A study tested the hypothesis that highly unrepresentative news stories about foreign nations would have a more damaging effect on the accuracy of inferences formed by United States audiences if the stories concerned developing countries than if they were about developed nations. Subjects, 76 college students, answered questions concerning a nation located in either a familiar or an unfamiliar world region (Europe or Africa), then read or did not read news stories containing highly nonrepresentative information concerning that nation. The results confirmed the hypothesis. They also showed that such stories would increase the confidence people place in their inferences only if the material and peoples' generalizations concerned unfamiliar nations. (FL)