This paper reviews the methodologies of both experimental structure and data analysis of recent studies of child learning of English morphology, in particular, Jean Berko's work of 1958. It concentrates on the child s acquisition of English plurals, noting that possibilities for the refinement of these studies lie in subject‐sample selection and the control of subject aptitude and socio‐economic status. It points out that increased reliability can result from a reduction of the number of linguistic parameters selected for study, concomitant with an increase in the number of stimulus items in an instrument. The procedural difficulties of item sequence and subject misperception of stimuli are explored with the suggestion that subject repetition of the stimuli is a way of separating error of misperception from actual errors in plural formation. The Kernan and Bount study (1966) is regarded as derivative from Berko's work, and (it is noted) that their interpretation of results concerning plural formation performance of given final segments is statistically unfounded. This paper suggests that researchers must take more care to present all the data relevant to the statistical analysis from which the study's conclusions are drawn so that discussion is diverted from interpretation of the findings to the evidence itself. © 1969 Language Learning Research Club, University of Michigan
