Abstract
The main purpose of this study was to investigate the reliability and validity of the Gender Perception Scale (GPS) for adolescents. Another purpose was to investigate the differences based on some socio demographic variables (sex, school type, family monthly income, working statue of mother and parents education levels differences) regarding with gender perception. The scale was administered to a total of 823 adolescents (451 girls, 372 boys) and the test-retest reliability analysis was carried out on 118 of the 823 adolescents. Confirmatory Factor Analysis (CFA) verified two-factor model as different from the original scale. Resulting factors were traditional gender perception and equalitarian gender perception. Goodness of fit indices of two factor hypothetic model was found as follows: χ2=500.15, df=168, χ2/df=2.97, GFI=.92, AGFI=.90, CFI=.91, RMSEA=0.06, SRMR= .05. Also, it was found that 27% of the scale’s up and down groups average was totally meaningful. The internal consistency coefficients were sufficient for each subscale (ranging from .80 to .86) and total scale (α= .89). In addition, the test-retest reliability coefficients for equalitarian gender perception subscale (.73), and traditional gender perception subscale (.80) were determined to be within acceptable limits. As a result, the GPS was found tobe valid and reliable and its psychometric characteristics were acceptable. In addition, both equalitarian and traditional gender perception subscale scores were significantly differed according to sex, school type, family monthly income, working statue of mother and parents education levels. The scale can be used to measure the behavior of Turkish adolescent populations with respect to gender perception.

Keywords: gender perception, high school students, validity, reliability.