Special notes for instructors and measurement classes

Students and instructors often find that the free "Mini" version of Lertap 5 will serve as an effective introduction to Lertap, giving students a chance to tackle the exercises suggested in the topics found below. There are also a number of other spots with possible exercises and pertinent discussions for classes; see, for example, the so-called "Erudite epistles" page, and the sample datasets website. The manual has a fairly extensive discussion of the various statistics used in classical test theory and how they're interpreted, particularly in Chapter 7.

Two "practical exhibits" showcase examples of Lertap5 in action, one for a cognitive test and another for an affective scale on depression. They are primarily introductory in nature but do delve a bit into more advanced topics. The cognitive test is called "Test13" and is available here. The depression inventory (or scale) is a very popular, true-to-life scale and may be accessed via this link.

A unique option, "To halve and hold", can be used to divide any dataset into two random halves, setting up the possibility of having classes look at parameter stability, or, perhaps, calibration and validation. A paper by Nelson (2008) reflects this idea in the context of Rasch scaling and the Winsteps program. Other related suggestions and challenges may be seen in this special exercises page.

The matter of linking Lertap 5 to other programs, such as SPSS and item-response theory (IRT) programs, is discussed in this topic. A theme popular in many classes and research endeavors is factor analysis. It might be suggested that students select one of the sample datasets, such as "Mente 2010", and then apply the "To halve and hold" option mentioned in the previous paragraph to create two samples. In each sample an inter-item correlation matrix could be formed using Lertap 5's "IStats" option. The two matrices could then be picked up by SPSS's factor analysis routines and results compared. This type of exercise readily leads to topics such as the number of common factors and dimensionality; it can be particularly interesting when students compare their results. (Exercises of this sort can keep a class going for 60 to 90 minutes, giving the instructor a chance to slip out for a coffee break and a check of last night's lotto results.)

Instructors may request that the Mini version be "unlocked" to the full version for use by their students during a specified study period. Write to lertap5@gmail.com with your request.

The following resources have handouts and exercises for use with classes. While they're largely designed to be used with Lertap 5, a system such as SPSS can be applied to some of the exercises, as can Excel just by itself (without Lertap 5). Nelson's 2004 paper, "Excel as an Aide in Teaching Measurement and Research Methods" references some of these resources and exercises at greater length. Nate Thompson's little "CITAS" Excel item analysis spreadsheet may be another useful aid; other CTT and IRT systems are covered in a very useful Wikipedia webpage.

As an additional resource, this paper discusses SAS University and jMetrik as options to Lertap 5 -- these two programs will work on Macintosh computers as well as Windows and they're both free.

- The Chemistry Quiz & the "CEQ" -

We've got two little exercises made especially for use in classes. Each set comes with actual answer sheets which may be printed to simulate an authentic mini-research project -- students enter data in Excel and then get the results needed to answer the questions posed by the exercises.

The first example is from a chemistry quiz, the second involves the "CEQ", a computers-in-education survey given to a small sample of school principals. Each example can take up to 60 minutes for students to complete; having them work in pairs may make the data entry phase easier for them.

- The Lertap Quiz (November 2006 version) -

No doubt your grandparents have mentioned the world-renowned Lertap Quiz data set -- it's been with us since 1973. The original version of the dataset is described in the appendix to the manual. Here we give you the chance to download a modified version, one whose second column contains a school type code. This modified version has been created in order to provide more challenges to beginning users; the presence of a school type code can lead to the use of, for example, the 'Breakout scores by groups' option on the Run menu (this dataset may be downloaded from this link: Excel xls file).

There is also a brief codebook which goes with the Excel workbook. It may downloaded by clicking here (Word document).

We've also created some tasks to help ward off brain addle. Try your hand, for example, at our Research Questions Set A.

Those tasks were too easy, you say? Fair enough, mate, avagowiththese: Research Questions Set B.

- The Ed 501 Data Set -

Here we've got a very practical exercise built on a real test from a university educational psychology class. Use this exercise to remind yourself (and students?) how a Lertap data set is formed when given just a worksheet of item responses, and a list of the correct answers to test items. (Includes a set of research questions to answer.)

- Cognitive Holding Power Questionnaire -

This sample stems from the work of John Stevenson of Griffith University (Australia). The CHPQ is an affective instrument with thirty items. Our sample includes an Excel data set. (Web page, will open in a new browser window.)

- Concurrent validity study -

Relates results from an authentic concurrent validity study undertaken by a masters student. Has numerous samples of Lertap output, and exemplifies how to apply some of Lertap's less-used options. Demonstrates use of an external criterion score. (Web page, will open in a new browser window.)

- Parallel-forms reliability study -

Discusses the development of an academic aptitude test for high school students, based on authentic data. Gets into some practical data processing problems; shows how to use Excel's Data Filter, a powerful utility. Talks about speeded tests, and practice effects (based on the two equivalent forms developed in the study). Shows how short-answer constructed-response items may be used in Lertap 5. This is a typical example of classical test development methods, with many snapshots of Lertap results. (Web page, will open in a new browser window.)

- Importing data -

A fairly technical discussion of downloading sample datasets from the Internet in typical "text" format, and making them ready for Excel and Lertap. Shows use of "The Spreader" . Discusses general matters related to importing data, using samples from Iowa University. (Web page, will open in a new browser window.)"