A study investigated the attitudes toward reading of third-grade students involved in a whole language program. The sample consisted of 71 students--48 students participated in the study group, and a control group consisted of 23 students. The independent variables were participation status, gender, socioeconomic status, race, qualification for Chapter 1 services, and participation in English Training services. The dependent variables were scores from the scales of the Elementary Reading Attitude Survey (ERAS). These were: attitudes toward recreational reading, attitudes toward academic reading, and total. Pretest scores from the scales of the ERAS were employed as covariant measures and included attitudes toward recreational reading, attitudes toward academic reading, and total. Six composite null hypotheses were tested, and a total of 18 comparisons were made. Results indicated that none of the 18 main effects were statistically significant. Recommendations for replication include: (1) use a large random sample; (2) examine treatment in detail; (3) investigate other dimensions of effectiveness; and (4) control for teacher variables. (Contains 24 references and 6 tables of data. Letters requesting or granting permission, the Elementary Reading Attitude Survey, instructions for using the ERAS, a discussion of the validity and reliability of ERAS, blank demographic recording sheets, and a description of the treatment program are attached.) (RS)
