Item Discrimination & Selection

¶ … discrimination, assessment and some of the related consequences. The first question asks what the general definition of item discrimination is. The second question asks how topic of item discrimination can be assessed. The problem then asks what the consequences for item selection when the criteria are total test scores as opposed to external scores. While there are going to be some consistencies and patterns and while this will even exist with external scores, one should be very careful to analyze items correctly and in the proper context.

Questions Answered McGrory et al. notes that performance on psychometric tests is key to diagnosis and the monitoring of dementia treatment, among other psychological treatments and assessments. Results are often assessed based on the total score. However, there is often additional information in individual items of tests which vary based on several values. These values are known as difficulty and discrimination. Item difficulty refers to the overall ability level at which the probability of responding correction is about fifty percent.

Item discrimination is an index of how well an item can be differentiated between patients of varying levels of things like severity. Using external scores as a guidepost can be problematic based on the differing of questions that are used, the severity of the patients involved, the item discrimination levels in both surveys and so forth. McGrory et al. goes on to note that Item Response Theory (IRT) analysis can be used to examine and refine measures related to cognitive functions.

In the McGrory study, the overall systematic review was aimed to identify all published literature which had applied the prior-mentioned IRT to instruments assessing global cognitive function in people with dementia (McGrory et al., 2014; McElhiney, Kang, Starkey & Ragan, 2014). Conclusion Per the above, item discrimination and item selection are items that could and should be assessed when it comes to any questions that are related to assessing cognitive function and performance.

This can be true in mentally healthy patients and students but it can also stretch across all other forms and realms of cognitive psychology. It should be noted that mixing and matching questions and results can be problematic because there are going to be differences and thus the "apples to apples" comparison yardstick will not be there. There can and might be situations where two tests can be compared side by side in terms of the questions and their results.

However, this will be rare because the tests will tend to be different, the pool of test-takers will tend to be different and so forth. Making conclusions based on unlike tests or pools of test-takers is less than wise because the conclusions garnered will probably be incorrect or at least unprovable. Question Two The author of this response is asked to review how a teacher could and should make an analysis of test results when a test is completed.

The question pertains to the test itself as well as the questions therein. The author is asked to decipher items and item responses should be revised and why. Indeed, test questions should be assessed one by one as well as in a pool based on the content of the questions themselves, what the questions mean, the common responses that are offered, whether the question is answered correction, which incorrect answers or answers are prevalent, why those incorrect answers might be prevalent, whether the question is flawed and so forth.

While testing science is not esoteric and abstract, it must still be done in a focused way so that the results of the associated tests are relevant and correct as possible. Analysis A teacher should make an analysis of a tests results and its questions because it has to be assessed whether the results are indicative of how adept (or not) the children who take the test are.

For example, many hold that African-American people do not perform on standardized tests as well as they could or should because their schooling or the test itself does not endear them to getting good test scores. One of the best tools to decipher what performance exists (or does not exist) and to what level is to look at the results of the test and compare to the results on other tests and in other academic spheres and arena.

Year to year progression can be useful, comparison to other standardized tests is useful and so forth. A real-world example of this was described by Foorman (2013) when it was noted that 2011 FCAT 2.0 scores were used to predict what would happen as it related to 2012 FCAT 2.0 scores. Later on, the actual results were compared and contrasted (Foorman, Kershaw & Petscher, 2013). Test scores alone are assailed by many as a bad method of assessing performance and attainment.

For example, many hold that active learning activities (which testing is not) is a better way to measure such performance attainment (Koksal, Yagisan & Aksoy, 2013). Conclusion Testing scores and assessment is a double-edged sword. It puts students in a controlled environment in.