While affective scales designed to measure students' attitudes toward reading have become an important part of teachers' assessments of student performance, little research has been conducted to determine whether these scales provide accurate information, nor have researchers adequately compared reading attitudes according to student achievement and socioeconomic levels. To provide evidence regarding these areas, two studies were undertaken involving more than 300 students, half from an inner city school and half from a suburban school. In the first study, the reading attitudes of the students were measured by the Estes Reading Attitude Scales and the results were compared according to grade, school, achievement level (low or average), and sex. In the second study, students who had indicated a positive attitude toward reading in the first investigation were given a polygraph test to determine if they had expressed their true attitudes. The results of the first study revealed that average achieving students and those in the inner city school had more positive attitudes toward reading than their peers. No effects were found for either grade or sex. Results of the second study revealed that half of the students in the low achieving group were not truthful in assessing their attitudes, while only one student in the average achieving group was not--with no student underestimating his or her reading attitude. (FL)