Relationship of stimulus and examinee parameters to performance on visual and tactile block design tasks.

摘要

Despite numerous attempts to produce a tactile analog to the block design (BD) test for use with persons who are blind or visually impaired, no single effort has produced an enduring test, due in part to a lack of basic cognitive research that would otherwise support clinical interpretation of test performance. The current research compared direct visual and tactile analogs of BD tasks with respect to stimulus and examinee variables. Stimulus parameters were set size uncertainty (U), perceptual cohesiveness (PC), presentation order (PO), and cueing of block edges (C). The examinee variable was field independence (FI). Data from eighty normally-sighted participants' were analyzed, including Form B of the individual Embedded Figures Test (EFT, a measure of FI), nine identical BDs that varied orthogonally with respect to U and PC (the within-subjects factors), and six coherent designs intended to estimate, and control for, motor speed. Between-subjects factors were PO and C, with cueing achieved by the superimposition of a visual or tactile grid to indicate block edges, and PO defined by whether the visual tests were administered first or second. Two mixed-design repeated measures analyses of covariance (EFT as covariate) were performed for each modality, and ratios of effect sizes were calculated for cross-modal comparison. Main effects and interactions for the visual tests were consistent with prior studies, except for U, which was range-restricted due to the elimination of the lowest (9-bit) level. Tactile designs, however, yielded smaller effect sizes across the board, and evidenced no interaction between PC and C, suggesting different processes underlying tactile BD than have been suggested in earlier models of visual BD solution. Three post hoc analyses were also performed. Results suggested that stimulus and examinee parameters function and interact differently in the two modalities, and that direct cross-modal interpretations of test performance are not supported. A cross-modal protomodel is proposed, in which differential task demands at each stage of a two-stage process, mediated by FI and stimulus characteristics, account for disparate results in visual and tactile tests.