Extending Imagery Research to Sounds: Is a Sound Also Worth a Thousand Words?

Citation:

April Atwood (1989) ,"Extending Imagery Research to Sounds: Is a Sound Also Worth a Thousand Words?", in NA - Advances in Consumer Research Volume 16, eds. Thomas K. Srull, Provo, UT : Association for Consumer Research, Pages: 587-594.

Advances in Consumer Research Volume 16, 1989      Pages 587-594

EXTENDING IMAGERY RESEARCH TO SOUNDS: IS A SOUND ALSO WORTH A THOUSAND WORDS?

April Atwood, University of Washington

ABSTRACT -

The imagery process has been the focus of considerable attention in the consumer behavior and psychology literatures in recent years. An issue that -has not been tested is how nonvisual imagery cues, and sounds in particular, affect the information processing of consumers. This paper reviews the imagery literature and the relevant literature on attention theories, and draws implications for the use of audio imagery cues. Research questions arising from consideration of the theoretical areas are suggested, and methodological concerns are raised.

INTRODUCTION

The topic of imagery is a research area that has enjoyed widespread attention in recent years in the consumer behavior literature and conferences. Most imagery studies to date have examined "visual imagery" or the effect of visual cues on imagery activity. A nearly completely overlooked issue is the possibility that other types of sensory cues may affect imagery activity. Audio imagery cues (sounds) would seem of particular interest to researchers in marketing and consumer behavior since they are easily controllable in test settings and they have obvious marketing and advertising applications.

The intent of this paper is to provide a selective review of imagery theories and research and to consider another theoretical area (attention) that has implications for the use of audio imagery cues. Research questions that emerge from consideration of the imagery and attention theory areas will be discussed, along with research methodology considerations.

IMAGERY

Imagery Defined

Imagery, or mental imagery, has been defined in various ways over the years. Bugelski (1971) reviewed the imagery literature and found that although images are commonly thought of as "pictures in the mind," different researchers have defined imagery in different ways: An early consideration of the meaning of imagery was Titchener's (1921) conceptualization of imagery as consisting of "impressions made upon a plastic substance," which treated imagery as the mental remnant of some perceptual process. More fully developed descriptions are currently available. Richardson (1969) defined imagery as follows: "Mental imagery refers to (1) all those quasi-sensory or quasi-perceptual experiences of which (2) we are self-consciously aware, and which (3) exist for us in the absence of those stimulus conditions that are known to produce their genuine sensory or perceptual counterparts, and which (4) may be expected to have different consequences from their sensory or perceptual counterparts." While definitions vary somewhat, an effort to combine their implications yields a conceptualization of mental images as mental experiences that resemble more or less closely an actual perceptual occurrence and that may take place even when the actual perceptual occurrence (the perception of sensory information) is not currently happening. When a person actively imagines sights, sounds. or other perceptual events, s/he may be said to be experiencing the phenomenon of mental imagery. It is necessary to distinguish between imagery activity and "imagery cue," which refers to the external image, that is, the picture, word, sound, etc, that acts as a cue to elicit the internal imagery process.

To date, the discussion of imagery has usually been limited to visual imagery and visual cues, although imagery is thought by some to be possible in any of the senses; Holbrook (1983) proposes that nonverbal imagery may consist of not only visual components, but gustatory, olfactory, tactile, and auditory components as well. Paivio (1971a) also allowed for types of imagery other than visual, cautioning researchers to be certain to specify them as such. Despite the acknowledgment of imagery in other senses, even the most recent reviews (e.g., MacInnis and Price, 1987) are notably lacking in their coverage of any imagery executions except visual. The vast majority of the work on imagery has dealt with the process thought to occur in response to external visual cues or instructions to create a visual image. The lone exception was work (Segal and Fusella, 1970; described in detail later) that examined interference effects when subjects were involved in visual versus sound imaging and given signal detection tasks.

An extensive review of the psychology literature and the consumer behavior literature revealed almost no studies that have examined responses to auditory imagery cues -- sounds. The possibility that sounds may act as external imagery cues to encourage the generation of images in listeners presents interesting research questions. It would be an extension of current knowledge to compare what we know about pictures as imagery cues to the analogous use of sounds as imagery cues in the virtually untested auditory mode. Findings that sounds are able to evoke images in listeners as pictures can in viewers would provide a valuable theoretical extension of imagery theories.

Experimental Approaches to the Study of Imagery

Several experimental approaches have been used in the examination of the imagery process. Three of the most commonly used techniques, and an overview of their typical findings, are presented in this section.

One general approach to studying imagery has been to examine the differential effects of high-imagery versus low-imagery words on learning and memory tasks. Generally, concrete words are considered to be examples of words that are high in imagery potential, while low-imagery words are abstract words. Ratings performed on over 900 nouns indicated that the imagery dimension and the abstract-concrete dimension of words, while not perfectly related, are highly correlated (Paivio, Yuille, and Madigan, 1968). Many studies have demonstrated the clear superiority of concrete over abstract words on a variety of measured outcomes: free recall of words presented in lists during the experimental session, recognition memory of words from experimental lists, and paired-associate learning of both nouns and adjectives (e.g., Gorman, 1961; Paivio, 1969; Paivio, 1971a). In addition, response times have been used to indicate that the ease with which an image may be attached to a stimulus word or sentence depends on the concreteness of that stimulus (Paivio, 1971a). Specifically, it was found that subjects were able to generate an image to concrete words and sentences significantly more quickly than to abstract words and sentences.

A second approach to manipulating imagery and testing outcomes has been to use pictures as imagery cues. The experimental outcomes of numerous comparisons between pictures and words resemble those for high- versus low-imagery words. Presentations of object pictures have been shown to result in better recognition and recall of the objects than presentations of object names (e.g., Jenkins, Neale, and Deno, 1967; Paivio, 1983; Paivio and Csapo, 1973).

A third approach to manipulating imagery has been through explicitly instructing subjects to generate images. Subjects report being able to do this when instructed, and experiments comparing imaginal processing to rote repetition of to-be-remembered items yield results similar to the outcomes of the other two approaches. Paired associate learning, recognition, and recall measures of memory have all been demonstrated to be higher when subjects have been instructed to engage in imagery than when they rotely repeat the stimulus information (Bower, 1970; Paivio, 1971a).

Theories of Imagery

There is no widespread agreement on the imagery process that has been studied using these approaches, but several models have been proposed to explain imagery effects. Paivio's dual code theory (1971b) has been widely referenced. In it, he asserts that there exist independent but interconnected imagery and verbal systems to handle different presentations of information. The verbal system deals with discrete linguistic units and structures, while the imagery system is thought to be specialized for encoding, storing, organizing, transforming, and retrieving nonverbal information concerning "concrete objects and events." These two systems are thought to be interconnected, so that with the appropriate instructions or cues, nonverbal information can be transformed into verbal information and vice vasa. The two systems are also considered to be independent, each activated separately by the stimuli most relevant to it. The imagery system represents our knowledge of the world in a form that is highly isomorphic with perceptual information. The presentation of pictures, or other nonverbal information, is thought to directly activate the imagery system, and result in the formation or retrieval of the appropriate perceptual code. This perceptual code is thought to consist of only perceptual information; the imagery system is activated and acts independently of verbal or semantic information. The presentation of verbal information triggers the verbal system.

Paivio cites as evidence for his theory the results of the research on the processing of concrete versus abstract words and the superiority of pictures over words in learning tasks. Additional support for the dual coding theory is found in the response time latency experiments that pitted object names against object pictures (Paivio, 1975). Paivio found that when subjects were presented with the names of two objects and asked to indicate which object was larger, reaction times' were fastest when the objects differed greatly in size. He also found that differences in the size of the words had no effect on size judgments, but differences in picture size did: when the size relationship of the pictures was incongruent with the objects' true size relations (for example, a picture of a rabbit was larger than a picture of an elephant), reaction time was slowed. The reason provided by the theory illustrates the essence of the dual coding theory. The meaning of words was thought to come from the processing of their semantic components, so there is no reason to expect letter size differences to affect judgments based on meaning. However, the pictures served to access stored image codes directly, and incongruencies in picture size and true object size had to be resolved before the judgments could be made. The result was that judgments were made less quickly in the incongruent picture condition.

An alternative to the dual coding model of Paivio is the "relational associations" theory proposed by Bower (1972b). Bower had thoroughly reviewed the paired associate learning paradigm studies and found that subjects using an imagery mnemonic outperformed subjects given instructions to use rote-rehearsal, subjects told to produce a meaningful sentence using the pair of words, and subjects told to use whatever learning strategy they typically used (cf., Bower, 1972a). Bower developed two explanations for the superior performance of the subjects who used imagery. One reason dealt with the distinctiveness of the encoding of the imagery information. The distinctiveness explanation was not in direct opposition to Paivio's dual coding view; it merely emphasized the strength or distinctiveness of the image code. Alternatively, the imagery formed about a word pair might have provided information about an association between the two words--it was possible that a spatial relationship could be found or described that rendered the two items part of a "perceptual whole." Thus, the recall of one word given the other would be higher. This perspective is in disagreement with Paivio's theory, which maintains that the imaged items have separate perceptual codes that render them more memorable individually; Bower's relational association theory argues that it is the relationship imagined between the items that is key to better recall.

Bower tested his ideas by comparing recall of word pairs that were learned under different instructions. Subjects were to use rote repetition, to construct an interactive image using the two words in each pair, or to imagine the two objects "noninteracting and separated in 'imaginal space'.' Recognition of the words was unaffected by the instructions, but recall of the word pairs was significantly better when interactive imagery was used than when either of the other strategies was employed. Such evidence was taken as support for the relational association explanation, since both imaging strategies should have resulted in codes that used the same picture elements and should have been equally distinctive in terms of the shapes and features involved. Other researchers have found evidence that supports this conclusion as well (Lutz and Lutz, 1977; Wollen and Lowry, 1971; Wollen, Weber and Lowry, 1972), although not all studies have documented the superiority of interacting imagery cues (Biron and McKelvie, 1984).

A third model of the imagery process was offered by Kieras (1978). His propositional representation model is not in disagreement with Bower's theory, but it is much more general. The propositional representation model conceptualizes images as internally constructed perceptual descriptions, separate and different from semantic information, yet handled by the same overall processing system that is responsible for semantic processing. The images, or perceptual representations, in his model are thought to be nodes in memory structure containing perceptual information that have more interconnections to other nodes in the memory network than most verbal or semantic information.

"Context availability" is a theoretically promising component of the perceptual representation model, according to Kieras (1978). According to this view, a key process in the comprehension and recall of a sentence is assigning a framework or context to the sentence. If a context can be assigned, the sentence is better comprehended and better learned as a result. This view maintains that high imagery sentences would be expected to be better remembered because their image content has more chance of being interconnected with information already in memory, and thus they are more easily assigned a relevant context. The availability of the relevant context is key, however--when such a context is not available, recall even of imagery information suffers.

Support for this view comes from research that has demonstrated that the availability of a contextual framework is a determinant of recall. Bransford and Johnson (1972) found significantly better comprehension and recall of an ambiguous verbal passage when a complete context was first provided. An important aspect of their finding was that providing the context after the information to be learned was significantly less effective than providing it before--the timing of the availability of a context was important.

Pedzak and Royer (1974) found that providing a to-be-remembered sentence at the end of a context-providing paragraph resulted in better recall of abstract but not concrete sentences. Bower, Karlin, and Dueck (1975) found that providing subjects with an interpretation of each nonsense picture at the time of exposure resulted in significantly more picture information being recalled than when subjects were not provided the context-supplying interpretation. Thus, whether the to-be-remembered information consists of picture or verbal material, it seems that a key to effective learning of the information is the availability of a relevant context or framework in which the information can be interpreted.

The three imagery theories are not all very closely aligned. yet the contributions of each to the understanding of the imagery process are quite complementary. The biggest discrepancy is the conceptualization of the form in which the image information is stored in memory. Even on that dimension, the implications of the dual coding theory can be accounted for by the perceptual representation approach as well. The implications of the models, and particularly of * perceptual representation model, which approaches the issue from a broader information processing perspective, will be the basis for research questions dealing with a particular way in which imagery may be evoked--the use of sounds--and its expected effects.

Related Processes

Elaboration

The idea that elaborative thinking facilitates learning has been widely studied and is generally accepted.Elaborative thinking refers to the active relating of the target information to concepts, events, and other information already stored in memory. The result is a more elaborate network connecting the target information-and existing information in memory--the new information is said to be more elaborately encoded than it would have been without these relational, connective thoughts. In general, elaboration has been found to result in enhanced learning, as measured by recall, of the stimulus information (cf., Bower and Clark, 1969). Elaboration works by enabling the to-be-learned information to be incorporated into the memory structure more easily, thanks to the links that have been established through elaboration, to existing knowledge in memory.

However, it has been found that the types of elaborations that are generated are important. Elaborations that are semantically congruous with the information to be learned have resulted in improved learning (Craik and Tulving, 1975; Fisher and Craik, 1980). Stein and Bransford (1979), however, found that the elaborations most directly related to effective learning were those that allowed the learner to understand the significance of the information. Bransford et al (1982) found that the generation of elaborations that enhanced the significance of relationships between parts of the verbal passages that were to be learned was the key to successful learning. They went on to speculate that past personal experience or stored relevant knowledge may enable learners to generate relevant elaborations more easily.

The concept of elaboration is thus clearly linked lo Kieras' propositions that imagery may be effective because of the more numerous connections between images and other information in memory. Images are thought to be coded more elaborately than non-image information, and so may be more frequently connected to stored information. It is the elaborate encoding of imagery information that provides the better-connected memory trace, which results in better memory and easier elaboration of the information. The notion of context availability is related to elaboration as well. Learners would be engaging in elaboration when they seek a relevant context for understanding and remembering stimulus information; to the extent that a contextual framework is readily available, the elaboration process is made easier.

Perception

Perception refers to the processing of incoming stimulus information--identification and interpretation of the features of the stimulus that lead to a meaning being attached to that stimulus. The imagery process has been conceptualized as being similar to perception. In fact, imagery processes have been demonstrated to involve the formation and use of the same qualitative types of information that are generated by perceptual processes (cf., Crowder, 1976). However, imagery may take place even when the physical stimulus information is not present in the individual's environment, while perception occurs when the person is processing information that is present in her/his environment. The finding that imagery and perception share the same processing system would imply that imagery and perception processes should interfere with one another if attempts are made to perform both simultaneously.

This interference has been documented experimentally. Segal and Fusella (1970) manipulated the perception task subjects performed, either an audio or a visual detection task, and also manipulated the imagery task, imagination of a visual cue or a sound, that the subjects were performing simultaneously with the perception task. Results indicated interference: performance on the signal detection task was poorer when imaging was going on, regardless of the sense mode of either operation. In addition, a significant interaction was found, such that when the perception task and the imaging were being performed in the same sense mode (iso-modal), performance was worse than when the perception task and imaging were taking place in different sense modes. This iso-modal interference finding provides support for the notion that the imaging process shares some of the same analyzers or capacity within the processing system as the perceptual process. Further support for modality-specific interference was provided by the experimental findings of den Heyer and Barrett (1971), Bower (1972), and Peterson and Graham (1974). These results support the expectation that a message or task that encouraged subjects to form images while they were perceiving information in the same modality would result in worse learning of the information than would the nonsimultaneous performance of imaging and perception.

Imagery and advertising

Two advertising researchers (Rossiter and Percy, 1983) suggest that advertising can stimulate visual imagery in consumers in any of three ways: (1) by using stimuli known to increase the probability of image generation, (2) by instructing consumers to form visual images, and (3) by trying to capitalize, through target audience selection, on individual differences in visual imaging ability. Several attempts to apply visual imagery notions to advertising have been undertaken over the years.

Research interest has been focused on pictures (visual image cues) and the effects of their inclusion in an advertising message on consumers' learning and evaluative reactions. Starch (1966) found that ads with pictures were significantly more likely to be recognized than ads without pictures. Kisielius and Sternthal (1984) hypothesized that the effect of pictures on the evaluative response of the consumer would depend on the valence of the information recalled by the individual, and could be difficult to predict a priori. As discussed earlier, Lutz and Lutz (1977) and Biron and McKelvie (1984) found recall of product class - brand name pairings to be better when the verbal information was accompanied by pictures.

Childers and Houston (1984) concluded that including pictures in an ad was likely to enhance recall when either the motivation or capability of consumers- to process the ad at a semantic level (attending to the verbal content) was reduced. Edell and Staelin (1983) found that in order to be effective in enhancing audience learning, pictures must create images that are somehow related to the advertised brand. Results of other studies (Mitchell and Olson, 1981; Mitchell, 1986) indicate that pictures may affect attitudes independently from beliefs. In summary, generally positive effects of related pictures in ads for recall of brand name or ad information have been found. Effects on evaluative reactions, either toward the advertised brand or the ad itself, have been mixed, with some studies reporting significant effects of pictures on attitudes toward the ad or the brand, and others finding no effects.

It is important to note that the findings of facilitative effects of pictures on recall were found in tests that used static pictures and words. It is not clear that the same positive effects should be expected when imagery occurs in response to cues presented in a dynamic message (e.g., radio or television advertising). As others have recognized (cf., Wright, 1974), dynamic presentations present more difficult processing situations for the message recipients. In light of the additional demands on the message recipient's processing capabilities caused by dynamic presentations, it would be expected that the effectiveness of imagery cues would depend on whether imagery activity would interfere with the other processing activities taking place concurrently. For this reason, consideration of the individual's attentional system and capacity is needed.

ATTENTION

The second major theoretical issue to be examined concerns the possibility that providing sounds as imagery cues in audio messages may interfere with the reception and/or processing of other message information. Thus, the findings on attention and processing capacity limitations merit consideration.

Attention is commonly defined as the allocation of processing capacity, or the conscious, controllable, selective aspects of perception and response (Triesman, 1969). Several theories have been developed to describe how the attentional system works (cf., Broadbent, 1958; Duncan, 1980; Kerr, 1973; Posner, 1982).

A potentially insightful model is the limited capacity central mechanism hypothesis (see Kerr, 1-973), which proposes that some but not all mental operations require space in a single limited capacity central mechanism. Processes that do not require the central system can be performed in parallel with other operations without interference, but processes that do require the central mechanism can not be performed simultaneously without interference. Results of secondary task studies (Posner and Keele, 1970; Johnston et al, 1972; Kerr, 1973) indicate that processes such as low level, single item encoding, monitoring, and routine memory look-up operations may be performed in parallel with other tasks without resulting in any secondary task performance decrements. However, to note or interpret a stimulus, or to do something with the incoming information, such as rehearse it, transform it, or respond to it, requires central capacity, so that the processing of multiple inputs is expected to result in interference (Duncan, 1980; Posner, 1982), less complete processing, and subsequent worse recall for either or both of the input stimulus streams.

The extent to which simultaneous stimuli will result in interference will depend, in part, on the number of different channels occupied by the incoming stimuli and the tasks involved. Studies suggest the degree to which operations interfere may depend on the extent to which they place overlapping demands on specific subsystems, such as the verbal system or the spatial system (Kerr, 1973; Kimchi, 1982; Segal and Fusella, 1970).

A test of audio interference was reported by Warshaw (1978), who tested subjects' recall of simultaneous nonredundant audio and video messages versus their recall of either channel alone. His findings that recall was lowest for the simultaneous audio and visual presentation supported the limited processing capacity conceptualization of the processing system. Another test of audio attention is the reading comprehension test performed by Zimmer and Brachulis-Raymond (1978), in which student subjects were tested for their comprehension of material that they had read under one of four noise conditions: no noise, music, speech unrelated to the reading material, and industrial noise. Significantly worse comprehension was found for the material that had been read under the industrial noise condition; no other differences were significant. The authors concluded that perhaps the findings were a result of "habituation," and they cited other work that provided evidence that students are able to quite effectively "gate out" or screen out familiar stimuli, such as music, when their attention is focused on a task such as studying.

These results support Triesman's (1969) findings that individuals can be very effective in selecting their focus of attention, so that one stream of sensory input is selected for processing and all other streams of incoming information are, for the most part, ignored. This does not imply that distraction will never occur, but it appears that individuals can to some extent screen out information streams that are of no interest to them, rather than allowing those information streams to be automatic causes of distraction.

Attention to Advertising

An important actor that distinguishes consumer learning of advertising information from individuals' learning of other information is the exposure setting. Exposure to and processing of advertisements usually takes place in a setting characterized as "noisy," with many streams of information competing for consumers' attention. In addition, the consumer is unable to control the pace at which much of the information is being presented, so that processing may be rushed or incomplete (Wright, 1974). Moreover, most consumers are not motivated to expend much cognitive effort on interpreting, processing, and integrating most advertising information (Batra and Ray, 1983).

This difficult exposure environment results in consumer processing that is different from the processing that results from most laboratory research tests, which generally produce high levels of subject involvement. Recent studies and related findings have proposed that "low involvement" processing may be a more ecologically valid approach to the study of advertising effects (cf., Batra and Ray, 1983; Mitchell, 1983; Mitchell, 1986; Park and Young, 1986).

The notion that consumers can select a target and focus their attention effectively and at will has found wide acceptance in the psychology literature (cf., Triesman, 1969; Glass, Holyoak and Santa, 1979). In advertising practice, it provides the rationale for "borrowed interest" techniques (i.e., using an attention-attracting stimulus in an ad, even if that stimulus has no close relation to the advertised product) for attracting attention from other environmental stimuli. In fact, Batra and Ray (1983) suggest that one way to convert the low interest level of a consumer advertising audience to a "pseudo high involvement" one is to add borrowed interest techniques that "artificially" raise the level of "natural" drive or interest in the situation. They caution that an advertiser must take care to ensure that the attention-attracter does not draw attention away from the advertised product itself, but rather draws attention from distracting elements outside the commercial.

RESEARCH ISSUES

In light of the research findings in the areas of imagery theory and attention, consideration of nonvisual imagery cues leads to several issues that merit research attention

1) Can non-visual cues stimulate imagery activity? The power of non-visual cues has not yet been established. So far, the non-visual imagery activity of subjects has been tested only under experimental conditions that included instructions to subjects to imagine particular sounds (Segal and Fusella, 1970). The question of interest here is whether sounds are capable of producing in subjects mental images when listeners are not instructed to imagine.

2) How does imagery evoked by non-visual cues affect learning and memory? Imagery research findings would indicate that positive effects of imagery may result from elaboration and context availability processes, while the findings on perception and attention indicate that negative effects of imagery might result from perceptual interference and the exceeding of processing capacity limits. A question of real interest, then, is under what conditions would each of these effects would be expected to dominate

3) Is there a way that image y cues could be presented or placed within the message so as to maximize their positive effects on learning and memory and minimize their negative effects? The work on interference and context availability indicates that a variable worth consideration is the ordering of verbal information and sounds within the message. Arranging for nonsimultaneous occurrence of sounds and words in the message would be expected to lighten the informational load facing the listener at an instant, and should reduce the likelihood of interference in the reception of the audio channel information. As well, sounds, as imagery cues, may serve as frames for the processing of subsequently presented information.

4) Does personal experience with the imagery cues affect the efficiency with which the cues evoke images? People often react very differently to the same sounds--e.g., the music of the '60s has been found to appeal much more strongly to the "baby boomers" than to the elderly. Imagery cues may be used effectively to segment the audience on the basis of their experiences.

5) How does the ambiguity of cues affect the listener's attention level, imagery activity, and learning and memory? It may be that ambiguous sounds are attention-attracting, but findings from research on elaboration suggest that imagery activity might be curtailed, with negative consequences for learning and memory.

Methodology Considerations

The method by which these issues are to be tested also merits consideration. Given the reality of noisy reception environments for many, if not most, mass media communications, a strong case can be made for research environments that are externally valid, or that at least include several interventions to render them more realistic. Beyond external validity concerns, there are theoretical considerations that should affect the research setting decision as well: important distinctions exist between the implications of high and low involvement processing. Another methodological consideration is the importance of the use of multiple operationalizations of the imagery cues. Since the images evoked by any particular sound, or other imagery cue, may be idiosyncratic and dependent on the individual's past experiences, any test of the imagery processes under consideration would require replication across different imagery cues.

REFERENCES

Batra, R. and Ray, M.L., "Advertising Situations: The Implications of Differential Involvement and Accompanying Affect Responses," in R.J. Harris, ed., Information Processing Research in Advertising, Hillsdale, N.J.: Lawrence Erlbaum Associates, Publishers, 1983, 127-151.

Biron, J. and McKelvie, S.J., "Effects of Interactive and Noninteractive Imagery on Recall of Advertisements," Perceptual and Motor Skills, Vol. 59, 1984, 799-805.

Bower, G.H., "Imagery as a Relational Organizer in Associative Learning," Journal of Verbal Learning and Verbal Behavior, Vol. 9, 1970, 529-533.

Bower, G.H., Cognition in Learning Memory, New York: John Wiley, 1972(a).

Bower, G.H., "Mental imagery and associative learning," in L.W.Gregg, ed., Cognition in learning and memory, New York: Wiley, 1972(b).

Bower, G.H. and Clark, M.C., "Narrative stories as mediators for serial learning," Psychonomic Science, Vol. 14, 1969, 141-182.

Bower, G.H., Karlin, M.B., and Dueck, A., "Comprehension and Memory for Pictures," Memory and Cognition, Vol. 3, 1975, 216-220.

Bransford, J.D. and Johnson, M.K., "Contextual Prerequisites for Understanding: Some Investigations of Comprehension and Recall," Journal of Verbal Learning and Verbal Behavior, Vol. 11, 1972, 717-726.

Broadbent, D.E., Perception and Communication, London: Pagamon Press, 1958.

Broadbent, D.E., "The Hidden Preattentive Processes," American Psychologist, Vol. 32(2), February, 1977, 109-118.

Bugelski, B., "The Definition of the Image," in S.J. Segal, ed., Imagery: Current Cognitive Approaches, New York: Academic Press, 1971, 49-68.

Childers, T.L. and Houston, M.J., "Conditions for a Picture-Superiority Effect on Consumer Memory," Journal of Consumer Research, Vol. 11 (September), 1984, 643-654.

Craik, F.I.M. and Tulving, E., "Depth of processing and the retention of words in episodic memory,' Journal of Experimental Psychology: General, Vol. 104, 1975, 268-294.

Crowder, R.G. Principles of Learning and Memory, Hillsdale, NJ: Lawrence Erlbaum Associates, 1976.

den Heyer, K. and Barrett, B., "Selective loss of visual and verbal information in STM by means of visual and verbal interpolated tasks," Psychonomic Science, Vol. 25, 1971, 100-102.

Duncan, J., ' The Locus of Interference in the Perception of Simultaneous Stimuli," Psychological Review, Vol. 87(3), 1980, 272-300.

Edell, J. and Staelin, R., "The Information Processing of Pictures in Print Advertisements," Journal of Consumer Research, Vol. 10, 1983, 45-61.

Fisher, R.P. and Craik, F.I.M., 'The effects of elaborations or recognition memory," Memory & Cognition, Vol. 8, 1980, 400-404.

Glass, A.L., Holyoak, K.J., and Santa, J.L., Cognition, Reading, MA: Addison-Wesley Publishing Company, 1979.

Gorman, A.M. "Recognition memory for nouns as a function of abstractness and frequency," Journal of Experimental Psychology, Vol. 61, 1961, 23-29.

Holbrook, M.B., "Product Imagery and the Illusion of Reality: Some Insights from Consumer Esthetics," in R. Bagozzi and A. Tybout, eds., Advances in Consumer Research, Vol. 10, 1983, 65-71.

Jenkins, J.R., Neale, D.C., and Deno, S.I., "Differential memory for picture and word stimuli," Journal of Educational Psychology, Vol. 58, 1967, 303-307.

Johnston, W.A., Griffith, D., and Wagstaff, R. "Speed, accuracy and ease of recall," Journal of Verbal Learning and Verbal Behavior, Vol. 11, 1972, 512-520.

Kahneman, D., Attention and effort, Englewood Cliffs, NJ: Prentice-Hall, 1973.

Kerr, B., "Processing demands during mental operations," Memory and Cognition, Vol. 1(4), 1973, 401-412.

Kieras, D., "Beyond Pictures and Words: Alternative Information-Processing Models for Imagery Effects in Verbal Memory," Psychological Bulletin, Vol. 85(3), 1978, 532-554.

Kimchi, R., "Mental effort and task interference in auditory attention," Perception & Psychophysics, Vol. 32(5), 1982, 473-480.

Kisielius, J. and Sternthal, B., "Detecting and Explaining Vividness Effects in Attitudinal Judgments," Journal of Marketing Research, Vol. 21 (February), 1984, 54-64.

Lutz, K.A. and Lutz, R.I., "Effects of Interactive Imagery on Learning: Application to Advertising," Journal of Applied Psychology, Vol. 62(4), 1977, 493-498.

MacInnis, D.J. and Price, L.L., "The Role of Imagery in Information Processing: Review and Extensions," Journal of Consumer Research, Vol. 13, 1987, 473-491.

Mitchell, A.A., "Cognitive Processes Initiated by Exposure to Advertising," in R.J. Harris, ed., Information Processing Research in Advertising, Hillsdale, N.J.: Lawrence Erlbaum Associates, Publishers, 1983, 13-42.

Mitchell, A.A., "The Effect of Verbal and Visual Components of Advertisements on Brand Attitudes and Attitude Toward the Advertisement," Journal of Consumer Research, Vol. 13(June), 1986, 12-24.

Mitchell, A.A. and Olson, J.C., "Are Product Attribute Beliefs the Only Mediator of Advertising Effects on Brand Attitudes?" Journal of Marketing Research, Vol. 18(August), 1981, 318-332.

Paivio, A.,"Mental Imagery in Associative Learning and Memory," Psychological Review, Vol. 76, 1969, 241-263.

Paivio, A., "Imagery and Language," in S.J. Segal, ed.,f Imagery: Current Cognitive Approaches, New York: Academic Press, 1971(a), 7-32.

Paivio, A., Imagery and Verbal Processes, New York: Holt, Rinehart & Winston, 1971(b).

Paivio, A., "Perceptual comparisons through the mind's eye," Memory and Cognition, Vol. 3, 1975, 635-647.

Paivio, A., "The Empirical Case for Dual Coding," in I.C. Yuille, ed., Imagery, Memory and Cognition: Essays in Honor of Allan Paivio, Hillsdale, N.J.: Lawrence Erlbaum Associates, Publishers, 1983,-.307-322.

Paivio, A. and Csapo, K. "Picture superiority in free recall: Imagery or dual coding?" Cognitive Psychology, Vol. 5, 1973, 176-206.

Paivio, A., Yuille, J.C., and Madigan, S. "Concreteness, imagery, and meaningfulness values for 925 nouns," Journal of Experimental Psychology Monograph, Vol. 76, 1968, Part Two.

Park, C.W. and Young, S.M."Consumer Response to Television Commercials: The Impact of Involvement and Background Music on Brand Attitude Formation," Journal of Marketing Research, Vol. 23, 1986, 11-24.

Pedzek, K. and Royer, J.M., "The role of comprehension in learning concrete and abstract sentences," Journal of Verbal Learning and Verbal Behavior, Vol. 13, 1974, 551-558.

Peterson, M.J. and Graham, S.E., "Visual detection and visual imagery," Journal of Experimental Psychology, Vol. 103, 1974, 509-514.

Posner, M., "Cumulative Development of Attentional Theory," American Psychologist, Vol. 37(2), February, 1982, 168-179.

Posner, M.I. and Keele, S.W. 'Time and space as measures of mental operations," Address, American Psychological Association, September, 1970.

Rossiter, J.R. and Percy, L., "Visual Communication in Advertising," in RJ. Harris, ed., Information Processing Research in Advertising, Hillsdale, N.J.: Lawrence Erlbaum Associates, Publishers, 1983, 83-125.

Segal, S.J. and Fusella, V., 'Influence of Imaged Pictures and Sounds on Detection of Visual and Auditory Signals," Journal of Experimental Psychology, Vol. 83(3), 1970, 458-464.

Starch, D., "How Does the Shape of Ads Affect Readership?" Media/Scope, Vol. 10, 1966, 83-85.

Stein, B.S. and Bransford, J.D., "Constraints on effective elaboration: Effects of precision and subject generation," Journal of Verbal Learning and Verbal Behavior, Vol. 18, 1979, 769-777.

Triesman, A., "Strategies and Models of Selective Attention," Psychological Review, Vol. 76(3), 1969, 282-299.

Warshaw, P.R., "Application of Selective Attention Theory to Television Advertising Displays," Journal of Applied Psychology, Vol. 63(3), 1978, 366-372.

Wollen, K.A. and Lowry, D.H., "Effects of Imagery on Paired-Associate Learning," Journal of Verbal Learning and Verbal Behavior, Vol. 10, 1971 (June), 276-284.

Wollen, KA., Weber, A., and Lowry, D.H., "Bizarreness versus Interaction of Mental Images as Determinants of Learning," Cognitive Psychology, Vol. 3, 1972 (July), 518-523.

Wright, P.L., "Analyzing Media Effects on Advertising Responses," Public Opinion Quarterly, Vol. 38, 1974, 192-205.

Zimmer J.W. and Brachulis-Raymond, J., "Effects of Distracting Stimuli on Complex Information Processing," Perceptual and Motor Skills, Vol. 46, 1978, 791-794.

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