Abstract
This study aimed to investigate the effect of different methods for handling missing data and its volume on detecting differential item functioning (DIF) for test items according to the gender variable, using the Generalized Mantel-Haenszel test (GMH). To achieve the study's aims, the responses of Jordanian eighth-grade students to items on the mathematics test in the International Study (TIMSS, 2015) were obtained. The sample of the study consisted of 2,261 students who responded to four test booklets (1, 7, 9, 13). Several methods were used for handling missing data, including treating missing data as nominal, corrected item mean substitution, the expectation-maximization algorithm, multiple imputation, and chained equations. GMH was also employed to detect DIF based on the gender variable after each treatment method. Additionally, the effect of the volume of missing data (5%, 10%) on the percentages of questions that showed DIF was assessed. The results indicated that there were no statistically significant differences at (α = 0.05) between the ratios of items that showed DIF across different missing data handling methods. In contrast, all five methods agreed on the existence of a statistically significant relationship at (α = 0.05) between the presence of DIF in favor of females in the field of algebra and in favor of males in the field of numbers. Regarding the volume of missing data, the findings revealed that there were no differences in the percentage of questions that showed DIF in favor of males or females due to the volume of missing data being 5% or 10%.
Recommended Citation
Ghazou, Iman and Al-Zoubi, Amal
(2022)
"The Effect of Missing Data Handling Methods and their Volume on Detecting Differential Item Functioning of Test Items,"
Jordan Journal of Applied Science-Humanities Series: Vol. 31:
Iss.
1, Article 6.
Available at:
https://digitalcommons.aaru.edu.jo/jjoas-h/vol31/iss1/6