A Script Theoretic Approach to Informatin Processing: an Energy Conservation Application

Lorne Bozinoff, The Pennsylvania State University
ABSTRACT - The paradox of how unconscious or routine response behavior can occur within an information processing paradigm is examined and resolved in terms of script theory. The empirical evidence concerning scripts is reviewed and an application in the area of energy conservation is suggested. An empirical study based on script theory yielded three findings. First, support for the existence of scripts is found. Second, scripts can be reliably measured. And third, scripts can provide insights into routine response behavior not available through traditional consumer research methods.
[ to cite ]:
Lorne Bozinoff (1982) ,"A Script Theoretic Approach to Informatin Processing: an Energy Conservation Application", in NA - Advances in Consumer Research Volume 09, eds. Andrew Mitchell, Ann Abor, MI : Association for Consumer Research, Pages: 481-486.

Advances in Consumer Research Volume 9, 1982      Pages 481-486

A SCRIPT THEORETIC APPROACH TO INFORMATIN PROCESSING: AN ENERGY CONSERVATION APPLICATION

Lorne Bozinoff, The Pennsylvania State University

[The author is now Marketing Research Manager at Bell Canada, Toronto, Ontario. This research was supported by the Center for Research, College of Business Administration, The Pennsylvania State University. The author thanks Jerry Olson and Arno Rethans for their insightful comments on an earlier draft of this paper.]

ABSTRACT -

The paradox of how unconscious or routine response behavior can occur within an information processing paradigm is examined and resolved in terms of script theory. The empirical evidence concerning scripts is reviewed and an application in the area of energy conservation is suggested. An empirical study based on script theory yielded three findings. First, support for the existence of scripts is found. Second, scripts can be reliably measured. And third, scripts can provide insights into routine response behavior not available through traditional consumer research methods.

INTRODUCTION

It has long been held that the amount of information processing undertaken by consumers varies across product classes. Consumers in the process of buying either a new or expensive product are hypothesized to engage in more information processing than consumers who are engaged in the process of buying an inexpensive, frequently purchased good. For example, there is evidence that as the price of the product increases, the extent of information search increases (Dommermuth, 1965). Also, there is evidence that as product class familiarity decreases, the extent of information search increases (Bennett and Mandell, 1969). In contrast, it is posited that for frequently purchased consumer goods, little information processing will occur. In such instances, brand loyalty and routine response behavior (RRB) are likely to occur (Howard, 1977). Taken in the extreme, this type of behavior may tend to be unconscious and similar to a habit

Routine or unconscious behavior appears to run counter to the more conscious extended problem solving (EPS) behavior studied by most consumer researchers (Howard, 1977). Although conscious decision making behavior is more widely studied than unconscious behavior, ironically, people are far more likely to engage in unconscious behavior than conscious behavior (Lachman, Lachman and Butterfield, 1979). A key issue then is how RRB can be explained within a more consciously oriented information processing paradigm.

SCRIPT THEORY

A common meeting ground can be found in the notion of scripts (Abelson, 1980; Shank and Abelson, 1977; Shank, 1980). According to most current theories of cognitive structure, knowledge is stored in long term memory (LTH) as schemata (Lindsay and Norman, 1973). Schemata are the large sets of well structured cognitions that have been learned over time as experience accumulates (Norman and Bobrow, 1975). Schemata consist of a framework for organizing the information about a concept into a meaningful structure. A script can be thought of as a temporally ordered schema. That is, a script is a knowledge structure stored in LTM which contains a sequential ordering. Specifically, scripts contain a series of actions, each arranged in terms of temporal order - hence the name "scripts." The most commonly cited script is the RESTAURANT script (Abelson, 1980). The RESTAURANT script contains a series of actions which are expected to occur when one goes to a restaurant, e.g., talking to the maitre'd, being shown to a table, reading the wine list and so forth. A restaurant schema on the other hand would not contain the information that these actions are expected to occur in a distinct temporal order

Scripts which have been stored in memory are used to direct behavior when activated. Activation is believed to occur automaticallY by the situational context. When one enters a restaurant, the RESTAURANT script will be activated and thus begin to guide behavior. On the other hand, when one enters a dentist's office, the DENTAL VISIT script becomes activated. It might contain such actions as checking in with the receptionist, sitting down, waiting, looking at other people, reading magazines and so forth. An activated script will result in unconscious or routine response behavior because while the script is initially developed or constructed in a conscious learning context, once learned, no new conscious effort is required. The situation acts as a cue to automatically trigger the appropriate script. Using a computer analogy, the script would be like a stored subroutine called by the central processor from time to time.

Scripts are inherently useful because of their ability to conserve the individual's limited information processing capacity. By retaining actions or procedures learned from previous situations, new actions need not be learned for new, but relatively similar, situations. The RESTAURANT script saves the consumer the trouble of deciding what should be done or what is expected upon entering a different restaurant. This information is already available in LTM in the form of scripts.

A script need not specify all the actions in complete detail. As Abelson (1980) notes, "free behaviors" may be allowed to occur in certain scripts. For example, the DENTAL VISIT script may prescribe all the actions up until sitting in the dentist's chair. However, the course of the conversation between the patient and the dentist may not be scripted - any subject of conversation may be permitted. At this point, the script temporarily ceases to guide behavior which is now directed by more conscious processes. Once the conversation with the dentist has ended, the script is activated again. In terms of a computer analogy, this is similar to a subroutine which returns control of the program back to the main part of the program and then is called again by the main program at some later point.

Scripts may also differ in terms of strength. Strong scripts are very explicit and contain a great number of "scene" constraints. That is, these scripts specify many actions or scenes which must occur. Weak scripts are vague and have fewer scene constraints. These scripts contain many scenes which do not have high probabilities of occurring. For example, a strong RESTAURANT script would always include a maitre'd but a weak script would not.

EMPIRICAL EVIDENCE OF SCRIPTS

Most research concerning the nature of scripts is very recent. However,there is also some older evidence. Bartlett (1932), for example, found that subjects when given stories to memorize tended to change the stories so that they conformed more to their own ideas or scripts. Brandsford, Barclay and Franks (1972) found that subjects in a memory task tend to "construct semantic descriptions of situations" rather than to remember the details of information presented. Baggett (1975) found also that subjects tend to fill in missing details in a memory task in order -to form z coherent story. In all of these studies the implication is clear that subjects can and do draw upon their own scripts to fill in missing details of stories, and that they tend to store these stories in the form of scripts.

More recently, Bower, Black and Turner (1979) conducted a series of studies on scripts. They found five significant results concerning scripts. First, they showed that there are script norms. That is, when asked to describe familiar activities, subjects tend to write similar descriptions. The subjects tended to use the same characters, props, actions and order of events.

Second, subjects tended to agree on how a series of action sequences should be grouped into segments or "scenes." Third, in a recall task, subjects tended to confuse stated actions with unstated actions implied by a script. Fourth, they found that given a series of actions scrambled out of order, subjects tended to use a common "natural" or "canonical" order in unscrambling the actions.

Finally, it was shown that subjects tended to remember exceptional actions rather than scripted actions. This finding has been replicated in other studies (e.g., Graesser, Gordon and Sawyer, 1979; Graesser, Woll, Kowalski and Smith, 1980) and is very important because it is direct evidence of the unconscious nature of scripts. Subjects do not tend to notice or be conscious of scripted activities. Thus the use of scripts tends to be unconscious. However, it was shown that subjects can consciously draw upon their scripts when asked to describe familiar activities. The net result is an unconscious process - the initial development of scripts is conscious but over time, through overlearning, the use of scripts becomes unconscious.

AN APPLICATION OF SCRIPT THEORY: ENERGY CONSERVATION

Energy consumption can be viewed from a "scripts" perspective. The use of energy in some situations has become so routine that it is possible that energy use actions have become parts of strong scripts. For example, a consumer may have, at one point in time, decided that the best method of traveling to work each day was by private automobile. Once the consumer has purchased an automobile, he will seldom reconsider his daily use of the automobile for commuting purposes. His continued automobile usage is now a matter of unconscious or script-directed behavior. Little additional information processing will occur in this context. A commuter going to work may have a GOING TO WORK script which contains such activities as 'getting into a car,' 'driving to work alone' and 'parking the car.' It may not contain activities like 'turning off all the lights' and 'turning down the heat.' This script will be triggered by certain early morning cues such as the time of day and day of the week.

This does not mean that energy usage is totally unconscious. If questioned directly, the consumer could report his use of the automobile each morning. Rather, his behavior is script-directed. The initial behavior was conscious but over time a GOING TO WORK script has been constructed and stored in LTM. It is this script which now guides behavior. When the appropriate early morning situational cues are present, this script becomes activated and pre-empts any conscious consideration of how to get to the office.

Script-directed energy use may explain why several attempts to reduce consumers' energy consumption have failed. For example, McClelland and Cook (1980) tried to reduce energy use by providing monetary rebates to consumers every two weeks. The amount of the rebate was determined by the amount of energy saved. It was found that the rebate plan did not significantly reduce energy consumption. In a similar study, Winett, Kagel, Battalis and Winkler (1978) found that neither monetary rewards nor energy use feedback information reduced energy consumption to any great extent. McNeill and Wilkie (1979) Looked at the impact of energy labels upon appliance selection and found that consumers do not actively seek energy information. The failure of monetary incentives to affect energy consumption may be due to the fact that if energy use is script-directed, then incentives must be provided at the time that the energy scripts are being activated. A monthly electric bill rebate will not be effective in getting consumers to turn down the heat each morning if this is not in the GOING TO WORK script. Again, scripts are activated automatically depending upon the situational context. The finding that consumer; to not actively seek energy information when buying appliances indicates that a 'get energy information' action is not in consumers BUYING APPLIANCE scripts.

RESEARCH PURPOSE

The study described below serves only as an initial, exploratory investigation of some of the ideas discussed above. The purpose of this study is to answer three questions. First, can scripts be measured for situations for interest to consumer researchers? In addition, what is the variability of scripts across consumers for the same situation? Second, how reliable are scripts? If scripts are to be of some use to consumer researchers, they must be reliably measured. Third, and most importantly, to scripts yield additional insights into consumer behavior which cannot be provided by traditional consumer research methods?

METHOD

Subjects

Forty-three subjects were recruited from two undergraduate marketing management classes. During regular class time, students were asked to participate in the study. Participation was voluntary, although all of the students agreed to participate. Approximately 60% of the sample were males. A convenience sample of students is defensible since the principal aim of the present research is exploratory in nature. In addition, Kruglanski (1975) has argued that homogeneous samples (such as students in particular classes at a given institution) are appropriate for theory-oriented research where individual differences are not of theoretical interest.

Measurement Instrument

The subjects were asked to generate a list of events or actions for each of three situations - getting up in the morning 9 preparing a Thanksgiving dinner and a day at the office. These situations were chosen because they each were hypothesized to contain several energy related activities. For example, a MORNING script might contain such activities as brushing teeth with an electric toothbrush, using an electric shavers blow trying wet hair, turning off the lights, turning town the heat when leaving the house and so forth. Energy related activities for a THANKSGIVING DINNER script might include using an electric can-opener and an electric carving knife. The OFFICE script was hypothesized to include driving to work and using an elevator to get to the office. These energy related activities are particularly interesting because for each activity there is an alternative which uses little or no energy. The specific instructions for generating the scripts were the same as those used by Bower, Black and Turner (1980). The instructions for the MORNING script were as follows:

"Write a list of actions describing what you normally to when you get up in the morning. We are interested in the common actions of a routine getting-up-in-the-morning stereotype. Start the list with waking up and end it with leaving the house. Include about 20 actions or events and put them in the order in which they would occur."

Because scripts are learned through experience with many similar situations, they contain information concerning a generalized or stereotypical situation. Consequently, this is the level at which they should be (ant were) measured. It is not absolutely necessary that a subject may never have actually cooked a Thanksgiving dinner or gone to work in an office. A subject should still have had some indirect experience upon which to construct a THANKSGIVING DINNER script and an OFFICE script. In fact, over 75% of the respondents were able to write THANKSGIVING DINNER and OFFICE scripts and all subjects were able to construct GETTING UP scripts.

After writing the scripts, the respondents answered twelve Likert-type items drawn from Antil and Bennett's (1979) Socially Responsible Consumption Behavior Scale (SRCB). Some items dealt specifically with attitudes towards energy conservation and some items dealt with general ecological issues. The SRCB was chosen because of its high reliability and convergent validity with other energy conservation scales (Antil and Bennett, 1979). Finally, respondents completed some energy use self-report measures.

RESULTS

Script Norms

The scripts were edited and tabulated according to frequency of citation of specific events and actions. Paraphrases and synonyms were lumped together. The first issue is whether respondents agree in the actions they mentioned. Because people draw on the same general experiences to construct scripts, it is hypothesized that there should be a core of actions common to all the respondents ' scripts.

Tables 1, 2 and 3 report the frequency of occurrence for actions listed in the three scripts. As hypothesized, there was a common core of actions cited by the respondents. In the GETTING UP script, over 75% of the respondents mentioned taking a shower, getting dressed, brushing teeth, getting books together and leaving the house. In addition, over 502 of the respondents mentioned waking up, turning off the alarm clock, getting up, going to the bathroom, trying hair and having breakfast. Another 252 mentioned such things as combing hair. shaving and so forth.

Similar findings were found for the other two scripts. For the Thanksgiving DINNER script, over 75% of the respondents mentioned stuffing the turkey, putting the turkey in the oven, and cooking the vegetables. In the OFFICE script, over 75% of the subjects mentioned driving to work, having coffee and having lunch.

Script Reliability

Given the nature of the script generation task, it is very difficult to conduct a test-retest reliability study. Respondents generating scripts on a second occasion may very well recall the scripts which they generated on the first occasion. Because of this, the reliability of the frequency with which a specific event or action is mentioned was calculated by dividing the sample in half and correlating the frequencies of actions mentioned by the two halves. The Pearson product moment correlation for the THANKSGIVING scripts was .76 and for the OFFICE script it was .81. (The reliability was not calculated for the remaining script.) Thus, for a homogeneous group like undergraduate students, there is a fairly high degree of script action frequency reliability.

Script Versus Self-Report Measurement

The remaining issue concerns whether script measures provide any additional insights into consumer behavior which are not available through the more traditional consumer research methods. Although the scripts which were generated contain a considerable amount of detail, most energy related actions which were listed were not detailed enough to allow for an analysis of whether an energy saving alternative was used. For example, in the THANKSGIVING DINNER script, over 50% of the respondents mentioned carving the turkey. However, not one respondent mentioned whether an electric knife was used. In the GETTING UP script, over 50% of the respondents mentioned drying their hair although very few respondents mentioned whether a blow dryer was used. This is indirect evidence of the unconscious nature of many energy related activities. Many energy related activities are so routine or automatic that even the process of generating a script fails to make these behaviors conscious.

TABLE 1

EMPIRICAL GETTING UP SCRIPT NORMS

TABLE 2

EMPIRICAL THANKSGIVING DINNER SCRIPT NORM

TABLE 3

EMPIRICAL OFFICE SCRIPT NORMS

One energy related behavior which was specified in considerable detail was found in the OFFICE script. More than 75% of the respondents reported driving to work while the remainder reported taking the bus, walking or joining a carpool. No one failed to report how they traveled to work. The method of commuting listed in the scripts was cross-tabulated with the method of commuting reported in the self-report measure of energy use. Table 4 shows that self-reported methods of commuting were not related to the script generated methods of commuting. (Because of the low cell frequencies in Table 4, this result should be interpreted with some caution.)

TABLE 4

FREQUENCIES OF SCRIPT GENERATED MEASURES BY SELF-REPORT MEASURES FOR COMMUTING METHODS

In another comparison of self-report and script generated measures of energy related activities, respondents who mentioned turning off the lights when leaving the house in the morning (in the GETTING UP script) were cross-tabulated with those in the self-report who indicated that they turned off the lights when not in the house. Table 5 indicates that again self-reported and script generated measures are not significantly related to each other. One possible conclusion is that energy conservation activities as well as energy use activities are primarily unconscious. (Because of the low cell frequencies in Table 5, this result should be interpreted with some caution)

TABLE 5

FREQUENCIES OF SCRIPT GENERATED MEASURES BY SELF-REPORT MEASURES FOR TURNING OFF LIGHTS

Script Versus Attitude Measurement

Respondents who claim to be interested in energy conservation and general ecological issues should be more conscious of energy related activities. Presumably someone who feels energy conservation is important should put more conscious effort into energy use behaviors. In order to test this hypothesis, the sample was divided into two halves based on total energy conservation and ecological concern attitude scores. Respondents who score low in terms of energy conservation concern should mention fewer energy related activities in their scripts than respondents who claim a higher concern for energy conservation. Table 6 reports the results of this analysis.

The degree to which a respondent claims to be interested in energy or ecological issues had little impact upon the likelihood that an energy related activity would be reported. Respondents who are interested in energy issues did not in general report more energy related activities than the respondents who were less interested in energy issues. The most notable exception occurred with the OFFICE scripts. Respondents who are interested in energy issues have a significantly greater propensity to mention driving their car to work and using the elevator than do the uninterested respondents.

DISCUSSION

The present study provides some tentative answers to the questions raised earlier. First, it has been shown that scripts can be generated by respondents for situations of interest to consumer researchers. Furthermore, there does appear to be a common core of activities found in the scripts. Second, the reliability of scripts appears to be well within an acceptable range.

TABLE 6

PROPORTION OF RESPONDENTS WHO MENTIONED ENERGY RELATED ACTIVITIES

More importantly, scripts do provide insights into consumer behavior which are not available through the use of the more traditional consumer research methods. For example, it was shown that self-reported behavior differs considerably from script generated behavior. One obvious reason for this may be social desirability. It is relatively difficult for a respondent to provide socially desirable scripts because the goals of the research are not easily discernable to the respondent. This is not the case for self-report measures.

Finally, this study provides some support for the scripts conceptualization provided above. It was reported that a great number of energy related activities were not reported by the respondents. One reason for this may be that these behaviors are generally unconscious and hence not easily recalled by the respondent. As expected however, those respondents who were interested in energy issues did tend to report more energy related activities. In no instance did the less interested subjects report significantly more energy related activities.

CONCLUSIONS AND IMPLICATIONS

As Bower, Black and Turner (1979) note, key issues remaining in script theory are how to access scripts, and how to learn new scripts. Script theory does offer some guidelines however. Because scripts are triggered by specific situational cues, efforts to change scripts must be aimed at these critical situations. For advertisers attempting to change existing scripts there are a couple of implications. First, in order to be effective, advertisements must be presented at the time that a script is being triggered. Second, in terms of ad copy, it is crucial to design ads which provide what Shank and Abelson (1977) call "interrupts" and act to make otherwise unconscious scripts conscious. By moving the consumers back to a conscious mode of thought, perhaps scripts can be altered. In order to do this, the consumer's scripts must be identified and new actions or scenes written into the scripts. A scripts framework can provide an information processing approach to routine response behavior.

REFERENCES

Abelson, Robert (1980), The Psychological Status of Scripts. Cognitive Science Technical Report #2, (Yale University).

Antil, John J., and Bennett, Peter D. (1979), "Construction and Validation of a Scale to Measure Socially Responsible Consumption Behavior," in The Conserver Society, Karl H. Henion II and Thomas C. Kinnear, eds., (Chicago, IL: American Marketing Association, 51-68).

Baggett, Patricia "Memory for Explicit and Implicit Information in Picture Stories," Journal of Verbal Learning and Verbal Behavior, 16, 538-548.

Bartlett, F. C. (1932), Remembering: An experimental and social study, (Cambridge: Cambridge University Press).

Bennett, Peter D. and Mandell, Robert (November 1969), "Prepurchsse Information Seeking Behavior of New Car Purchasers - The Learning Hypothesis," Journal of Marketing Research, 6, 430-433.

Bettman, James R. (1979), An Information Processing Theory of Consumer Choice, (Reading, MA: Addison-Wesley).

Bransford, John D., Barclay, Richard J., and Franks; Jeffrey J. (1972), "Sentence Memory: A Constructive Versus Interpretative Approach," Cognitive Psychology, 3, 193-209.

Bower, Gordon H. (1975), "Cognitive Psychology: An Introduction," in Handbook of Learning and Cognitive Processes, William R. Estes, ed., (Hillsdale, NJ- Erlbaum and Associates).

Bower, Gordon H., Black, John B., and Turner, Terrence J. (1979), "Scripts in Memory for Test," Cognitive Psychology, 11, 177-220.

Dommermuth, William P. (May 1965), "The Shopping Matrix and Marketing Strategy," Journal of Marketing Re- search, 2, 128-132.

Graesser, Arthur C., Gordon, Sallie E., and Sawyer, John D. (1979), "Recognition Memory for Typical and Atypical Actions in Scripted Activities: Tests of a Script Pointer and Tag Hypothesis," Journal of Verbal Learning and Verbal Behavior. 48, 319-332.

Graesser, Arthur C., Woll, Stanley B., Kowalski, Daniel J., and Smith, Donald A. (1980), "Memory for Typical and Atypical Actions in Scripted Activities," Journal of Experimental Psychology: Human Learning and Memory, 6, 5, 503-515.

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

Kruglanski, Arie (1975), "The Two Meanings of External Validity," Human Relations, 28, 653-659.

Lachman, Roy, Lachman, Janet L., and Butterfield, Earl C. (1979), Cognitive Psychology and Information Processing, (Hillsdale, NJ: Lawrence Erlbaum).

Lindsay, Peter H., and Norman, Donald E. (1977), Human Information Processing, (New York: Academic Press).

McClelland, Lou, and Cook, Stuart W. (January-February 1980), "Promoting Energy Conservation in Master-Metered Apartments through Group Financial Incentives," Journal of Applied Social Psychology, 10, 20-31.

McNeill, Dennis L., and Wilkie, William L. (June 1979), "Public Policy and Consumer Information: Impact of the new Energy Labels," Journal of Consumer Research, 6, 1-11.

Norman, Donald A., and Bobrow, Daniel F. (1975), "On the Role of Active Memory Processes in Perception and Cognition," in The Structure of Human Memory, Charles N. Cofer (ed.), (San Francisco: W. H. Freeman).

Sachs, J. (1967), "Recognition Memory for Syntactic and Semantic Aspects of Connected Discourse," Perception and Psychophysics, 2, 437-442.

Schank, Roger C. (1980), "Language and Memory," Cognitive Science, 4 243-284.

Schank, Roger C. and Abelson, R. P. (1977), Scripts. Plans, Goals and Understanding, (Hillsdale, NJ: Lawrence Erlbaum).

Winnett, Richard A., Kagel, John R., Battalis, Raymond C. and Winkler, Robin CA (February, 1978), "Effects of Monetary Rebates, Feedback and Information on Residential Electricity Conservation," Journal of Applied Psychology, 63, 73-80.

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