What are the steps in a Lertap analysis? Create a new Lertap workbook. Enter item responses in the Data worksheet. Make up the "control cards" for the CCs worksheet.

 

Go to the Run menu on the Lertap toolbar.  Click on "Interpret".

 

What happens?  Lertap has a squiz of the CCs cards, reading down the rows of the CCs sheet until it encounters a row whose first column is empty.

 

If an error is found in the CCs lines, Lertap stops and makes an effort to tell you what the error is. You'll need to fix the error, and then click once again on the Interpret option.

 

If the CCs lines appear to be error-free, Lertap then starts to read the records in the Data worksheet, going down the Data rows until it encounters a row whose first column is either empty or contains a zero.

 

This process initiates the production of Lertap output, the secondary worksheets often referred to as Lertap "reports". The first of these is called "Freqs"; simultaneously, Lertap makes the behind-the-scenes series of "Sub" worksheets. There will be one Sub worksheet for each subtest, that is, for each *col card found in the CCs worksheet.

 

It then hides the Sub worksheets, brings Freqs to the fore, and announces that it's ready for you to squiz the Freqs. If you're satisfied with the squiz, you return to the Run menu, and select the "Elmillon" option. Note that there are settings in the System worksheet which get Lertap to roll from the Interpret option to Elmillon automatically, non-stop. Lertap is said to be in "production mode" when this happens.

 

Once Elmillon starts up, what happens? Quite a bit. Lertap returns to the Data worksheet and reads all its records, extracting the responses corresponding to the first *col card in CCs, forming item statistics, and making a subtest score for each respondent.

 

It writes the subtest scores "on the fly", that is, as it reads the Data records (you can sometimes see it doing this). Writes them to where? To the "Scores" worksheet, another new sheet which Lertap adds to the workbook.

 

Then Lertap usually creates its two main reports with item statistics. Each of these reports is a new worksheet. They're called "Stats1f" and "Stats1b", respectfully containing "full" and "brief" item statistics. If the subtest is a cognitive one, Lertap usually creates another new worksheet, "Stats1ul", with upper-lower discrimination and difficulty estimates. (In the process of making the Stats1ul report, Lertap creates a temporary worksheet called "Scratch". It deletes this worksheet on completing the Stats1ul report.)

 

What happens next? If there's more than one *col card in the CCs worksheet, Lertap repeats this process. Each *col card defines a Lertap subtest. For each and every subtest, Lertap adds a score to the Scores worksheet, and creates the appropriate series of Stats reports.

 

As to nomenclature, the Stats reports for the first subtest are Stats1f and Stats1b; for the second subtest they're Stats2f and Stats2b; ... and so on.

 

It is possible to control the number of reports made by Lertap. The Stats1f report is always standard, but the Stats1b and Stats1ul may be turned off. How? By making alterations in lines 9 and 10 of the System worksheet. The computational resources used to make the Stats1ul report are rather extensive; turning off this report will usually save a noticeable amount of processing time, especially when there are more than 500 records in the Data worksheet.

 

Freqs, Scores, and the suite of Stats worksheets comprise Lertap's main output. But it's possible to get Lertap to deliver more. For example, it'll make histograms, scatterplots, item response charts,  a worksheet with item scores, subtest correlation matrices, and even another Stats report, ECStats1f (resulting from an external-criterion analysis).