Thesis (Ph. D.)--University of North Texas, Aug., 2009. / Title from title page display. Includes bibliographical references.

Thesis (Ph. D.)--University of Texas at Austin, 2003. / Vita. Includes bibliographical references. Available also from UMI Company.

A multi-stage test (MST) design is an alternative design for the delivery of automated tests. While computerized adaptive tests (CAT) have dominated testing for the past three decades, increasing interest has been focused on the MST because it offers two advantages that CAT does not: Test sponsors and test developers can see an entire test before administration because it is pre-constructed from sets of modules of test items, and within a module examinees may skip forward and back through test items and make changes to previously answered items. Due to the dominance of CAT, little research has been devoted to differing MST designs with regard to the number of items per stage and routing rules that direct the selection of the next module after a previous module has been completed. This research used simulated response data for a large national test and the generalized partial credit model to compare a CAT to one of three MST designs that had either decreasing numbers of items per stage, increasing number of items per stage, or the same number of items per stage, and one of three routing rules, maximum information, fixed [theta], or number-right routing. As anticipated, CAT had the best performance with respect to estimating proficiency and item pool use. Among the MSTs, the MST with increasing numbers of items per stage performed the best with respect to estimating proficiency, followed by the MST with decreasing number of items per stage, and equal numbers of items per stage. By routing rule, maximum information performed the best and number-right routing performed the worst. Only one panel was constructed per MST design, so only limited comparisons of item pool use could be made. Although the MST designs did not perform as well as CAT, the differences in estimating proficiency were not large, implying that the MST design is a viable alternative to CAT. / text

The multistage testing (MST) has drawn increasing attention as a balanced format of adaptive testing that takes advantages of both fully-adaptive computerized adaptive testing (CAT) and paper-and-pencil (P\&P) tests. Most previous studies on MST have focused on purely dichotomous or polytomous item formats although the mixture of two item types (i.e., mixed-format) provides desirable psychometric properties by combining the strength of both item types. Given the dearth of studies investigating the characteristics of mixed-format MST, the current study conducted a simulation to identify important design factors impacting the measurement precision of mixed-format MST. The study considered several factors-namely, total points (40 and 60), MST structures (1-2-2 and 1-3-3), the proportion of polytomous items (10%, 30%, 50% and 70%), and the routing module design (purely dichotomous and a mixture of dichotomous and polytomous items) resulting in 32 total conditions. A total of 100 replications were performed, and 1,000 normally distributed examinees were generated in each replication. The performance of MST was evaluated in terms of the precision of ability estimation across the wide range of the scale. The study found that the longer test produced greater measurement precision while the 1-3-3 structure performed better than 1-2-2 structure. In addition, a larger proportion of polytomous items resulted in lower measurement precision through the reduced test information during the test construction. The interaction between the large proportion of polytomous items and the purely dichotomous routing module design was identified. Overall, the two factors of test length and the MST structure impacted the ability estimation, whereas the impact of the proportion of polytomous items and routing module design mirrored the item pool characteristic. / text

Not available / text

Not available / text

Computer adaptive testing (CAT) has grown both in research and implementation. Test construction and security issues, however, have led many to reconsider the merits of CAT. Multistage testing (MST) is an alternative adaptive test design that purportedly addresses CAT's shortcomings. Yet considerably less research has been conducted on MST. Also, most research in adaptive testing has been based on item response theory (IRT). Many tests now make use of testlets -- bundles of items administered together, often based on a common stimulus. The use of testlets violates local independence, a fundamental assumptions of IRT. Testlet response theory (TRT) is a relatively new measurement model designed to measure testlet-based tests. Few studies though have examined its use in testlet-based CAT and MST designs. This dissertation investigated the performance of testlet-based CATs and MSTs measured using the TRT model. The test designs compared included a CAT that is adaptive at the testlet level only (testlet-level CAT), a CAT that is adaptive at both the testlet and item levels (item-level CAT) and a MST design (MST). Test conditions manipulated included test length, item pool size, and examinee ability distribution. Examinee data were generated using TRT-calibrated item parameters based on data from a large-scale reading assessment. The three test designs were evaluated based on measurement effectiveness and exposure control properties. The study found that all three adaptive test designs yielded similar and good measurement accuracy. Overall, the item-level CAT produced better measurement precision, followed by the MST design. However, the MST and CAT designs yielded better measurement precision at different areas of the ability scale. All three test designs yielded acceptable exposure control properties at the testlet level. At the item level, the testlet-level CAT produced the best overall result. The item-level CAT had less than ideal pool utilization, but was able to meet its pre-specified maximum exposure control rate and maintain low item exposure rates. The MST had excellent pool utilization, but a higher percentage of items with high exposure rates. Skewing the underlying ability distribution also had a particularly notable negative effect on the exposure control properties of the MST. / text

Not available / text

Thesis (Ph. D.)--University of Texas at Austin, 2008. / Vita. Includes bibliographical references.

Thesis (Ph. D.)--University of Texas at Austin, 2004. / Supervisor: Hua-Hua Chang. Vita. Includes bibliographical references.