We've given you two really straightforward examples.  In our experience, the two simple examples you've seen would be very much like what maybe a quarter of Lertap users regularly live and breathe.  But Lertap was designed to handle more complex situations.  Those little CCs lines can pack more punch than what's been on display in the previous two samples.

 

As an example, look at this CCs worksheet:

 

 

There are 11 lines in use in this CCs example.  Four (4) of the lines are comments; these are the lines which do not begin with an asterisk.  The use of comments is entirely optional, but they can be real helpful.

 

There are two *col lines above.  Each of these defines a group of items which will be processed together, as a unit.  Such units are generally referred to in Lertap as "subtests".

 

A subtest may be comprised of cognitive items, or it may be comprised of affective items.

 

You've already seen examples of the CCs lines used with these two types of subtests.  We've pointed out that cognitive subtests will always have a *col line and a *key line, while affective subtests will always have a *col line and a *sub line.

 

In the example above, a *sub line has been used with the cognitive subtest for several reasons.  In this example, some of the items had as many as six possible responses -- that is, some of the multiple-choice items used in this subtest had six choices, or responses.

 

The Res=(A,B,C,D,E,F) declaration tells Lertap this.  Without an Res= declaration, Lertap assumes Res=(A,B,C,D), the default set of item response codes for cognitive items.  (The default set for affective items is Res=(1,2,3,4,5).)

 

The *key line gives the right answer for each of the 25 items in the first subtest.  The *alt line tells Lertap that the items used a different number of the six possible responses.  For example, the first item used just 3 responses: they would be A, B, and C, the first three characters found in the Res= declaration.  The second item used 5 responses: A, B, C, D, and E.  Only two of the 25 items made use of all 6 possible responses.

 

The Name= declaration provides a brief description of each subtest, while the Title= declaration gives a short title.  When Name is used, some of Lertap's reports display Name on their top line.  Title, when used, appears in some reports as a label for subtest scores.

 

Subtest scores?  What are they?  Well, as an example, on a multiple-choice test people usually get one point for each right answer.  On a 25-item test, or "subtest", it would be possible to get a score of 25, assuming one point for each correct answer (Lertap permits the right answer to have any number of points, and it even allows the wrong answers to have points too -- sometimes wrong answers are penalised by assigning them negative points).

 

Let's look now at the affective subtest defined above.  The *col line points to 10 (ten) columns, c28 through c37.  A *sub line is required for affective subtests, and it must have the "Aff", or "Affective", control word on it.  This is seen above.  In this case, the *sub line has also been used to assign a Name and a Title.

 

What's that *pol line doing?  To answer this question, we have to return to the matter of scoring.  Affective items do not have a correct answer.  It's customary to give a certain number of points for each of an affective item's answers.

 

What were the possible response choices for this set of 10 affective items?  Very good question.  There is no Res= declaration on the *sub line, and, in this (common) case, Lertap assumes Res=(1,2,3,4,5).  Each affective item had five possible responses.  Unless you say otherwise, Lertap will give one point if someone selects 1 as their response; two points when someone selects 2; and so on.  This is called "forward scoring".  On a 10-item affective subtest with five response choices per item, the maximum possible score would be 50; the minimum possible would be 10.

 

The *pol line allows each affective item to be reverse-scored, if wanted.  On a reverse-scored item, the first possible response will get 5 points, not 1.  The last (fifth) response will get just 1 point, not 5.  Reverse scoring for affective items is pretty common.

 

The *pol line above indicates that the first item is to be forward-scored, while the next four items are to be reverse-scored.  This subtest has 10 items, so there are 10 + or - (minus) symbols shown on the *pol line.  (In fact, for this subtest, four items are forward-scored, while six are reverse-scored.)

 

Lertap's forte is in the flexibility it provides for item scoring.  Any response to any item can have any "weight", that is, any number of points.  In the three examples we've presented thus far, including the one above, items are being scored in a conventional manner.  Departures from normal are supported by the use of other CCs control lines, such as *wts and *mws.

 

Are we going to get away without talking about the Wt= declaration seen on the two *sub lines? No siree Bob; here goes: whenever multiple subtests are scored, Lertap will add up all the subtest scores to make a "Total", or "composite", score for each person.  Each subtest ordinarily comes into the composite with a weight of one (1); to keep a subtest out of the total score, Wt=0 is used.  In the example above, both subtests have been given a weight of zero, and Lertap will not make its Total score.

 

One final point which people often ask about ... there are spaces in the *key line above, in the *alt line, and also in the *pol line.  There's a space after every five characters in each of these lines.  Why?  Simply to make the line a bit more legible.  The spaces are not required.

 

Is the example above a common one?  Yes and no.  Yes in the sense that Lertap users frequently have more than one subtest to process, no in the sense of mixing subtest types -- this example has a cognitive subtest, and an affective subtest: a mix of subtest types -- that's quite uncommon.  If you browse on into the following topics, you'll see a couple of other examples.

 

---

Related tidbit:

 

For a really bonza example of a job which worked Lertap's CCs lines close to the limit, have a look at "Using Lertap in a Parallel-Forms Reliability Study", a 16-page Word document available via the Internet: click here if you're connected.

 

The Total score, a composite formed by summing subtest scores, gets further mention here.