisition of New Product Knowledge of Complex Service Products. Results of an Experiment on the Effect of Prior Product Knowledge", in AP - Asia Pacific Advances in Consumer Research Volume 2, eds. Russel Belk and Ronald Groves, Provo, UT : Association for Consumer Research, Pages: 47-59.

Asia Pacific Advances in Consumer Research Volume 2, 1996      Pages 47-59

CONSUMER’S ACQUISITION OF NEW PRODUCT KNOWLEDGE OF COMPLEX SERVICE PRODUCTS. RESULTS OF AN EXPERIMENT ON THE EFFECT OF PRIOR PRODUCT KNOWLEDGE

Tony Ward, Central Queenslands University

ABSTRACT -

This paper describes an experiment to measure the ability of buyers to acquire new product information describinga complex service product from material written in the style of a sales brochure. The literature shows that measurements of product knowledge should be separated into two types of knowledge: declarative and procedural. The results of the research indicate that people with low levels of prior product knowledge acquired more new knowledge than people with higher levels of prior product knowledge, while people acquired an equal amount of new product declarative and procedural knowledge in percentage terms. A model is developed for the buyer’s acquisition of new product knowledge.

INTRODUCTION

Many buyers find difficulty in understanding just what they are buying when confronted with written details of complex service products. Such situations give rise to the question: #What is the ability of readers to acquire the product information contained in sales brochures?’ There are many variables applicable to each individual associated with this question, such as: individual ability to understand the text, prior knowledge, learning ability, and ability to acquire a large amount of new knowledge. This research, using personal superannuation in Australia as the product, is focused on just one of these variables, prior product knowledge.

This paper commences with a discussion of the background to the topic, then proceeds to justification for the research, boundaries of the research, the method used, the test instrument, results, and discussion of the results. The research was conducted in cooperation with the MLC corporation (a life insurance company) who provided industry and product expertise, and the use of their customer base. The research reported here was a part of a more extensive study.

BACKGROUND

Most marketing theory has evolved in the context of consumer goods (Gummesson 1978; 1979; Gilly & Dean 1985), with the majority of marketing textbooks reflecting this approach. In the case of the marketing of service products there is a large volume of academic literature which, upon close inspection, contains a number of anomalies, which has concerned many authors, including: Lovelock (1983); Gershuny (1987); Gummesson (1987); and Beaven and Scotti (1990). Akehurst (1989, p. 4), is most positive in his observations, stating:

There is a serious lack of statistical information and the difficulties of defining, classifying and measuring services is a major obstacle to innovation and prosperity.

In most of the current academic literature, the root cause for many of these anomalies is identified as the #intangibility’ of services, both in the absence of a material form, and the difficulty people experience in #mentally grasping’ the concept of a service, the #cognitive aspect’ of intangibility. This intangibility was first referred to by Adam Smith (1776) and, more recently, by Rathmell (1966); Levitt (1981); Shostack (1982 & 1984); Cowell (1984); Berry (1987 & 1989); Jackson and Cooper (1988); Kenny and Fisk (1990); and Dahringer and Muhlbacher (1991), and remains dominant in the literature. By comparison, goods have the following general characteristics which aid comprehension: material entity (physical appearance); units of length; units of colour; materials specifications; performance specifications; and production drawings. These general characteristics provide the means to visualise, describe and specify almost all goods in very precise terms.

In the case of service products the above are not available. Conceptualising and marketing a service poduct is thus rather more difficult than marketing a good (Wilson 1972; Kotler & Connor 1977; Kotler & Bloom 1984; Van Doren & Smith 1987; and Lovelock 1991). With regard to the cognitive concept of intangibility, the importance to a company of the #need to know a product’ is fundamental to marketing, and is well documented (Cowell 1984; Baker 1987; and Kotler 1991). A similar difficulty applies from a buyer’s perspective. The need for buyers to search for and acquire product information is fundamental in consumer behaviour (Malhotra 1982 & 1984; Baker 1987; Engel; Blackwell & Miniard 1990; and Kotler 1991).

Thus, it is probable that if a service is difficult to #know’, it is difficult both to market and to purchase effectively.

A number of authors (for example, Barry 1986; Cravens & Woodruff 1986; and Kotler & Armstrong 1994) agree on the overall process from problem recognition to post purchase behaviour shown in figure 1. The model in figure 1 shows the five step sequence that a buyer goes through during the decision making process: firstly the recognition of a problem and/or need; secondly a search for information of the options available to solve the problem or meet the need; thirdly an evaluation of the available alternatives, leading to; fourthly the decision to purchase a product; and finally, post purchase behaviour which includes an evaluation of the product.

There has been much research directed at understanding the effects of knowledge on consumer behaviour, such as information search and decision making (Feick, Park & Mothersbaugh 1992). However, rather than studying the effects of product knowledge on consumer behaviour this research is more precisely directed at the buyer’s acquisition of product knowledge, a research area which has received little attention. Indeed, the role of knowledge acquisition is omitted from the simple model in figure 1, because it is implicitly assumed that the acquisition of product knowledge takes place within the second box, information search, and is used in the third box, evaluation of alternatives, which takes product knowledge into account.

Definitions

For this research information is defined as an item of knowledge about a service product. Two types of knowledge were identified and separately tested in this research, and their definitions are taken from the cognitive psychology literature. Firstly, declarative (product) knowledge is defined as factual knowledge of an entity of a service product, and secondly, procedural (product) knowledge is defined as an item of understanding of a service product. The definition of prior knowledge used for this research is knowledge of a domain as an understanding of its basic contents, as well as its goals, rules and/or principles (developed from Chiesi, Spilich & Voss 1979).

FIGURE 1

BUYER DICISION MAKING PROCESS: A SIMPLE MODEL

Justification

Justification for this research is provided from three perspectives. Firstly, the importance of service industries to the Australian economy. In 1991 services accounted for 70 per cent of GDP, employed 80 per cent of the workforce, and comprised 21 per cent of Australia’s exports (Austrade 1994).

Secondly, the general lack of research concerning the acquisition of product knowledge. As is shown below there is a general lack of research concerning the acquisition of product knowledge within the marketing discipline. This research has the potential to contribute to an important gap in the purchase decision making process in regard to the acquisition of product knowledge.

Finally, seeking ways to reduce the perceived intangibility of service products. As discussed above, a number of researchers has recognised that service products are #difficult to grasp mentally’. There appears to be little previous empirical researh which measures the ability of people to acquire product knowledge using objective measurement techniques.

Boundaries of the research

The research was concerned with investigating whether potential customers with different levels of prior knowledge would be able to acquire product information from sales brochure written material. There was no intention to investigate the effect of this factor on the overall #marketability’ of the product, the purchasing intentions of potential customers, or the perceptions of product providers. Comparisons of knowledge acquisition between different types of product, such as #simple’ service products and goods, were specifically excluded.

Prior knowledge

It is generally accepted that prior knowledge results in a schema, or several linked schemata, onto which new knowledge can be built, and without such schemata the acquisition of new knowledge is generally poor (Alba & Hasher 1983). Further, it has been empirically shown that providing ways to link new information with existing knowledge greatly enhances memory and retention in memory (Royer & Perkins 1977).

Thus, as new information is integrated with existing knowledge which is encoded in memory the schemata become enriched and enlarged, and are thus subsequently more likely to be remembered (Alba & Hasher 1983; Celsi & Olson 1988). Alba and Hasher likened this activity to a mapping process which is dependent upon a sufficiently well developed knowledge base. However, there are some indications that as people acquire more knowledge about a domain their subsequent motivation to acquire additional knowledge is reduced, thus resulting in lower levels of knowledge acquisition (Johnson and Russo, as reported in Srull 1983).

The general consensus of opinion appears to be that subjects with prior knowledge of a domain:

- are able to assimilate more complex information;

- are able to identify the information which will be of most use to acquire, and will thus not waste processing effort on acquiring non useful information;

- may not search for additional information due to the lack of a recognised need to acquire additional information;

- may acquire the ability to process complex information more quickly if initial processing objectives establish the need to gain additional knowledge.

Thus there are two conflicting forces identified here which influence knowledge acquisition in opposite directions: firstly, as people acquire more richly developed schema of a domain they have the ability to acquire more knowledge of that domain; and secondly, accompanying this greater knowledge there is often an associated reduction in motivation to acquire additional domain knowledge. This research addressed this conflict.

The buyers’ acquisition of product knowledge

This research is concerned with the acquisition of service product knowledge by buyers from written information. From the cognitive psychology literature (for example, Fitt 1964; Anderson 1983, 1990; and Best 1989) it was established that the accepted process for acquiring knowledge was through:

- acquiring short-term decarative (factual) knowledge;

- the retention of declarative knowledge into long-term memory;

- an associative stage in which a method for performing a skill is worked out, and which includes elaboration, congruence, and more complex schemata (procedural knowledge, or understanding); and,- an autonomous stage, in which a skill is further developed.

Figure 2 shows how the buyer behaviour and cognitive psychology literature can be combined to form a simple model of how product description and product understanding fit into the overall purchasing process. In figure 2 the product is described by the provider, and other factors, such as, need, awareness, and problem recognition, all contribute to the buyer’s acquisition of product knowledge.

Thus, the buyer’s product knowledge will be acquired in declarative and procedural forms, and this knowledge will subsequently be used in the remainder of the purchasing process. The work of Leinberger (1993), and Engel, Blackwell and Miniard (1990) shows that the buyer’s knowledge of the product is an important part of the decision making process, and that both declarative and procedural knowledge play an important role in this process.

RESEARCH QUESTION

The review of the literature on prior knowledge clearly showed that the level of prior knowledge has an effect upon a person’s ability to acquire new knowledge, and to relate that newly acquired knowledge to existing knowledge in such a way that it expands the overall level of knowledge on a given topic. However, as identified above there is conflicting evidence concerning whether a high, medium or low level of prior knowledge results in the most new knowledge being acquired from presented new product information (Bettman and Park 1981; Alba 1983; and Srull et al 1983). The research question is whether there is a relationship between the level of prior knowledge and the amount of new knowledge acquired. From the review above both declarative and procedural knowledge are identified as two different kinds of knowledge which it would be appropriate to measure. Thus, two propositions were formulated as follows:

P1  Subjects with a medium prior knowledge of the personal superannuation product acquire a greater declarative product knowledge than subjects with a low or high prior declarative prior knowledge of the personal superannuation product.

P2  Subjects with a medium prior knowledge of the personal superannuation product acquire a greater procedural product knowledge than subjects with a low or high prior procedural prior knowledge of the personal superannuation product.

FIGURE 2

THE BUYERS' ACQUISITION OF PRODUCT KNOWLEDGE

METHOD

This section reviews the requirements and selection of the experimental design and justifies the design chosen, the Solomon four group design. The features of the Solomon design are then reviewed, including validity and reliability considerations.

The independent variable, prior knowledge, for this research was uncontrolled, while the two dependent variables to be measured were the acquisition of declarative and procedural knowledge respectively. Having identified the research method as an experiment in which peoples’ knowledge was to be measured, review was undertaken which confirmed that this could be carried out in an ethical manner.

The requirement to measure the acquisition of product knowledge immediately placed a requirement on the experimental design to allow for a pre test of respondent prior product knowledge, exposure to a treatment and a post test of respondent product knowledge. A review of available experimental designs showed that the Solomon four group design provided the required control of extraneous variables, provided the method of data collection provided sufficient scope for the inclusion of control experiments (Kidder 1981; Green, Tull, & Albaum 1988; and Malhotra 1993). Helmstadter (1970, p. 110) considered that the Solomon was "the most desirable of all the...basic experimental designs", while Malhotra (1993, p. 235) described the Solomon as a "conceptual ideal". The Solomon four part design offered the following features:

- control of all internal extraneous variables, including pre test sensitisation;

- only a half of the respondents would be required to conduct both pre and post tests;

- the direct correlation of prior knowledge with post treatment knowledge could be undertaken for individual respondents.

The Solomon design thus includes the features required for a high degree of validity and reliability. Four groups of the test instrument were required (Solomon 1949):

EG1     R     O₁      X     O₂     )

                                              ) Type A

CG1     R     O₃               0₄     )

EG2      R               X       0₅    )

                                              ) Type B

CG2     R                         0₆     )

The two groups which contained a pre test were termed type A and the two groups without a pre test were termed type B test instruments.

TABLE 1

TEST INSTRUMENT PAGE LAYOUT

Test instrument structure

The main requirement for the test instrument was to allow a pre test of prior knowledge of the personal superannuation product, exposure to a treatment, followed by a post test of knowledge of the product. In addition, various instructions were required to request respondents to self administer the test instrumen in a certain way. Administering the test instrument by mail was selected as the most practical method of data collection. The mail piece contained three elements, a covering letter, the test instrument, and a reply paid envelope. The test instrument comprised 8 pages, as shown in Table 1.

The test instrument was professionally prepared and printed on two sheets of folded A3, light blue paper. The treatment material contained a general description of the personal superannuation product in Australia, using both declarative and procedural items of information to describe the product. There were 70 items of product information provided in the treatment.

Measurement design

The measurement elements (questions) of the test instrument contained both declarative and procedural questions. In type A test instruments the pre test measurement of prior product knowledge comprised four demographic questions and a set of 27 pre test questions (15 declarative and 12 procedural). Preliminary testing indicated a need for #don’t know’ boxes for many of the questions, as many respondents were not comfortable with leaving a blank when they did not know an answer. The pre-test questions are given at appendix A.

The design of individual questions was reviewed between open-ended, dichotomous and multiple choice. To measure product knowledge it was considered that multiple choice questions alone would not be generally appropriate as they test only recognition, and not recall. Thus, open-ended and dichotomous questions were therefore employed. Dichotomous questions using #true’, #false’, and #don’t know’ responses, and #yes’, #no’, and #don’t know’ responses were used. Two types of open-ended questions were used, firstly where there was a single correct answer to a specific question, and secondly, where there were a number of correct answers. The use of these two types of question was justified by Brucks (1985) who found a high correlation of 0.80 between the two types. The measure of declarative and procedural knowledge was calculated as the additive score obtained by respondents for each set of questions. For both type A and B test instruments the post test measurement commenced at page 6 and was the same as the pre test questions, except that there were no demographic questions, and the question order was changed to reduce the degree of similarity with the pre test section.

Instructions were given in the test instrument requesting respondents to work through the instrument in sequence, and especially not to refer to the treatment material when answering the questions.

The use of the MLC customer base did, however, result in a pre-selection bias of the sample population to previous purchasers of an MLC investment product. Thus, members of the whole population of Australia who had not previously purchased an investment product from MLC were excluded. However, this situation was considered acceptable as previous customers were taken to be representative of people with an interest in purchasing this type of product.

Mailing

Following pilot testing a sample of 4,000 subjects was drawn randomly from the MLC customer base (of over 400,000 people) across the whole of Australia, 250 for the each of the two control groups, and 1,750 for each of the two treatment groups.

In all, 355 responses were received from the main experiment. Of these seven were received after the cut off date and a further 28 were considered to be invalid. The remaining 320 test responses represented a response rate of 8 per cent. The low response rate was anticipated as the test instrument was in effect an examination of the respondents product knowledge, hence the high mail out number. Sufficient valid responses were available for the analysis phase to proceed.

RESULTS

The preliminary analysis phase was designed as an initial inspection of the data, to investigate whether there were any variables which would affect the main analysis, and also to check for non response bias. The preliminary analysis comprised: individual question analysis; individual difference score analysis; measurement of means; establishment of the nature of the data (normality tests); test for pre test sensitisation; demographic analysis; and a non-response check (first half versus second half of responses received as recommended by Armstrong and Overton (1977)). The results of the respondent’s scores for declarative and procedural measures are shown in tables 2 and 3 respectively.

Two findings of note were found in the preliminary analysis. Firstly, a non response check, which compared the first 160 responses received with the second 160 responses received, indicated no significant response bias (z=0.0206, and p=0.9836). Secondly, by comparing the means of the post test scores of groups EG 1 and EG 2 (z=0.5754, and p=0.5650), there was no indication of pre test sensitisation due to respondents answering the pre test questions in the type A test instruments prior to reading the treatment material. The two propositions are now considered in turn.

TABLE 2

RESULTS-DECLARATIVE SCORES

TABLE 3

RESULTS-PROCEDURAL SCORES

PROPOSITION 1BNULL HYPOTHESIS

H1  There will be no difference in the gain of declarative product knowledge of the personal superannuation product between subjects with low, medium and high pre test scores of declarative knowledge.

Prior knowledge of the personal superannuation product was an uncontrolled independent variable in the experiment, and was measured in the type A test instruments by the 15 declarative pre test questions in group EG 1. The pre test scores of the 128 type A respondents in EG 1 are given at figure 3.

Only 13 respondents scored over 6, so this formed a #high’ category, and the #medium’ category was then chosen as a score of 3 to 6, leaving the #low’ category with scores of 0, 1 and 2. The selection of these three categories by pre test score is acknowledged as arbitrary, however, this method does reflect the focus of the experiment which was to test the rate at which these different categories of respondents scored. The number of respondents in each category, the gain means, and the standard deviations are shown in table 4.

Comparisons were conducted between the three categories using a t-test when the sample sizes were in excess of 30, and a Mann-Whitney test when one of the sample sizes was under 30, and when the distribution of that sample was non normal.

The comparison between the medium and high categories indicates that there is a significant difference (z=2.0169, p=0.0437), while the tests between the low and medium (t=4.17, p=0.0000), and the low and high (z=3.5408, p=0.0004) categories indicate a very highly significant difference. The means in table 4 indicate that respondents with a high level of prior knowledge acquired less declarative product knowledge than respondents with medium or low prior knowledge, and so a linear regression was performed which varied the gain in declarative scores by individual (not grouped into the three categories) pre test score.

The relationship between the pre test score and the gain in declarative knowledge was very highly significant (p=0.0000), and the correlation coefficient of -0.47335, explained 22.406 per cent of the variance (R squared) betwen pre test score and gain. The intercept with the vertical axis was 5.43, and the slope of the line was -0.434.

FIGURE 3

DISTRIBUTION OF PRE TEST DECLARATIVE SCORES

TABLE 4

COMPARISON OF DECLARATIVE MEANS

The results of the linear regression are thus very highly significant, and indicate a very strong treatment negative relationship between the pre test score and the gain in declarative knowledge. Thus the null hypothesis that there is no difference in the gain of declarative knowledge between subjects with low, medium and high prior product knowledge cannot be accepted.

PROPOSITION 2BNULL HYPOTHESIS

H2  There will be no difference in the gain of procedural product knowledge of the personal superannuation product between subjects with low, medium and high pre test scores of procedural knowledge.

The analysis procedure used to test hypothesis 2 was the same as for hypothesis 1. The procedural pre test scores of the 128 type A respondents are given at figure 4. Twenty six respondents scored over 7, so this formed a #high’ category, and the #medium’ category was then chosen as a score of 5 to 7, leaving the #low’ category with scores of 0 to 4.

Table 5 shows the number of respondents in each category, the gain means, and the standard deviations. Tests to compare the three categories showed a significant difference between the low and medium categories (t=2.21, p=0.031), and a very highly significant difference between the low and high categories (z=4.5034, p=0.0000) and the medium and high categories (z=3.9459, p=0.0001).

A linear regression was then performed which varied the gain in procedural scores by the pre test score. The relationship between the pre test score and the gain in procedural knowledge was very highly significant (p=0.0000), and the correlation coefficient of -0.45460, explained 20.666 per cent of the variance between pre test score and gain. The intercept with the vertical axis was 4.15, and the slope of the line was -0.361.

The results of the linear regression are thus very highly significant, and indicate a very strong treatment negative relationship between the pre test score and the gain in procedural knowledge. Thus the null hypothesis that there is no difference in the gain of procedural knowledge between subjects with low, medium and high prior product knowledge cannot be accepted.

Further analysis of the relationship between prior knowledge and the acquisition of product knowledge

The results of the analysis of the testing of the two propositions above indicated a strong negative relationship between prior product knowledge and the acquisition of additional product knowledge, which warranted further analysis. Figure 5 shows the relationship between the pre test scores (x-axis) and gains (y-axis) for declarative and procedural knowledge, and clearly illustrates the negative relationship between pre test score and gain in knowledge.

One possibility for the nature of this negative relationship is due to the bounded nature of the experiment. The gain scores were recalculated as percentages of the questions answered incorrectly in the pre test (the percentage gain available), thus taking the bounded nature of the experiment into account, and are plotted in figure 6.

FIGURE 4

DISTRIBUTION OF PRE TEST PROCEDURAL SCORES

TABLE 5

COMPARISON OF MEANS FOR THE PROCEDURAL SCORES

Figure 6 shows that the amount of learning relating pre test score and gain (when expressed as a percentage of the available gain) is generally in the range between 20 and 50 per cent. The means for the declarative and procedural knowledge percentage gains available (irrespective of pre test score) were 33.6 and 32.6 per cent respectively. A one sample chi-square test was used to determine whether there was a correlation between the two distributions of the perentage gain available and pre test score. The result obtained was very highly significant (c²₌1758, df=1584, p=0.001), indicating a very high correlation between the percentage available gains in declarative and procedural knowledge with pre test score.

Thus, although people with a low pre test score learnt the most from the treatments, the percentage of available learning gain was generally even with respect to low, medium and high pre test scores. In addition to figures 5 and 6, linear regression tests of pre test score against percentage gain available was conducted for declarative and procedural knowledge, but neither of these regressions was significant (F=1.17, p=0.2818; F=0.67, p=0.4140 respectively), indicating no linear relationship between the percentage gain available and pre test score.

In conclusion, the results of the testing of the propositions of declarative and procedural product knowledge show a very strong negative relationship between the pre test score and the gain in product knowledge acquired from the treatments. However, when this acquisition of additional product knowledge is expressed as a percentage of the gain available there is no significant linear relationship found with the pre test score. The acquisition in product knowledge averaged about 33 per cent of available learning, and showed a very high degree of correlation between declarative and procedural knowledge with the pre test score.

DISCUSSION

The results of this research are not in accordance with the relationship found by Alba and Hasher (1983), Srull et al (1983), or Celsi and Olsen (1988) who all postulated that subjects who had a high level of prior knowledge would be able to acquire new knowledge more quickly than subjects with a low or medium level of prior knowledge. Rather, the findings reflect the concerns of, firstly, Freedle and Carroll (1972) who noted difficulties of expressing the assumptions and presuppositions underlying written material, and secondly, Brucks (1985) who found that the relationship between experience and the search for information was inconsistent. Johnson and Russo (as reported by Srull 1983) found that the effect of prior knowledge on memory was highly dependent on the objectives of the subject, which could account for some of the large degree of unexplained variance associated with the linear regression tests.

The results of this research indicate that for the acquisition of additional product knowledge of a complex service product the motivation to acquire additional knowledge is dominant over the ability to acquire additional product knowledge. This finding therefore indicates that customers will process product information to a certain level and will then not be generally willing to engage in the extra cognitive effort required to fully use their ability to acquire additional product knowledge.

FIGURE 5

RELATIONSHIP BETWEEN PRE TEST SCORE AND GAIN IN KNOWLEDGE

FIGURE 6

RELATIONSHIP BETWEEN PRE TEST SCORE AND PERCENTAGE GAIN AVAILABLE

FIGURE 7

MODIFIED MODEL - BUYER'S ACQUISITION OF PRODUCT KNOWLEDGE

It should be noted that very little literature could be found which directly addressed the acquisition of product knowledge, either of goods or service products, a factor which indicates the embryonic nature of this field of research. It is thus difficult to identify other factors which could account for this finding from the literature, but four possibilities have been identified.

Firstly, it is possible that the way in which potential customers acquire product knowledge is different from other forms of knowledge, which would have significant repercussions for both marketing and cognitive psychology disciplines.

Secondly, there is a possibility that a threshold exits, whereby subjects acquire product knowledge to a certain level, and then either lose motivation to acquire further product knowledge, or encounter a degree of difficulty with the details of the subject, whereupon the acquisition of additional knowledge becomes slowed.

Thirdly, the use of a sample of existing MLC customers may have had aneffect upon the prior knowledge of respondents, possibly with respect to financial products other than personal superannuation, but sufficient to provide a more general prior knowledge of similar products, that is, respondents already had a mapping capability (c.f., Alba and Hasher 1983).

Finally, the results prompted four main modifications to the model shown in figure 2:

1) To delete declarative knowledge leading directly to procedural knowledge,

2) the use of marketing terms in place of cognitive terms in the acquisition of declarative and procedural knowledge boxes,

3) the inclusion of a step for procedural knowledge being acquired through the cognitive effort of the buyer, and

4) the inclusion of prior knowledge.

Thus, this modified model shown at figure 7 was developed from an analysis of the experiment results.

The modified model shown in figure 7 shows the product, product description and #other factors’ as before. The strong influence of prior knowledge is also recognised, and has been separated from #other factors’ because of its importance. The learning process has been divided into three, rather than two, elements. Firstly, the buyer’s acquisition of declaratively encoded knowledge, which is the buyer’s ability to remember declarative information provided by the marketer. Secondly, the buyer’s acquisition of procedural knowledge which is declaratively encoded, and is the buyer’s ability to remember procedural product information related by the marketer. Finally, a new stage has been added, in which buyer’s create procedural knowledge through their own cognitive effort; that is, acquire an understanding of the product in a procedurally encoded format, and thus gain an understanding of how it could be of benefit to themselves in their own specific context. This last stage was not included in the experiment in this research. The improvements to this model make it easier to test in future (though this will remain a difficult task), and it is now more consistent with the results than the original model developed in figure 2.

APPENDIX 1

PRE TEST QUESTIONS

CONCLUSIONS

This research has contributed to existing knowledge in two ways, firstly, by showing that customers do not necessarily acquire new knowledge about products in the same way as they acquire knowledge about other entities. The very strong negative relationship between the level of prior knowledge and the gain in knowledge resulting from the treatments clearly indicate that there are processes involved which, at present, cannot be fully explained. Secondly, a refined model has been developed which indicates the complex nature of the acquisition of product knowledge of complex service products and the need to separately measure the acquisition of customer declarative and procedural knowledge.

Much more research is required to identify possible additional variables, and to seek more data on the attitudes and opinions of subjects with respect to product knowledge. In addition, comparisons between the ability of subjects to acquire product knowledge of service products of varying degrees of complexity should be conducted, especially taking into account the complexity of certain service products. Such additional research will assist marketing practitioners to better understand factors which affect the cognitive aspect of the intangibility of service products.

REFERENCES

Akehurst, G.P. 1989, "Service industries", in Management in Service Industries, ed Jones, P., Pitman, London.

Alba, J.W. & Hasher, L. 1983, "Is memory schematic?", Psychological Bulletin, vol. 93, no. 2, pp. 203-31.

Alba, J.W. 1983, "The effects of product knowledge on the comprehension, retention, and evaluation of product information", Advances in Consumer Research, vol. 10, pp. 577-80.

Anderson, J.R. 1983, The Architecture of Cognition, Harvard University Press, Mass.

Anderson, J.R. 1990, Cognitive Psychology and Its Implications, Freeman, New York.

Armstrong, J.S. & Overton, T.S. 1977, "Estimating nonresponse bias in mail surveys", Journal of Marketing Research, vol. 14, August, pp. 396-402.

Austrade 1994, Intelligent Exports: And the Silent Revolution in Services, Australian Trade Commission, Sydney.

Baker, M.J. 1987, The Marketing Book, Heinemann, London.

Barry, T.E. 1986, Marketing: An Integrated Approach, Dryden, New York.

Bartlett, F.C. 1932, Remembering, Cambridge University Press, Cambridge.

Beaven, M.H. & Scotti, D.J. 1990, "Service-oriented thinking and its implications for the marketing mix", The Journal of Services Marketing, vol. 4, no. 4, Fall, pp. 5-19.

Berry, L.L. 1987, "Big ideas in services marketing," The Journal of Services Marketing, vol. 1, no. 1, Summer, pp. 5-9.

Berry, L.L. 1989, "4 keys to success with new services", Supermarket Business, vol. 44, no. 12, December, pp.19a-21a.

Best, J.B. 1989, Cognitive Psychology, 2nd edn, West Publishing, New York.

Bettman, J.R. and Park, C.W. 1980, "Effects of prior knowledge and experience and phase of the choice process on consumer decision processes: A protocol analysis", Journal of Consumer Research, vol. 7, December, pp. 234-48.

Brucks, M. 1985, "The effects of product class knowledge on information search behavior", Journal of Consumer Research, vol. 12, June, pp. 1-16.

Campbell, D. & Stanley, J. 1963, Experimental and Quasi-experimental Designs for Research, Rand McNally, Chicago.

Celsi, R.L. & Olson, J.C. 1988, "The role of involvement in attention and comprehension processes", Journal of Consumer Research, vol. 15, September, pp. 210-24.

Chiesi, H., Spilich, G. & Voss, J.T. 1979, "Acquisition of domain related information in relation to high and low domain knowledge", Journal of Verbal Learning and Verbal Behavior, vol. 18, pp. 257-73.

Cowell, D.W. 1984, The Marketing of Services, Heinemann, Oxford.

Cravens, D.W. & Woodruff, R.B. 1986, Marketing, Addison-Wesley, Massachusetts.

Dahringer, L.D. & Muhlbacher, H. 1991, International Marketing: A Global Perspective, Addison-Wesley, Massachusetts.

Donnelly, J.H. 1976, "Marketing notes and communications: Marketing intermediaries in channels of distribution for services," Journal of Marketing, vol. 40, no. 1, pp. 55-70.

Engel, J.F., Blackwell, R.D. & Miniard, P.W. 1990, Consumer Behavior, 6th edn, Dryden Press, Fort Worth.

Feick, L., Park, C.W. & Mothersbaugh, D.L. 1992, "Knowledge and knowledge of knowledge: What we know, what we think we know, and why the difference makes a difference", Advances in Consumer Research, vol. 19, pp. 190-92.

Fitt, P.M. 1964, "Perceptual-motor skill learning", in Melton (ed), Categories of Human Learning, Academic Press, New York, pp. 243-85.

Freedle, R.O. & Carroll, J.B. 1972, "Language comprehension and the acquisition of knowledge: Reflections", in Freedle and Carroll (eds), Language comprehension and the acquisition of knowledge: Reflections, Wiley, New York, pp. 359-67.

Gershuny, J. 1987, #Time use and the dynamics of the service sector’, Services Industry Journal, vol. 7, no. 4, October, pp. 56-71.

Gilly, M.C. & Dean, D.L. 1985, "A market-oriented taxonomy of public services: Implications for marketing management", in Services Marketing in a Changing Environment, American Marketing Association Proceedings, Chicago, pp. 94-97.

Green, P.E., Tull, D.S. & Albaum, G. 1988, Research for Marketing Decisions, 5th edn., Prentice-Hall, New Jersey.

Gummesson, E. 1978, "Toward a theory of professional service marketing", Industrial Marketing Management, vol. 7, pp. 89-95.

Gummesson, E. 1979, "The marketing of professional services-An organisational dilemma", European Journal of Marketing, vol. 13, no. 5, pp. 308-18.

Gummesson, E. 1987, "Lip service-a neglected area in services marketing: A commentary", The Journal of Services Marketing, vol. 1, no. 1, pp. 19-23.

Helmstadter, G.C. 1970, Research Concepts in Human Behavior: Education, Psychology, Sociology, Appleton-Century-Crofts, New York.

Jackson, R.W. & Cooper, P.D. 1988, "Unique aspects of marketing industrial services", Industrial Marketing Management, vol. 17, pp. 111-18.

Kenny, J.T. & Fisk, R.P. 1990, "Why services promotion differs from goods promotion", in Marketing Theory and Applications, American Marketing Society Winter Educators’ Conference, Chicago, vol.1, pp. 177-81.

Kidder, L.H. 1981, Selltiz, Wrightman and Cook’s Research Methods in Social Relations, 4th edn, Holt, Reinhart and Winston, New York.

Kotler, P. & Armstrong, G. 1994, Principles of Marketing, 6th edn, Prentice-Hall, Englewood Cliffs.

Kotler, P. & Bloom, P.N. 1984, Marketing of Professional Services, Prentice-Hall, Englewood Cliffs.

Kotler, P. & Connor, Jr. R.A. 1977, "Marketing professional services", Journal of Marketing, January, pp. 71-76.

Kotler, P. 1991, Marketing Management: Analysis, Planning, Implementation and Control, 7th edn, Prentice-Hall, Englewood Cliffs.

Leinberger, P. 1993, "Clued up on the intelligent consumer", Australian Professional Marketing, February, pp. 38-9.

Levitt, T. 1981, "Marketing intangible products and product intangibles," Harvard Business Review, May/June, pp. 94-102.

Lovelock, C.H. 1983, "Classifying services to gain strategic marketing insights," Journal of Marketing, vol. 47, no. 3, Summer, pp. 9-20.

Lovelock, C.H. 1991, Services Marketing, Second edn, Prentice-Hall, Englewood Cliffs.

Malhotra, N.K. 1982, "Information load and consumer decision making", quoted in Owen, R.S. 1992, "Clarifying the simple assumption of the information load paradigm", Advances in Consumer Research, vol. 19, pp. 770-6.

Malhotra, N.K. 1984, "Reflections of the information overload paradigm in consumer decision making", quoted in Owen, R.S. 1992, #Clarifying the simple assumption of the information load paradigm, Advances in Consumer Research, vol. 19, pp. 770-6.

Malhotra, N.K. 1993, Marketing Research: An Applied Orientation, Prentice Hall, Englewood Cliffs.

Minsky, M.L. 1975, "A framework of representing knowledge", in Winston, (ed), The Psychology of Computer Vision, McGraw Hill, New York, pp. 211-77.

Rathmell, J.M. 1966, "What is meantby services?," Journal of Marketing, vol. 30, no. 4, October, pp. 32-36.

Royer, J.M. & Perkins, M.R. 1977, "Facilitative transfer in prose learning over an extended period of time", quoted in Alba, J.W. & Hasher, L. 1983, "Is memory schematic?", Psychological Bulletin, vol. 93, no. 2, pp. 203-31.

Shostack, G.L. 1982, "How to design a service," European Journal of Marketing, vol. 16, no. 1, pp. 49-63.

Shostack, G.L. 1984 ,’Designing services that deliver’, Harvard Business Review, vol. 62, no. 1, January/February, pp. 133-39.

Smith Adam, 1776, An Inquiry into the Nature and Causes of the Wealth of Nations, reprinted 1937, Modern Library, New York.

Solomon, R.L. 1949, "An extension of control group design", Psychological Bulletin, vol. 46, pp. 137-50.

Srull, T.K. 1983, #The role of prior knowledge in the acquisition, retention, and use of new information’, Advances in Consumer Research, vol. 10, pp. 572-6.

Van Doren, D.C. & Smith, L.W. 1987, "Marketing in the restructured professional services field", The Journal of Services Marketing, vol. 1, no. 1, Summer, pp. 67-75.

Wilson, A. 1972, The Marketing of Professional Services, McGraw-Hill, London.

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