Development of Prototype Software for Item Analysis and Item-Banking System for Likert Scales

Abstract

Before a researcher distributes a survey to measure certain attributes of a sample, it is necessary that the survey undergoes the process of item analysis to determine the effectiveness of each item and to build quality item pool for later use. The process of filtering item statements forreliability characteristics entails computing statistics that is tedious especially when done by hand.
When using commercially available statistical software’s, the scale constructor has to know how to run the software and find the appropriate command for each statistics for item analysis. Most of these statistical software’s are made for a wide-range of purposes but there are only a few software’s tailored solely for item analysis. Hogan (2007) pointed out that “Availability of item analysis program is somewhat erratic. They tend to come and go. More exactly, the companies that produce them tend to come and go.” Again, the problem that may arise from this issue is the stability of the company that supplies the software to support their clients when technical failure occurs or when the clients need assistance.
Another alternative is the use of electronic spreadsheet for item analysis. Setting-up the worksheet and the formula also requires time and experience. This includes testing the result for accuracy in the worksheet with another computation reference.
Test constructors may opt to purchase software license but the cost of ownership is relatively expensive. Engaging the services of a statistician to do the item analysis is another choice, but again it entails expense.
Thus, there is a need to develop software solely for item-analysis and itembanking that is more practical, less-costly, and compact.