You always know class when you see it

For classes getting into item and test analysis at an introductory level, these resources might be useful:

• Two sample data sets:"Class quiz" and "CEQ Survey". The first of these seems particularly popular, judging by number of downloads. Free. See the Samples topic.
• Some challenging exercises for aspiring data sleuths may be found under "The Lertap Quiz (November 2006)" heading over on the Samples page. Also, see the "La Florida Data Set" on the same page -- it involves item responses from over ten thousand students -- can be fun to play with. (Es tuyo, totalmente libre de impuestos.)
• A special samples website came on line in April 2011. It has 6 sample datasets which may be downloaded, and it also has a fresh look at the "Cook's Tour", a time-tested introduction to Lertap with hands-on things to do.
• The Student Version of Lertap. Free— see the Software topic. If this version's limit of 44 data records is too restrictive, perhaps have students use it to learn with, timing themselves carefully, and, when the time is right, downloading the 30-day trial version for the final bash of their data. Note that students can enter hundreds / thousands of data records in the Student Version; only the initial forty-four are processed, true, but if they do download the 30-day trial, it'll pick up and run with all the records. (Students can enter all of their data in the Student Version; they don't have to stop after 44 records.)
• The latter chapters of the manual involve a fairly thorough discussion of interpreting classical item and test statistics, and have numerous references to the literature, especially to texts used in this field. Free. See the Manual topic.
• For classes which use or make reference to Allen & Yen's still excellent 1979 text (see references), Lertap fits in at numerous spots. One particulary nifty niche relates to Allen & Yen's discussion of item characteristic curves Section 6.5 (p.127) -- Lertap's quintile plots match this section well. This topic can lead, of course, right into IRT discussion (item response theory).
• The Crocker & Algina 1986 text (see references) and Lertap also match up at various spots. The text's discussion of reliability coefficients for criterion-referenced tests, Chapter 9, p.192, closely matches Lertap's mastery test analysis features (refer to Chapter 10, p. 167 of the Lertap manual, and to one of the Lertap erudite epistles on cut-off scores). The presentation of item response theory, Chapter 15, links with Lertap in more than one spot -- for example, the discussion of the relationship between ICC parameters and classical items statistics (p.350) directly matches features discussed in another Lertap paper.

For some suggestions on using Excel in introductory classes, possibly try Nelson (2004). A copy of this paper is available via a click here (will open a Word doc of about 250 KB in a new window).

Again for introductory classes, those getting into statistics and data analysis in general, a sample data set used successfully for years in New Zealand, the "Highlands Adolescent Data Base", may be accessed here (Word doc of about 35 KB; leads on to an Excel worksheet of 85 KB). The HADB was designed to be used with calculators, but it lends itself well to Excel. Excel has some general statistical procedures, such as t-tests, ANOVA, and linear regression built in. (Treat your students to a holiday in Platisla.)

Coming back to test and measurement classes, and moving beyond the introductory level, these resources could be helpful:

• The University of Illinois IRT Modeling Lab is a particularly comprehensive website with sample data, having references to classical test theory, as well as item-response theory.
• The website maintained by Assessment Systems Corporation has an extensive list of resources in this area; most are commercial, but not all.
• The Foxes Group in Europe has some very useful matrix manipulation routines freely available. These are referenced in Lelp, where there are some suggestions as to how to use the Foxes routines for factor analysis.
• There are numerous Excel-based resources available on the Internet; for example, there are several papers on how to make stem-and-leaf diagrams, and box plots.
• Some of the references on this site's "erudite epistles" list may be useful with advanced measurement classes. The same might be said of the samples list.