Students’ attitudes toward school chemistry: The effect of interaction between gender and grade level

Sample

The sample of the study was assigned through a convenience sampling method. The sample was composed of 197 high school students from different grades that ranged from Grades 9 to 11 (Years 16-18), from a public high school in Isparta, Turkey. There were 184 valid cases and among them 111 of the students were female and the remaining 73 of them were male. 83 of the students were attending to Grade 9, 36 of them to Grade 10, and 65 of them to Grade 11.

Instrument

The Attitude Scale toward Chemistry (ASTC), developed by Geban, Ertepınar, Yılmaz, Altın & Şahbaz in 1994, was administered by the instructor during regular chemistry lessons. The scale consists of 15 items of the 5-point Likert type that ranges from “Completely Disagree” to “Completely Agree”. The items of the scale involve both positive (10 items) and negative sentences (5 items) in nature. The possible minimum score of a student is 15 and the maximum is 75. The reliability of the scale was found as .83 by the developers. However, the researchers computed the Cronbach Alpha internal consistency coefficient of the general scale as .93 for the present data.

Data Analysis

The negatively worded items were recoded before performing statistical analyses which enables researchers to comment on the results, smoothly; that is, lower scores show negative attitudes and higher scores, on the contrary, show positive attitudes toward chemistry as a school subject. In order to check dimensionality of the scale, as suggested by Rennie & Parker (1987) and Gardner (1995), the Principal Component Analysis was used. The analysis pointed out two dimensions of the scale for the present data (detailed information related to the dimensions reported in the “Results and Discussion” section). Moreover, reliability analyses were computed for both of these dimensions. Two-Way Multivariate Analysis of Variance (Two-Way MANOVA) was conducted to test whether there is an interaction between gender and grade levels on the two dimensions of attitudes toward chemistry lessons. Finally, in-depth analyses was carried out to locate where the interaction effect was significant if there was a significant interaction effect according to the results of MANOVA. All of the mentioned analyses were performed by SPSS.