If you have come here and yet have no idea of what "Lertap" is, see the link to introductory slides at point 2 below, and also have a look here, if you please.

 

I had several objectives in mind when I set about to create this document/website.

 

It is designed to be very practical, using real data and presenting actual issues which were faced by a very active testing site in 2012.

 

There are messages and suggestions.  Foremost among them: examining data quality.  It's easy to get Lertap to output lots of tables and graphs, but these will be of questionable validity if they're based on data which are not trustworthy.  Excel and Lertap both have tools which let us look into the matter of data quality; these are demonstrated in the first topics.  By and large these tools are easy to use.

 

Once the data have been poked and prodded, cleaned and caressed, it's on to looking at item and test quality before scores are used for reporting student results.  In the dataset you're about to visit in this document, the responses students gave to two items signaled that they may have found the items to be ambiguous.  There are steps that may be taken to counterbalance the ramifications of item ambiguity; you'll see them.

 

At times I make some effort in this document to cover matters which I suspect many Lertap users may be unaware of.  From Excel, for example, I look thoroughly at the use of the data Filter, and make mention of pivot tables.  From Lertap: the use of blank lines in CCs and Data worksheets; a tricky use of res= to count the number of unanswered questions; getting the histogrammer to work from a Breaks report; using packed quintile plots; interpreting item response flags; and a bit more.

 

The dataset itself is from a source which will remain anonymous.  I should explain one term that may seem odd if you haven't seen it before: some countries refer to primary school grade levels as "Standards", and to high school grade levels as "Forms".  The item responses contained in this dataset are from Standard 4 and Standard 5, which would generally correspond to what other countries might call Grade 4 and Grade 5.

 

The data were collected via the use of a scanner and imported into Excel.

 

Almost all of the screen shots in this document were taken while using Excel 2010; one or two are from Excel 2013.  Lertap version 5.10.2 was used exclusively.

 

Links to other Lertap resources are listed below.  Note that you can pick up some sample data from 4 -- the "M.Nursing" dataset, for example, exhibits problems similar to those found in this document's dataset.  Lertap 5.10.2 is readily available at 5 and may be used for practice.

 

 

Please direct questions or comments to: lertap5@gmail.com

---

Last update: 27 May 2021