This thesis investigated the personality trait of absorption (Tellegen & Atkinson; 1 974) using reaction-time performance on an analog of the Stroop task. The Stroop task examines the degree to which an incongruent colour word interferes with the naming of the ink colour in which that word is printed (e.g., saying "green" to the word blue in green ink). Review of the evidence traces several lines of research on absorption, and related phenomena such as hypnosis and imagery, and focuses concern on the questions of whether trait-related performance remains relatively impervious of situational and task-adaptive factors, and whether trait capacity interacts with these factors in an (as yet) unspecified manner.
In pursuit of these questions, the present program of research sets out a series of experiments, designed cumulatively to offer successive refinements in hypotheses and methodological procedures. This research begins with the findings of Qualls and Sheehan (1979; 1981 a, 1981 b, 1981c) who have suggested that attentional deployment mediates trait-related performance, though alternative explanations have been offered for their data (Tellegen, 1981). The results of the present research suggest that a synthesis of these viewpoints is possible, and a model of absorption is developed to account for trait-related differences in performance found in this research.
The first experiment was conducted to examine whether the trait of absorption is associated with a predisposition for sustained imaginal attending. An analog of the Stroop task, the Response-Change task, was designed to test this hypothesis. The Response-Change task uses the normal Stroop interference effect, where an incongruent word interferes with colour-naming, but in addition, alternates blocks of these colour-naming trials with blocks of word-naming trials, so that the target stimulus is quickly changed from one block of trials to the next. . It was argued that if the target stimulus was quickly changed, then the predisposition for sustained attending on the previous stimulus would increase the time required to attend to the new stimulus.
Consistent with this hypothesis, results indicated that performance was correlated negatively with absorption capacity; high-absorption individuals demonstrated slower performance than low-absorption individuals. Effects appeared to be mediated by age, however, since they occurred in an over-35 age group, but not in an under-20 age group. Further analysis suggested that the unanticipated contrast in results between the two age groups was due to an interaction between processing difficulty and trait capacity (formulated thereafter as the difficulty x absorption hypothesis). If the results of Experiment 1 reflected variations in subjects' experiences of task difficulty, then it was argued that absorption-related effects could be obtained for younger subjects by increasing the level of difficulty.
In test of this hypothesis, the second experiment employed a more compelling version of the Response-Change task, and focused on the performance of the under-20 age group. A postexperimental enquiry was also conducted, to examine subjects' use of response strategies. Results revealed that high-absorption subjects were inclined to use Positional strategies, while a significantly greater proportion of low-absorption subjects used Rehearsal strategies. Positional strategies were more perceptual in their functioning, while Rehearsal strategies relied heavily on a cognitive-memory activity. No significant differences in performance, however, were found between high- and low-absorption subjects. This null result was explained by appealing to the strategy data.
It was believed that subjects ' use of their preferred strategies might be especially effective in mitigating task demands, thereby reducing the level of difficulty, and thus the likelihood of trait-related effects. It was reasoned that such effects could be obtained only if the task manipulation of the difficulty variable was such that subjects were compelled to adopt response strategies which were not their preferred ones (the difficulty x ·strategy hypothesis). Self-report data showed that buffer trials and longer interstimulus intervals enabled subjects to prepare their strategy for the next stimulus. Removing buffer trials and shortening interstimulus intervals should therefore reduce subjects ' use of their preferred strategies. These modifications were thus incorporated in the design of the third experiment.
Results from the third experiment showed no group differences in the use of response strategies, and data from the study suggested that subjects ' use of their preferred strategies was reduced by appropriate changes to the demand structure of the task. Consistent with this position, a main effect of absorption was found, as well as an absorption x strategy interaction: high-absorption subjects recorded slower performances using Rehearsal strategies than did low-absorption subjects using the same strategies. But when high-absorption subjects used a No-Strategy approach, their performance was significantly faster than similar subjects using Rehearsal strategies.
Tellegen's (1981) notion of contrasting- and congruent-sets was introduced at this point to account for the pattern of data. High-absorption subjects have an experiential set, and low-absorption subjects an instrumental set. The enactment of Rehearsal strategies is congruent with an instrumental set, but contrasts with the operation of an experiential set; while a reverse pattern is obtained with No-Strategy responses. If strategy and set are congruent, performance may be facilitated; if strategy and set are contrasting, then performance is likely to be inhibited. In this way, data from the first three experiments were held to be consistent with a difficulty x strategy x congruence specification for effects. This specification was tested in Experiment 4.
The final experiment in the series of experiments not only involved the Response-Change task, but also introduced a second task (the No-Change task) to provide further refinement of the theory under test. In the No-Change task, subjects were asked to name colours throughout rather than changing between colour-naming and word-naming, as required by the Response-Change task; in all other respects, the demand structure was equivalent for both tasks. The design of this fourth study sought to examine whether change in response requirements was also implicated in trait-related differences in performance. Results showed a replication of effects in the Response-Change task, but no significant differences were found with the No-Change task. As well, Rehearsal strategies were the predominant strategy used in the Response-Change task by both high-and low-absorption subjects, but hardly used at all in the No-Change task. On the basis of these results, it was concluded that low-absorption subjects have an enhanced ability to discriminate between discrete informational cues, and incorporate these cues as an adaptive strategy. This ability is required by the categorization demands of the Response- Change task, but not by the No-Change task. It was also argued that high-absorption subjects are less proficient at utilizing strategies which require the discrimination of detailed cues. High-absorption subjects may perform effectively, however, when they use a mode of processing that relies on a "unified representation of the attentional object" (Tellegen & Atkinson, 1 974, p.274}, that is, a No-Strategy approach. Experiment 4 also employed an instructional manipulation of strategies and a repeated-measures design to examine the performance of both absorption groups, when using Rehearsal and No-Strategy approaches. Results provided a final test of the difficulty x strategy x congruence hypothesis, and confirmed the hypothesis regarding differential utilization of strategies.
The present program of research, as it developed, moved the emphasis away from attentional constructs such as the predisposition to sustain · attention towards a broader information-processing framework, as a means of describing the complexity existing in the data. A model of absorption based on differences in information encoding systems was argued to account for the pattern of obtained results: within this information-processing framework, strategic and non-strategic responses were shown to be associated with different modes of representing stimuli for categorization. Specifically, the findings of Experiment 4 suggested that strategic/instrumental responses called for the discrimination and utilization of detailed cues, while non-strategic/experiential responses relied on the subjects ' direct experience of the attentional object as a unified representation (including its affective and imagery-related qualities). Low-absorption subjects are more disposed towards strategic/instrumental responses, while high-absorption subjects are more likely to use non-strategic/experiential responses. Relevant research in the related fields of hypnosis and imagery was also discussed as converging on this position. Thus, the model of absorption involving different modes of stimulus representation, as developed in this research, was found to be consistent with Crawford and Allen's (1983) conception of detail-oriented and holistic-oriented encoding systems.
It was concluded in this thesis that absorption is primarily an interactive rather than a latent trait, and effects arising from this trait require consideration of a particular mix of situational and adaptive influences. These influences are expressed operationally in the difficulty x strategy x congruence formulation, which provides a close specification of test conditions for the operation of the personality trait of absorption.
Finally in the thesis, some speculative implications were discussed, that highlight the importance of the State/Trait distinction in other theoretical treatments of absorption. Using salient examples from the literature, particular focus was placed on problems of circularity when inferring Trait dispositions in processing from subjects' experiences associated with so-called altered states of consciousness. Directions for future work were examined, and a research design was described which seeks to establish the limits of influence of response strategies on absorption-related performance. The notion that absorption may have psychophysiological correlates was also introduced, with particular reference made to the utility of the construct of Sensation-seeking.