Example 9: Grouping and modeling items |
Top Up Down
A A |
A 20 item test. Items 1, 2 are dichotomous items, coded "Y", "N". Item 3 is a "Partial Credit" item. Items 4-10 are all ratings on one Andrich scale or test-level partial-credit scale (Grouping 1), and items 11-20 are all ratings on another Andrich scale or test-level partial-credit scale (Grouping 2). These are grouped with ISGROUPS=. Winsteps discovers from the data what the item structure is. Items 3-20 have response codes "A", "B", "C", "D", "E" or "a", "b", "c", "d", "e".
; This file is EXAM9.TXT
TITLE="Grouping and Modeling"
ITEM1=11 ; Item responses start in column 11
NI=20 ; There are 20 items
ISGROUPS=DD011111112222222222 ; The groupings in item order
IREFER=DDAAAAAAAAAAAAAAAAAA ; for recoding
CODES = 10ABCDEabcde ; Response codes to all items
IVALUED= 10********** ; for Items 1 & 2
IVALUEA= **1234512345 ; for Items 3-20
DATA=EXAM9DAT.TXT ; Location of data file
IWEIGHT=* ; Item weighting file list
3 2.5 ; Item 3 has weight of 2.5
*
&END
RD Prompt 1 ; Item id's remind us of MODELS= and ISGROUPS=
RD Prompt 2 ; MODELS= and ISGROUPS= are shown in item measure Tables 10, 13, 14, 15
S0 Logic
R1 Grammar 1
R1 Grammar 2
R1 Grammar 3
R1 Grammar 4
R1 Grammar 5
R1 Grammar 6
R1 Grammar 7
R2 Meaning 1
R2 Meaning 2
R2 Meaning 3
R2 Meaning 4
R2 Meaning 5
R2 Meaning 6
R2 Meaning 7
R2 Meaning 8
R2 Meaning 9
R2 Meaning 10
END NAMES
The data is in file EXAM9DAT.TXT:
Richard M 00bCDCDddCDddddCDccE
Tracie F 00BcBABBccbBbbBbBBBb
.
James M 00ccaBbabBAcbacbaBbb
Joe M 10BdBBBBccBccbbccbcC