The purpose of this study is to compare the effectiveness of three types of practices applied in Korea in enhancing the validity and equivalency of test instruments when cross-cultural adaptation of attitude measures is necessary. The three types of practices are: (1) translation and review (translation version); (2) translation, back translation, and review (back translation version); and (3) translation, back translation, review, and empirical validation study (validation version). The focus was on the relative effectiveness of back translation applied to the construction of Korean versions of instruments. Participants were 734 fifth graders from 3 public elementary schools in Seoul (Korea). Responses on the three test versions and two other motivation scales were collected within a 3-week period at approximately 1-week intervals. Results show that the back translation version is superior to the translation version in terms of its similarity to the validation version and construct-related evidence. However, results from item-response theory analyses reveal that the quality of the translated items is similar. The nature of adapted attitude scales is discussed. Appendixes contain the Academic Failure Tolerance Scale (M. Clifford 1988, 19991) and two back translation versions. (Contains 13 tables and 33 references.) (Author/SLD)