Abstract

Presented here is a brief report of a study along with an extensive criticism of this and other studies that use contrived tasks for investigating children's humor and writing. The original study had hoped to anwer questions related to how child-produced humor might vary with the sex and age of the intended audience, how writing would figure in child-produced humor, and any relationships that might exist between production of and talk about humor. Two middle-class fourth-grade classrooms were told a contrived story and asked to produce something funny for some sick children. Children in one class produced something for two sick children, both male, one in first and one in eighth grade. The other class produced something for two sick children, both female, one in first and one in eighth grade. Productions (N = 136) were collected and analyzed. Of the fourth graders, 9 were then interviewed about their conceptions of humor and of their productions. Children's productions did vary according to the age and sex of the intended audience. There were no particular relationships between the humor of variability of the productions and the talk about humor. The major criticism in this and similar studies is that the data are flawed. Despite efforts in the opposite direction, the contrived task produced a confused pragmatic context. Once the pragmatics were distorted, the data no longer represented the phenomena of interest—humor and writing. Without extensive observations and interviews, there is little evidence that these findings represent what the chosen variables make them appear to represent. An argument is thus made for increased sensitivity to what phenomena research data actually represent.