An investigation of some correlates of mature moral judgement : a thesis presented in partial fulfilment of the requirements for the degree of Master of Arts in Education at Massey University

Abstract

From Kohlberg's cognitive developmental perspective, moral judgement has been viewed as an aspect of ego development which is related to other aspects of ego development. The purpose of this research was to investigate some possible correlates of mature moral judgement: although some research had been conducted with children practically no research into correlates in adult populations has been reported. A sample of 50 male and 32 female first year teachers' college volunteer subjects was used. Of these, a random subsample of 10 males was used to test the congruent validity of the written questionnaire form of 4 of Kohlberg's (1958) moral dilemma situations devised for this research, with the original interview technique. Using a design counterbalanced for order of presentation with one week retest interval the difference between the measuring devices was found to be non-significant (t = -0.13, p>.55) . The remaining subjects wore randomly assigned to one of two testing orders - 20 males and 16 females to each group. In three testing sessions at 3-4 day intervals both groups were administered the: Kohlberg Moral Judgement Scale (written form). Study of Ethical Attitudes, C.P.I., P.O.I., Study of Values, E.P.I., Californian F-test and Mill Hill Vocabulary Scales. The order of presentation was counterbalanced. Two-way ANOVAs were performed on the data to test for sex or order effects. In the majority of cases no significant effects wore found and so the data was contained to provide a sample n of 72. On scales on which significant differences were found separate analysis of the subgroups was performed. A measure of Piagetian cognitive level was obtained 2 months later by individual testing. Analysis of the Kohlberg scale revealed a high level of stage mixture: only 5.5 percent of subjects had a range as low as 4 stages. Evidence was put forward that indicates that high levels of stage mixture have also been found by others and the conflict between this finding and Kohlberg's theory was discussed. A factor analysis of the item scores revealed 6 factors with eigen values greater than one. Five of these could be interpreted meaningfully in terms of Kohlberg's theory but the second factor was difficult to interpret; it seemed because of its correlations with other measures to be concerned with general sensitivity to ethical issues. The factors were identified as: factor 1 (conflict between obedience to the law and the value of human life), factor 3 (duty vs contractual obligation), factor 4 (spirit of the law), factor 5 (property rights) and factor 6 (power rights vs moral rights). Contrary to Kohlberg's finding but in line with a number of other studies no major moral judgement factor emerged. Correlations between Moral maturity scores and other personality measures were largely non-significant. A correlation of -.30 (p<.05) confirmed the negative correlation between these two measures reported by Kohlberg (1964). The correlation of .34 (p<.05) with the Study of Values religious scale for the male subsample was in the opposite direction to that reported by Klinger et al (1964). The only other significant correlation was that of .34 (p<.05) with Piagetian cognitive level in the females. Failure to find a similar correlation in the males was attributed to the high level of male responses resulting in little variability over the critical range. Reanalysis involving correlations between Kohlberg scale factor scores and the other personological scales revealed 11 significant correlations; i.e. 4.8 percent of those calculated reached the .05 level or higher. 4 of these correlations were with the first factor (conflict between obedience to the law and the value of human life), and seemed to show that conventional level judgements related to the acceptance of conventional Christian attitudes toward life and work and to authoritarian tendencies. Reanalysis of the data, from subsamples in which clearer personological differences could be expected, was conducted. t-tests between Ss whose moral maturity scores were in the extreme quartiles were calculated for those C.P.I. and P.O.I. scales measuring traits expected to be most closely related to moral maturity. No significant differences were found. Correlations between moral maturity and all other scales were calculated for those 14 Ss with the most homogeneous moral judgement profiles and again those for 34 Ss with the most homogeneous moral judgement profiles. In both of these subsamples the only significant correlations were with the Californian F test (r xy = -.80, p<.01, n = 14: r xy = -.37, p<.05, n = 34). 2-way ANOVA of the effects of empathy and autonomy on moral maturity revealed do significant effects. A factor analysis of all scales including moral maturity, but excluding the ipsative Study of Values scales, extracted 10 factors with eigen values greater than 1. The eighth factor had a loading of -.90 on moral maturity and was clearly a Kohlberg scale factor. Canonical correlation between Kohlberg scale item scores and the other scales revealed 15 significant correlations, the first of which was a canonical correlation of .97. Inspection of the normalized weights showed that meaningful interpretation of these linear combinations within existing theoretical and conceptual frameworks was not possible. A considerable number of variables were measured and intensive analysis of results conducted but the only variable that was found to consistently relate to moral maturity was authoritarianism. Evidence obtained from this sample pointed to the possibility that a positive correlation between moral maturity and cognitive development may be found in samples more representative of adult variability on this dimension.