Computerized classification tests (CCT) have been used in high-stakes assessment settings where the express purpose of the testing is to assign a classification decision (e.g. pass/fail). One key feature of sequential probability ratio test-type procedures is that items are selected to maximize information around the cutscore region of the examinee ability distribution as opposed to common features of CATs where items are selected to maximize information at examinees' interim estimates. Previous research has examined the effectiveness of computerized adaptive tests (CAT) utilizing classification testing procedures a single cutscore as well as multiple cutscores (e.g. below basic/proficient/advanced). Several variations of the SPRT procedure have been advanced recently including a generalized likelihood ratio (GLR). While the GLR procedure has shown evidences of improved average test length while reasonably maintaining classification accuracy, it also introduces unnecessary error. The purpose of this dissertation was to propose and investigate the functionality of a modified GLR procedure which does not incorporate the unnecessary error inherent in the GLR procedure. Additionally this dissertation explored the use of the multiple cutscores and the use of ability-based item selection. This dissertation investigated the performance of three classification procedures (SPRT, GLR, and modified GLR), multiple cutscores, and two test lengths. An additional set of conditions were developed in which an ability-based item selection method was used with the modified GLR. A simulation study was performed to gather evidences of the effectiveness and efficiency of a modified GLR procedure by comparing it to the SPRT and GLR procedures. The study found that the GLR and mGLR procedures were able to yield shorter test lengths as anticipated. Additionally, the mGLR procedure using ability-based item selection produced even shorter test lengths than the cutscore-based mGLR method. Overall, the classification accuracy of the procedures were reasonably close. Examination of conditional classification accuracy in the multiple-cutscore conditions showed unexpectedly low values for each of the procedures. Implications and future research are discussed herein. / text

Not available / text

Auditory Processing Disorder (APD) is an auditory-specific perceptual deficit in the processing of auditory stimuli that occurs in spite of normal peripheral hearing thresholds and normal intellectual capacity American Speech-Language-Hearing Association (ASHA, 2005). The diagnostic process of APD typically involves a test battery consisting of sub-tests designed to examine the integrity of various auditory processes of the central auditory nervous system (CANS). One category of these sub-tests is the low-pass filtered speech test (FST), whereby a speech signal is distorted by using filtering to modify its frequency content. One limitation of the various versions of the FST currently available is that they are administered using a constant level of low-pass filtering (e.g. a fixed 1 kHz corner frequency) which makes them prone to ceiling and floor effects (Farrer & Keith, 1981). As a consequence, the efficacy and accuracy of these tests is significantly compromised (Martin & Clark, 1977). The purpose of the present study was to counter these effects by utilising the University of Canterbury Adaptive Filtered Speech Test (UCAST) which uses a computer-based adaptive procedure intended to improve the efficiency and sensitivity of the test over its constant-level counterparts. A comprehensive APD test battery was carried out on 18 children aged 7-13 suspected of APD and on an aged-matched control group of 10 children. Fifteen of the APD suspected children were diagnosed with APD based on their performance on a traditional APD test battery, comprising the Compressed and Reverberated Words Test (CRWT), the Double Digits test (DDT), the Frequency Pattern test (FPT) and the Random Gap Detection test (RGDT). In addition, the UCAST was administered to examine whether the low-pass filter limit at which children score 62.5% of words correct i) differed significantly between children who either passed or failed the APD test battery; ii) correlated with their score on the traditional APD battery (TAPDB); and iii) correlated with their score on a commercially available low-pass filtered speech test, the Filtered Words Subtest of SCAN-C (Keith, 2000b). Results demonstrated a significant difference between the UCAST low-pass filter limit at which APD and control children scored 62.5% words correct (two-way repeated measures ANOVA, p < 0.01). Significant correlations were found between the UCAST and three of the four tests used in the TAPD - the DDT, the RGDT and the FPT (Pearson Correlation coefficient, p < 0.01). No correlation was found between the UCAST and the CRWT or between the UCAST and the SCAN-C (FW) test (p > 0.05). These findings provide evidence that an adaptive filtered speech test may discriminate between children with and without APD with greater sensitivity and specificity than its constant-level counterparts.

Thesis (Ph. D.)--Florida State University, 2005. / Advisor: Dr. Joseph K. Torgesen, Florida State University, College of Arts and Sciences, Dept. of Psychology. Title and description from dissertation home page (viewed June 9, 2005). Document formatted into pages; contains vii, 56 pages. Includes bibliographical references.

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

Thesis (Ph. D.)--University of Texas at Austin, 2005. / Supervisor: Hua-Hua Chang. Vita. Includes bibliographical references.

This thesis investigates Goodness-of-Fit tests for parametric regression models. With the help of sufficient dimension reduction techniques, we develop adaptive-to-model tests using projection in both the fixed dimension settings and the diverging dimension settings. The first part of the thesis develops a globally smoothing test in the fixed dimension settings for a parametric single index model. When the dimension p of covariates is larger than 1, existing empirical process-based tests either have non-tractable limiting null distributions or are not omnibus. To attack this problem, we propose a projected adaptive-to-model approach. If the null hypothesis is a parametric single index model, our method can fully utilize the dimension reduction structure under the null as if the regressors were one-dimensional. Then a martingale transformation proposed by Stute, Thies, and Zhu (1998) leads our test to be asymptotically distribution-free. Moreover, our test can automatically adapt to the underlying alternative models such that it can be omnibus and thus detect all alternative models departing from the null at the fastest possible convergence rate in hypothesis testing. A comparative simulation is conducted to check the performance of our test. We also apply our test to a self-noise mechanisms data set for illustration. The second part of the thesis proposes a globally smoothing test for parametric single-index models in the diverging dimension settings. In high dimensional data analysis, the dimension p of covariates is often large even though it may be still small compared with the sample size n. Thus we should regard p as a diverging number as n goes to infinity. With this in mind, we develop an adaptive-to-model empirical process as the basis of our test statistic, when the dimension p of covariates diverges to infinity as the sample size n tends to infinity. We also show that the martingale transformation proposed by Stute, Thies, and Zhu (1998) still work in the diverging dimension settings. The limiting distributions of the adaptive-to-model empirical process under both the null and the alternative are discussed in this new situation. Simulation examples are conducted to show the performance of this test when p grows with the sample size n. The last Chapter of the thesis considers the same problem as in the second part. Bierens's (1982) first constructed tests based on projection pursuit techniques and obtained an integrated conditional moment (ICM) test. We notice that Bierens's (1982) test performs very badly for large p, although it may be viewed as a globally smoothing test. With the help of sufficient dimension techniques, we propose an adaptive-to-model integrated conditional moment test for regression models in the diverging dimension setting. We also give the asymptotic properties of the new tests under both the null and alternative hypotheses in this new situation. When p grows with the sample size n, simulation studies show that our new tests perform much better than Bierens's (1982) original test.

The problem with which this study is concerned is determining the effectiveness of a computer adaptive test as compared to the effectiveness of using the entire test. The study has a twofold purpose. The first is to determine whether the two test versions generate equivalent scores, despite being of different lengths. The second is to determine whether the difference in time needed to take the computer adaptive test is significantly shorter than the computer administered full test.

文剑冰. / 论文(哲学博士)--香港中文大学, 2003. / 参考文献 (p. 138-147). / 中英文摘要. / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Mode of access: World Wide Web. / Wen Jianbing. / Zhong Ying wen zhai yao. / Lun wen (zhe xue bo shi)--Xianggang Zhong wen da xue, 2003. / Can kao wen xian (p. 138-147).

Incorporating off-grade items within an on-grade item pool is often seen in K-12 testing programs. Incorporating off-grade items may provide improvements in measurement precision, test length, and content blueprint fulfillment, especially for high- and low-performing examinees, but it may also identify some concerns when using too many off-grade items on tests that are primarily designed to measure grade-level standards. This dissertation investigates how practical constraints such as the number of on-grade items, the proportion, and range of off-grade items, and the stopping rules affect item pool characteristics and item pool performance in adaptive testing. This study includes simulation conditions with four study factors: (1) three on-grade pool sizes (150, 300, and 500 items), (2) three proportions of off-grade items in the item pool (small, moderate, and large), (3) two ranges of off-grade items (one grade level and two grade levels), and (4) two stopping rules (variable- and fixed-length stopping rule) with two SE threshold levels. All the results are averaged across 200 replications for each simulation condition. The item pool characteristics are summarized using descriptive statistics and histograms of item difficulty (the b-parameters), descriptive statistics and plots of test information functions (TIFs), and the standard errors of the ability estimate (SEEs). The item pool performance is evaluated based on the descriptive statistics of measurement precision, test length and exposure properties, content blueprint fulfillment, and mean proportion of off-grade items for each test. The results show that there are some situations in which incorporating off-grade items would be beneficial. For example, a testing organization with a small item pool attempting to improve item pool performance for high- and low-performing examinees. The results also show that practical constraints of incorporating off-grade items, organized here from most impact to least impact in item pool characteristics and item pool performance, are: 1) incorporating off-grade items into small baseline pool or large baseline pool; 2) broadening the range of off-grade items from one grade level to two grade levels; 3) increasing the proportion of off-grade items in the item pool; and 4) applying variable- or fixed-length CAT. The results indicated that broadening the range of off-grade items yields improvements in measurement precision and content blueprint fulfillment when compared to increasing the proportion of off-grade items. This study could serve as guidance for test organizations when considering the benefits and limitations of incorporating off-grade items into on-grade item pools.