Responses to Commercials in Laboratory Versus Natural Settings: a Conceptual Framework

Meryl P. Gardner, New York University
S.P. Raj, Syracuse University
ABSTRACT - Laboratory experiments are frequently used to investigate consumer reactions to commercials because they permit a high level of internal validity. Their external validity, however, is frequently questioned. The goal of this study is to investigate factors which limit the generalizability of findings from laboratory investigations of consumer reactions to commercials. A framework is presented which can help evaluate the external validity of particular laboratory settings.
[ to cite ]:
Meryl P. Gardner and S.P. Raj (1983) ,"Responses to Commercials in Laboratory Versus Natural Settings: a Conceptual Framework", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 142-146.

Advances in Consumer Research Volume 10, 1983      Pages 142-146


Meryl P. Gardner, New York University

S.P. Raj, Syracuse University

[The authors thank Peter Webb and two ACR conference reviewers for their helpful comments on an earlier draft.]


Laboratory experiments are frequently used to investigate consumer reactions to commercials because they permit a high level of internal validity. Their external validity, however, is frequently questioned. The goal of this study is to investigate factors which limit the generalizability of findings from laboratory investigations of consumer reactions to commercials. A framework is presented which can help evaluate the external validity of particular laboratory settings.


What is really going through viewers' minds during a television commercial? The importance of consumer reactions and responses to commercials has long been recognized. They have been investigated with experimental and non-experimental methods in laboratory and field settings; Each method has its advantages and disadvantages. The selection of the best method and setting to use depends on the specific research questions addressed and the trade-offs the researcher is willing to make. The two common combinations of method and setting used to study consumer reactions to commercials are: 1) experimental method - laboratory setting, and 2) non-experimental method - field setting. The former is often employed due to the enhanced internal validity it offers.

The goal of this study is to investigate factors which limit the generalizability of findings obtained from laboratory studies of consumer reactions to commercials. To our knowledge, no previous research has directly considered this issue. It has been common to merely acknowledge that there are many real-world conditions which will modify the conclusions drawn from laboratory studies. Our concern here is how they will be modified; we seek to identify and understand potential threats to external validity .

In order to do so, we must first model consumers' responses to commercials. Our framework is limited to the effects of six independent variables: advertisement content, program content, product interest, situational distraction, clutter and repetition. These factors were selected because they have often been investigated individually and have been found to affect responses to commercials. Our model enables us to evaluate the external validity of a study by judging whether the experimental setting differs from real world settings of interest on factors likely to interact with the treatment variable. Before presenting the model, we will discuss some paradigms commonly used to investigate the effects of advertisements.


Previous research has provided information about consumers' thoughts under various laboratory conditions: the cognitive response approach, the audience reaction approach, and the "home-like" laboratory.

Cognitive Response Approach

The basic tenet of the cognitive response approach is that conscious thoughts mediate the effects of a communication. Theoretical models linking cognitive responses to elements of cognitive structure have been postulated (Lutz and Swasy 1977; Olson, Toy and Dover 1979) and some empirical support for this approach has been demonstrated (e.g. ,. Olson, Toy, and Dover 1978; Sternthal, Dholakia and Leavitt 1978; Patty and Cacioppo 1979).

Cognitive responses are typically investigated with the following experimental procedure: Subjects are provided with stimuli designed to elicit various kinds and/or quantities of cognitive responses and are asked to list everything that went through their minds while reading the ads. Their responses are then coded as counterarguments, support arguments, and source derogations and are analyzed as mediators of persuasion, attitude formation or attitude change (e.g., Wright 1973, 1980).

The focus of research using the cognitive response approach has been to investigate the impact of cognitive responses on attitudes under conditions of high internal validity. To achieve this goal, the laboratory manipulations and settings under which responses are elicited have varied widely in-external validity and in relevance to marketing. Typically, experimenters have manipulated one or two factors and have controlled all others. This approach has provided us with exciting insights into the independent variables' main effects, but has not investigated the complex interactions caused by their concurrently varying levels in the real world. Our model attempts to provide a framework for assessing which of the many feasible interactions are most likely to increase the external validity of a given study.

Audience Reaction Approach

The audience reaction paradigm has concentrated on the emotional component of consumers' reactions to ads. This non-experimental approach is typically investigated using measurement techniques developed to assess consumers' feelings about the advertisements themselves (Wells, Leavitt and McConville 1971; Schlinger 1979). Responses are collected by showing people scales or adjectives designed to inform advertisers of consumer perceptions. For example, an ad intended to be funny may be funny or silly or meaningless to a consumer depending on his sense of humor. Subjects' overall responses can also be coded to designate ads that received a primarily positive or negative response. In sum, the goal of audience reaction research is to provide the advertiser with an estimate of the emotional impact of his ad. This impact, however, is not due solely to the advertisement, but is a function of the idiosyncratic reactions thoughts produced in a particular situation, advertisement, product, consumer interaction. Our model postulates that the emotional impact or mood induced by a commercial will be affected by the surrounding program content and the consistency of that content over repeated exposures.

"Home-like" laboratories

In order to have high external validity without sacrificing the high internal validity of the laboratory setting, researchers may create a "home-like atmosphere" in the lab. For example, Ray and Webb (1978, p. 9) provide the following description of their experimental setting:

The room containing the TV monitor was set up to look like a living room. In addition to a sofa and several comfortable chairs, a coffee table containing current magazines was placed in the sitting area. Refreshments were available in the back of the room and in a kitchen adjacent to the living room...

(Subjects) were instructed to-act as they would watching TV at home - to talk with each other if they wished (every participant knew at least one other person in the room before the viewing session, and 20 percent of the groups consisted of actual families);-to get refreshments at any time; in short, to relax and enjoy themselves. Earlier studies in the program had demonstrated the effectiveness of this setting and these instructions in creating as natural a viewing situation as is possible outside the home.

The setting described is both impressive and expensive. It seems very realistic - to us and to the subjects (Ray and Webb 1978, p. 9). For example, it could permit situational distractions and clutter to interact with the content of commercials and television programs. Its most serious limitation involves investigations of repetition. This shortcoming is not specific to this setting, but applies to most laboratory studies of ad effects. Although repetition is spaced in the real-world, time pressures often cause it to be massed in the Laboratory (Sawyer 1974). Results of a field study by Zielske and Henry (1980) indicate that this difference may be critical. A given number of exposures spread over a 13-week period yielded a higher and earlier peak recall level than the same number of exposures spread over a 59-week period. The home-like setting described above raises the question: Is it always necessary to permit all independent variables to vary simultaneously in an elaborate, expensive setting? Greater efficiency may be obtained by holding constant variables which do not interact with the treatment variables in the real-world. To determine which variables meet this qualification, we must have a model of the underlying process.


The model specified in Figure 1 provides a framework for examining the effects of a commercial. We believe that the independent variables (rectangles) affect a set of conceptual variables (triangles) which, in turn, mediate the changes in awareness, knowledge, attitude and behavior induced by commercials. We will first examine how the conceptual variables moderate ad effects and then how they are affected by the independent variables.




Consumers' moods may affect their perceptions, interpretations, and evaluations of advertised brands. Information may be viewed more favorably by people in pleasant moods. Bower (1981) has hypothesized-that subjects will actively attend to material which is in tune with their current feelings, leading to selective attention. Isen et al. (1978) found that consumers in good moods tend to evaluate products they own more favorably than those in neutral moods. The researchers propose that people in good moods are more likely to retrieve positive than negative material from memory and that this affects their decision making. Bower (1981) has found that emotions powerfully influence snap judgments about others' personalities, e.g., angry subjects are prone to find fault with others. In addition, mood during exposure to a commercial may become associated with the advertised brand and affect brand attitudes through affect transfer.

Schema Activated

The schema activated during exposure to an advertisement may affect the content and organization of acquired information as well as the attitude formation process. [A memory schema is defined here as an organized set or information about a concept stored in long term memory (e.g., Rumelhart and Ortony 1977).] Activating an appropriate schema during exposure to a message facilitates recall (e.g., Bradsford and Johnson 1979, 1973; Dooling and Mullet 1973). Taylor and Crocker (1979) report that an activated schema may direct a search for data. When data along some dimension is missing, the schema may fill in the missing values with best guesses. We might postulate that a commercial which does not provide information along an attribute presents an analogous situation. If consumers activate their schemata for the product class, they may use the average value of that attribute for all brands of the product to fill in the missing information for the advertised brand.

In addition. schema activation is closely associated with attitudes and affect. Making careful, rational evaluations of advertised brands may involve full activation of extensive. information-laden schemata. For instance, to determine if a product costs too much relative to others. individuals must activate schemata containing knowledge about the prices of similar products (e.g., Gardner, Mitchell and Russo 1982). Generating emotional, gut-level affective reactions to advertised brands may involve the activation of schemata which include pictures, fantasies or self-projections.

Attention Level

Level of attention upon exposure to a commercial affects information acquisition and attitude formation. If attention is too low, comprehension of a complex message may be inhibited (Regan and Cheng 1973) or parts of the message may be missed. Consumers may fill in the missed information with schema-based best guesses. Petty, Wells and Brock (1976) have manipulated distraction to investigate processing with low levels of attention. They propose that distraction disrupts the dominant cognitive response to a persuasive message (whether support arguing or counterarguing) and so, can increase or reduce yielding.

The above discussion has focused on the effects of three conceptual variables: mood, schema activated and attention level. We now turn to the effects of the independent variables on those conceptual variables.

Factors Affecting Mood

Consumers' moods while viewing a commercial may largely reflect their general-emotional states and overall mental health. Nonetheless, both commercials and surrounding television programs are designed to affect viewer's moods. These factors may interact to affect mood upon exposure to a commercial in ways which modify ad effects. The impact of this interaction may be mediated by repetition.

Although there is little evidence that the mood created by a TV program can affect viewer response to an advertisement, Yuspeh (1979) found that advertising effect is influenced by the T.V. program in which it is shown. The study used on-air testing methods in a field experiment. Commercials were embedded in different television shows. Placing commercials in different shows results in significant differences in brand recall, playback, buying intention and brand perceptions. We postulate that some of these differences may be due to the effect of program content on viewers' moods. There is, however, a dearth of research on the strength and durability of moods created by television programs. Indirect support for our conjecture is provided by Axelrod's (1963) findings. In that study, he found that consumers' beliefs that a particular product would lead to a desired mood was significantly shifted after viewing a film which induced moods.

Moods induced by television programs may interact with those created by commercials. Some of them attempt to elicit strong feelings while others are informational and neutral in mood (Schlinger 1979). Mood advertisements try to evoke specific feelings in the viewer in an attempt to associate that feeling with the product. However, if the mood created by the advertisement is different from that created by the TV program, a clash or moods may occur. If the mood created by the program is stronger and is in direct contrast to the ad-created mood, the former may overwhelm the latter. In such cases, the advertiser's attempt to associate desired feelings with the brand might be in vain. In addition, selective attention is more likely to occur if moods contrast (Bower 1981) rather than in a situation where the advertisement is informational and neutral in mood.

If the program and ad induced different moods, but these moods are not in direct conflict, a blending of emotions may occur. Plutchik (1980) has developed the notion of opposing and similar primary emotions in an "emotion circle" and the blending of emotions, for example, anger and surprise leading to sadness. Further, Bower (1981) showed supporting evidence when he found that subjects recalled word lists best when the mood they are in when recalling the words is the same mood they originally were in when first learning the words. Worst recall occurred when their moods during learning and recall were in direct contrast. Intermediate recall occurred if the moods were adjacent.

If a commercial is aired repeatedly, the effects of mood may be enhanced or diminished. If the viewer sees the ad in contexts which differ in mood, the impact of mood may diminish. If an advertiser regularly sponsors a television series, different episodes may be similar in mood and so, the impact of mood may be accentuated

Factors Affecting Schema Activated

We postulate that program content, ad content, and the importance of the advertised product to the consumer interact to affect the schema activated upon exposure to a commercial. However, we note that this conjecture is difficult to falsify; it is hard to identify schemata activated by experimental manipulations and to explore their contents. (For a discussion of methodological issues in schema research, see Fiske and Linville 1980).

Program content may activate schemata containing compatible material. Heavy, intellectual shows may be more likely to fully activate data-laden schemata. Mindless, sensual shows may be more likely to fully activate fantasy or feeling-laden schemata.

Advertisement content may also activate schemata containing compatible material. Using print advertisements. Edell (1981) found that prose formats generated more verbal statements than pictorial formats for the same information. We note that this finding is compatible with, but only indirectly related to, the role of ad content in schema activation.

The importance of the advertised product to the consumer may affect schema activation. If consumers are interested in the advertised products, their schemata for the brand or product category may be activated. If consumers are not interested in the advertised product, they may direct their processing toward something else, such as enjoyment or censorship. Such individuals may not fully-activate their schemata for the brand or product category during processing. They may not make inferences about product attributes and may not form brand attitudes. (For a more complete discussion, see Gardner, Mitchell and Russo 1982).

Factors Affecting Attention Level

Attention level may be decreased by situational distractions and clutter and increased by consumer interest in advertised products and "catchy" advertisements. Repetition may mediate these effects.

Distraction and clutter divide consumers' attention between commercials and other stimuli. In lab settings, subjects usually fully attend to commercials, but distractions are high during at-home television viewing (Comstock 1975). In the context of consumer information processing, interrupt mechanism have been suggested by Bettman (1979). Ray and Webb (1978) have extensively and explicitly investigated the effects of clutter, but it has often been neglected in studies involving other independent variables. The enormous number of commercials during television programs is often a source of diminished attention and confusion to consumers. Their minds are cluttered with the messages preceding and following the target ad.

Interest in the advertised product and "catchy" commercials may elicit higher levels of attention. An interested consumer may devote full attention to the advertisement, especially those aspects of the ad that enable him to evaluate the advertised brand. Someone who is not interested may not pay full attention to the at, or he may pay attention to aspects of it that are not relevant to brand evaluation (Gardner 1981). Thus, an interested consumer may deliberately concentrate on a commercial in an attempt to overcome the effects of distractions and clutter. Ray and Webb (1978) found that "high involvement" commercials were less vulnerable to the effects of clutter than "low involvement" commercials. In addition, advertisements which are novel or shocking may be able to overcome the attention-attenuating effects of distraction and clutter

The effect of repetition is uncertain. On one hand, it might serve to lessen the consequences of low attention since the viewer has many opportunities to see the commercial. Over repeated exposures, the consumer may be more likely to fully absorb the information it contains despite environmental distractions and the clutter of surrounding messages. On the other hand, advertising wear-out and negative accompanying effects may prevent repetition from lessening the effects of low attention (e.g., Sawyer 1974). Some commercials are able to sustain viewer interest even after many repetitions due to their excellent creative advertising. Others, however, are soon subject to wear-out. In the first distracted viewing, the consumer has a token awareness of the commercial and because there is no newness to it in future repetitions, they are more likely to be distracted during the commercial thereby negating the effectiveness of repetition. Consumers also mentally tuned out commercials if they were repeated in a pattern during their favorite TV shows (Raw and Webb. 1978).


Studies investigating advertising affects typically involve the manipulation of one or two experimental variables and the control of other factors, usually by maintaining them at constant levels over all treatments. These experiments frequently achieve a high level of internal validity, but their external validity may often be enhanced by allowing selected additional factors to vary over realistic ranges. These factors should be those independent variables that interact with the treatment variables. Varying levels of independent variables that do not interact with the experimental variables should be omitted to maximize efficiency and minimize costs. The model presented explicitly discusses expected interactions and so, may aid in the selection of variables to enhance external validity.


Axelrod, Joel N. (1963), "Induced Moods and Attitudes Towards Products," Journal If Advertising Research, 3,2.

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

Bower, Gordon H. (1981), "Mood and Memory," American Psychologist, 36, 2.

Bradsford, J.D. and Johnson, M.K., (1972), "Contextual Prerequisites ,or Understanding: Some Investigations of Comprehension & Recall," Journal of Verbal Learning and Verbal Behavior, 11, 717-726.

Bradsford, J.D. and Johnson, M.K., (1973), "Considerations of Some Problems of Comprehension, " in W.T. Chase, ed., Visual Information Processing, New York : Academic Press .

Comstock, George (1975), "Television and Human Behavior: the Key Studies," Report no. R-1747-CF, the Rand Corporation, CA.

Dooling, D.J. and Mullet, R.L. (1973), "Locus of Thematic Effects in Retention of Prose," Journal of Experimental Psychology, 97, (3), 404-406.

Edell, A., (1981), "The Information Processing of Pictures in Print Advertisements," Unpublished doctoral dissertation, Carnegie-Mellon University.

Fiske, S.T. & Linville, P.W. (1980), "What Does the Term Schema Buy Us?," Personality and Social Psychology Bulletin, 6, 543-557.

Gardner, Meryl, Mitchell, Andrew and Russo, J. Edward (1982), "Strategy-Induced Low Involvement with Advertising", New York University Working Paper Series.

Isen, Alice M., Shalker, Thomas E., Clark, Margaret, and Karp, Lynn (1978), "Affect, Accessibility of Material in Memory, and Behavior: A Cognitive Loop?", Journal of Personality and Social Psychology, 36, l.

Lutz, Richard, and Swasy, John (1977), "Integrating Cognitive Structure and Cognitive Response Approaches to Measuring Communication Effects," in Advances in Consumer Research, Vol. 4, ed. William Perrault, Atlanta: Association for Consumer Research.

Olson, Jerry C., Toy, Daniel R., and Dover, Phillip A. (1979), "Do Cognitive Responses Mediate the Effects of Advertising Content on Cognitive Structure?" Working Paper, No. 82, College of Business Administration, The Pennsylvania State University, University Park, PA.

Olson, Jerry C., Toy, Daniel R., and Dover, Phillip A. (1978), "Mediating Effects of Cognitive Responses to Advertising on Cognitive Structure," in Advances in Consumer Research, Vol. 5, ed., K. Hunt, Ann Arbor, MI.. Association for Consumer Research.

Petty, Richard and Cacioppo, John (1979), "Issue Involvement Can Increase or Decrease Persuasion by Enhancing Message - Relevant Cognitive Responses," Journal of Personality and Social Psychology, Vol. 37, No. 10, 1915-1926.

Petty, Richard E., Gary L. Wells and Timothy C. Brock (1976), "Distraction Can Enhance or Reduce Yielding to Propaganda: Thought Disruption Versus Effort Justification," Journal of Personality and Social Psychology, 34, 5.

Plutchik, Robert (1980), "A Language for Emotions," Psychology Today, February 1980.

Michael, Ray and Webb, Peter (1978), "Advertising Effectiveness in a Crowded Television Environment," Report No. 78-113, Marketing Science Institute, Cambridge, MA.

Regan, Dennis T. and Cheng, John B. (1973), "Distraction and Attitude Change: .N Resolution," Journal of Experimental Social Psychology, 9.

Sawyer, Alan G. (1974), "The Effects of Repetition: Conclusions and Suggestions About Experimental Laboratory Research," in David G. Hughes and Michael L. Ray (Eds.), Consumer Information Processing, University of North Carolina Press, Chapel Hill, N.C.

Schlinger, Mary Jane (1979), "A Profile of Responses to Commercials," Journal of Advertising Research, Vol. 19, No. 2, pp. 37-46.

Sternthal, Brian, Dholakia, Ruby, and Leavitt, Clark (1978), "The Persuasive Effect of Source Credibility: Tests of Cognitive Response," Journal of Consumer Re- search, 4, 252-60.

Taylor S. and Crocker, J., (1981), Schematic Bases of Social Processing in E.T. Higgins C.P. Herman and M.P. Zanna, eds., Social Cognition: The Ontario Symposium, Vol. 1, Hillsdale, N.J.: Lawrence Erlbaum Association, 89-134.

Wells, William, Leavitt, Clark and McConville, Maureen (1971), "A Reaction Profile for TV Commercials," Journal of Advertising Research, Vol. 11, No. 6, pp. 11-17.

Wright, Peter (1973), "The Cognitive Processes Mediating Acceptance of Advertising," Journal of Marketing Re- search, 10, 53-67.

Wright, Peter (1980), "Message - Evoked Thoughts: Persuasion Research Using Thought Verbalizations," Journal of Consumer Research, 7, 151-175.

Yuspeh, Sonia (1979), "The Medium Versus the Message," in George B. Hafer (Ed.), A Look Back, A Look Ahead, Proceedings Series, 10th National Attitude Research Conference, American Marketing Association, Chicago.

Zielske, Hubert A., and Henry, Walter A. (1980), "Remembering and Forgetting Television Ads," Journal of Advertising Research, 90, 9.