SAS Dataset and Analysis
Source: http://web.archive.org/web/20061230031910/http://support.sas.com/ctx/samples/index.jsp?sid=478&tab=output
EXAMPLE:
These examples include the CORR procedure to
validate the summary statistics for the test score and the item-total
correlations. The ALPHA option is also used to validate KR20 (which
is equivalent to Cronbach's alpha) and the item-remainder correlations.
PROC FORMAT; VALUE $Q
'0'='1' '1'='2' '2'='3' '3'='4' '4'='5'
'5'='1' '6'='2' '7'='3' '8'='4' '9'='5';
RUN;
DATA TEST1;
INPUT SSN $ 12-20 (Q1-Q33) (33*$CHAR1. );
BY = 1 + (_N_ > 21);
FORMAT Q1-Q33 $Q.;
CARDS;
101 999999990372715151617250838251535163517182
101 200465598350507250617350638261536171517183
101 233868923373525163817250738063517251538173
101 372527163618252828251516161517352
101 236909800362505173617250718263837161537361
101 235884995360505282717350836051715161538181
101 270606129370835151817153736051835071537171
101 188447596372517162617252738251535151537152
101 054389259372515250817350738351535071538153
101 130463838373825172618250838050818171537183
101 165463286371517253616253838251516253836251
101 2176696113728 5383608353836261737051538153
101 3528351526151508382518151618171 2
101 228761635382825261717272835151525171538351
101 236869018372525062817251725251715250538182
101 206422209372825263817250838053515170537182
101 081481581363815151717250825251635161517172
101 236562064372635263618170725252525271537363
101 296549834373515050817153738061815273518151
101 234908052372515162817150836280535071517151
101 407681435372815163617150738261815071517151
101 271535281817251 38251515161537161
101 21728907337261"151618 50725281535171507352
101 37252"151617252728151737161807360
101 234171262370735162817152725150737071518373
101 360505062817052838081815251537361
101 059486239372818262617252838051515061517152
101 232721382372718051617253828051735173537083
101 17638728 372835173617270838050516361537152
101 371527162617250838251816371707383
101 194381114372537151617252828251715161537162
101 198424787372835061617272838051738161517162
101 190521297372718162817153718251835171537152
;
%inc "<location of your file containing the ITEM macro>";
title 'ITEM macro: Original test data';
title2 'Test 1';
%item(
data=test1,
var=q1-q33,
id=ssn);
title2 'Test 2';
%item(
data=test1(firstobs=2),
var=q1-q33,
response='12345',
key='43332121222331144'||
'4312141224123243',
id=ssn,
out=outtest,
outitem=outitem,
outbin=outbin,
score=right,
missing=blank,
invalid=bad,
percent=pcnt,
prophecy=.75,
uplo=.27);
proc print data=outitem; id _item_ _name_ _respon_; run;
proc print data=outtest; run;
proc corr data=outbin alpha nomiss nocorr nosimple; var bin:; run;
proc corr data=outbin nomiss; var bin:; with right; run;
*****************************************************************;
title2 'ITEM macro: Subtests';
%item(
data=test1,
var=q1-q33,
subtest=_all_ 'English Final'/
q1-q10 'Spelling Proficiency'/
q11-q33 'Grammar Proficiency',
subname=_sub_,
out=outsub,
id=ssn);
proc print data=outsub; id _sub_ ssn; run;
proc corr data=outbin alpha nomiss nocorr nosimple;
var bin1-bin10; run;
proc corr data=outbin alpha nomiss nocorr nosimple;
var bin11-bin33; run;
*****************************************************************;
title 'ITEM macro: BY groups';
title2 'Two BY groups';
%item(data=test1,var=q1-q33,by=by,id=ssn);
*****************************************************************;
title 'ITEM macro: Subtests and BY groups';
%item(by=by,
data=test1,
var=q1-q33,
subtest=_all_ 'English Final'/
q1-q10 'Spelling Proficiency'/
q11-q33 'Grammar Proficiency',
id=ssn);
RESULTS:
Following is abbreviated output from the first run of the %ITEM macro in this example. The output would continue through item Q33.
ITEM macro: Original test data
Test 1
Grade Report
SSN Score #Missing #Invalid Percent
200465598 19 0 0 58%
233868923 14 0 0 42%
20 0 0 61%
236909800 17 0 0 52%
235884995 16 0 0 48%
270606129 17 0 0 52%
188447596 23 0 0 70%
054389259 19 0 0 58%
130463838 18 0 0 55%
165463286 16 0 0 48%
217669611 13 1 0 39%
22 1 0 67%
228761635 15 0 0 45%
236869018 16 0 0 48%
206422209 21 0 0 64%
081481581 24 0 0 73%
236562064 13 0 0 39%
296549834 16 0 0 48%
234908052 20 0 0 61%
407681435 20 0 0 61%
19 1 0 58%
217289073 20 1 1 61%
18 0 1 55%
234171262 12 0 0 36%
12 0 0 36%
059486239 22 0 0 67%
232721382 23 0 0 70%
17638728 20 0 0 61%
17 0 0 52%
194381114 23 0 0 70%
198424787 22 0 0 67%
190521297 20 0 0 61%
ITEM macro: Original test data
Test 1
Summary of Test Statistics
Number of Items 33
Number of Subjects 32
Mean Score 18.344
Variance of Scores 11.523
Standard Deviation 3.395
Kuder-Richardson 20 0.433
Kuder-Richardson 21 0.279
Standard Error (from KR20) 2.555
Standard Error (from KR21) 2.883
Minimum Score 12
Maximum Score 24
Total Number of Answers 1056
Number Missing 4
Number Invalid 2
Spearman-Brown Prophecy (from KR21):
To obtain a reliability of 0.90
the test should contain 769 items
ITEM macro: Original test data
Test 1
Test Frequency Distribution
Raw Standard Per- Cum
Score Score Centile Percent Freq Freq Each X represents 1 score
24 667 98 3.1% 1 32 X
23 637 92 9.4% 3 31 XXX
22 608 82 9.4% 3 28 XXX
21 578 76 3.1% 1 25 X
20 549 65 18.8% 6 24 XXXXXX
19 519 51 9.4% 3 18 XXX
18 490 43 6.3% 2 15 XX
17 460 35 9.4% 3 13 XXX
16 431 25 12.5% 4 10 XXXX
15 401 17 3.1% 1 6 X
14 372 14 3.1% 1 5 X
13 343 9 6.3% 2 4 XX
12 313 3 6.3% 2 2 XX
ITEM macro: Original test data
Test 1
Item 1: * is keyed 1 2 3 4* 5 Omit Invalid
Q1 Responses 0.0% 0.0% 3.1% 96.9% 0.0% 0.0% 0.0%
Upper 33% 0.0% 0.0% 0.0% 100.0% 0.0% 0.0% 0.0%
Lower 33% 0.0% 0.0% 0.0% 100.0% 0.0% 0.0% 0.0%
Item-Total:
Point Bis 0.000 0.000 0.035 -0.035 0.000
Biserial 0.000 0.000 0.087 -0.087 0.000
Item-Remainder:
Point Bis 0.000 0.000 0.087 -0.087 0.000
Biserial 0.000 0.000 0.215 -0.215 0.000
t test 0.000 0.000 0.479 -0.479 0.000
Prob>|t| 1.0000 1.0000 0.6356 0.6356 1.0000
Item 2: * is keyed 1 2 3* 4 5 Omit Invalid
Q2 Responses 6.3% 12.5% 78.1% 3.1% 0.0% 0.0% 0.0%
Upper 33% 12.5% 12.5% 75.0% 0.0% 0.0% 0.0% 0.0%
Lower 33% 0.0% 20.0% 70.0% 10.0% 0.0% 0.0% 0.0%
Item-Total:
Point Bis 0.167 -0.124 0.077 -0.180 0.000
Biserial 0.328 -0.199 0.108 -0.444 0.000
Item-Remainder:
Point Bis 0.227 -0.035 -0.047 -0.138 0.000
Biserial 0.448 -0.057 -0.065 -0.341 0.000
t test 1.279 -0.194 -0.256 -0.763 0.000
Prob>|t| 0.2106 0.8474 0.7994 0.4513 1.0000
... etc.