There are three main "themes" to this document, organized as "Parts".

 

Part 1 is given to data "snooping".

 

The dataset has a few categorical variables, such as age and gender.  I show how to use Excel and Lertap to check on data integrity.

 

Part 2 involves snooping of another sort.

 

In this section I take a look at missing data, especially the many unanswered questions that are a "feature" of this particular dataset.  There are so many unanswered questions that test reliability was affected, but in a way you might not have anticipated.  This section also looks at some grade-level and age-level differences in test scores, and in item responses too.

 

Part 3 gets into Lertap's real forte: item analysis.

 

This part exemplifies how I myself go about item and test analysis.  As is very common, the test I've looked at had some items which did not perform well, having some apparent ambiguity.  In "double keying items", I show how to fine tune test scoring so that students are not disadvantaged by this oft-encountered problem.

 

I suspect that many readers will want to focus on the third part.  But I suggest that the other parts are at least worthy of a good, relaxed browse.  I say this as I've attempted to point out some Excel and Lertap tools which may well come in handy someday.

 

Get yourself up and running by first having a look at the Data and CCs worksheets, coming up next.