Methods for Detecting and Representing the Item-Position Effect and the Exploration of its Cognitive Sources

本研究的主要目的是通过研究能力测试领域中的项目-位置效应，即偏向测量，促进心理测评的发展。为实现这一目标，研究建议建立一种方法，用于可靠地检测和识别能力数据中的项目位置效应，在考虑天花板和相似性效应的干扰时优化项目位置效应的表征，将现有的非加速测试数据的方法扩展到快速测试数据。此外，拟议的研究将有助于进一步澄清项目-位置效应的来源。因此，需要澄清的是，项目位置效应是否源于学习，以及与难度引导的项目安排是否存在交互作用。研究工作的主要部分必须通过模拟数据进行。模拟数据能够控制数据的一般特征，因此特别适合于找出物品位置效应的可检测性和物品排列的特殊性。考察快速测验对项目位置效应的影响需要实证数据，也需要考察项目位置效应的来源。对来源的考察主要集中在学习假说上。因此，数据收集除了能力测量外，还必须包括学习测量的应用。最后，应该有一种经过修订的方法，能够控制物品位置效应，并最终将其作为另一种有价值的信息来源加以利用。

The major aim of the proposed research is the advancement of psychological assessment by researching the item-position effect that is biasing measurement in the area of ability testing. For achieving this aim the research proposal suggests the construction of a methodology for the reliable detection and identification of the item-position effect in ability data, the optimization of the representation of the item-position effect in considering disturbances due to ceiling and similarity effects, the extension of the available method for data obtained in non-speeded testing to data obtained in speeded testing. Furthermore, the proposed research shall contribute to the further clarification of the sources of the item-position effect. Accordingly it is to be clarified whether the item-position effect is due to learning and whether there is an interaction with the difficulty-guided arrangement of items. A major part of the research work has to be conducted by means of simulated data. Simulated data enable the control over the general characteristics of data and, therefore, are especially well suited for finding out about the detectability of the item-position effect and about the specificities of the arrangement of items. The investigation of the impact of speeded testing on the item-position effect requires empirical data and also the investigation of the sources of the item-position effect. The investigation of the sources concentrates on the learning hypothesis. Therefore, the data collection must include the application of measures of learning besides ability measures. In the end there should be a revised method that enables the control of the item-position effect and eventually its exploitation as another source of valuable information.