We are grateful to Roberts (1996) for raising the issue of the potential hazards of using imagery and multiple retrieval attempts in the cognitive interview (CI), especially when the witnesses are children. Our reply attempts to clarify the issues surrounding the use of multiple retrieval attempts in general, and imagery in particular. We also suggest some fruitful areas for future research.
1. The cognitive interview (CI), developed by Geiselman and colleagues (1984) and later revised by Fisher and Geiselman (1992), has been adopted as an invaluable tool in forensic investigative procedures. It has provided interviewers with a means of maximising the amount that a witness can recall whilst operating within a structured and repeated interview setting. Recently, however, evidence has begun to suggest that the CI may be a vehicle for increased errors as well as increased information. The target article by Memon and Stevenage (1996a) urges researchers and practitioners to be aware of the potential disadvantages of the CI, and Roberts' (1996) commentary further extends this cautionary note. In particular, Roberts has focussed on the potential difficulties involved when using imagery, especially when multiple techniques are being used within a single interview, and when witnesses are young. We find Roberts' perspective to be a valuable one and the present reply focusses on some of the more theoretical implications of the evidence that Roberts has reviewed.
2. The encouragement to use imagery as part of the CI was based on a wide literature suggesting that imagery facilitated the retrieval of information. For instance, Paivio's (1971) work clearly showed how one's memory for word pairs was enhanced when those word pairs scored high on concreteness and imagery and when participants were encouraged to picture the two objects together. Using rather less artificial to-be-remembered information, imagery has again been found to be helpful. When encouraged to imagine a crime scene, for example, witnesses' recollections of that scene are typically richer than if no imagery is used (see Bekerian, Dennett, Hill & Hitchcock, 1990; Dobson & Markham, 1993). However, as Roberts notes, the use of imagery is problematic, not least for the fact that huge individual differences exist in people's ability to picture something.
3. What is more worrying, however, is that the instruction to imagine something may actually be quite counter-productive to the generation of an accurate recall. The difficulty lies in maintaining a distinction between what is "actual" and what is "imagined", and failure to do this can lead to source-monitoring errors (Johnson, Hashtroudi & Lindsay, 1993) and cryptomnesia (Brown & Murphy, 1989) - the assimilation of information from another source into one's account of an incident. In a recent review of this literature, Johnson (in press) suggested that imagery may increase the likelihood of source monitoring errors particularly when conditions dictate strong demand characteristics or an "undifferentiated feeling of familiarity" for the information that is retrieved. Accordingly, it seems important to evaluate the conditions under which retrieval attempts are being made before blindly making use of a technique which may or may not be helpful.
4. Both the issues of individual differences and of memory accuracy are ones in which further research is needed. It is possible that these two areas can be investigated in tandem and several instruments exist which can be used to examine the qualitative characteristics of memories of internal/external events (i.e., the source monitoring test - Lindsay & Johnson, 1989; the Memory Characteristics Questionnaire (MCQ) - Mather, Henkel & Johnson, in press). For example, the MCQ elicits ratings from participants of their confidence, perceptual/auditory detail, feelings and reactions, as well as the extent to which they recalled thinking about the event. As such, it can be used to gather information about the individual differences apparent through perceptual, affective and semantic/associative processes. In addition, these tests may force participants to subject their memories to a more stringent evaluation and this may have the effect of minimising the possibility of recall errors.
5. An alternative technique that has been discussed when investigating individual differences and accuracy of recall is to record "remember/know" judgments. "Remember" judgments reflect the conscious recollection of the event while "know" judgments reflect the feeling of familiarity and confidence in a memory in the absence of conscious recollection (see Higham & Roberts, 1996). It is possible that participants instructed to use imagery techniques will show a different pattern of remember/know judgments compared to a control group where no explicit imagery instructions are provided, and again, scrutiny of one's mnemonic processes by forcing a distinction between "remember/know" judgments may buffer the rememberer against recall errors. This hypothesis may be worth exploring in future research studies.
6. Roberts presents evidence in support of three very interesting general effects with imagery-aided recall. First, she highlights the distinction between the effectiveness of spontaneous versus demanded imagery. Roberts cites Markham and Hynes (1993) who have shown that imagery is successful in facilitating memory when it is invoked spontaneously, but when imagery is explicitly demanded, participants had trouble distinguishing what was actually perceived from what was imagined. That is, demanded imagery resulted in more source-monitoring errors.
7. It may be profitable to link these results with the literature on automaticity of skills with expertise. Consider the skill of driving a car; initially a learner-driver has to pay close attention to every aspect of controlling the car. However, as expertise develops, the routine of controlling the car becomes more and more automatic and attention can be given over to other tasks such as reading road signs, switching on the radio, or holding a conversation. In the expert driver, some skills have become "automatic" and are no longer under conscious control. We talk of acting "under autopilot". The important thing for the present discussion is what happens when attention is suddenly diverted back to these automated processes. According to Reason's (1990) classification of human errors, the switching of attention back to ordinarily automatic tasks often results in errors. Specifically, switching of attention from automatic control to conscious control results in a high likelihood of "mistakes" - errors of knowledge, monitoring or planning. In the same way, switching one's attention back to the otherwise spontaneous task of imaging may actually interfere with one's ability to distinguish between what is real and what is imagined. Conscious effort is suddenly involved and knowledge/planning errors are more likely. Perhaps more effort is also involved as participants try harder to create an image in response to an explicit instruction. Consciousness and effort may reasonably be expected to combine and facilitate the creation of a source monitoring error.
8. To our minds, the picture is actually rather more complex than merely whether imagery is spontaneous or demanded. Task demands have a bearing on the effectiveness of imagery as well. For instance, Markham and Hynes (1993) asked high and low imagers to either rate a half-shape for complexity, or to imagine what it would look like whole. They found that high imagers were more confused about which shapes were initially presented whole when they were instructed to use imagery than when instructed to rate complexity. Bekerian and Dennett (1997) report similar results, with demanded imagery instructions producing more errors than when imagery is not demanded. Interestingly, Bekerian and Dennett's effects seemed to be stronger when participants were required to give a spoken recall rather than a written recall: thus imagery instructions gave rise to more correct information but more source monitoring errors as well. Foley, Durso, Wilder and Friedman (1991), however, found the opposite result. In their study, children were required to either describe what an object was used for (function condition), or they were explicitly instructed to imagine it (demanded imagery condition). The results suggested that participants were more able to distinguish objects if they had specifically been told to imagine them than if images had been generated spontaneously (function condition). In this case, demanded imagery appears to be beneficial, and Foley et al. suggest that this may be because of the "cognitive operations" associated with deliberate imagery generation which may help in the task of source monitoring.
9. In addition to the distinction between spontaneous and demanded imagery, a second distinction also besets the imagery literature - explicit/implicit imagery. Much of the literature seems to use these terms interchangeably, however, it may be useful to think of spontaneous/demanded imagery as an indication of whether or not the participant is told to use imagery, and explicit/implicit imagery as an indication of whether or not the participant is told what to image. In these terms, it appears to us that Markham and Hynes's participants used demanded but explicit imagery whereas Foley et al.'s participants used demanded but implicit imagery. A recent study by Henkel and Franklin (in press) outlines the importance of the implicit/explicit distinction in a well-controlled design. They asked participants to either view simple line drawings of objects or to imagine them. The test object either perceptually resembled a second perceived item (e.g., asked to imagine a lollipop and shown a magnifying glass), was semantically related to a perceived item (e.g., imagined a shirt and shown trousers), or the two items were unrelated (e.g., imagined a book and shown an axe). Henkel and Franklin's results suggest that if imagined and perceived items are related, either perceptually or semantically, participants are impaired in their ability to determine the source of the test item. Consequently, there is some evidence to suggest that the nature of the image may affect the effectiveness of imagery as a retrieval aid, and thus, demanded explicit imagery instructions need to be clear about what to ask the participants to imagine. Confusion of perceptually, semantically and perhaps, temporally, related items may account for instances in which an innocent bystander may be mistakenly identified instead of the criminal. Careful use of context reinstatement instructions within the CI may help minimise such errors by guiding the participant to image the appropriate material.
10. Further work is clearly required to tease apart the various factors determining the effectiveness of imagery instructions. However, at present, these mixed results do not allow us to make an unambiguous statement about the use of imagery techniques in a CI. On the one hand, the information about cognitive operations associated with explicit instructions to image should recommend the use of demanded imagery within a CI. However, Markham and Hynes' results, together with the problems of individual differences mentioned above, would appear to caution against the use of demanded imagery until further work has led us to a better understanding of its effectiveness.
11. The second general effect that Roberts discusses involves the differential influence of imagery on perceptive and apperceptive information. Roberts presents evidence from Suengas and Johnson (1988, cited in Roberts, 1996) which suggests that imagery is effective for apperceptive information (thoughts and feelings) but is detrimental for perceptive information. This again, is an interesting finding and one which deserves replication given the obvious difficulty of interpreting thoughts and feelings of actors in a witnessed scene. It may also be profitable to examine whether there is any distinction between the effectiveness of imagery techniques for information that can be classified as central or peripheral to an incident. The distinction between memory for central and peripheral details is well documented and suggests that central event details are more accurately preserved in memory and are more durable over time (see Christianson & Hubinette, 1993; Heuer & Reisberg, 1990; Burke, Heuer & Reisberg, 1992; Stevenage & Sutherland, submitted). It would be interesting to see whether imagery instructions are equally effective for each type of detail or whether the memory for peripheral details may benefit more. Of course, the effectiveness of imagery for central detail information may be restricted by the presence of a ceiling effect. Nevertheless, given that what appears to be peripheral to a witness may actually be of central importance to an investigating officer, the possibility of raising peripheral detail memory to the level of central detail memory appears to offer a fruitful line for further investigation.
12. Finally, Roberts raises the important point of verbal rehearsal on the effectiveness of imagery over repeated retrieval attempts. Roberts reviews evidence which suggests that it is harder to distinguish actual from imagined details when one rehearses the information verbally. This may be due to the possibility that verbal rehearsal itself increases one's confidence that the memory is real. Alternatively, if we view memory as a continuous reconstructive process where a memory is modified by previous remembering instances, then it is easy to see how initial source-monitoring errors resulting from imagery techniques may become cemented into one's recollection of an instance (see Morton, Hammersley & Bekerian's (1985) Headed Records Theory for a theoretical explanation of this).
13. Recent research has raised two further points that have a bearing on the usefulness of imagery as a retrieval technique - plausibility and confidence. For example, there is mounting evidence to suggest that memory for an event may be influenced by the subjective likelihood that the event actually occurred. Sherman, Cialdini, Schwartzman and Reynolds (1985) illustrated the mediating effects of ease of imagery in determining the perceived likelihood that a disease could occur. Participants were asked to imagine a disease with symptoms that were easy or difficult to imagine. When symptoms were judged to be easy to imagine, the likelihood of getting the disease was perceived to be higher than when the symptoms were hard to imagine. More recently, a study of memory for childhood events examined how imagery may lead to source monitoring errors while actually increasing confidence that the memory is accurate (Garry, Manning, Loftus & Sherman, 1996). The authors state that it is not entirely clear why imagery inflates confidence. It could be because imagery techniques increase the accessibility of event information, or produce additional genuine memories (hypermnesia). Alternatively, imagery techniques may enhance confidence about the reality of false memories, that is, it might be supposed that if imagery is easy then the event must have been real. This latter suggestion is in line with Tversky and Kahneman's (1974) "availability heuristic" and suggests that the attribution process, or subjective experience of remembering, is an important factor. Further research on this issue is awaited.
14. Roberts moves on to consider the effectiveness of the CI in general and of imagery techniques with young children in particular (see also Douglas's (1996) commentary on our target article). When considering imagery in terms of the instruction to "change perspective" it may be beneficial to examine children's abilities with reference to Premack and Woodruff's (1978) Theory of Mind. Following a Piagetian influence, the "theory of mind" suggests that young children develop through a stage of egocentrism (2-5 years), in which they cannot take account of the perspective of a third party, to a stage (7-8 years) in which children are aware that another person may have a different perspective to themselves. The instruction to "change perspective" may therefore be especially difficult for young children who have not yet developed a functional theory of mind.
15. In terms of the evidence on young children's use of imagery techniques, the evidence is actually somewhat mixed. Roberts cites studies which demonstrate that imagery may be a useful technique to use with young children: Children can maintain a distinction between actual and imagined events on immediate interview (see Gordon, Jens, Hollings & Watson, 1994, cited in Roberts, 1996). Moreover, any confusion between actual and imagined details appears to decrease over time (see Huffman, Crossman & Ceci, 1996; Poole & Lindsay, 1996, cited in Roberts, 1996). However, there is also evidence to suggest that children find it difficult to distinguish between events they performed and events they imagined performing (Foley & Johnson, 1985) or words said versus words they imagined saying (Foley, Johnson & Raye, 1983). Taking a slightly different approach, Ackil and Zaragoza (1995) demonstrate that participants are not only susceptible to making source monitoring errors in response to suggestion, but that they come to believe that they saw the suggested details. Critically, Ackil and Zaragoza find that the tendency of participants to believe they saw the suggested information increases with younger participants - younger participants are less able to separate the apparent sight of actual and suggested details. It is likely that age differences in the encoding of source-relevant information may have contributed to these effects, however, young children are also less likely to use retrieval strategies spontaneously and this may handicap their apparent source monitoring abilities (see our reply to Douglas' commentary: Memon & Stevenage, 1996b)
16. In summary, the evidence on the sensitivity of children to disruptive effects of imagery techniques is far from clear. What is apparent is that children, as adults, do have the ability to benefit from imagery techniques. It is also clear that children, as adults, have a susceptibility to source-monitoring errors and that imagery may increase this susceptibility perhaps by increasing confidence and plausibility. However, it is by no means clear, at present, that children are differentially disadvantaged by instructions to use imagery to aid recall (but see Ceci, Huffman, Smith & Loftus, 1994; Ceci, Loftus, Leichtman & Bruck, 1994). Consequently, the separation of adults and young children with respect to the use of imagery techniques in the CI may, at present, be based on a false distinction. Clearly, further work in this area is urgently required, and should proceed alongside the current developments of the CI for use with young children. Research, particularly by Karen Saywitz and colleagues (i.e.., Saywitz & Geiselman, in press; Geiselman, 1996) is being carried out to modify the CI for use with very young children. Until we know more, we echo Roberts' call for caution with the use of imagery techniques with both adults and children. Imagery, at present, emerges as a useful technique for situations in which it is possible to check for accuracy of recall. However, in the absence of any other corroborating evidence, we must be careful how imagery is used and how imagery-assisted recall is interpreted.
We are grateful to Kim Roberts for a stimulating set of discussions about the ideas presented here.
Ackil, J.K. & Zaragoza, M.S. (1995). Developmental differences in eyewitness suggestibility and memory for source. Journal of Experimental Child Psychology, 60, 57-83.
Bekerian, D.A. & Dennett, J.L. (1997). Imagery effects in spoken and written recall. In D. Payne & F. Conrad (Eds.), Intersections in Basic and Applied Memory Research, pp 279-290, Mahwah: Lawrence Erlbaum Associates.
Bekerian, D.A., Dennett, J.L., Hill, K. & Hitchcock, R. (1990). Effects of detailed imagery on simulated witness recall. Paper presented at the Second European Conference on Law and Psychology, Nuremberg, Germany.
Brown, A.S. & Murphy, D.R. (1989). Cryptomnesia: Delineating inadvertent plagiarism. Journal of Experimental Psychology: Learning, Memory and Cognition, 15(3), 432-442.
Burke, A., Heuer, F. & Reisberg, D. (1992). Remembering emotional events. Memory and Cognition, 20(3), 277-290.
Ceci, S.J., Huffman, M.L.C., Smith, E. & Loftus, E.F. (1994). Repeatedly thinking about a non-event: Source misattributions among preschoolers. Consciousness and Cognition, 3, 388-407.
Ceci, S.J., Loftus, E.F., Leichtman, M.D. & Bruck, M. (1994). The possible role of source misattributions in the creation of false beliefs among preschoolers. International Journal of Clinical and Experimental Hypnosis, 42, 304-320.
Christianson, S. A. & Hubinette, B. (1993). Hands up! A study of witnesses' emotional reactions and memories associated with bank robberies. Applied Cognitive Psychology, 7, 365-379.
Dobson, M. & Markham, R. (1993). Imagery ability and source monitoring: Implications for eyewitness memory. British Journal of Psychology, 84(1), 111-118.
Douglas, R. (1996). Considering the Witness in Interviews. PSYCOLOQUY 7(21) witness-memory.7.douglas.
Fisher, R.P. & Geiselman, R.E. (1992). Memory enhancing techniques for investigative interviewing: The Cognitive Interview. Springfield III: Charles C. Thomas.
Foley, M.A., Durso, F.T., Wilder, A. & Friedman, R. (1991). Developmental comparisons of explicit versus implicit imagery and reality monitoring. Journal of Experimental Child Psychology, 51, 1-13.
Foley, M.A. & Johnson, M.K. (1985). Confusions between memories for performed and imagined actions: A developmental comparison. Child Development, 56, 1145-1155
Foley, M.A., Johnson, M.K. & Raye, C.L (1983). Age-related changes in confusion between memories for thoughts and memories for speech. Child Development, 54, 51-60.
Garry, M., Manning, C.G., Loftus, E. & Sherman, S. (1996). Imagination inflation: Imagining a childhood event inflates confidence that is occurred. Psychonomic Bulletin and Review, 3, 208-214.
Geiselman, R.E. (1996). On the Use and Efficacy of the Cognitive Interview. PSYCOLOQUY 7(11) witness-memory.2.geiselman.
Geiselman, R.E., Fisher, R.P., Firstenberg, I., Hutton, L.A., Sullivan, S.J., Avetissian, I.V. & Prosk, A.L. (1984). Enhancement of eyewitness memory: an empirical evaluation of the cognitive interview, Journal of Police Science and Administration, 12, 74-79.
Henkel, L.A. & Franklin, N. (in press). Reality monitoring of physically similar and conceptually related objects. Memory and Cognition.
Heuer, F. & Reisberg, D. (1990). Vivid memories of emotional events: The accuracy of remembered minutiae. Memory and Cognition, 18(5), 496-506.
Higham, P.A. & Roberts, W.T. (1996). Analyzing States of Consciousness During Retrieval as a Way to Improve the Cognitive Interview. PSYCOLOQUY 7(17) witness-memory.4.higham.
Johnson, M.K. (in press). Source monitoring and memory distortion. In L Squire & D Schacter (Eds.), Biological and Psychological Perspectives on Memory and Memory Disorders. New York: American Psychiatric Press.
Johnson, M.K. Hashtroudi, S. & Lindsay, D.S. (1993). Source monitoring. Psychological Bulletin, 114, 3-28.
Lindsay, D.S. & Johnson, M. (1989). The eyewitness suggestibility effect and memory for source. Memory and Cognition, 17, 349-358.
Mather, M., Henkel, L. & Johnson, M. (in press). Evaluating characteristics of false memories: Remember/know judgments and memory characteristics questionnaire compared. Memory and Cognition.
Memon, A. & Stevenage, S.V. (1996a). Interviewing Witnesses: What Works and What Doesn't? PSYCOLOQUY 7(6) witness-memory.1.memon.
Memon, A. & Stevenage, S.V. (1996b). A Consideration of Individual Differences and Imagery Ability Amongst Cognitive Interviewees. PSYCOLOQUY 7(24) witness-memory.10.memon.
Morton, J., Hammersley, R.H. & Bekerian, D.A. (1985). Headed records: A model for memory and its failure. Cognition, 20, 1-23.
Paivio, A. (1971). Imagery and verbal processes. New York: Holt, Rinehart and Winston.
Premack, D. & Woodruff, G. (1978). Does the chimpanzee have a "theory of mind"? Behavioral and Brain Sciences, 4, 515-526.
Reason, J. (1990). Human Error. Cambridge: Cambridge University Press.
Roberts, K. (1996). How Research on Source Monitoring Can Inform Cognitive Interview Techniques. PSYCOLOQUY 7(44) witness-memory.15.roberts.
Saywitz, K.J. & Geiselman, R.E. (in press). Interviewing the child witness: Maximising completeness and minimising error. In S. Lynn (Ed.), Truth and Memory. New York: Guilford.
Sherman, S., Cialdini, R., Schwartzman, D. & Reynolds, K. (1985). Imagining can heighten or lower the perceived likelihood of contracting a disease: the mediating effects of ease of imagery. Personality and Social Psychology Bulletin, 11, 118-127.
Stevenage, S.V. & Sutherland, R. (submitted). The good, the bad and the forgotten: Are positive memories equivalent to negative ones?
Tversky, A. & Kahneman, D. (1974). Judgments under uncertainty: Heuristics and biases. Science, 85, 1124-1131.