Measuring Script Development: an Evaluation of Alternative Approaches

ABSTRACT - A common limitation in script research, both in the consumer context and in other settings, is the failure to quantify subjects' level of script development. The majority of studies simply assume a homogeneous script across all subjects for an event selected at the researcher's discretion. This assumption clearly hampers developmental studies of scripts, and also restricts the ability to make unambiguous statements about the effects of scripts on consumer information processing. This paper discusses various approaches to the measurement of script development that will reveal the degree to which the structural features of such representations are present in consumer memory.

Citation:

Ruth Ann Smith and Michael J. Houston (1986) ,"Measuring Script Development: an Evaluation of Alternative Approaches", in NA - Advances in Consumer Research Volume 13, eds. Richard J. Lutz, Provo, UT : Association for Consumer Research, Pages: 504-508.

Advances in Consumer Research Volume 13, 1986      Pages 504-508

MEASURING SCRIPT DEVELOPMENT: AN EVALUATION OF ALTERNATIVE APPROACHES

Ruth Ann Smith, Virginia Polytechnic Institute and State University

Michael J. Houston, University of Wisconsin - Madison

ABSTRACT -

A common limitation in script research, both in the consumer context and in other settings, is the failure to quantify subjects' level of script development. The majority of studies simply assume a homogeneous script across all subjects for an event selected at the researcher's discretion. This assumption clearly hampers developmental studies of scripts, and also restricts the ability to make unambiguous statements about the effects of scripts on consumer information processing. This paper discusses various approaches to the measurement of script development that will reveal the degree to which the structural features of such representations are present in consumer memory.

INTRODUCTION

The impact of knowledge on product evaluations and purchase decisions is a growing concern in the consumer behavior literature (Brucks 1985; Johnson and Russo 1984; Sujan 1985). The need to measure consumer knowledge structures, and assess their influence on information processing, is acknowledged by Mitchell (1982), who suggests that "in order to develop valid measures, you must first have a theory of memory" (p. 45). The objective of this paper is to describe and evaluate several techniques for the measurement of consumers' knowledge structures which are derived from schema theories of memory.

The first part of the paper will briefly review the basic propositions of such theories, and describe the structural characteristics of schematic memory, with particular emphasis on scripts, or event schemas. Given this theoretical framework, attention will then focus on alternative stimuli and dependent measures that may be employed to quantify an individual's level of script development, as indicated by the presence or absence of these structural features.

SCHEMA THEORY

Although the organization of human memory may assume various forms, memory schemas appear to perform an important role in the structure of consumer knowledge (John and Whitney 1982; Leigh and Rethans 1983; Roedder John and Whitney 1985; Whitney and John 1983). Event schemas, or scripts, consist of abstract knowledge about the sequence of actions typical of a commonplace event (Abelson 1976, 1980; Schank and Abelson 1977). Scripts, like all schemas, are propositional, verbally or semantically encoded knowledge structures (Lord 1980). However, they possess several structural characteristics that differentiate them from other memory representations, including concept abstraction, a set quality, causal or temporal sequence, and hierarchical structure (Abbott and Black 1980; Thorndyke and Yekovich 1980).

Concept abstraction refers to the nature of the knowledge represented in a script. An event schema does now contain the specific details of any particular occurrence of an event. Rather, it consists of a prototypical event composed of the common actions abstracted from repeated experiences. Thus, the structure of a script is a network of generic actions and rules for combining them. In addition, this network possesses a set quality in that activation of any component action provides access to the entire group of actions constituting an event. The temporal or causal sequence typical of a script's organization refers to the characteristic that certain actions in this set must precede other actions in time in order to provide the enabling conditions necessary for the latter action to occur.

Finally, scripts possess a hierarchical structure which defines the relationships among actions of varying levels of specificity or centrality to an event. As described by Abbott and Black (1980), this hierarchy is composed of three levels including a script header, scene headers, and scene actions. A script header represents the most general, or macro level of the hierarchy, and summarizes an entire event (e.g. dining in a restaurant). This event is composed of several scenes which are identified by scene headers such as entering the restaurant, ordering the meal, and eating. The scene headers exist at an intermediate level of the hierarchy, and contain more detailed information about the event. At the lowest level of the hierarchy are scene actions, which are the individual actions relevant to each scene. Thus the ordering scene may consist of reading the menu, discussing the choices, and placing the order.

Given these structural features of event schemas, the task of quantifying the level of script development may be clarified. As an individual's script for an event becomes better articulated, it would be expected chat their response to some stimulus would exhibit a higher degree of concept abstraction, the set quality, remporal/causal sequence, and hierarchical structure. Thus the measurement task involves devising stimuli that will elicit responses that can be quantified using various dependent measures to assess the presence or absence of these structural features. It should be emphasized that measuring script development should focus on assessing the structure of knowledge, rather than its processing effects. The influence of scripts on consumer information processing can only be demonstrated once it has been determined chat consumers have actually organized their knowledge of an event in a schematic structure.

MEASURING SCRIPT DEVELOPMENT

The script literature is replace with a variety of measures that are used to elicit event schemas or to evaluate their impact on information processing. Only rarely, however, have these measures been employed to assess respondents' level of script development, although many are adaptable for this purpose. In the following discussion, verbal and visual stimuli appropriate to this task will first be considered, and various dependent measures useful in assessing responses to these stimuli will then be discussed.

Stimulus Materials

Verbal Stimuli. Inasmuch as scripts are composed of semantically or verbally encoded propositions (Lord 1980), verbal stimuli are naturally suited to measuring script development. A wide variety of measures employing this approach are found in the script literature, although by far the most common verbal stimulus utilizes a probe to elicit a written or spoken description of an event. Typically referred to as protocols or retrospective self-reports, stimuli employing such free-recall tasks have served several purposes in script investigations, including identifying the component actions of an event (Bower, Black, and Turner 1979; Bozinoff 1982; Fivush 1981, Graesser, Gordon, and Sawyer 1979; John and Whitney 1982; Rethans and Hastak 1982; Whitney and John 1983), evaluating the influence of a script on information processing (Black and Bern 1981; Bower, Black, and Turner 1979; Brewer and Dupree 1983; Graesser, Woll, Kowalski, and Smith 1980; Lichtenstein and Brewer 1980: Martin, Harrod, and Siehl 1980; Salmaso, Baroni, Job, and Peron 1983; Whitney and John 1983), and assessing level of script development (Roedder John and Whitney 1985; Smith and Houston 1985).

Although only rarely used in this final context, it seems reasonable to expect that the responses of subjects with various levels of schematic memory for an event would vary systematically with respect to all four of the structural properties noted previously. That is, an individual with a highly developed script would be more likely to generate an abstract description of a complete set of actions, ordered in their typical temporal sequence and reflecting the hierarchical structure of component scenes than a respondent lacking a script. A potential limitation on this approach, however, is that generating a response to the stimulus imposes a substantial cognitive burden, and one which would seem to be inconsistent with the automatic, mindless nature of script-based processing. Some preliminary evidence that responses may not adequately reflect script development is offered by Smith and Houston (1985) who observed substantial instability over time in the responses of individuals with well-developed scripts. The lack of test-retest reliability suggests that this type of verbal stimulus should be used cautiously when assessing script development.

Verbal stimuli requiring cued recall may be designed, however, that minimize this problem of cognitive load. For example, subjects might be presented with a single action in an event, and be asked to recall the action that precedes or follows it. Such responses would obviously be informative with respect to the temporal sequence quality of scripts, but might indicate the set quality as well. That is, as the set of actions stored in memory was less complete, one would expect larger "gaps" between the stimulus action and the action given in response. Moreover, a respondent's use of abstract versus episodic language would offer a clue to the degree of concept abstraction. Finally, findings by Abbott and Black (1980) imply that the degree of hierarchical structure would also be evident in responses to such stimuli. These authors found that subjects with scripts exhibited an asymmetrical recall pattern, such that the probability of recalling a scene action given a scene header was less than the probability of recalling a scene header given a scene action. This systematic pattern would not be expected however, among individuals whose knowledge of an event was not hierarchically structured.

A third type of verbal stimulus chat also minimizes the problem of excessive cognitive load consists of a list of randomly ordered actions, only some of which are relevant to the event. The response to such a stimulus involves a recognition task in which subjects select the actions they feel are relevant to the event, and number those chosen to indicate the normal order of occurrence. While a number of studies have employed verbal stimuli which require discrimination, rather than reconstruction of an event, to assess the processing effects of scripts (Abbott and Black 1980; Bower, Black, and Turner 1979; Brewer and Dupree 1983; Graesser, Woll, Kowalski, and Smith 1980; Lichtenstein and Brewer 1980; Salmaso, Baroni, Job, and Peron 1983), their use in evaluating script development is less common. However, Smith and Houston (1985) found that this approach was effective in determining the presence of the set quality and temporal sequence characteristics of scripts. Moreover, the measure employed in their study exhibited reliability superior to that of a retrospective self-report measure. This type of verbal stimulus offers a promising approach to measuring script development, and one whose potential should be further explored.

A fourth type of verbal stimulus chat has been employed to assess script development consists of a multi-item scale of experience/knowledge about an event. A variant of this procedure is a single-item scale consisting of a self-rating of familiarity. A higher score (as indicated by the scale value of a subject's responses) is presumed to indicate a more highly developed script. Although such measures have been used in a number of studies (Alba 1983; Conover 1982; Schurr and Calder in press; Smith and Houston 1985; Smith, Houston, and Childers 1984, in press), they suffer from a major conceptual flaw in that responses indicate only the level of knowledge, but not the structure of that knowledge. Therefore, a response to this type of stimulus may be only indirectly related to the level of script development. Close correspondence has been observed, however, between questionnaire scores and those obtained on the dependent measures used to quantify responses to order verbal stimuli (Smith and Houston 1985; Smith, Houston, and Childers 1984, in press). Therefore, self-ratings and questionnaires may be useful in validating other measures of script development.

Visual Stimuli. In spite of the strongly propositional and semantic character of scripts, prior research has demonstrated a strong link between script development and the ability to generate visual images of an event (Smith, Houston, and Childers 1984; in press). Therefore, visual stimulus materials are also appropriate in assessing script development, and offer certain advantages over verbal stimuli. In particular, they avoid the vargaries of verbal stimuli, a feature that is particularly desirable in research settings where certain age groups (i.e. children or the elderly) may experience difficulties encoding verbal presentations.

While verbal stimuli dominate script research, a few investigations have utilized visual approaches to ascertain the component actions of a script and assess its effects on information processing (Brewer and Dupree 1983; Goodman 1980; Lichtenstein and Brewer 1980) as well as evaluate script development (Fivush 1981). In the latter context, Fivush (1981) used a visual stimulus consisting of a set of pictures depicting the component actions of an event in a study of the development of children's scripts. The stimuli contained pictures of actions that were both relevant and irrelevant to the event. The children responded by selecting the relevant pictures, arranging chem in the proper sequential order, and then telling a story based on the selected pictures.

Although such responses would reasonably be expected to reflect the presence of all the structural characteristics of scripts, Fivush concluded that they may be misleading with respect to script development. Compared to the responses of children who were simply asked to describe an event, the children performing the picture sorting task were more prone to include irrelevant actions and to improperly order the chosen action pictures. It appears that children relied on existing scripts to guide their spontaneous construction of a verbal description of an event, but not their selection of pictures. While tentatively suggesting the inappropriateness of picture sorting for measuring script development, the generalizability of these results to adult subjects remains an empirical issue.

Studies of the processing effects of scripts in which visual stimuli have been used suggest that videotaped presentations offer a more promising approach to assessing the level of development of consumers' scripts. Lichtenstein and Brewer (1980) reported that videotapes of goal-directed events elicited patterns of memory effects similar to those observed when verbal descriptions served as the stimuli. Specifically, recall for goal-directed actions was generally superior to that for non-goal-directed actions, a pattern that was also observed in subsequent studies by Brewer and Dupree (1983).

Although videotapes have never been used specifically to evaluate script development, these findings suggest that they may offer an attractive alternative to consumer researchers. The utility-seeking nature of consumer events implies that they are goal-directed. Videotaped presentations afford an opportunity to represent such action sequences in a more comprehensive and externally valid manner than verbal descriptions. Moreover, if subjects are asked to respond to such presentations by describing what they observed, their responses may reflect the structural properties of scripts in much the same way that responses to verbal probes would, but without the attendant task-related effects.

Dependent Measures

Regardless of whether verbal or visual stimuli are employed to evaluate script development, a variety of dependent measures may be used to analyze responses and to determine the degree to which the structural characteristics of scripts are present. Since any single measure is unlikely to reflect all the structural properties of a script, multiple indicators will often be desirable.

For those stimuli that elicit responses indicative of concept abstraction, two types of dependent measures are available. The first is a content analysis system developed by Martin, Harrod, and Siehl (1980), in which various language characteristics are analyzed. The system permits a subject's description of an event to be categorized as representing an episodic (most concrete), categorical, or hypothetical (most abstract) script (Abelson 1976). A simpler adaptation of this measure was developed by Roedder John and Whitney (1985), in which coders were given examples of the language typical of these three developmental stages, and assigned an overall rating of script development based on their reading of the response as a whole.

A second indicator of concept abstraction is the relative number of central and incidental actions in a response. As the event representation becomes more highly developed, responses will reflect less frequent mention of actions incidental to the main theme, relative to the occurrence of those actions central to the continuation of the event. Moreover, the central actions will be described abstractly, rather than episodically, as script development increases. Roedder John and Whitney (1985) used both these measures in a study of the development of children's scripts, and as expected found that older and more experienced children described a consumer event more abstractly and reported fewer incidental actions.

To assess the degree to which responses reflect the set quality of scripts, a different approach is necessary. Since activation of any pare of a script permits access to the entire action set that comprises the event, it would be expected that the responses of subjects with well-developed scripts would be more complete than those of individuals without schematic representations. Therefore, the number of central actions (independent of the occurrence of incidental actions) will be indicative of this aspect of a script's structure. Moreover, the order in which these central actions appear in a response offers a clue about the temporal sequence characteristic. That is, if a script were used in retrieving these actions from memory, the order of retrieval should correspond to the generic sequence typical of the event.

Evidence concerning both these points is provided in an investigation by Smith and Houston (1985). A measure was constructed that evaluated both the completeness and sequence of actions in responses to two types of verbal stimuli (a free recall protocol and a recognition task). Variations of this measure, which incorporated corrections for omitted actions and for false alarms (in the recognition task), were also constructed. All versions of the measure successfully differentiated between individuals with varying levels of script development.

Clustering indices represent another type of dependent measure adaptable to assessing script developments and may be used to indicate the degree of hierarchical structure in a memory representation. Such measures are only useful for responses that involve free recall, however. A clustering index is designed to "measure the degree to which a series of responses provided by a subject conforms to a hypothesized structure within the see consisting of all possible responses" (Hubert and Levin 1976).

Clustering indices have usually been employed to assess categorical, rather than schematic memory. However, if the action set comprising a script is viewed as the total set of responses, the actions comprising each scene in the event constitute its hypothesized structure (Abbott and Black 1980). Therefore, the degree to which actions related to each scene cluster together during recall will be indicative of the presence of this structural quality.

A variety of clustering indices are available, and their strengths and weaknesses have been reviewed elsewhere (Frankel and Cole 1971, Roenker, Thompson, and Brown (1971). Further consideration of the individual indices is beyond the scope of this paper, but it should be noted that one measure, the adjusted ratio of clustering index (ARC) was used by Srull (1983) to evaluate the structure of consumers' brand knowledge. While there is no empirical evidence of its ability to differentiate levels of schematic knowledge about consumer events, it appears to overcome the shortcomings of certain other indices, and bears further consideration in this context.

A final dependent measure suitable for use with either verbal or visual stimuli is response time. In a few investigations, chronometric devices have been used to assess the processing effects of scripts (Bower, Black, and Turner 1979; Galambos and Rips 1982; Nottenburg and Shoben 1980). These findings indicate that response latencies may also be informative with respect to the temporal sequence and hierarchical structure of event schemas.

Bower, Black, and Turner (1979) found that subjects with scripts more quickly recognized the proper order of action pairs that were closer together in the script than those further apart. They also found quicker comprehension of actions later rather than earlier in the event, perhaps because of their temporal nearness to the goal-outcome of the script. Among individuals lacking a schematic representation of an event, no such systematic relationship between response latencies and temporal position of actions would be expected.

Other investigations have reported that reaction times may be related to the hierarchical structure of a script. Nottenburg and Shoben (1980) observed lower response times to recognize typical or important actions among subjects with scripts. Such actions would occupy higher positions within a script's hierarchy, while less central actions, for which response time is greater, would be located at lower levels. Responses latencies for individuals lacking a script should be independent of the hierarchical position of component actions.

It should be noted, however, that teasing out the separate effects of hierarchical structure and temporal sequence on response time may be problematic since these characteristics may nor always be independent. Galambos and Rips (1982) reported that the centrality of actions more strongly influenced response latencies than did temporal position. Thus, in using chronometric measures to assess the presence of these script qualities, it is important to ensure their independence, in so far as possible. In the consumer context, however, centrality may be the more important consideration in evaluating script development since central actions are likely to be associated with greater utility.

SUMMARY

This paper has described various stimulus materials and dependent measures available to assess the level of script development for consumer events. All of the techniques evaluated have the potential to provide insight as to the presence or absence of various structural characteristics of event schemas. Since it is nor possible to demonstrate the effects of scripts on consumer information processing without first ascertaining the degree to which knowledge has assumed these structural features, the use of these measures, either individually or in combination, is a vital component of research on consumers' scripts.

REFERENCES

Abbott, Valerie and John B. Black (1980), "The Representation of Scripts in Memory," Cognitive Science Technical Report #5. New Haven, CT: Yale University.

Abelson, Robert P. (1976), "Script Processing in Attitude Formation and Decision Making," in J.S. Carroll and J.W. Payne (eds.), Cognition and Social Behavior, Hillsdale, N.J.: Lawrence Erlbaum Associates. 33-45.

Abelson, Robert P. (1980), "The Psychological Status of the Script Concept," Cognitive Science Technical Report #2. New Haven, CT: Yale University.

Alba, Joseph W. (1983), "The Effects of Product Knowledge on the Comprehension, Retention, and Evaluation of Product Information," in R.P. Bagozzi and A.M. Tybout (eds.), Advances in Consumer Research, Vol. 10, Ann Arbor: Association for Consumer Research. 577-580

Black, John B. and Hyman Bern (1981), "Causal Coherence and Memory for Events in Narratives," Journal of Verbal Learning and Verbal Behavior, 20 (3), 267-275.

Bower, Gordon H., John B. Black, and Terrence J. Turner (1979), "Scripts in Memory for Text," Cognitive PsychologY>" 11 (2), 177-220.

Bozinoff, Lorne (1982), "A Script Theoretic Approach to Information Processing: An Energy Conservation Application," in A. Mitchell (ed.), Advances in Consumer Research, Vol. 9, Ann Arbor: Association for Consumer Research, 481-86.

Brewer, William F. and David A. Dupree (1983), "Use of Plan Schemata in the Recall and Recognition of Goal-Directed Actions," Journal of Experimental Psychology: Learning, Memory and Cognition, 9 (1), 117-129.

Brucks, Merrie (1985), "The Effects of Product Class Knowledge on Information Search Behavior," Journal of Consumer Research, 12 (1), 1-16.

Conover, Jerry N. (1982), "Familiarity and the Structure of Knowledge," in A. Mitchell (ed.), Advances in Consumer Research, Vol. 9, Ann Arbor: Association for Consumer Research, 494-98.

Fivush, Robyn (1981), "Young Children's Knowledge About Familiar Events," Paper presented at the meeting of the American Psychological Association, Los Angeles, CA.

Frankel, Frederick and Michael Cole (1971), "Measures of Category Clustering in Free Recall," Psychological Bulletin, 76 (1), 39-44.

Galambos, James A. and Lance J. Rips (1982), "Memory for Routines," Journal of Verbal Learning and Verbal Behavior, 21 (3), 260-281.

Goodwin, Gail S. (1980), "Picture Memory: How the Action Schema Affects Retention," Cognitive Psychology, 12 (4), 473-495.

Graesser, Arthur, Sallie E. Gordon, and John D. Sawyer (1979), "Recognition Memory for Typical and Atypical Actions in Scripted Activities: Tests of a Script Pointer + Tag Hypothesis," Journal of Verbal Learning and Verbal Behavior, 18 (June). 319-332.

Graesser, Arthur, Stanley B. Woll, David J. Kowalski, and Donald A. Smith (1980), "Memory for Typical and Atypical Actions in Scripted Activities," Journal of Experimental Psychology: Human Learning 2-d Memory, 6 (5), 503-515.

Hubert, Lawrence J. and Joel R. Levin (1976), "A General Statistical Framework for Assessing Categorical Clustering in Free Recall," Psychological Bulletin, 83 (6), 1072-1080.

John, George and John C. Whitney, Jr. (1982), "An Empirical Investigation of the Serial Structure of Scripts," in B.J. Walker, W.O. Bearden, W.R. Darden, P.E. Murphy, J.R. Nevin, J.C. Olson and B.A. Weitz (eds.), An Assessment of Marketing Thought and Practice, Chicago: American Marketing Association, 75-79.

Johnson, Eric J. and J. Edward Russo (1984), "Product: Familiarity and Learning New Information," Journal of Consumer Research, 11 (1), 542-563.

Leigh, Thomas W. and Arno J. Rethans (1983), "Experiences With Script Elicitation Within Consumer Decision Making Contexts," in R.P. Bagozzi and A.M. Tybout (eds.,) Advances in Consumer Research, Vol. 10, Ann Arbor: Association for Consumer Research, 667-672.

Lichtenstein, Edward H. and William F. Brewer (1980), "Memory for Goal-Directed Events," Cognitive Psychology, 12 (3), 412-445.

Lord, Charles G. (1980), "Schemas and Images as Memory Aids: Two Modes of Processing Social Information," Journal of Personality and Social Psychology, 38 (2), 257-269.

Martin, Joanne, Wendy Harrod, and Caren Siehl (1980), "The Development of Knowledge Structures," Research paper No. 557, Graduate School of Business, Stanford University.

Mitchell, Andrew A. (1982), "Models of Memory: Implications for Measuring Knowledge Structures," in A. Mitchell (ed.), Advances in Consumer Research, Vol. 9, Ann Arbor: Association for Consumer Research, 45-51.

Nottenburg, Gail and Edward J. Shoben (1980), "Scripts as Linear Orders," Journal of Experimental Social Psychology, 16 (4), 329-347.

Rethans, Arno J. and Manoj Hastak (1982), "Representation of Product Hazards in Consumer Memory," in A. MitcheLl (ed.), Advances in Consumer Research, Vol. 9, Ann Arbor: Association for Consumer Research. 487-493.

Roedder John, Deborah and John C. Whitney, Jr. (1985), "The Development of Consumer Knowledge in Children: A Cognitive Structure Approach," Working paper, Graduate School of Business, University of Wisconsin, Madison, WI.

Roenker, David L., Charles P. Thompson, and Sam C. Brown (1971), "Comparison of Measures of the Estimation of Clustering in Free Recall," Psychological Bulletin, 76 (1), 45-48.

Salmaso, Paola, Maria Rosa Baroni, Remo Job, and Erminielda Mainardi Peron (1983), "Schematic Information, Attention, and Memory for Places," Journal of Experimental Psychology: Learning, Memory and Cognition, 9 (2), 263-268.

Schank, Roger C. and Robert C. Abelson (1977), Scripts, Plans, Goals, and Understanding, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Schurr, Paul H. and Bobby J. Calder (in press), "Psychological Effects of Restaurant Meetings on Industrial Buyers," Journal of Marketing.

Smith, Ruth Ann and Michael J. Houston (1985), "A Psychometric Assessment of Measures of Scripts in Consumer Memory," Journal of Consumer Research, 12 (2).

Smith, Ruth Ann, Michael J. Houston, and Terry L. Childers (in press), "The Effects of Schematic Memory on Imaginal Information Processing: An Empirical Assessment," Psychology and Marketing.

Smith, Ruth Ann, Michael J. Houston, and Terry L. Childers (1984), "Verbal Versus Visual Processing Modes: An Empirical Test of the Cyclical Processing Hypothesis," in T. Kinnear (ed.) Advances in Consumer Research, Vol. 11, Provo, UT: Association for Consumer Research, 75-80.

Srull, Thomas K. (1983), "The Role of Prior Knowledge in the Acquisition, Retention, and Use of New Information," in R.P. Bagozzi and A.M. Tybout (eds.), Advances in Consumer Research, Vol. 10, Ann Arbor: Association for Consumer Research, 572-76.

Sujan, Mita (1985), "Consumer Knowledge: Effects on Evaluation Strategies Mediating Consumer Judgments," Journal Of Consumer Research, 12 (1), 31-46.

Thorndyke, Perry W. and Frank R. Yekovich (1980), "A Critique of Schema-Based Theories of Human Story Memory," Poetics, 9 (1), 23-49.

Whitney, John C., Jr. and George John (1983), "An Experimental Investigation of Intrusion Errors in Memory for Script Narratives," in R.P. Bagozzi and A.M. Tybout (eds.), Advances in Consumer Research, Vol. 10, Ann Arbor: Association for Consumer Research, 661-666.

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