排序方式: 共有44条查询结果,搜索用时 0 毫秒
41.
Marie Wiberg Wim J. van der Linden Alina A. von Davier 《Journal of Educational Measurement》2014,51(1):57-74
Three local observed‐score kernel equating methods that integrate methods from the local equating and kernel equating frameworks are proposed. The new methods were compared with their earlier counterparts with respect to such measures as bias—as defined by Lord's criterion of equity—and percent relative error. The local kernel item response theory observed‐score equating method, which can be used for any of the common equating designs, had a small amount of bias, a low percent relative error, and a relatively low kernel standard error of equating, even when the accuracy of the test was reduced. The local kernel equating methods for the nonequivalent groups with anchor test generally had low bias and were quite stable against changes in the accuracy or length of the anchor test. Although all proposed methods showed small percent relative errors, the local kernel equating methods for the nonequivalent groups with anchor test design had somewhat larger standard error of equating than their kernel method counterparts. 相似文献
42.
We investigate the current bandwidth selection methods in kernel equating and propose a method based on Silverman's rule of thumb for selecting the bandwidth parameters. In kernel equating, the bandwidth parameters have previously been obtained by minimizing a penalty function. This minimization process has been criticized by practitioners for being too complex and that it does not offer sufficient smoothing in certain cases. In addition, the bandwidth parameters have been treated as constants in the derivation of the standard error of equating even when they were selected by considering the observed data. Here, the bandwidth selection is simplified, and modified standard errors of equating (SEEs) that reflect the bandwidth selection method are derived. The method is illustrated with real data examples and simulated data. 相似文献
43.
Alina Ivanova Elena Kardanova Christine Merrell Peter Tymms David Hawker 《Assessment in Education: Principles, Policy & Practice》2018,25(2):141-159
Is it possible to compare the results in assessments of mathematics across countries with different curricula, traditions and age of starting school? As part of the iPIPS project, a Russian version of the iPIPS baseline assessment was developed and trial data were available from about 300 Russian children at the start and end of their first year at school. These were matched with parallel data from representative samples of equal numbers of children from England and Scotland. The equating of the scales was explored using Rasch measurement. A unified scale was easiest to create for England and Scotland at the start and end of their first year at school when children only differ by a half a year in age, and live in adjacent countries with a common language. Although fewer items showed invariance across the three countries, it was possible to link iPIPS scores in mathematics from the start and end of the first year at school across Scotland, England and Russia. The findings of this study suggest that, despite the apparent difficulties, meaningful comparisons of mathematics attainment and development can be made. These will allow for substantive interpretations with policy implications. 相似文献
44.
Judith Zimmermann Alina A. von Davier Joachim M. Buhmann Hans R. Heinimann 《European Journal of Engineering Education》2018,43(1):144-165
Graduate admission has become a critical process in tertiary education, whereby selecting valid admissions instruments is key. This study assessed the validity of Graduate Record Examination (GRE) General Test scores for admission to Master’s programmes at a technical university in Europe. We investigated the indicative value of GRE scores for the Master’s programme grade point average (GGPA) with and without the addition of the undergraduate GPA (UGPA) and the TOEFL score, and of GRE scores for study completion and Master’s thesis performance. GRE scores explained 20% of the variation in the GGPA, while additional 7% were explained by the TOEFL score and 3% by the UGPA. Contrary to common belief, the GRE quantitative reasoning score showed only little explanatory power. GRE scores were also weakly related to study progress but not to thesis performance. Nevertheless, GRE and TOEFL scores were found to be sensible admissions instruments. Rigorous methodology was used to obtain highly reliable results. 相似文献