The study aimed to examine whether an experimental imagery intervention, which elicited ecological representations of self, was more effective in enhancing vividness of movement imagery, action control and performance ability than a conventional imagery intervention and/or that of a standard, instruction intervention. In order to test this hypothesis, a sample of 36 young ballet dancers (11-13 years) were secured from five reputable ballet schools and randomly assigned to three groups (n = 12) for the purposes of conducting a true experiment: a standard control group, an imagery control group and an experimental imagery group. Descriptive statistics and frequencies were conducted for all instruments used to describe and represent single variables of interest. Parametric and nonparametric statistical procedures were used to determine whether significant differences existed between the three groups for measures of vividness of movement imagery, action control and performance ability. Additional emerging relationships and trends were explored through bivariate correlational analyses, multiple stepwise regression procedures, a factor analysis as well as MANOVA and ANOVA statistical procedures. The results of the study were varied and some interesting trends were observed. Contrary to what was expected, the results revealed significant differences across the three groups for mean difference scores of performance ability, in favour of the standard control group. In particular, the standard control group revealed positive benefits in performance ability while the imagery control and experimental imagery groups revealed deteriorations in performance ability. The experimental imagery group was found to be significantly different from the imagery control group for mean difference measures of internal vividness ofmovement imagery. However, the study failed to reveal significant variations in mean difference scores for action control across the three groups. Furthermore, the results indicated that external and internal vividness of movement imagery were found to be significantly correlated with performance ability for the entire sample (N = 36) and across the two imagery groups (N = 24) respectively. Finally, numerous significant and near-significant relationships were observed between measures of vividness of movement imagery, action control and performance ability. The results were discussed in relation to the literature on mental imagery, ecological self, action control and development. / Thesis (M.A.)-University of Natal, Pietermaritzburg, 2001.