﻿ Pre-scored items

# Pre-scored items

It is usually the case that the entries found in the columns of the Data worksheet correspond to the response codes selected by each person.

For example, have a squiz at this snippet from rows 2 and 3 of a Data worksheet:

On Q20, this person selected the response which had been coded as A.  Apparently s/he did not answer Q21.  On Q22, the person selected the response coded as E.

The column entries change to digits from Q26 on, but the meaning is probably the same: on Q26 the person selected the response coded as 1 (one), while on the next item, Q27, s/he chose the response which had been coded as 3.

What we're looking at here is part of the Lertap Quiz data set, described in Appendix A of the manual.  This quiz consisted of 25 cognitive items, Q1-Q25, followed by 10 affective items, Q26-Q35.  The cognitive items used letters as response codes, while the affective items were of the Likert style, with 1 the code for "strongly agree", and 5 the code for "strongly disagree".

How many points did this person get for her/his answer of A on Q20?  We don't know; we can't tell just by looking at the data above.  And, even though there are digits in some of the columns, we can't assume that a "1" for Q28 means that the person got one point for his/her answer.

These item responses have yet to be scored.

Okay?  Consider now another case.  Suppose a cognitive test included the following question:

33) Read the five sentences below, and place a tick next to those sentences which use the pluperfect tense.

Student answers to a question like this one have to be scored by hand.  How?  Well, if there were three pluperfect sentences, and the student found and ticked each, then the student would probably get 3 points.  If the student found two of the three, s/he'd get 2 points.  A student might get 0, 1, 2, or 3 points on this item.

Look now at a snippet from another Data worksheet, would you?:

In this case, language students listened to an audio tape with 10 short sentences, and they also read ten brief paragraphs.  The digits in the boxes are now real numbers -- they're not response codes -- the digits represent the number of points the student earned on each of the ten listening items, L1-L10, and the number of points earned on each of ten reading items, R1-R10.  The items in this test are said to have been "pre-scored", or "pre-coded".

The following CCs lines were used to process results for the ten Listening items:

*col (c2-c11)

*sub res=(0,1,2,3), name=(Listening AARP test), Title=(L-AARP)

*mws  c2, 0, 1

*mws  c3, 0, 1

*mws  c4, 0, 1, 2

*mws  c5, 0, 1, 2

*mws  c6, 0, 1, 2

*mws  c7, 0, 1, 2, 3

*mws  c8, 0, 1, 2, 3

*mws  c9, 0, 1, 2

*mws c10, 0, 1

*mws c11, 0, 1, 2, 3

The *mws lines indicate that the number of item points possible varied from item to item.  For the first two items, scores were limited to 0 or 1; for the items found in columns 4, 5, 6, and 9 possible scores ranged from 0 to 2.  Three items, those in columns 7, 8, and 11, had possible scores ranging from 0 to 3.

Fewer lines would have been required had it been possible to get the same number of points on each item:

*col (c2-c11)

*sub res=(0,1,2,3), name=(Listening AARP quiz), Title=(L-AARP)

*mws call, 0, 1, 2, 3

Here, the "call" form of the *mws line has been used -- "call" means "columns all".  According to these CCs lines, it's possible for a student to get up to 3 points on each item.