Abstract
No planetarium research had established the superiority of particular methods of teaching and learning in the planetarium over other methods with adolescent students. A large body of educational theory indicated that the Karplus Learning Cycle (based on the work of Piaget), student model manipulation, and student drawing would be superior to a traditional method, if students were able to master the astronomy concepts. The basic astronomy topics studied were the celestial sphere and earth roation, the seasons, lunar phases, and planet positions and motions. These topics are frequently taught to adolescent students in planetariums, and they stress projective spatial ability . A research design which is a variation of the posttest only- control group true experimental design was employed. The subjects were 98 eighth grade students of one suburban junior high school in Northeastern Ohio. Students in each of three science classes taught by the same teacher were randomly assigned to the developed experimen tal or the traditional planetarium unit treatment, each unit consisting of eight instructional visits and each following identical learning objectives. The investigator presented all lessons, while adult raters judged there was no difference in teacher enthusiasm or stu dent treatment. Immediate topic posttests and immediate and delayed unit posttests were given, developed for this study with the guidance a panel of experts. Cronbach alpha reliability coefficients for the tests ranged between .83 and .71. The first two topic posttests and questions on the two unit posttests requiring a view of the sky were administered in the planetarium. Student anxiety was reduced by telling groups that good scores could help the term science grade, while poor scores could not affect the grade. A control group (no treatment) of a single intact eighth grade science class was post tested for unit learning. All students were pre- and posttested for spatial ability with the DAT Space Relations Test. The £ test was used to test significance. A .05 significance level was adopted, and a .10 level was adopted as a significant trend. Multiple Linear Regression was utilized to analyze data. There was significantly greater retention between the immedi ate and the delayed unit posttest by students who had experienced the experimental unit in excess of the variance due to initial spatial ability, gender, and known model manipulation experience two years prior to treatment. Boys improved significantly more than girls from the initial to the final spatial ability test when both experienced the experimental unit, but not when both experienced the traditional unit. Boys performed significantly better than girls on the initial spatial ability test when an n of 98 was examined. Students who had experienced either type of planetarium unit performed significantly better on a unit test than the no-treatment group in excess of variance due to spatial ability, gender, and known model manipulation experience. Spatial ability was found significantly correlated with performance on the astronomy posttests. Significant rends toward correlation were found between experimental treatment and performance on the delayed unitposttest and between known model manipulation experience and performance on the most difficult topic posttest. Students in treatment groups completed a questionnaire following all treatment and testing. Students did not generally realize that different treatments were given. Students in the experimental group generally enjoyed the unit more. A majority of students in the experimental group were unable to answer correctly 75%> of the questions on any of the six astronomy posttests. The superiority of the developed experimental unitwas shown, but it was also shown that most students were incapable ofmastering the astronomy concepts with either method. Spatial ability , gender, and previous model manipulation experience were found to be important variables affecting performance.
