Read on Twitter
Item analysis (IA) is the process of analyzing the performance of a MCQ after it has appeared in a test. It is a quality control tool used to complete a key validation step prior to final scoring.
Careful review of IA data can help to improve the reliability & validity of scores
Careful review of IA data can help to improve the reliability & validity of scores
Item Difficulty (ID) refers to the proportion of examinees who answer an item correctly. It is the most basic essential information to evaluate in relation to the performance of a test item
ID ranges from 0-1, in which 0.6 means 60% of the test-takers answered the item correctly
ID ranges from 0-1, in which 0.6 means 60% of the test-takers answered the item correctly
Item Discrimination (ID) differentiates high-ability examinees from low-ability examinees. This is a fundamental principle of all educational measurement & a basic validity concept.
ID is the MOST important info to evaluate the performance of a test item.
ID is the MOST important info to evaluate the performance of a test item.
Treat any item difficulty or discrimination index cautiously if the statistics are based on a test administration with fewer than 100 examinees.
About 200 examinees are needed to have stable item analysis statistics. All item analysis data based on CTT are sample dependant.
About 200 examinees are needed to have stable item analysis statistics. All item analysis data based on CTT are sample dependant.
For small samples the results may still provide some useful guidance for item improvement.
In general, the smaller the sample size on which an item analysis is based, the greater the sampling error & the larger the standard errors around the sample statistics.
In general, the smaller the sample size on which an item analysis is based, the greater the sampling error & the larger the standard errors around the sample statistics.
