To explore whether different balances of fiction/non-fiction reading and challenge might help explain differences in reading achievement between genders, data on 45,670 pupils who independently read over 3 million books were analysed. Moderate (rather than high or low) levels of challenge were positively associated with achievement gain, but…

Chang and Zhang (2002, 2003) proposed several baseline criteria for assessing the severity of possible test security violations for computerized tests with high-stakes outcomes. However, these criteria were obtained from theoretical derivations that assumed uniformly randomized item selection. The current study investigated potential damage caused…

This paper examines two empirical studies that used online surveys to collect data to measure satisfaction in job training and workforce development. A description of each study, findings related to response rate, the processes used in online survey development and implementation, as well as recommendations for the future use of online surveys…

The pedagogy underlying online learning and teaching is being reconceptualised to incorporate the opportunities being offered by the development of online educational settings. The pedagogy of constructivism and in particular socio-constructivism is underpinning much of the online learning and teaching developments currently being developed. The…

In two experiments, naive participants took turns telling each other to click on a target picture while gaze was monitored. Critical trials included a "contrast" picture that differed from the target only in size. In both experiments, the timing of speakers' fixations on the contrast predicted whether the contrast was encoded in a phrase with a…

This paper deals with optimal partitioning of limited testing time in order to achieve maximum total test score. Nonlinear optimization theory was used to analyze this problem. A general case using a generic item response model is first presented. A special case that applies a response time model proposed by Wang and Hanson (2005) is also…

In addition to the potential that computer-based testing (CBT) offers, empirical evidence has found that identical computerized and paper-and-pencil tests have not produced equivalent test-taker performance. Referred to as the "mode effect," previous literature has identified many factors that may be responsible for such differential performance.…

This study examined the equivalency of computer and conventional versions of the Test of Workplace Essential Skills (TOWES), a test of adult literacy skills in Reading Text, Document Use and Numeracy. Seventy-three college students completed the computer version, and their scores were compared with those who had taken the test in the conventional…

This study investigated the validity of a statewide alternate assessment program, IASEP (Indiana Assessment System of Educational Proficiencies) by examining supporting profile patterns on the 100 core IASEP items of individuals with significant disabilities. Participants were 5,192 students ranging in age from 7-21 years with special education…

This study explores the potential for using a computer-based scoring procedure for the formulating-hypotheses (F-H) item. This item type presents a situation and asks the examinee to generate explanations for it. Each explanation is judged right or wrong, and the number of creditable explanations is summed to produce an item score. Scores were…

The relationship of several individual differences variables to Computer Adaptive Testing (CAT) as compared with traditional written tests are explored. Seven hundred sixty-five examinees took a Computer Adaptive Test and two fixed-length written tests. Each examinee also answered a computer literacy inventory, a satisfaction questionnaire, and a…

The average Kullback-Keibler (K-L) information index (H. Chang and Z. Ying, in press) is a newly proposed statistic in Computerized Adaptive Testing (CAT) item selection based on the global information function. The objectives of this study were to improve understanding of the K-L index with various parameters and to compare the performance of the…

This study focused on the robustness of unidimensional item response theory (UIRT) models in computerized classification testing against violation of the unidimensionality assumption. The study addressed whether UIRT models remain acceptable under various testing conditions and dimensionality strengths. Monte Carlo simulation techniques were used…

This research investigates whether the Computer-Based Academic Assessment System (CAAS), a battery of tests examining speed and accuracy at performing component reading skills, would be useful for identifying reading disability in college students. One important question was whether CAAS could distinguish reading disability from non-disability or…

When items are selected during a computerized adaptive test (CAT) solely with regard to their measurement properties, it is commonly found that certain items are administered to nearly every examinee, and that a small number of the available items will account for a large proportion of the item administrations. This presents a clear security risk…