Constructs and Measures of Individual Differences in Imagery Processing: a Review

Deborah J. MacInnis, University of Arizona
ABSTRACT - This paper reviews characteristics of measures designed to test individual differences in imagery processing. Scales in three domains are reviewed: (1) measures of imagery ability (2) measures of processing style and (3) measures of daydreaming/fantasy content and frequency. Awareness of the content and psychometric properties of these scales is important for future research on imagery.
[ to cite ]:
Deborah J. MacInnis (1987) ,"Constructs and Measures of Individual Differences in Imagery Processing: a Review", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 88-92.

Advances in Consumer Research Volume 14, 1987      Pages 88-92


Deborah J. MacInnis, University of Arizona


This paper reviews characteristics of measures designed to test individual differences in imagery processing. Scales in three domains are reviewed: (1) measures of imagery ability (2) measures of processing style and (3) measures of daydreaming/fantasy content and frequency. Awareness of the content and psychometric properties of these scales is important for future research on imagery.


Imagery is a process by which sensory information is represented in working memory (MacInnis and Price 1986). In a consumer context, the use of imagery in processing both visual and verbal information can enhance memory for product related information (Rossiter 1982; Childers and Houston 1983, 1984; Kisielius 1982; Lutz and Lutz 1977 Rossiter and Percy 1983). Imagery may also serve as a form of cognitive elaboration (Greenwald and Leavitt 1984), and play an important role in forming product attitudes (Rossiter and Percy 1978; Percy and Rossiter 1980; Kisielius and Sternthal 1984, 1986). Hypotheses concerning the role of imagery in forming purchase related expectations and its subsequent impact on post-purchase satisfaction have also been advanced (MacInnis and Price 1986). Finally, imagery has been hypothesized to accompany hedonic consumption activities (Hirschman and Holbrook 1982; Holbrook and Hirschman 1982).



As research on imagery continues, identifying adequate measures of imagery processing becomes particularly important. One area that has generated considerable interest elsewhere is the study of individual differences in imagery processing. Three areas of interest are studies of individual differences in (1) imagery ability (2) imagery processing style and (3) imagery content. These areas are of interest both for their direct and moderating roles on the processing on consumption-related information. This paper reviews measures of individual differences that assess these dimensions. This paper's focus is on the psychometric properties of individual scales, specifically their reliability, convergent validity, and discriminant validity. These empirical criteria, while necessary contributors to the scales' construct validity, -should not. however, be taken as evidence of their construct validity. Additional nonempirical criteria such as the logical relationships between constructs and measures in a network of relationships must also be considered (Peter and Churchill 1986). Table 1 summarizes the paper and serves as a framework for discussion.


Measures of Imagery Vividness

Imagery vividness refers to the ability with which one can evoke clear images (Harks 1972). Imagery vividness has a known moderator effect on incidental learning (see Forisha 1978, review), discriminative reaction time (Gur and Hilgard 1975), and emotional and physiological arousal (Drummond. White and Ashton 1978, Grossberg and Wilson 1968). Two scales commonly used to assess individual differences in imagery vividness are Betts Questionnaire Upon Mental Imagery and Marks Vividness of Visual Imagery Questionnaire.

Betts Questionnaire Upon Mental Imagery. The Betts Questionnaire Upon Mental Imagery (QMI) assesses imagery vividness in each of seven sensory modalities (visual, auditory, cutaneous, kinaesthetic, gustatory, olfactory and organic). Five items assess vividness in each modality. For example, respondents rate the visual clarity of the sun as it sinks below the horizon (visual modality). the smell of paint (olfactory modality), the touch of fur (cutaneous modality), the taste of salt (gustatory modality), the muscular feeling of running upstairs (kinaesthetic modality). the sound of a mewing cat (auditory modality) and the sensation of hunger (organic modality). Using a seven point response scale, respondents indicate the extent to which images are (1) perfectly clear and as vivid as the actual experience to (7) no imagery present. Initially developed as a 150 item questionnaire (Betts 1909). the scale was modified and reduced to 35 items by Sheehan (1967) (see Sheehan 1967 for a reproduction of the scale).

As Table 1 indicates. the internal consistency of the scale is fairly high, with reliability coefficients ranging from .91 (Westcott and Rosenstock 1976) to .95 (Juhasz 1972). Women tend to score consistently higher than men (Sheehan 1967; White, Ashton and Brown 1978; Durndell and Wetherick 1978). Moreover test-retest reliability declines substantially as time between the tests increases. Some (Divesta et al 1971; Durndell and Wetherick 1976) have also reported social desirability effects with this scale. particularly with males (Richardson 1978). Inconsistent results have also been found in determining the factor structure of the scale. Richardson (1969) proposes that the scale measures a single factor, however, White, Ashton and Law (1976) identified several f actors . One was comprised of sensations that are chemically activated (taste and smell). while another was comprised of mechanically activated sensations (touch and hearing). A third factor labeled suggestibility and social desirability was also found. Use of the scale needs to be made with the precise needs of the investigator in mind. The QMI is a scale of multisensory imagery,-however it is often (improperly) used to predict performance on visual tasks.

Marks VVIQ. While the Betts QMI assesses imagery vividness as a multisensory construct, Marks Vividness of visual imagery Questionnaire IQ) (Marks 1973) is designed to assess the vividness of visual images. Respondents use a five point scale to indicate the extent to which they can clearly visualize four variations on four familiar scenes. For example, one question asks respondents to visualize a shop they often frequent. Respondents imagine (1) the overall appearance of the shop from the opposite side of the road (2) a window display with individual items for sale (3) the color, shape and details of the door (4) images of entering the shop. going to the counter, having the salesperson's assistance and exchanging money. Marks (1973) originally advocated that respondents complete the scale twice, once with their eyes open and once with their eyes closed. However Dowling (1973) found that the 2 methods produced identical scores. The scale correlates significantly with the Betts QMI (r-.55; Rossi and Fingeret 1977). suggesting that both scales converge on the vividness dimension.

Gur and Hilgard (1975) believe the process of completing this scale is more interesting than the QMI. Moreover, questions for this test may generate more vivid images than questions for the Betts scale because they deal with familiar (and hence highly imageable) scenes. Questions also provide respondents with a richer set of cues which may make imagery more concrete and hence more vivid. On the other hand, because questions ask subjects to hold in mind and alter images in a sequential format, the scale may also tap different imagery ability construct; imagery control (Gur and Hilgard 1975). Thus, the discriminant validity of the scale may be problematic.

The high internal consistency of the VVIQ has been demonstrated repeatedly (McKelvie and Gingras 1974; Marks 1973; Rossi 1977; Dowling 1983; Childers, Houston and Heckler 1985) (see Table 1). Rossi (1977) conducted a principle component analysis on scale items and found that a single factor accounted for more than 50% of the variance. The test-retest reliability of the scale is adequate, although not high, with reported reliabilities ranging from .67 for 3 weeks to .73 for 7 weeks (McKelvie and Gingras 1974 and Rossi 1977, respectively). Interestingly, Rossi (1977) found that although test-retest reliability for the entire scale was .73, the coefficient increased to .86 when three items were removed. He suggests that the scale may be reliable for most items, but unreliable for a few. The scale has also been found to correlate moderately with the Crowne and Marlowe (1964) scale of social desirability (Childers, Houston and Heckler 1985).

Tests of Imagery Control

Imagery control refers to the extent to which one can manipulate, transform. and hold images in mind at will. It is distinct from the vividness construct since one can presumably imagine stimuli vividly without being able to manipulate thee at will (Richardson 1972). The control of imagery has been shown to influence a number of cognitive processes. including mental rotations (Ernest 1977, review), and performance on divergent thinking tasks (Forisha 1978). Gordon's test of Visual Imagery Control (VIC) is a frequently used measure of imagery control.

Gordon's Test of Visual Imagery Control. Gordon (1949) developed a 12 item scale assessing individuals ' abilities to control their imagery. Respondents use a yes-no response format to indicate whether or not they can "see" various modifications on images of an automobile. For example, respondents are asked to picture a car standing at a garden gate. They are then asked to picture the same car lying upside down. Nest they are asked to see the car running along a road. Twelve items constitute the scale. Due to problems inherent in the initial scale, Richardson (1969) developed 8 modified version, changing the response format from a 2 to a 3 point format and reducing the scale from 12 to 11 items.

The internal consistency of the VIC has been demonstrated repeatedly (see Table l), and the test-retest reliability has been consistently found to be high. Westcott and Rosenstock (1976), however found relatively low reliability coefficients, and hypothesized that the scale may not be unidimensional. Most of the variance in imagery control is attributed to those items that relate to physical manipulation or movement (Morrison and White 1984). A new scale which focuses more specifically on these physical movement dimensions is likely to have a simpler, more coherent factor structure. Although few studies have examined the social desirability effects of the Gordon scale. a recent study by Childers et al (1985) found that the scale was uncorrelated (r=.05) with the Crowne and Marlowe (1964) social desirability scale. Empirically, the discriminant validity between the Gordon scale and the QMI has been demonstrated (r=.22, Morris and Gale 1974; r=.36, Starker 1974). However, the discriminant validity between the Gordon scale and the VVIQ is less clear. Childers, Houston and Heckler (1985) and McKelvie and Gingras (1974) both found that the Gordon scale was highly correlated with the W IQ (r-.56, and r=.67, respectively). These results confirm Gur and Hilgard's (1975) hypothesis that the W IQ taps both ability and control dimensions.


While the above scales were designed to assess individual differences in abilities to imagine. several scales have been designed to assess individual differences in individuals' preferences and ProPensities for using imagery. Paivio (1971) and Richardson (1978) hypothesize that individual differences exist in individualize' tendencies and preferences for processing information visually or verbally. Richardson (1978) conceptualizes a visualizer as one who prefers literal encoding (e.g., preferences for seeing, feeling physical features in relating information to prior knowledge), and one who uses daydreams or mental pictures as the accompaniment to thinking. A verbalizer is conceptualized as one who prefers linguistic encoding (e.g., reading how to do something), and one who uses inner speech as a conscious accompaniment to thinking. Processing style has also been found to influence recall and recognition of verbal and visual material (Hiscock 1976, Childers et al 1985).

The Visualizer/Verbalizer Questionnaire. Richardson (1977) developed a 15 item scale called the VVQ designed to differentiate verbalizers from visualizers. Respondents use a true-false format to indicate the extent to which visual vs. verbal processes are normally engaged in various activities. The questionnaire asks such questions as: "I enjoy doing work that requires the use of words" (verbalizer) and "my daydreams are sometimes so vivid I feel as though I actually experience the scene" (visualizer).

Unfortunately, internal consistency of the scale is relatively low. Cronboch's alpha ranges from .54 to .66 (Childers, Houston and Heckler 1985). Childers, Houston and Heckler (1985) hypothesize that several factors may account for the low internal consistency. First, the visualizer/verbalizer construct is conceptualized as a preference and a propensity dimension (not an ability dimension), however the visualizer items more adequately measure ability than preference. The mixture of items relating to preference and ability makes the scale multidimensional, and reduces its internal consistency. Second, the true-false response format restricts variance of items which in turn influences the scale's reliability. Although only a few tests of the scale's test-retest reliability have been conducted, wide variations in stability are reported. Richardson (1977) reported test-retest reliabilities of .91 after 1 week. Warren and Good (1979), however, found test-retest to be quite low after 3 weeks, particularly for women (: =.29) .

The Style of Processing Questionnaire. Recognizing the deficiencies inherent in the VVQ, Childers, Houston and Heckler (1985) developed a new measure of processing style, the Style of Processing Questionnaire (SOP). The 22 item questionnaire substantially modified VVQ visualizer items so that they more adequately reflect the preference dimension. In addition, the authors added a 4 point true-false response format. Items on this scale include such statements as: "I like to daydream, "I like to think of synonyms for words," and "I find it helps to think in terms of mental pictures when doing many things." Internal consistency of the scale was substantially improved over the VVQ (Cronbach's alpha = .81 to .86) (Childers, Houston and Heckler 1985). Moreover, a confirmatory factor analysis demonstrated the independence of visual and verbal components. The discriminant validity of the scale was also demonstrated, as the scale was shown to have zero correlations with tests of imagery vividness (r=.01 with VVIQ) and tests of imagery control (r--.03 with VIC). In addition, correlations between the SOP and the (1964) Marlowe Crowne Social Desirability scale were negligible (r=-.00). Partial predictive validity of the scale was also demonstrated. Verbalizers had higher recall and recognition scores of advertisements (r=.34 and .31 respectively) compared to visualizers. Childers et al (1985) note that although one may score high on both visualizer and verbalizer dimensions, the term processing style is used to reflect a preference for one mode over the other. Thus one who scores either very high or very low on both dimensions has no preference for one mode over the other.

Tests of Daydreaming and Fantasy Content and Frequency

A final category of scales are general inventories of imagery content and frequency. These inventories measure a variety of constructs. including the vividness of daydreams and fantasies, the frequencies of daydreams and fantasies, and frequency, emotional involvement in imagery, and control over daydreams. Hence they have subscales which tap both vividness and control dimensions, as well as more general imagery content.

Wilson-Barber Inventors. Wilson and Barber (1978; 1983) developed a series of interview questions designed to assess imaginative involvement in fantasy. A modified version of the inventory was developed by Myers (1983) and is geared specifically for children and adolescents. Using a yes-no format, individuals indicate the extent to which they experience fantasies and feeling-related fantasies. Items assess both current imaginal experiences as well as childhood experiences. Using a large sample (N--1337) Myers (1983) found internal consistency of the inventory to be .89. Test-retest reliability for a subset of the sample was also found to be high (r=.87; N=104), however the period between testing was relatively short (5 hours to l week). An exploratory factor analysis, yielded a complex factor solution. Fourteen factors were identified from the inventory. These include "living in a mike believe world," "vivid pretending," "vivid memories," and "mental adventures." Many of the factors identified by this inventory may be relevant for identifying individuals who enjoy fantasy imagery through engaging in hedonic consumption activities. However, the extremely complex factor structure of this relatively small inventory suggests that further refinements are needed. Caution must also be exercised in assessing the validity of the inventory. Specifically it is not clear whether respondents, particularly children, can accurately and adequately report on inner experiences from their past.

The Imaginal Process Inventory. Several general inventories assess individual differences in a more specific part of imagery; daydreaming. Daydreaming is conceptualized as spontaneous and often task-irrelevant inner dialogue involving multisensory images. The content of these inner dialogues often reflect "current concerns" or "unfinished business." Singer and Antrobus (1963) developed a massive 344 item inventory (29 subscales) designed to assess individual differences in daydreaming content, frequency, vividness, controllability, and general feelings about daydreaming. A modified version of the scale was developed by Singer and Antrobus (1972). A shortened version containing 45 items (7 subscales) was developed by Huba et al 1982. Internal consistencies for the scales are generally quite good, with mean internal consistency scores approximately .81. Test-retest reliabilities are also adequate (Cundiff and Gold 1979) (See Table 1).

The initial IPI was designed to relate individual differences in daydreaming to individual differences in personality. Three "types" of individuals have consistently been identified using IPI and personality measures: (1) a positive vivid daydreamer (2) a guilty-dysphoric daydreamer and (3) a mind-wandering, highly distractible daydreamer. The latter two types indicate a psychologically maladaptive personality type (Singer and Antrobus 1972). The majority of adults studied report engaging in some sort of daydreaming every day and most of these are classified as positive-vivid daydreamers (Singer & McGaven 1961). Positive vivid daydreamers most frequently report daydreams of the future (anticipatory imagery). positive reactions to daydreams, visual imagery, and sensual daydreams. Daydreaming frequency and content change across the life-span. (Giambra 1977). Sensual and anticipatory daydreams are greatest in the early years of adulthood, while daydreams related to past memories predominate as people age.

An interesting and untapped dimension of daydreaming in a consumption contest is the use of anticipatory imagery. According to Singer and Antrobus. anticipation of the future is an important and prevalent dimension of daydreaming for many adults. The extent to which anticipation of the future includes consumption experiences is unknown. Moreover, little is known about the types of marketing stimuli most likely to produce daydreams. An interesting extension the choice literature is the study of daydreaming patterns in solving consumption problems. Imagery of this sort represents a form of spontaneous. internal search. perhaps accompanied by emotional experiences.

The Imagery Processing Scale. While the IPI has properties that make it a useful inventory for studying daydreaming, it may be less useful in a consumer contest since it was initially developed to assess both normal and pathological functioning. MacInnis (1985) developed an inventory designed to measure imagery and daydreaming more specific to a consumer contest. The Imagery Processing Scale is a 45 item inventory modified and extended from the IPI. The inventory assesses the use of fantasy imagery in owning products and experiencing events. the use of anticipatory imagery. the use of imagery as a substitute for consumption experiences. the vividness of imagery, the use of imagery to solve problems. the use of imagery in reducing needs for stimulation. and the perceived general utility of imagery. Items include. "I like to read and daydream about exciting vacations even though I'll probably never take them." "Compared to most people I believe my dreams and fantasies are much more vivid." "I sometimes dream about owning a lot of expensive things and i,<4,pe all the status they would give me," "I often daydream about all the great things I am going to buy in the future." Respondents use a 1-7 point scale to indicate their level of agreement with each statement.

MacInnis (1985) found internal consistency reliability for individual scales of the inventory to be relatively high (Cronbach's alpha ranges from .71 to .89 in two studies). Partial predictive validity of the scale was established using a scale of optimum simulation level. delay of gratification. and risk taking as criteria.


Many of the scales and inventories reviewed here have adequate psychometric properties. However. several factors should be considered when using these scales in a consumer behavior contest. First. most were designed to assess general imagery processes. The use of these scales to predict specific consumer behaviors requires careful consideration. Relatedly. state characteristics may have an equal or greater influence in imagery processing than trait characteristics in some consumption contests. For example. an individual with a highly integrated. well developed knowledge base for a specific consumption domain may be able to generate more vivid imagery for that domain than another individual. despite the fact that both have equivalent scores on a general imagery vividness scale. In sum. the results presented here provide preliminary. not confirmatory evidence for the construct validity of these scales in a consumption contest.

These precautions should not. however. discourage the use of these individual difference scales in a consumption context. Several studies have demonstrated that these individual difference variables have a main and moderating effect on the processing of-consumption-related information (Childers and Houston 1984; Childers et al 1985. Rossiter and Percy 1978). Moreover, important and interesting questions can be raised about the effects of various combinations of imagery ability, processing style and content on consumption outcomes. For example. individuals with vivid. controllable imagery oriented toward the future may make considerable use of imagery in planning purchases; using imagery as a mode of visual problem solving in selecting among product alternatives. Those with vivid imagery oriented toward the past may be able to prolong the pleasure of post consumption activities. Understanding how individual differences in imagery processing interact. influence. and moderate information processes represents an important direction for future research on imagery.


Betts. G. H. (1909). The Distribution and Functions of Mental Imagery, Teachers College, Columbia University, New York, NY.

Childers, T. and M. Houston (1983). "Imagery Paradigms for Consumer Research: Alternative Perspectives from Cognitive Psychology," in A. Tybout and R. Bagozzi (eds)., Advances in Consumer Research, Vol. 10. 59-64.

Childers. T. and M. Houston (1984), "Conditions for a Picture Superiority Effect on Consumer Memory," Journal of Consumer Research. 11 (Sept). 643-655.

Childers. T.. M. Houston and S. Heckler (1985). "Measurement of Individual Differences in Visual vs. Verbal Information Processing." Journal of Consumer Research. 12 (September). 125-134.

Crowne D. and D. Marlowe (1964), The Approval Motive, MY: John Wiley and Sons.

Cundiff, G. S. and R. Gold (1979), "Daydreaming: A Measurable Concept," Perceptual and Motor Skills, 49, 347-353.

Divesta, F. I., G. Ingersoll and P. Sunshine (1971), "A Factor Analysis of Imagery Tests," Journal of Verbal Learning and Verbal Behavior, 10, 471-479.

Dowling, P. (1973), "Validation of the Vividness of Visual Imagery Questionnaire," Unpublished Manuscript, Psychology Department, University of Queensland.

Drummond, P., K. White and R. Ashton (1978), "Imagery Vividness and Habituation Rate," Psychophysiology, 15 (May), 193-195.

Durndell, A. J. and N. E. Wetherick (1976), "The Relation of Reported Imagery to Cognitive Performance," British Journal of Psychology, 67, 501-506.

Ernest, C. (1977), "Imagery Ability and Cognition: A Critical Review," Journal of Mental Imagery, 2, 181-216.

Evans, I. M. and W. S. Kamemoto (1973), "Reliability of the Short Form of Betts Questionnaire Upon Mental Imagery: Replication," Psychological Reports, 33, 281-282.

Forisha, B. L. (1978), "Imagery and Creativity: Review and Speculations," Journal of Mental Imagery, 2, 209-238.

Giambra, L. M. (1977), "A factor Analytic Study of Daydreaming, Imaginal Processing and Temperament: A Replication on an Adult Male Life-Span Sample," Journal of Gerontology, 32, 675-680.

Gordon, R. A. (1949), "An Investigation Into Some of the Factors that Favor the Formation of Stereotyped Images," British Journal of Psychology, 39, 156-167.

Greewald, A. G. and C. Leavitt (1984), "Audience Involvement in Advertising: Four Levels," Journal of Consumer Research, 11 (June), 581-592.

Grossberg, J. M. and H. F. Wilson (1968), "Physiological Changes Accompanying the Visualization of Fearful and Neutral Situations," Journal of Personality and Social Psychology, 10, 124-133.

Gur, R. C. and E. R. Hilgard (1975), "Visual Imagery and the Discrimination of Differences between Altered Pictures Simultaneously and Successively Presented," British Journal of Psychology, 66 (3), 341-345.

Hilgard, J. R. (1978), Personality and Hypnosis: A Study of Imaginative Involvement, Chicago, IL: University of Chicago Press.

Hirschman, E. C. and M. B. Holbrook (1982), "Hedonic Consumption: Emerging Concepts, Methods and Propositions," Journal of Marketing, 46 (Summer), 92-101.

Hiscock, M. (1986), "Effects of Adjective Memory on Recall from Prose," Journal of General Psychology, 94. 295-299.

Holbrook, M. B. and E.C. Hirschman (1982), "The Experiential Aspects of Consumption: Consumer Fantasies, Feelings and Fun," Journal of Consume Research, 9 (September), 132-140

Juhasz, J. B. (1972), "On the Reliability of Two Measures of Imagery," Perceptual and Motor Skills, 35, 874.

Remaining references available upon request.