Week 9
WEEK 9: Item response theory (Mokken scaling)
Hi, everyone
I feel that the assignment in this week is an important topic (and it is really difficult to understand) in the health science research. We will discuss about the item response theory (Mokken scaling) in terms of how this being used in the social sciences. First of all, in the blog I will explain how Item response theory differs from Classical test theory .
I feel that the assignment in this week is an important topic (and it is really difficult to understand) in the health science research. We will discuss about the item response theory (Mokken scaling) in terms of how this being used in the social sciences. First of all, in the blog I will explain how Item response theory differs from Classical test theory .
Classical test theory
Classical test theory (CTT) is a traditional quantitative approach to test the reliability and validity of a scale based on its items, usually set up only one version of the test that covers all elements of the feature or content that you want to measure with various of the difficulty level of the test. There is no definite proportion of difficulty level. Moreover, we have to set up a large number of questions (To Find the quality of the test requires the difficulty level (p) and the discrimination (r)) to ensure the test more accurate and reliable. These items will be sorted according to the test content. They usually consist of intermediate difficulty level items which is between 0.2-0.8. Regardless of whether the examiner is high, middle, or low, everyone must take the same test with equal numbers and all the same. There will be errors which can occur through mistakes within the process of testing, as well as everyday malfunctions such as being tired, hungry, etc; but if a standard error can be found then it becomes easier to factor this out of the equation.
There are many different models for CTT:
💧 Factor analysis
💧 Internal consistency
Item response theory
Item response theory (IRT) is a new test theory, a collection of measurement models that attempt to explain the connection between observed item responses on a scale and an underlying construct. The difference of assessment the quality of the test between the item response theory and the classical test theory is providing parameter by discrimination (a), difficulty level (b), and probability (c) with improving measurement accuracy and reliability. The numbers of questions and duration for taking the test will be decreased such as there are 10 questions, each examiner does not need to do all 10 questions, but it depends on the ability of each examiner to do it.
The item characteristic curve (ICC) is the fundamental unit in IRT and can be understood as the probability of endorsing an item (for a dichotomous response) or responding to a particular category of an item (for a polytomous response) for individuals with a given level of the attribute.
Feature of models
💧 Item parameter estimates are independent of the group of examinee used from the population of examinees for whom the test was designed.
💧Examinee ability estimates are independent of particular choice of test items used from the population of items which were calibrated.
💧 Precision of ability estimated are known.
These are different models for IRT:
1. Two- parameter Normal Ogive Model
2. Two- parameter Logistic Model; proposed an item response model in which the item characteristic curves take th eform of two- parameter logistic distribution functions.
3. Tree- parameter Logistic Model
4. One- parameter Logistic Model (Rasch Model) (By Georg Rasch, a Danish mathematician). Rasch Model has been performed some special properties that make it especially attractive to users. One reason is that it is easy to work with because the model involves fewer item parameters. Second, the problem with parameter estimation are considerably fewer in number than for the more general models.
5. Four- parameter Logistic model
6. One-, Three-, and Four- parameter Normal Ogive Model
7. Normal Response Model
7. Normal Response Model
8. Graded Response Model; this model was introduced to handle the testing situation where item responses are contained in two or more ordered categories.
It can be seen that there are several models of IRT. Therefore, we need to choose which model is common used and and be applied to educational and psychological test data.
References
1. https://www.jstor.org/stable/25791718?seq=1#metadata_info_tab_contents
2. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4096146/pdf/nihms594145.pdf
3. Hambleton, R. K. & Swaminathan, H. (1985) Item response theory : principles and applications. Boston: Klumer-Nijhoff.
4. Lord, F. M. (1981) Applications of item response theory to practical testing problems. Hillsdale, N.J.: Lawrence Earlbaum Associates.
ความคิดเห็น
แสดงความคิดเห็น