The research was carried out in a secondary school in Malawi, Central Africa, using four programs in biology. The experiment compared multiple-choice and constructed responding across four categories of learning (verbal association, generalisation, multiple discrimination and chaining). Analysis of covariance showed no significant differences on the post-test scores between the two response modes in any of the four categories. There was a small number of significant differences in the analysis of covariance of the retention test scores, but no general pattern emerged. However, pupils who used the multiple-choice versions of the programs took significantly less time to work through them.