New version of CITAS released

The Classical Item and Test Analysis Spreadsheet (CITAS) is one of our most popular products, for two reasons:

1.  It is the best software available for basic psychometric analysis, if you don’t want/need to learn complex products like Iteman
2. It is free.

I recently released a new version of CITAS based on some feedback I’ve received. It hasn’t been updated for about 5 years, so it was about time.

What’s New?

The biggest change is that it has been expanded from 50 items/examinees to 100 items/examinees. This allows it to be used for larger samples, such as medium-sized university courses or small credentialing exams. Classical test theory statistics become more stable as sample sizes approach 100, so this makes sense

I also added a deeper distractor analysis. Previously, CITAS did not include option point-biserials, which are considered essential by most psychometricians. I’ve added that, and also have implemented conditional formatting in Excel so that cells <0 are red and values 0<0.2 are yellow.

A score distribution graph has been added to the primary output tab.

A number of smaller changes were implemented, including an update to informational links.

I haven’t used it yet. What is CITAS?

CITAS is a spreadsheet I built in 2009 as part of a course I was teaching in Measurement and Assessment. I didn’t want students to have to learn how to run professional-level analysis software like Iteman or SPSS; instead, I wanted them to be able to get results as quickly and easily as possible, so they could focus on learning how to interpret statistics rather than navigate complex software.

CITAS filled this need. All a user needs to do is paste in their data, and the statistics will automatically populate. There is no data file formatting or “running” a program with various specifications to choose. This of course makes CITAS of little use for professionals that need large data files or complex specs, but is perfect for someone who is new to psychometrics or just wants a basic analysis on a small data set.

An example of the output is below. CITAS calculates essential statistics like item P, item Rpbis, KR-20 reliability, and standard error of measurement (SEM).

I’m totally new to assessment. What is classical test theory?

For starters, psychometrics is the Science of Assessment. That is, a body of scientific research on how to make assessments better, regardless of whether they are in Education, Credentialing, Employment/HR, or other applications.

There are two dominant paradigms in psychometrics. Classical test theory (CTT) is, as the name would suggest, much older and based on simple mathematics like means, proportions, and correlations. Item response theory (IRT) is far more complex, based on sophisticated nonlinear mathematical models. It is, however, far superior in many respects so most big-time assessment programs rely on IRT.