Adaptive testing

Wainer, H., Computerized Adaptive Testing. 2nd ed, Lawrence Erlbaum Associates Mahwah, New Jersey, 2000. 

In the United States and in most countries of the world, the GRE is now administered as a computer-based test (CBT), also called a computer-adaptive test (CAT). You make an appointment to go to a testing center, where you take the exam. You may either choose a center from the list in the GRE Registration Bulletin, or ETS will assign you to the nearest location. The exam is offered during the first three weeks of every month, year-round. 

A CBT/CAT is a test administered at a computer. CBT stands for computer-based test and CAT stands for computer-adaptive test. Special software enables the computer to accurately determine your proficiency at a given task within a short time. In brief, every test taker gets a different set of questions. The computer uses your performance on the early questions to give either easier or more difficult subsequent questions. This is the adaptive part. The harder the questions you successfully answer, the higher your score. You will find more information and specific strategies for CBT/CATs on the following pages. 

It is important to remember that a computer-adaptive test (CAT) is tailored to your performance level. The test will begin with a question of medium difficulty. Each question that follows is based on how you responded to earlier questions. If you answer a question correctly, the next question will be more difficult. If you answer a question incorrectly, the next question will be easier. The test is designed to analyze every answer you give as you take the test to determine the next question that will be presented. This is done to ascertain a precise measure of your quantitative abilities, using fewer test questions than traditional paper tests would use. 

General Considerations and Certain Restrictions in Adapting Tests to the Autoanalyzer... 

The essential components of the Autoanalyzer have been covered in some detail. The clinical chemical methods which have successfully been or possibly will be adapted for use by this device have been described. Finally some of the problems encountered in adapting test methods for use with the analyzer and some current limitations were discussed. It is to be expected that the manufacturer if not the users will explore every means of increasing the versatility of the machine and that with time and development some of the current restrictions will be removed. It is certain that the instrument has the potentiality of performing a greater variety of tests than has currently been discussed. It has demonstrated its capability in a sufficient number of different tests to make one optimistic that, even though the present test methods may undergo modification, the instrument will ultimately (and does now) produce reliable chemical results in many laboratories. 

Thus, using these techniques, it has been suggested that a sex-pheromone CHC of Drosophila stimulates the bitter taste receptor neuron in a CHC concentration-dependent manner. Moreover, the cross adaptation test between the sex-pheromone CHC and an ordinary bitter substance supported the idea that the sex-pheromone CHC tastes bitter (Lacaille el al., 2007). 

Analogously, Voss and Wang (2005) proposed individual and simultaneous adaptive tests based on p,/dmm for i = l,. h, which are similar to the adaptive confidence intervals of Section 4.2, but with <7min computed from all h estimators Pk rather than by setting aside /3, when testing 77b,i /3, = 0. The more powerful version of this test is discussed in Section 5.2. 

Voss, D. T. and Wang, W. (2005). On adaptive testing in orthogonal saturated designs. Statistica Sinica, in press. 

An early sign of vitamin A deficiency is impaired dark adaptation - an increase in the time taken to adapt to seeing in dim light. The apparatus required is not suitable for use in field studies, or for use with children (the group most at risk from deficiency), and the dark adaptation test is largely of historical interest. Balance, colorvision, and the senses oftaste and smell are also affected in early deficiency, but none of these provides a sensitive or specific test of status. 

Modern test theory offers the potential for individualized, comparable assessments for the careful examination and application of different health status measures. One such theory is item response theory (IRT). Researchers report that IRT has a number of potential advantages over the currently employed classical test theory in assessing self-reported health outcomes. Applications of the IRT models are ideally suited for implementing computer adaptive testing. IRT methods are also reported to be helpful in developing better health outcome measures and in assessing change over time. ° ... 

Ware, J.E., Jr. Bjorner, J.B. Kosinski, M. Practical implications of item response theory and computerized adaptive testing A brief summary of ongoing studies of widely used headache impact scales. Med. Care 2000, 38, II73-II82. 

Tests of hierarchical knowledge are best designed as adaptive tests, so that one has the flexibility to select the most appropriate items based on students previous responses. It is enormously difficult to determine how many items to use in the assessment of a hierarchy. I have elaborated on this issue elsewhere (see Marshall 1981), where I developed two heuristics for approximating optimal item selection. I will not repeat the details here. The point is that flexibility is required, so that one can go back and forth as needed between items testing composites and items testing individual elements. 

Dark adaption test Pupillary response test Plasma retinol concentration Relative dose response... 

Glud, R.N., Klimant, I., Holst, G, Kohls, O., Meyer, V., Kuhl, M. and Gundersen, J.K., 1999. Adaptation, Test and in situ Measurements with Micropt(r)o-des on Benthic Landers. Deep Sea Research, 46 171-183. 

Ventilatory Response to Test Odours in Pre-Adaptation Tests... 

Standardized test protocols and routines must be also further developed to make data subtracted from the fuel cell more reliable and comparable. Specific long-term test protocols should be also developed to investigate durability of individual fuel cell components and interactions among themselves. Besides, each fuel cell application may require its own adapted testing protocols. 

Night blindness is the earliest symptom of vitamin A deficiency in humans. It first manifests itself as a slow, dark adaptation, thence it progresses to total night blindness. In man, the dark adaptation test, which measures the eyes ability to recover visual activity in dim light, is used as a means of... 

The choice of the test functions may considerably affect the test result. Therefore, they should be chosen with care. The better a test problem is adapted to the characteristics of a numerical procedure, the better the test result will be. Thus, a serious testing should not only rely on one (artificial, well-adapted) test function, but on some different molecular systems of different size (number of nuclei) with different characteristic (deep minima, clear saddles, flat minima/saddles, etc.). 

---