Latent inhibition as a measure of learned inattention: some problems and solutions

https://doi.org/10.1016/S0166-4328(97)02307-3Get rights and content

Abstract

The latent inhibition (LI) paradigm has been used to assess attentional dysfunction in various pathological groups. The rationale is based on the assumption that passive stimulus exposure results in the acquisition of an inattentional response to that stimulus. Consequently, compared to a novel stimulus in the same new learning situation, the preexposed stimulus is at a disadvantage. It is argued that methodological and conceptual problems in constructing procedures and designs have created obstacles in relating disrupted LI to psychopathology. Specifically, issues associated with within- and between-subject designs, dichotomous dependent variables, ceiling effects, converging operations, and possible mis-attribution of the LI effect are addressed. Designs and data from several new human-LI paradigms, with normal, de novo Parkinson, and schizophrenic subjects are examined. Results from a multi-condition, within-subject visual search procedure suggest that LI, heretofore attributed only to a deficit in the stimulus preexposed group, may, in part, be due to enhanced performance in the nonpreexposed group. Implications for the design and interpretation of LI experiments, particularly with pathologic groups are discussed.

Introduction

The latent inhibition (LI) paradigm, originally developed within the framework of classical learning theory, has recently been applied to the assessment of attentional dysfunction, first in animal-models of schizophrenia, then to pathological groups, including hyperactive children, Parkinson patients, and schizophrenics, to schizotypal normals, and to normal humans under different dopaminergic-related drug regimens (for reviews, see Refs. 10, 28). The rationale for employing LI to assess attentional processes is based on the assumption that non-reinforced, incidental, stimulus exposure results in the acquisition of an inattentional response to that stimulus, and that this inattentional response must be countered before any new associations to that stimulus can occur. Consequently, at least at the beginning of the test phase, the preexposed stimulus is at a disadvantage compared to a novel one.

The basic LI phenomenon, reported extensively in the animal literature, has proven to be ubiquitous, yet theoretically puzzling [20]. In human subjects, for whom the acquisition of LI requires that the preexposed stimulus be presented such that it is not the focus of attention, the theoretical structure of the LI effect is even more problematic. The absence of an adequate theory of LI has been a major impediment in relating disrupted LI to psychopathology.

The problems for creating a good theory of LI can be grouped into two overlapping categories, methodological and conceptual. Methodological issues in human LI research include: dichotomous data, ceiling effects, inappropriate data analysis, individual versus group data, and the needs for within-subject designs and parametric studies. Among the conceptual issues, there is the need to develop experimental designs which allow one to choose between alternative sources of differences between the stimulus preexposed (PE) and nonpreexposed (NPE) groups (i.e. is the difference in performance to be attributed to poorer performance in the PE group, or better performance in the NPE group, or both).

A similar problem arises in regard to the masking task, a necessary condition for producing LI in adults in instrumental learning tests. The masking task, administered in the preexposure phase, is used in the human LI literature to denote the task to which the subject is instructed to attend. The word `masking' is meant to suggest that, during preexposure, the significance of the critical, to-be-tested stimuli, is not accessible to the subject. In this regard, designs are needed which can differentiate between failure to maintain attention to the masking task (i.e. distraction induced by the preexposed irrelevant stimuli) and an opposing process which would interfere with any processing of the preexposed irrelevant stimuli. Although the loss of LI in high schizotypals and schizophrenics has been attributed to elevated distractibility in those groups, given the experimental designs that have been used, complete failure to process the incidental stimuli remains a viable alternative, or even the possibility that LI attenuation is a result of different processes in different schizophrenia sub-groups.

Finally, there are two more concerns, general, but still theoretically critical: the identification and use of converging operations for the hypothetical variable of inattention; and, a theoretically guided effort to explore the similarities and differences between animal and human LI. The former has been dealt with in several studies which have examined the relationship between LI and pre-pulse inhibition (PPI; reduction in the magnitude of the startle response when the eliciting stimulus is preceded by 30–500 ms by a weaker stimulus 2, 7, 17, 38). However, the presumed theoretical link between PPI and LI is not entirely clear.

The present paper addresses several of these problems, specifically as they relate to LI experiments with special populations. In particular, four topics will be examined: (1) between and within-subject designs; (2) dependent variables; (3) convergent operations for LI-related attention; and (4) the possibility that the LI effect is, at least in part, mis-attributed to the stimulus preexposed group. These problems are addressed in a historical review and critique of between-subject LI designs, and in a discussion of the failures of such designs to differentiate between normal and de novo PDs. In addition, data is presented from a within-subject visual LI study with normals, and from a new within-subject visual search procedure. The latter was designed to provide a converging operation for the attentional manipulation in the typical LI experiment. The new procedure not only uncovers significant differences between PDs and normal controls, but also provides data for assessing a possible mis-identification of the source of the LI effect.

Section snippets

History

The typical LI experiment consists of two stages, stimulus preexposure and test, and two groups. In the preexposure stage, one group is preexposed to the to-be-critical stimulus (PE). The second group, otherwise treated identically to the first one, is not given such preexposure (NPE). In the second, or test, stage, both groups are given a learning task in which the preexposed stimulus is the critical stimulus. With this procedure, LI is defined as the difference between the groups in stage-two

Normal subjects

To overcome these problems, we developed a within-subject procedure in which subjects were preexposed to attended and unattended visual stimuli (different nonsense figures). The subjects were then administered a visual search task, in which they had to identify the presence or absence of a unique figure, the target, amongst a group of 19 other homogeneous figures, the distractors [29]. On any given trial, the target could be present or absent, and if present it could be either the attended or

Possible mis-attribution of the LI effect

The problem can be characterized in the following way. Given two groups, and a significant difference in some dependent measure, to what is the difference in scores attributed? Assuming that the groups are drawn from the same population, we almost always attribute the effect to the so-called treated group. Although this may often be a reasonable conclusion, labelling one group as treated and the other as untreated carries with it assumptions that may not be valid. The problem is particularly

Summary and conclusions

A number of problems have been identified which impede the search for a relationship between dysfunctional attentional processes, as represented by atypical LI scores, and psychopathology. Several solutions have been suggested. In particular, there are considerable benefits to be gained by using within-subject designs with dependent variables whose scores are normally distributed, as opposed to the present practice of using between-subject designs with dichotomous distributions of scores.

A

Acknowledgements

This research was supported by the Israel Science Foundation administered by the Israel Academy of Sciences and Humanities. The author thanks Regina Dressler, Nehemia Friedland, Nina Gilead-Roelofs, Yonathan Goshen-Gotstein, and Armonit Roter for helpful comments on this paper.

References (40)

  • Allan, L.M., Williams, J.H., Wellman, N.A., Tonin, J., Taylor, E., Feldon, J. and Rawlins, J.N.P., Effects of tobacco...
  • Bakshi, V.P., Geyer, M.A., Taaid, N. and Swerdlow, N.R., A comparison of the effects of amphetamine, strychnine and...
  • Baruch, I., Hemsley, D.R. and Gray, J.A., Differential performance of acute and chronic schizophrenics in a latent...
  • Baruch, I., Hemsley, D.R. and Gray, J.A., Latent inhibition and `psychotic proneness' in normal subjects, Pers. Indiv....
  • De la Casa, L.G., Ruiz, G. and Lubow, R.E., Latent inhibition and recall/recognition of irrelevant stimuli as a...
  • Dressler, R., Melamed, E., Kaplan, O. and Lubow, R.E., Visual search in de novo Parkinson's disease patient and normal...
  • Ellenbroek, B.A., Geyer, M.A. and Cools, A.R., The behavior of APO-SUS rats in animal models with construct validity...
  • Gibson, E.J., Principles of Perceptual Learning and Development, Appleton-Century-Crofts, New York,...
  • Ginton, A., Urca, G. and Lubow, R.E., The effects of preexposure to a nonattended stimulus on subsequent learning:...
  • Gray, J.A., Feldon, J., Rawlins, J.N.P., Hemsley, D.R. and Smith, A.D., The neuropsychology of schizophrenia, Behav....
  • Gray, N.S., Hemsley, D.R. and Gray, J.A., Abolition of latent inhibition in acute, but not chronic, schizophrenics,...
  • Gray, N.S., Pilowsky, L.S., Gray, J.A. and Kerwin, R.W., Latent inhibition in drug-naive schizophrenics: relationship...
  • Hall, G., Perceptual and Associative Learning, Clarendon Press, Oxford, England,...
  • Johnston, W.A. and Hawley, K.J., Perceptual inhibition of expected inputs: The key that opens closed minds, Psychonom....
  • Johnston, W.A., Hawley, K.J. and Farnham, J.M., Novel popout: Empirical boundaries and tentative theory, J. Exp....
  • Killcross, A.S., Robbins, T.W., Differential effects of intra-accumbens and systemic amphetamine on latent inhibition...
  • Lipp, O.V., Siddle, D.A.T. and Arnold, S.L., Psychosis proneness in a non-clinical sample II: A multi-experimental...
  • Lipp, O.V. and Vaitl, D., Latent inhibition in human Pavlovian differential conditioning: Effect of additional...
  • Lovibond, P.F., Preston, G.C. and Mackintosh, N., Context specificity of conditioning, extinction, and latent...
  • Lubow, R.E., Latent Inhibition and Conditioned Attention Theory, Cambridge University Press, New York,...
  • Cited by (0)

    View full text