Imagery Paradigms For Consumer Research: Alternative Perspectives From Cognitive Psychology

Terry L. Childers, University of Minnesota
Michael J. Houston, University of Wisconsin
ABSTRACT - Recent research in cognitive psychology and consumer behavior has focused increasingly on the effects of nonverbal information processing. This paper explores the nature of mental imagery as a facilitator of learning and memorial retention. Several paradigms developed in cognitive psychology dealing with-picture-word processing differences are discussed in terms of their explanation of the pictorial superiority effect. Finally, areas of further research are proposed within the context of these conceptual paradigms in order to further .our understanding of the effects of imaginal information processing on marketing communications.
[ to cite ]:
Terry L. Childers and Michael J. Houston (1983) ,"Imagery Paradigms For Consumer Research: Alternative Perspectives From Cognitive Psychology", in NA - Advances in Consumer Research Volume 10, eds. Richard P. Bagozzi and Alice M. Tybout, Ann Abor, MI : Association for Consumer Research, Pages: 59-64.

Advances in Consumer Research Volume 10, 1983      Pages 59-64


Terry L. Childers, University of Minnesota

Michael J. Houston, University of Wisconsin


Recent research in cognitive psychology and consumer behavior has focused increasingly on the effects of nonverbal information processing. This paper explores the nature of mental imagery as a facilitator of learning and memorial retention. Several paradigms developed in cognitive psychology dealing with-picture-word processing differences are discussed in terms of their explanation of the pictorial superiority effect. Finally, areas of further research are proposed within the context of these conceptual paradigms in order to further .our understanding of the effects of imaginal information processing on marketing communications.


The literature on picture versus word processing in psychology and consumer behavior has established in reasonably strong fashion that the learning of pictures occurs more readily than their verbal counterparts (Paivio and Foth 1970; Yuille 1973; Bower 1970; Peterson and Murrey 1973; Peterson and McGee 1974; Lippmann and Shanahan 1973; Lutz and Lutz 1977; Childers and Houston 1982). This effect is at least partially attributable to the superior ability of pictures to evoke images of perceived stimuli in the mind's eye of the receiver (Paivio 1969). Our understanding of imaginal processing and its effect on consumer choice processes however is visibly lacking in the consumer behavior literature as most recent research has concentrated on the verbal concept system (Bettman 1979). The general purpose of this paper is to examine the conceptual underpinnings of mental imagery drawing from recent literature in cognitive psychology. More specifically, the paper will (1) present a conceptual discussion of the nature of mental imagery, (9) discuss three paradigms dealing with picture versus word information processing and, (3) examine the implications of these conceptual frameworks for conducting consumer behavior research.


Not unlike many constructs, imagery responses must be inferred, are hypothetical, implicit, internal, and therefore unobservable (Bugelski 1971). Given these characteristics, the study of mental imagery is at best a difficult endeavor. Its popularity dates back to the early 1900's after which disfavor towards it developed as "behaviorism" (Watson 1913) gripped the study of psychology following the "imageless thought controversy." With the loosening of this grip and the resurgence of the study of conscious experience, mental imagery has reemerged as a popular area of study (see Holt 1964 for an informative historical review). However, the reemergence of mental imagery has occurred with a somewhat broader meaning attached to it which has probably lent to the confusion surrounding its study.

Taxonomy of Imagery

Imagery should not be considered a singular concept but rather as an umbrella of concepts grouped under one common label. Richardson (1969) offers one taxonomy that differentiates the construct into (1) after imagery; (2) eidetic imagery; (3) imagination imagery; and (4) memory imagery. This taxonomy centers on the extent to which images are vivid and controllable . After imagery is relatively vivid (as when a camera flashbulb goes off) but impossible to control. Eidetic imagery is relatively vivid and subject to somewhat more control than the after image. The imagination image parallels the eidetic image in vividness but it is typically found that this type--is less controllable. The ordinary memory image involves the reconstruction of a past percept and is to this extent a more centrally initiated event not dependent upon the persistence of sensory after-effects. It is characteristically less vivid and more controllable than the after-image or the immediate eidetic image (Richardson 1969, p. 128). The memory image thus involves an individual ability to visualize in one's mind an event that has been previously experienced and stored in memory. It is towards an understanding of the memory image that this paper is focused.

Forms of Mental Imagery

The memory image is also not singular in terms of the nature of information that is represented from past experience. Images can be visual, tactual, auditory, gustatory, and olfactory (Bourne, Dominowski, and Loftus 1979). Although there are important implications for understanding consumer choice from each type of mental imagery the objective of this paper is on understanding picture versus word processing differences and hence will concentrate on mental visual imagery although some consideration will be given to our need to more fully understand the nature of each type of mental imagery.

Nature of Mental Visual Imagery

In perhaps the seminal work in the area, Paivio (1971, p. 135-36) defines imagery as, "a memory code or associative mediator that provides spatially parallel information that can mediate overt responses without necessarily being consciously experienced as a visual image." In his text, Richardson (1969) reviews some classical definitions from the prebehavioral era of psychology and offers the following definition:

Mental imagery refers to (1) all chose quasi-sensory or quasi-perceptual experiences of which (2) we are 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 (Richardson 1969, p. 2).

By quasi-sensory or quasi-perceptual experiences, Richardson is referring to any concrete re-presentation of sensory, perceptual affective, or other experimental states. When someone blushes with shame at the memory of a certain event, that person is assumed to experience the same emotion originally experienced even though the antecedent conditions differ across the two occasions.

According to Segal (1971, p. 97), the percept is simply the picture we build of the world around us and while the image may vary more based upon one's expectancies or needs the image also has a close relationship to the external world. Shephard (1978) would also appear in agreement with this view of imagery. He argues that the brain processes underlying a mental image must necessarily contain the information that could in principle permit the reconstruction of a picture with a high degree of isomorphism to the external object imagined. This characterization of imagery has been called "second-order isomorphic" since an image has a deep (interpreted meaning) as well as a surface (appearance) structure due to the cognitive interpretation of the image by the receiver.

Lutz and Lutz (1978) indicate in their review of mental imagery that mental visual imagery is a way to process information perhaps in a parallel fashion versus the more sequential processing of verbally based information. Yuille and Catchpole (1977) make a similar observation drawing upon the writings of Piaget and Inhelder 21973). An important distinction can be made between the operative level of thought which is the storehouse of an individual's basic knowledge and the figurative plane which, "is the servant of operative thought and permits internal representations (of which the image is a common type) that support operative activities such as remembering and problem solving," (Yuille and Catchpole 1977, p. 175). Images thus are the tools of thought that provide a temporary representation of our memories and ideas and, as such, can perform a functional role. Yuille and Catchpole (1977) offer a model which incorporates these two cognitive planes of operation as the representational (figurative and working) and abstract (operative and storage) planes. One principle function of the representational plane is to permit the reconstruction of concrete representations of memories, ideas, and feelings that have been incorporated into the abstract plane. This notion of reconstruction is important since it casts imaging as dynamic, active, and interpretative which are important facets of a cognitive process-oriented perspective of human behavior (Bettman 1979).


Given the representational nature of mental visual imagery several alternative perspectives exist as to the source of the functional superiority of imaginal based encoding processes. These frameworks stem from attempts to explain differences in picture versus word processing which as noted have at their roots differing capacities for eliciting mental images. Each paradigm will first be described individually, and then a comparison will be made and implications drawn for understanding the effects o' imaginal processing of pictures versus words on learning and memory.

Dual Coding Model

Paivio (1971; 1978) has proposed a model that distinguishes between the nonverbal imagery system and a verbal or linguistic system and is thus referred to as the dual coding model. The two systems, "are assumed to involve independent but partially interconnected systems for encoding, storage, organization, and retrieval of stimulus information" (Paivio and Csapo 1973, p. 177). Independence between the two coding systems is assumed and used to indicate that either of the codes can be activated in varying degrees and that the codes can produce a cumulative effect on retrieval. The codes are also interconnected in that one code can be transformed into the other (i.e., pictures can be named and words can evoke images).

Within the dual coding model when a stimulus is received, it first passes through the representational level where an image is activated or a verbal representation is acquired. Next the stimulus is processed at the referential level where pictures are named and images are generated for words (dependent upon the concrete-abstract nature of the verbal cue). Finally, associative processing may occur where connections may be made between the nonverbal and verbal codes and prior stored information. An important assumption of the dual coding model is that the imaginal code is specialized for parallel processing of stimulus features while the verbal system is conceived of as being more suitable for sequential processing of information.

Paivio attempts to account for picture-word differences in memorability in terms of the differential imagery evoking ability of the stimulus. "Image arousal in the case of pictures and verbal coding in the case of words have the highest availability, the verbal code to pictures second, imagery to concrete words third, and images to abstract words fourth" (Paivio 1969, p. 257). Memorability is thus a function of the greater probability that dual codes will be available to aid in the retrieval of processed pictures compared to concrete or abstract nouns. The model essentially attributes the picture superiority effect to coding redundancy (two codes are better than one). Paivio, however, differs from a strict dual coding approach in that the analogue, imagery code is considered to be qualitatively superior to the verbal memory code (Paivio and Csapo 1973). The nature of the superior imagery code, however, is left open to question although speculation is provided regarding the possible superior associative or organizational qualities of pictures versus words.

The Sensory-Semantic Model

A second conceptual framework, the sensory-semantic model, (Nelson, Reed and McEvoy 1977; Nelson 1979) is similar to levels of processing theory (Craik and Lockhart 1972) in that it as well acknowledges that stimuli are represented in memory in terms of three distinctive features; visual (sensory or appearance features), phonemic (acoustic or sound), and semantic (meaning or significance). The model, however, differs from a levels approach in terms of several of its underlying assumptions regarding the nature of picture versus word processing.

One assumption of the sensory-semantic model is that these features are processed in different orders dependent upon whether the stimulus is a picture or a word. When processing pictures one goes from visual (appearance) features to accessing meaning from memory and then, if necessary, phonemic features can be accessed which provide a name or label for the picture. In contrast, words are processed differently by progressing from visual features then either to phonemic features or to meaning features (Nelson and Brooks 1973; Nelson and Reed 1976). Thus, a word might be named without accessing meaning while a picture would require access to meaning prior to assigning a name.

Another assumption of the model relates to the type of information encoded at different levels of processing. The model assumes that encoding efforts can be selectively directed (Nelson 1979). Encoding efforts can be selectively directed toward any of the three types of features by instructions (Paivio 1971), requiring subjects to answer questions (Craik and Tulving 1975), or by varying stimulus cues (Nelson, Reid, and McEvoy 1977; Nelson, Reed, and Walling 1976). However, even though encoding efforts may be directed toward a specific focal code other attributes of the stimulus event may also be activated (Nelson 1979). For example, even though the focal codes may be names for pictures, attributes activated prior to these name codes (visual and meaning features) may also affect performance (Nelson, Reed, and McEvoy 1977). In other words, operations directed at encoding meaning features do not result in only meaning codes in memory and operations directed toward sensory features do not result in only sensory codes.

A final assumption of the model relates to the relative distinctiveness and interactive qualities of activated

features or codes. Consistent with the levels of processing theory, retention of encoded material is considered to be a function of the unique and unified nature of the stimulus at encoding and the degree to which the retrieval environment recaptures this encoding (Nelson 1979). Distinctive encodings are potentially easier to remember because they are subject to less interference from competitive encodings at retrieval.

The sensory-semantic model attempts to account for the picture superiority effect through the more distinctive encodings obtained when processing pictures as compared to words.

The sensory code for a picture is apparently more differentiating and less susceptible to interference from successively occurring items The present findings suggest that pictorial stimuli provide a qualitatively superior sensory code and that at least for simple pictures, the semantic representations for pictures and their corresponding labels may be identical (Nelson, Reed, and Walling 1976, pp. 527-8).

Pictures are constructed from a visual alphabet that contains many more shapes and styles than does the verbal alphabet. 'When encoded, the distinctive pictorial stimulus produces a unique sensory code which then aids in the later recollection of the stimulus.

Prototypical Model of Picture and Word Processing

Snodgrass (1980) has proposed a model of picture and word processing, which we have designated as a prototypical model. The model contains three levels of processing each of which has a memory score. Thus the model combines the features of both process models of memory (e.g., Craik and Lockhart 1972) and structural models of memory (e.g., Atkinson and Shiffrin, 1971).

Level I consists of the new verbal or visual codes resulting from processing the physical characteristics of the external stimulus (e.g., pitch, orientation, or color). Level II consists of prototypical information about acoustical and visual images that has been acquired through experience. This prototypical information is accessed as a potential or generated image and then through either a feature type analysis similar to the model proposed by Rips, Shoben, and Smith (1973) or a template comparison, a mismatch processor assists in pattern recognition. Level III is a propositional store of semantic knowledge which is accessible by the acoustic or visual image prototypes.

Snodgrass identifies two basic differences between picture and word processing. First, there is greater variability in the appearance of objects or pictures compared to the appearance or sound of words resulting in greater variability in prototypical visual images than prototypical acoustical images. Second, there is greater ambiguity of reference for words than for pictures. Words having the same surface form will have multiple representations in semantic or propositional memory. For example, the homograph bank may be at the side of a river or a place for financial transactions. As a result, pictures will access fewer propositional modes than their verbal counterparts.

In attempting to account for the pictorial superiority effect a precise explanation stemming from the Snodgrass model is perhaps not as well developed as in the previous models. Research by Snodgrass and Asiaghi (1977) however, suggests that pictorial superiority stems from superior semantic codes as well as superior sensory codes. Within the framework of the present model this might be addressed on two accounts. First, the greater variability in prototypical visual images allows or a greater accumulation of mismatch features between the processed stimulus and the stored prototypical image. Hence, in a recognition task, more discriminating ability is available, perhaps through the sensory features of the image. In a recall task, pictorial superiority may arise due to the organizational quaLities of the picture and its corollary impact on semantic associations. With fewer propositional linkages occurring through picture stimuli, the organizational qualities of the stimulus may produce less response competition (Lupker 1979) as well as strengthen existing propositional networks, both of which would be expected to facilitate recall.

Comparison of Models

The three models take several different perspectives in their description of picture versus word processing. At the very basic level, models of cognitive processing, particularly with respect to memory, have followed a dichotomy of either a structural or a process orientation. The classic memory store model proposed by Atkinson and Shiffrin (1971) in which shore-term and long-term memory are distinguished is exemplary of the structural approach, while the levels of processing model (Craik and Lockhart 1972) assumes a process orientation.

Of the three models, Paivio (1971) takes a more structural approach to explaining picture-word processing, the sensory-semantic is a process oriented model, while the prototypical model is a combination of these approaches. One consequence of these approaches is the conceptual nature in--which imagery is created within each-framework. Drawing from the previous definition, Paivio regards imagery as a "memory code" whereas the other models consider imagery as a cognitive response that can be separated from the type of abstract memorial coding system discussed by Yuille and Catchpole (1977); Anderson (1978); and Pylyshyn (1973). Additionally, the dual coding approach acknowledges the possibility of consciously experienced imagery but its occurrence is given a second order nonessential status (Marks 1977), which is again inconsistent with the more process oriented models of imagery. A key issue therefore is whether imagery is a memory code or whether it affects memory codes.

The implications of this issue are grounded in the approach that each model takes in accounting for the pictorial superiority effect. The dual coding hypothesis relies on "coding redundancy" for its explanation of picture superiority. Pictures are more likely to generate an image and verbal code and thus are more likely to provide multiple access paths for the subsequent retrieval of memorial information. The other models, although incorporating coding redundancy, attempt to extend this concept by dealing with the qualitative nature of the coding system rather than its mere existence. In particular, the sensory-semantic model attributes picture superiority to a qualitatively superior sensory code. This model even asserts that under paired associate learning picture naming doesn't occur and therefore dual codes do not even necessarily exist in memory -- a totally opposite prediction from Paivio's model.

The prototypical model proposed by Snodgrass (1980) would appear to incorporate characteristics of the other models. Processing and storage at Level II (acoustic and visual image stores) is akin to Paivio's dual codes. However, interconnections between the two systems typically are made through the more abstract propositional store at Level III which is more consistent with the sensory-semantic model.

In contrast with the dual coding and sensory-semantic model, the prototypical model also appears to take a more flexible task-dependent approach to explaining the picture superiority effect. Comparing recognition to recall, recognition is more oriented towards matching a processed and stored stimulus, while recall is more heavily retrieval oriented. Only the Snodgrass model appears to vary the source of the pictorial superiority effect to accommodate these two memorial tasks. Acknowledging that both sensory and semantic attributes of the stimulus are important to fully understand pictorial superiority, the former perhaps plays a larger role in recognition and the latter a more substantial role in recall. In the sensory-semantic model, sensory attributes are hypothesized to differentiate pictures from words regardless of the task while coding redundancy would also not appear to be as sensitive to the processing task.


A consideration of the impact pictures and mental imagery have on the processing of information has important implications for improving the effectiveness of marketing communications. This discussion will attempt to draw on some of chose implications which stem from utilizing one or more of the models to guide research into picture-word processing differences.

If pictures consistently are found to aid in the learning and retention of product consumption information then a first objective for consumer researchers is to try to understand why the inclusion of a picture in an advertisement may improve the effectiveness of the ad. Advertisers often have multiple objectives when targeting an ad which focus on several stages of the hierarchy of objectives (Lavidge and Steiner 1961). Principle among these is of ten an increase in "top of the mind" recall of a brand or company name and the communication of product attribute information which serves to Position and differentiate a brand.

In trying to understand the role that pictures and imagery play in accomplishing these objectives the models provide us with different insights. Based upon the research of Nelson (1979) and his associates within the context of the sensory-semantic model, one might conclude .hat advertising dollars spent on appearance (sensory) attributes of the ad (e.g., color, size, graphics) are the basis for improved advertising effectiveness. Drawing from the Paivio (1971) paradigm, coding redundancy might imply that pictures should be incorporated into ads and that these pictures should convey as many product related cues as possible. These cues, after being processed imaginally under a cognitive elaboration strategy (Jacoby and Craik 1979), would provide multiple retrieval paths for assessing stored information. In contrast, within the framework of the prototypical model offered by Snodgrass (1980), research into the picture superiority effect should focus on the interplay between both cue accessibility and the qualitative nature of the cues. The accessibility of redundant imaginal and linguistic cues is perhaps an important condition for picture superiority but the nature of the cues is also important. Research guided by the prototypical model might then focus on the sensory versus semantic basis for picture superiority within the context of instore shopping behavior (more recognition oriented) versus out of store shopping (more recall oriented). Perhaps the accessibility of sensory attributes of the processed communication might be more helpful in explaining memorial recovery under instore shopping conditions while the semantic features of image might play a greater role in explaining consumer usage of consumption information outside of the store environment.

Another important consideration when eliciting imagery through pictures in advertisements would be the consistency between the copy and the visual component of the communication. Bransford and Johnson (1972) have shown that a picture which provided a meaningful context for a passage enhanced recall of the passage compared to a picture that was not as consistent. The dual coding model would not appear to deal with this finding as readily as the other models since it primarily focuses on the accessibility of multiple codes and not the qualitative nature of the interconnections among the codes. The consistency among the visual and verbal components of a marketing communication and its impact on memory is an important question that might be investigated within the context of either the sensory-semantic or the prototypical models of picture-word processing.

An additional area of needed research focuses on a more complete understanding of the role that the components of mental imagery have on the communication process. As Holbrook (1981) has suggested, research on the effects of imagery should be expanded to include for example, acoustical, olfactory, tactual, and gustatory imagery. Gorn (1982) recently found that the association of music with a product had a significant effect on product preference. Research is lacking however, on the degree to which music of this nature might elicit acoustical imagery and in turn its effect on product preference and choice. Within the context of the present models, both the sensory-semantic and prototypical models deal with acoustical processing with the latter more oriented towards acoustical imagery. None of the models however, being developed for picture-word processing, deal with other facets of mental imagery but it would seem possible to expand the scope of the Snodgrass model to incorporate their effects.


This paper has attempted to provide an overview of the psychological construct of mental imagery. The theoretical nature of mental imagery as a psychological construct was discussed. Three alternative models of picture-word processing were described and contrasted in terms of their bases for explaining the imagery-eliciting superiority of pictures over words. While each model has much to offer consumer research, the sensory-semantic model and, especially, the prototypical model provide more extensive paradigms for investigating the effects of imaginal processing. Several areas in which further research is needed were discussed within the context of increasing our understanding of the role that imagery plays in mediating marketing communications.


Anderson, John R. (1978), "Arguments Concerning Representations for Mental Imagery," Psychological Review, 85 (July), 249-75.

Atkinson, Richard C. and Richard M. Shiffrin (1971), "The Control of Short Term Memory," Scientific American, 295 (August), 82-90.

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

Bourne, Lyle E., Jr., Roger L. Dominowski, and Elizabeth F. Loftus (1979), Cognitive Processes, New Jersey: Prentice-Hall, Inc.

Bower, Gordon H. (1970), "Imagery As a Relational Organizer in Associative Learning," Journal of Verbal Learning and Verbal Behavior, 9, 529-33.

Bransford, J.D. and M.K. Johnson (1972), "Contextual Prerequisites for Understanding: Some Investigations Or Comprehension and Recall," Journal of Verbal Learning and Verbal Behavior, 11, 717-26.

Bugelski, B.R. (1971), "The Definition of the Image," Imagery: Current Cognitive Approaches, (Sydney Segal, ed. ), New York: Academic Press.

Childers, Terry L. and Michael J. Houston (1982), "Sensory and Semantic Bases of Interactive Imagery In An Advertising Context," Unpublished working paper '5-89-26, University of Wisconsin-Madison.

Craik, F. I. M. and R.S. Lockhart (1972), "Levels of Processing: A Framework for Memory Research," Journal Or Verbal Learning and Verbal Behavior, 11, 671-84.

Gorn, Gerald J. (1982), "The Effects of Music in Advertising on Choice Behavior: A Classical Conditioning Approach," Journal of Marketing, 46, 94-101.

Holbrook, Morris B. (1981), "Some Further Dimensions of Psycholinguistics, Imagery, and Consumer Response." in Advances in Consumer Research: Vol. IX (A.A. Mitchell, cd.), 112-117.

Holt, R.R. (1964), "Imagery: The Return of the Ostracized,' American Psychologist, 19, 254-264.

Jacoby, Larry L. and Fergus I.M. Craik (1979), "Effects of Elaboration of Processing at Encoding and Retrieval: Trace Distinctiveness and Recovery or Initial Context," in Levels of Processing in Human Memory, (Laird S. Cermak and Fergus I.M. Craik eds.) New Jersey: Lawrence Erlbaum Associates.

Lavidge, Robert L. and Cary A. Steiner (1961), "A Model for Predictive Measurements of Advertising Effectiveness," Journal of Marketing, (October), 59-63

Lippman, Marcia Z. and Morris W. Shanahan (1973), "Pictorial Facilitation of Paired-Associate Learning: Implications for Vocabulary Training," Journal of Educational Psychology, 64, 216-29.

Lupker, Stephen J. (1979), "The Semantic Nature of Response Competition in the Picture-Word Interference Task," Memory and Cognition, 7, 485-495.

Lutz, Kathy A. and Richard J. Lutz (1977), "Effects or Interactive Imagery on Learning: Application to Advertising," Journal or Applied Psychology, 62, 493-98.

Lutz, Kathy A. and Richard J. Lutz (1978), "Imagery-Eliciting Strategies: Review and Implications of Research," in Advances in Consumer Research, (H. Keith Hunt ed.), 5, 611-20.

Marks, David F. (1977), "Imagery and Consciousness: A Theoretical Review from an Individual Differences Perspective," Journal or Mental Imagery, 2, 275-90.

Nelson, D.L. and D.H. Brooks (1973), "Functional Independence of Pictures and Verbal Memory Codes," Journal of Experimental Psychology, 98, 44-48.

Nelson, D.L. and V.S. Reed (1976), "On the Nature or Pictorial Encoding: Levels of Processing Analysis," Journal of Experimental Psychology: Human Learning and Memory, 2, 49-57.

Nelson, D.L., V.S. Reed, and J.R. Walling (1976), "The Pictorial Superiority Effect," Journal of Experimental Psychology: Human Learning and Memory, 9, 523-78.

Nelson, D.L., V.S. Reed, and C.L. McEvoy (1977), "Encoding Strategy and Sensory and Semantic Interference," Memory and Cognition, 5, 462-67.

Nelson, D.L. (1979), "Remembering Picture and Words: Appearance, Significance, and Name," in Levels of Processing in Human Memory, (L. Cermak and F.I.M. Craik eds.), Hillsdale. New Jersey: Lawrence Erlbaum Associates.

Paivio, Allan (1969), "Mental Imagery in Associative Learning and Memory," Psychological-Reviews 76 ('lay), 241-63.

Paivio, Allan and Dennis Foth (1970), "Imaginal and Verbal Mediators and Noun Concreteness in Paired-Associate Learning: The Elusive Interaction," Journal of Verbal Learning and Verbal Behavior, 9, 384-90.

Paivio, Allan (1971), Imagery and Verbal Processes, New York: Holt, Rinehart and Winston, Inc.

Paivio, Allan and Kelman Csapo (1973), "Picture Superiority in Free Recall: Imagery or Dual Coding," Cognitive Psychology, 5, 176-906.

Paivio, Allan (1978), "A Dual Coding Approach to Perception and Cognition," in 'lodes of Perceiving and Processing Information, (H.L. Pick and E. Saltzman, eds.), Hillsdale New Jersey: Lawrence Erlbaum Associates.

Peterson, M.J. and A. Murray (1973), "Enhancement or Items and Associations," Journal or Experimental Psychology, 101, 82-89.

Peterson, M.J. and S.H. McGee (1974), "Effects of Imagery Instructions, Imagery Ratings, and Number of Dictionary Meanings Upon Recognition and Recall," Journal or Experimental Psychology, 102, 1007-14.

Piaget, J. and B. Inhelder (1973), Memory and Intelligence, New York: Basic Books.

Pylyshyn, Zenon W. (1973), "What the Mind's Eye Tells the Mind's Brain: A Critique of Mental Imagery." Psychological Bulletin, 30, (July), 1-99.

Richardson, Alan (1969), Mental Imagery, London: Routledge and Kegan Paul.

Rips, L.J., E.J. Shoben, and E.E. Smith (1973), "Semantic Distance and the Verification of Semantic Relations," Journal of Verbal Learning and Verbal Behavior, 19, 1-20.

Segal, Sydney (1971), Imagery: Current Cognitive Approaches, New York: Academic Press.

Shepard, Roger N. (1978), "The Mental Image," American Psychologist, (February), 125-36.

Snodgrass, Joan Gay and Anthony Asieghi (1977), The Pictorial Superiority Effect in Recognition Memory," Bulletin of the Psychonomic Society, 10, 1-4.

Snodgrass, Joan Gay (1979), "Toward a Model for Picture and Word Processing," in Processing of Visual Language, Vol.2, (Paul A. Kolers, Merald E. Wrolsted, and Herman Bourne eds.), New York, NY: Plenum Publishing Corp.

Watson, J.B. (1913), "Psychology as the Behaviorist Views It," Psychological Review, 20, 158-177.

Yuille, John C. (1973), "A Detailed Examination of Mediation in PA Learning," Memory and Cognition, 1, 333-42.

Yuille, John C. and Michael J. Catchpole (1977), " The Role of Imagery in Models of Cognition," Journal of Mental Imagery, 1, 171-80.