The formulation and development of instruments to measure field dependence-field independence using spatial and verbal modalities

• Main Author: Pearson, Frank
• Other Authors: Riding, Richard ; Rayner, Stephen ; McGregor, Debra
• Published: Oxford Brookes University 2013
• Subjects: 370
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.757985

The thesis gives a description of the formulation and development of an Embedded Shapes Test which was subsequently used as a method of measuring the construct/cognitive style of field dependence – field independence, using a spatial modality. Two additional methods of measuring field dependence – field independence were also formulated and developed, which used a verbal modality, in the form of a Sense Word Test (SWT) and Non-Sense Word Test (NSWT). Each of the above field dependence – field independence 'tests' were used in a series of pilot studies and studies, as part of their development and application. With two of the studies, of which there were four, additional ‘tests’ were used to measure cognitive attributes considered to have some association with those inherent in the measurement of field dependence – field independence. These included the Chronological Order Integration Test (COIT) and the Gestalt Picture Completion Test (GPCT) in Study Three; and the GPCT; British Ability Scales (BASI)/Short Form IQ; and the Cognitive Styles Analysis (CSA) in Study Four. The COIT in Study Three provided a framework to measure field dependence – field independence using a verbal modality through the medium of text. Therefore, this enabled comparisons to be made within a verbal modality through the medium of words and non-sense words by the SWT and NSWT, as well as a spatial modality by the EST. The inclusion of the GPCT, in Studies Three and Four enabled the process of measuring field dependence – field independence to be compared with a process that appeared to be opposite to it, i.e. an integration – synthesis process, or putting together pieces of information, instead of a segregated – analytical process, or taking apart a piece of information, respectively. The BASI/Short Form IQ was included in Study Four to provide a measure of cognitive ability. Since an argument directed at the concept and measurement of field dependence – field independence is that it is measuring cognitive ability rather than cognitive style, a comparison was made between field dependence – field independence and cognitive ability/intelligence as measured by the BASI/Short Form IQ. A further comparison with measurements of field dependence – field independence was made possible by the inclusion of the CSA in terms of a Wholistic or Analytic perception of information, as well as a propensity, on the part of the individual, to think spatially or verbally, i.e. in the Imager or Verbaliser mode of the CSA. The inclusion of school subject performance/attainment levels, in each of the four studies, was to enable comparisons to be made between the various measurements of field dependence – field independence, i.e. EST, SWT, NSWT and COIT, and GPCT, BASI/Short form IQ, CSA, and the characteristics of the different school subjects themselves. This enabled the exploration as to whether or not the characteristics of particular school subjects have an affinity to field dependence – field independence in terms of cognitive style and/or cognitive ability. This exploration was augmented in Study Three by the inclusion of GCSE subject grades for the sample members. In Chapter 14 (Comparisons of the Outcomes from the Four Studies), Witkin’s Group Embedded Figures Test (GEFT) was introduced to provide a comparison of reliability between the EST3, SWT2 and NSWT1, using the Cronbach Alpha statistic. The sample used for each of the pilot studies and studies consisted of Year 8, male and female, students of mixed ability, from Comprehensive and Middle Schools. The sample sizes for each of the two pilot studies were thirty, but varied for the four studies from 62 to 224 to increase the statistical validity and to decrease the standard error. Within each of the four studies, means; standard deviations; Multivariate Analysis of Variance (MANOVA); correlations (Pearson Product Moment) and Cronbach Alpha statistics were used to analyse the data. The EST3, SWT2 and NSWT1 were used in each of the four studies and produced a high level of construct validity across the four studies. A number of significant F ratios and correlations were obtained between the variables in each of the four studies. In particular, significant F ratios were obtained for EST3; Sex; Test Type (SWT2 and NSWT1), e.g. (F = 4.00; df 2, 56; p = 0.05 – Study 1); School Subjects and GCSE Subjects; and significant correlations between EST3 and Science (General), e.g. r = 0.4006; p = 0.05 – Study 1; Modern Language (French); COIT A and S; COIT B and S; and IQ within the corresponding that included these variables. Correlations (r) between EST3, SWT2 and NSWT1 within each study produced a number of high reliability values (See Chapter 14, Table 14.4 for full details). In addition, the Cronbach Alpha values for the EST3, SWT2 and NSWT1 gave a high reliability, i.e. 0.84263, 0.92799, 0.91946 for Study One; 0.90281, 0.93720, 0.88725 for Study Two; 0.89441, 0.94493, 0.92713 for Study Three; and 0.84942, 0.93779, 0.91099 for Study Four, respectively. The Cronbach Alpha value for the EST3 was greater than that for the GEFT, i.e. 0.93544 and 0.89234, respectively. Similar values were obtained for the SWT2 and NSWT1 when compared to the GEFT, i.e. 0.97891 and 0.96411, respectively.