摘要：This article sketches the framework for assessing reading competence across the lifespan in the German National Educational Panel Study (NEPS). It gives a detailed presentation of the two central dimensions in the framework: (a) text functions and text types and (b) the cognitive requirements of reading tasks. These are discussed against the background of relevant theoretical models and research findings. A pilot study of 447 adults is reported that analyzed the dimensionality and difficulty of the reading competence test for adults. Results indicated that the test meets the NEPS research goals. The article focuses particularly on whether text types and cognitive requirements prove to be appropriate structural elements for the framework, that is, whether each distinguishes sufficiently between good and poor readers. Results also report on the dimensionality of the reading competence test. A comparison between one unidimensional and two different multidimensional models examined whether the text types and/or cognitive requirements of the items/tasks are separate dimensions of reading competence, or whether the reading competence measured in NEPS can – as intended – be conceived as a unidimensional construct. Results are discussed against the background of the scope and limitations of the NEPS reading competence test and the relevant research literature on reading competence.
其他摘要：This article sketches the framework for assessing reading competence across the lifespan in the German National Educational Panel Study (NEPS). It gives a detailed presentation of the two central dimensions in the framework: (a) text functions and text types and (b) the cognitive requirements of reading tasks. These are discussed against the background of relevant theoretical models and research findings. A pilot study of 447 adults is reported that analyzed the dimensionality and difficulty of the reading competence test for adults. Results indicated that the test meets the NEPS research goals. The article focuses particularly on whether text types and cognitive requirements prove to be appropriate structural elements for the framework, that is, whether each distinguishes sufficiently between good and poor readers. Results also report on the dimensionality of the reading competence test. A comparison between one unidimensional and two different multidimensional models examined whether the text types and/or cognitive requirements of the items/tasks are separate dimensions of reading competence, or whether the reading competence measured in NEPS can – as intended – be conceived as a unidimensional construct. Results are discussed against the background of the scope and limitations of the NEPS reading competence test and the relevant research literature on reading competence.
