Goodness of Item Fit in TKTM by Linguistic Ability

Thursday, April 3, 2014
Exhibit Hall Poster Area 1 (Convention Center)
Chang-Hwan Choi, Eun-Hye Jo, Hyojun Yun, Seul-Ki Woo and Jae-Hyeon Park, Korea National Sport University, Seoul, South Korea
Background/Purpose:

TKTM stands for Taekwondo Knowledge Test based on Multimedia. In the physical education and sports science, certain level of subject's linguistic ability is required in paper and pencil test (PPT). The purpose of this study is to confirm how the goodness of item fit changes according to the linguistic ability, when the PPT is converted to a multimedia test (TKTM).

Method:

The participants of this study are 141 black-belt trainees in Taekwondo, undergraduates (age mean=22.5, sd=1.8, average training=6.6years). To control artificially the language ability of the subject, both mother language (Korean Version) and foreign language (English Version) are performed in the test. Taekwondo Knowledge test used in this study is 'Multilingual Taekwondo Knowledge Test', developed by Choi et al (2012). Multilingual Taekwondo Knowledge Test is an equivalent test which proposed the language differently, consisting of 17 items. The goodness of item fit for TKTM was compared with the PPT, and tests in mother language and foreign language were fulfilled repeatedly. Rasch model of item response theory is applied, and Facets (Lincere, 2012) program was used to analyze the data. Goodness of fitness index was determined by Infit and Outfit statistics (≥0.75 and ≤1.30).

Analysis/Results:

Using PPT, the analysis of the difference in the goodness of item fit in Taekwondo Knowledge Test according to linguistic ability shows that all of 17 items had goodness of fitness index with acceptable fit statistics (infit range 0.8 to 1.1, outfit range 0.8 to 1.1) in mother language, and 5 items of PPT in foreign language are out of range in goodness of fitness index (out of infit 1 item, out of outfit 5 items). When TKTM is used, the goodness of item fit in the tests in mother language and foreign language showed no out of fitness range.

Conclusions:

Therefore, compared to PPT, TKTM can reduce misfit of item, according to linguistic ability of the subject. To conclude, the results show that TKTM is able to minimize language bias.