AUDITR will scan raw case data input files looking for any errors in the formation of the records for each case. These errors include too many or too few cards for a given case as well as any case containing check characters which do not match. Any character found on the format cards other than a 'T', 'V', 'Z', '.', ' ' or 'B...' will be considered check characters. The input case data must contain these characters in the corresponding locations or the case will be flagged as invalid. All invalid cases will be written to the error output file. Valid output files are used only by INPSTD.
AUDITR also range checks the task responses. Any task responses which do not fall into a user specified range will be reset to a value equal to a 'non-response.' It should be noted that a 'non-response' is not equal to a zero response. Cases which have fewer than the user-specified minimum or more than the maximum percentage of valid task responses will also be written to the error file. AUDITR will read any card image raw case data input file. Each case may be composed of up to 200 records, each record containing 80 characters. The records must be in proper sequence as defined by the format cards. Each case may contain a maximum of 2000 computer words of background data. Note that a one digit background response and a 4 digit response both use one computer word each. The maximum number of task responses is 3000.
The input file should already be sorted by case control number! Otherwise, AUDITR is unable to detect duplicate entries.
opt: N NOGO option, scan control cards for errors
C Create Corrected File by Merging corrections
R Set default maximum percentage of valid task responses to 100
(used for rater data)
Z Change blank responses to zeros
ssss: The study number
CI1: Input raw case data or the valid raw data file from a previous
AUDITR run.
RDf: Output valid raw case data file
CI2: Output invalid raw case data file
CI3: Input corrected version of an invalid raw data file (CI2) from a
previous AUDITR run. If specified, <CI1> must be the
valid raw data file from the same run.
These include the AUDITR card and the format cards for AUDITR. These cards are only needed when <CI1 raw data> from the program invocation line is NOT a valid raw data file from a previous AUDITR run.
AUDITR Control Card
AUDITR: cc 01-06 The literal 'AUDITR'
study: cc 08-11 The study number for this run
nc: cc 13-15 The number of cards per case
max: cc 17-22 The maximum acceptable task response
min: cc 24-29 The minimum acceptable task response
ntask: cc 31-34 The number of tasks
nhist: cc 36-39 The number of background variables
maxp: cc 41-44 The maximum percentage of valid task responses for a
case to be considered valid. Default = 95.0, 100.0
for the 'R' option.
mint: cc 46-49 The minimum number of valid task responses for a case
to be valid. Default = 5.
Format Cards
filespec: The file containing the format cards
A description of the format cards can be found in the Common Cards document.
Examples
@codap auditr - ssss ji100 rd100 er100 AUDITR ssss 13 9 1 363 367 @add fmtcds.ji
The good output file (rd100) will receive all records for cases which are found not to be in error. The other output file (er100) will receive any cases which are found to be invalid. All cases written on the good file will have their task responses range checked and reset any which are found to be out of range. This operation will not be performed on any cases written on the error file.
@codap auditr c ssss rd100 rd101 er101 er100
This example will take the corrected version of the invalid raw data file (er100) from the first example and merge it with the valid raw data file from the same run (rd100). No control cards are required when using a valid raw data file previously created by AUDITR. The merged files will be written to the new valid raw data file (rd101) and the new invalid file (er101) will contain any cases from the previous run that were not properly corrected.