A
achievement tests
about 117
multiple-choice (MC) items 118–121
American Educational Research Association 2
American Psychological Association 2
analysis, classical item
combining item statistics in single tables 66–72
combining mean score with frequency counts 61–63
computing proportion correct by quartile 63–66
displaying mean score 55–60
interpreting item statistics 72–73
point-biserial correlation coefficient 51–53
producing attractive reports 53–54
restructuring data set 54–55
on 30-item physics test 112
analyzing tests with multiple versions 80–86
answer frequencies
about 31
developing automated programs to score tests and produce item frequencies 34–36
displaying in graphical form 36–40
displaying in tabular form 31–32
modifying programs to display answers in frequency tables 33–34
answer keys, reading from separate files 16–17
ARRAY statements 12, 33, 81
assessing test reliability
See test reliability
B
bar charts, producing 39–40
Birnbaum, Allan 98
BY statement 25
C
Cartesian products, creating 46
cheating
detecting on multiple-choice tests 123–140
programs to detect 163–170
programs to search for possible 170–173
classical item analysis
combining item statistics in single tables 66–72
combining mean score with frequency counts 61–63
computing proportion correct by quartile 63–66
displaying mean score 55–60
interpreting item statistics 72–73
point-biserial correlation coefficient 51–53
producing attractive reports 53–54
restructuring data set 54–55
on 30-item physics test 112
classical test theory (CTT), compared with Item Response Theory (IRT) 99–100
Cody, Ron
Learning SAS By Example: A Programmer's Guide 3
COLUMNS statement 27
combining
item statistics in single tables 66–72
mean score with frequency counts 61–63
comma-delimited data (CSV file), reading 12–13
computing
correlations between odd/even scores 90
Cronbach's Alpha 93
Kuder-Richardson Formula 20 (KR-20) 92–93, 161–162
mean score 57–58, 134
proportion correct by quartile 63–66
Spearman-Brown Adjusted Split-Half Correlation 90–91
split-half reliability 88–91
standard deviation 134
constructed response 119
CORR procedure
computing biserial correlations 71
computing correlations between odd/even scores 90
computing Cronbach's Alpha 93
computing Spearman-Brown Adjusted Split-Half Correlation 91
point-biserial correlation coefficient 51–53
producing attractive reports 53
CORRESPONDANCE array 81–82
creating
Cartesian products 46
rosters using REPORT procedure 27
Cronbach's Alpha, computing 93
CSV (comma-delimited) file, reading 12–13
CTT (classical test theory), compared with Item Response Theory (IRT) 99–100
D
data
comma-delimited 12–13
reading directly from Excel workbooks 14–16
reading from text files 6–10
space-delimited 10–12
data checking program 150–151
data sets
preparing for IRT procedure 102–103
restructuring 54–55
DBMS= option 28
DEFINE statement 27
deleting items from scoring programs 78–80
Delwiche, Laura
The Little SAS Book: A Primer, Fifth Edition 3
demonstrating
effect of item discrimination on test reliability 94–95
effect of test length on test reliability 95
detecting
cheating on multiple-choice tests 123–140
invalid IDs and answer choices 42–43
developing automated programs to score tests and produce item frequencies 34–36
displaying
answer frequencies in graphical form 36–40
answer frequencies in tabular form 31–32
histograms of test scores 20–24
mean score 55–60
DO loop 9, 79
duplicate records, checking for and eliminating 47–48
E
eliminating duplicate records 47–48
error checking
detecting invalid IDs and answer choices 42–43
duplicate records 47–48
for ID errors 43–45
identifying invalid IDs with "fuzzy" matching 45–47
test data 41–49
Excel files
analyzing tests with multiple versions 82–86
exporting student rosters to 28
reading data directly from workbooks 14–16
scoring multiple test versions from 157–161
scoring tests from 146–148
EXPORT procedure 28
exporting student rosters to Excel 28
F
FILE PRINT statement 43
files
Excel 14–16, 82–86, 146–148, 157–161
reading answer keys from separate 16–17
text 6–10, 144–145, 155–157
FINDC function 77
FREQ procedure
determining answer frequencies 56–57
displaying answer frequencies 31–32
inspecting variable score 104–105
modifying programs to display answers in frequency tables 34
frequency counts, combining with mean score 61–63
frequency tables, modifying programs to display answers in 33–34
functions
See specific functions
"fuzzy" matching 45–47
G
graphical form, displaying answer frequencies in 36–40
GROUPS= option 64
H
HBAR statement 36, 40
HISTOGRAM statement 20, 24
histograms
displaying of test scores 20–24
plotting 134
producing 128
I
ICC (item characteristic curve) 100
IDs
error checking 43–45
identifying invalid with "fuzzy" matching 45–47
matching student names with 24–26
indirect addressing 82
INFILE statement 9, 12, 17, 24, 44, 77, 82
INPUT statement 12, 77
interpreting item statistics 72–73
IRT procedure, running 104–110
See also Item Response Theory (IRT)
item analysis
about 1–2
program for 151–153
item baggage 120
item characteristic curve (ICC) 100
item discrimination, demonstrating effect of on test reliability 94–95
item frequencies, producing 34–36
Item Response Theory (IRT)
about 97–99
advanced 102
basics of 99–100
classical item analysis on 30-item physics test 112
compared with classical test theory (CTT) 99–100
preparing data sets for IRT procedure 102–103
results 100–101
running IRT procedure 104–110
running other models 111
item statistics
combining in single tables 66–72
interpreting 72–73
J
joint-wrongs 123, 131
K
Kane, M.T.
"Validation" 2
KEEP statement 89
Kuder-Richardson Formula 20 (KR-20), computing 92–93, 161–162
kurtosis statistic 21
L
LABEL statement 10
latent trait theory
See Item Response Theory (IRT)
Learning SAS By Example: A Programmer's Guide (Cody) 3
LENGTH statement 12, 25
%LET statement 68, 79
LIBNAME statement 15, 86
libref 15
list input 11
The Little SAS Book: A Primer, Fifth Edition (Delwiche and Slaughter) 3
Lord, Frederick 98
M
matching student names with student IDs 24–26
mean score
combining with frequency counts 61–63
computing 134
displaying 55–60
MEANS procedure
computing Kuder-Richardson Formula 20 (KR-20) 92–93
computing mean score 57–58, 134
computing proportion correct by quartile 65
computing standard deviation 134
MERGE statement 26
MISSOVER option 12
MODEL statement 111
modern test theory
See Item Response Theory (IRT)
modifying
programs to display answers in frequency tables 33–34
programs to score tests of arbitrary numbers of items 17–19
scoring program to accept alternate correct answers 76–78
multiple-choice (MC) items, writing
about 115, 118–120
achievement tests 117–121
getting organized 116–117
taxonomy of objectives and items 116–117
test blueprints 116
multiple-choice tests
detecting cheating on 123–140
searching for matches 136–140
N
named literal 15
National Council on Measurement in Education 2
NOCUM option 32
NOPRINT option 53
NOTDIGIT function 43
O
one-parameter model (1PL) 98
options
See specific options
OUTFILE= option 28
P
PLOTS= procedure 105
plotting histograms 134
point-biserial correlation coefficient 51–53
preparing data sets for IRT procedure 102–103
PRINT procedure 10, 19
printing rosters 148–149
procedures
See specific procedures
producing
attractive reports 53–54
bar charts 39–40
histograms 128
item frequencies 34–36
programs
See also scoring program
data checking 150–151
to delete items 154–155
to detect cheating 163–170
developing to score tests and produce item frequencies 34–36
item analysis 151–153
modifying to display answers in frequency tables 33–34
modifying to score tests of arbitrary numbers of items 17–19
to rescore tests 154–155
to search for possible cheating 170–173
PUT statement 43, 86, 91
Q
quartile, computing proportion correct by 63–66
R
RANK procedure 63–64, 68
Rasch, George 98
Rasch model 98
reading
answer keys from separate files 16–17
comma-delimited data (CSV file) 12–13
data directly from Excel workbooks 14–16
data from text files 6–10
space-delimited data 10–12
recognition format 117
reliability
See test reliability
REPLACE option 28
REPORT procedure 27
reports, producing attractive 53–54
%RESCORE macro 94
rescoring tests 78–80, 154–155
RESFUNC= option 111
restructuring data sets 54–55
RETAIN statement 9
Review of Research in Education (Shepard) 2
rosters
creating using REPORT procedure 27
exporting to Excel 28
printing 148–149
RUN statement 10
running
IRT procedure 104–110
other models 111
S
SAS 2–3
See also specific topics
SAS 9.4 Language Reference: Concepts 3
SAS 9.4 Macro Language Reference 3
SCAN function 24
%SCORE macro 20
scoring
analyzing tests with multiple versions 82–84
multiple test versions from Excel files 157–161
multiple test versions from text files 155–157
tests 6–10, 34–36, 144–148
tests from Excel files 146–148
tests from text files 144–145
scoring program
about 75
analyzing tests with multiple versions 80–82, 84–86
deleting items 78–80
modifying to accept alternate correct answers 76–78
rescoring test 78–80
scree plot approach 108
searching for matches in multiple-choice tests 136–140
SET statement 15
SGPLOT procedure
displaying answer frequencies in graphical form 36
plotting histograms 134
producing bar charts 39–40
producing histograms 128
SHEET= option 28
Shepard, L.A.
Review of Research in Education 2
skewness 21
Slaughter, Susan
The Little SAS Book: A Primer, Fifth Edition 3
SORT procedure 10, 25, 38, 47–48
space-delimited data, reading 10–12
Spearman-Brown Adjusted Split-Half Correlation, computing 90–91
Spearman-Brown formula 88
SPEDIS function 46–47
split-half reliability, computing 88–91
SQL procedure 46
standard deviation, computing 134
Standards for educational and psychological testing 2
statements
See specific statements
student names, matching with student IDs 24–26
SUBSTR function 33–34
swap-and-drop operation 86
T
tables, combining item statistics in single 66–72
TABLES statement 32, 61
tabular form, displaying answer frequencies in 31–32
TABULATE procedure
combining mean score with frequency counts 61
computing mean value 72
taxonomy, of objectives and items 116–117
test reliability
about 1–2, 87
computing Cronbach's Alpha 93
computing Kuder-Richardson Formula 20 (KR-20) 92–93
computing split-half reliability 88–91
demonstrating effect of item discrimination on 94–95
demonstrating effect of test length on 95
test-retest reliability 87
tests
achievement 117–121
analyzing with multiple versions 80–86
blueprints for 116
demonstrating effect of length of on test reliability 95
displaying histograms of scores 20–24
error checking data 41–49
multiple-choice (MC) 115–120, 123–140
rescoring 78–80, 154–155
scoring 6–10, 34–36, 144–148
scoring from Excel files 146–148
scoring from text files 144–145
text files
analyzing tests with multiple versions 80–82
reading data from 6–10
scoring multiple test versions from 155–157
scoring tests from 144–145
three-parameter model (3PL) 98
See also Item Response Theory (IRT)
TRANSPOSE procedure 54–55
two-parameter model (2PL) 98–99
See also Item Response Theory (IRT)
U
unidimensionality 98
UNIVARIATE procedure 23
V
"Validation" (Kane) 2
VAR statement 20, 52, 61
VBAR statement 36
W
WITH statement 52
workbooks (Excel), reading data directly from 14–16
Wright, Benjamin 98
writing multiple-choice (MC) items
See multiple-choice (MC) items, writing
3.21.246.223