The Effects of Test Anxiety and Success/Failure on Mathematics Performance in Grade Eight

Abstract

Scales of general anxiety and test anxiety have been employed to study effects of anxiety on cognitive performance in numerous studies. Kirkland (1971) asserts that one reason why anxiety studies sustain the interest of researchers is the conflicting findings of these studies. Age level, intelligence level, task difficulty, degree of stress situation, and sex differences combined with type of anxiety measuring instrument are some of the factors that contribute to such conflicting findings. The problem is compounded by contrasting conceptions of anxiety as a trait or state variable as observed by Sieber (1969). The instability of anxiety as a factor in learning is illustrated in three studies of anxiety and intelligence. Denny (1966), using the Taylor Manifest Anxiety Scale, found that at the high intelligence level the high-anxiety group deduced attributes of a concept significantly better than the low-anxiety group, but at the low intelligence level the highanxiety group performed less ably than the low-anxiety group. Mazzei and Goulet (1969) replicated Denny's study but employed the Test Anxiety Questionnaire instead of the Taylor scale. They also obtained a significant interaction between anxiety and intelligence but in a direction opposite to that of Denny's findings. In contrast to both of these significant interactions, Sarason and Hill (1966) conducted a five-year longitudinal study beginning with children in grades one and two and found that only 3 of 58 interactions between anxiety (Test Anxiety Scale for Children) and intelligence on measures of academic performance were significant. The importance of stress situations in anxiety studies is underlined by Katchmar, Ross, and Andrews (1958). They found that anxiety had no significant effect on performance during prestress trials, but after stress was introduced by failure on a task, high-anxiety subjects' performance was poorer, while low-anxiety subjects' performance improved. The absence of stress in the learning of mathematics with programed