Extending a Practical Method for Developing Alternate Test Forms Using Independent Sets of Items
研究扩展了Gibson和Weiner的方法，通过随机抽样和基于效标关联效度选择，为预测大学表现的情境判断量表生成了144个替代测试形式。
This study describes alternate test form development for a Situational Judgment Inventory (SJI) predicting college performance. Principal axis factor analysis of responses to the SJI lent support for a general factor, yet each SJI form sampled items across 12 distinct rationally derived content areas. The first step of developing alternate forms involved random and representative sampling of SJI items across each content area, creating a large number of preliminary 36-item SJI test forms. Gibson and Weiner (1998) provided criteria for selecting alternate forms; however, the authors of the present study extended this approach in the next step of selecting alternate forms based on their estimated criterion-related validity with grade point average. Results provide initial support for the 144 alternate forms generated. This general approach reflects a practical and methodologically sound means of developing alternate forms of types of measures that are rationally heterogeneous yet empirically homogeneous.