CORTAS compares the tasks performed by different groups of cases. It determines the "core" set of tasks which are performed by each group based on percent members performing, and identifies other tasks which help to discriminate each group from all the others.
Automatic Process Generation
If an output FACSET file is specified on the program invocation line, a SETSXX will be run automatically, listing all factors created.
Program Invocation
The specific syntax for program invocation varies between operating systems. This document diplays the code for the AIX version.
opt: D Sort descending on performance emphasis or OMC
discrimination (default = task discrimination)
G Do not print group summary report
L Print super/subgroup summary matrix
O Print OMC discrimination values instead of cumulative
percent time spent values
P Print performance emphasis values instead of cumulative
percent time spent values
Q Print "quick" task discrimination instead of cumulative
percent time spent values
T Do not print task level report
V Write output task discrimination task factor
ssss: The study number
TFf: Input/Output Task Factor file
TT: Input Task Title file
FS: Output FACSET file (optional)
CORTAS Control Card
CORTAS: cc 01-06 The literal 'CORTAS' study: cc 08-11 The study number for this run grpid: cc 13-18 The Group ID for the total sample group title: cc 20-79 The report title for this run.
CORTAS Cutoffs Control Card (optional)
CORTAS: cc 01-06 The literal 'CORTAS'
ccut: cc 08-13 Core task cutoff (minimum PMP, default = 66.66)
dcut: cc 15-20 Task discrimination cutoff (default=33.33)
ucut: cc 22-25 Task uniqueness cutoff (default=0.50).
To be unique, a task must have a uniqueness
value of at least <unicut> and a
chi-square value > 10.83. Also, the number
of group members times the number of members
in the total sample who perform the task must
be at least 25.
scut: cc 27-32 Supergroup PMP cutoff (default=0.01)
srt: cc 34 Sort column for PTS/SupSub table
(1-6, default=4)
col: cc 36 Column of PTS/SupSub table to place in summary
matrix (5-6, default = 5)
omc1: cc 38-43 Weight 1 for OMC value (default=1.0)
omc2: cc 45-50 Weight 2 for OMC value (default = maximum task
discrimination/100.0)
Complementary Group Card (optional)
This card is used only when the complementary method is desired for computing task discrimination values (as opposed to the contextual method). The group specified on this card should include all members contained in the group selection cards.
COMP: cc 01-04 The literal 'COMP'group: cc 06-11 The complementary Group ID (default is the total sample group).
Group Selection Cards
These cards specify the groups to be compared. Up to 200 groups may be specified, with 1 - 11 groups per card. There must be a PTS and a PMP vector available on the input Task Factor file for each group specified.
group1: cc 01-06 Group ID for this group group2: cc 08-13 Group ID for this group (optional) group3: cc 15-20 Group ID for this group (optional) group4: cc 22-27 Group ID for this group (optional) group5: cc 29-34 Group ID for this group (optional) group6: cc 36-41 Group ID for this group (optional) group7: cc 43-48 Group ID for this group (optional) group8: cc 50-55 Group ID for this group (optional) group9: cc 57-62 Group ID for this group (optional) group10: cc 64-69 Group ID for this group (optional) group11: cc 71-76 Group ID for this group (optional)
These cards are terminated by an end-of-file or an @eof
image.
@codap cortas - ssss tf100 tt001 CORTAS ssss ST0001 CORTAS run for job cluster task comparisons COMP ST0241 ST0196 ST0074 ST0031 ST0097 ST0046 @eof
This CORTAS uses all default values. Six task level reports will be printed using the complementary method. The group summary report will also include PMP values for ST0001. The task discrimination vectors will not be saved.
@codap cortas - ssss tf100 tt001 CORTAS ssss ST0001 CORTAS run for job cluster task comparisons CORTAS 50.00 50.00 1.00 20.00 1 6 ST0241 ST0196 ST0074 ST0031 ST0097 ST0046 @eof
This example uses the contextual method with new cutoff values and different sort columns.