Trailing-Edge
-
PDP-10 Archives
-
decuslib20-02
-
decus/20-0047/survey.txt
There are 2 other files named survey.txt in the archive. Click here to see a list.
SURVEY
------
THE FORTRAN PROGRAM SURVEY IS DESIGNED TO BE USED IN CONNECTION
WITH THE PROGRAMS SORTER AND CROSS. IT READS CODED OBSERVATIONS
OR RESPONSES INTO A DATA FILE SO THAT THEY CAN BE TABULATED OR
ANALYZED BY THE LATTER PROGRAMS. WHILE IT IS EXPECTED THAT THE
DATA WILL BE READ FROM CARDS, THIS IS NOT ABSOLUTELY NECESSARY.
PROGRAM SURVEY
DESCRIPTION
THIS PROGRAM READS CODED RESPONSES TO A QUESTIONNAIRE AND WRITES
THEM INTO AN ASCII DATA FILE.
SOURCE
NORMAN W. JOHNSON, DEPARTMENT OF MATHEMATICS, WHEATON COLLEGE,
NORTON, MASS.
INSTRUCTIONS
THERE SHOULD BE NO MORE THAN 128 QUESTIONS. RESPONSES MAY BE
CODED BY ANY DIGIT FROM 0 TO 9. ADDITIONAL SYMBOLS THAT MAY BE
USED ARE THE MINUS SIGN (-), THE AMPERSAND (&), AND THE BLANK,
INTERPRETED RESPECTIVELY AS THE NUMBERS 11, 12, AND 13. THESE
SHOULD BE RESERVED FOR SUCH CATEGORIES AS "NOT APPLICABLE",
"DON'T KNOW", AND "NO RESPONSE".
THE FIRST 5 COLUMNS OF EACH DATA CARD ARE FOR IDENTIFICATION,
WITH THE FIRST 4 COLUMNS BEING UNIQUE TO THE RESPONDENT. THE
NEXT 67 COLUMNS OF THE CARD ARE FOR THE CODED RESPONSES, ONE
COLUMN PER QUESTION. THE LAST 8 COLUMNS ARE NOT USED. IF THERE
ARE MORE THAN 67 QUESTIONS, A SECOND CARD FOR EACH RESPONDENT
MAY BE PREPARED IN THE SAME MANNER.
THE SET OF DATA CARDS MUST BE PRECEDED BY FIVE OR SEVEN PROGRAM
CARDS. THE FIRST OF THESE HAS PUNCHED IN ITS FIRST 5 COLUMNS
THE NAME THAT THE DATA FILE IS TO HAVE. THE SECOND MAY BE USED
FOR A DESCRIPTION OF THE SURVEY (NOT MORE THAN 48 CHARACTERS).
THE THIRD CARD SPECIFIES THE NUMBER OF QUESTIONS (ANY FORMAT
IS ACCEPTABLE). THE LAST TWO CARDS OR PAIRS OF CARDS, PUNCHED
IN THE SAME FORMAT AS THE DATA CARDS, INDICATE RESPECTIVELY THE
LOWEST NUMBER AND THE HIGHEST NUMBER TO BE COUNTED AS A RESPONSE
TO EACH QUESTION. NUMBERS OUTSIDE THIS RANGE ARE TREATED AS "NO
RESPONSE" AND EXCLUDED FROM THE TOTAL. ON THESE CARDS BLANKS
AND MINUS SIGNS ARE EQUIVALENT TO ZEROS, AND AMPERSANDS MAY NOT
BE USED.
THE LAST DATA CARD MUST BE FOLLOWED BY A SINGLE BLANK CARD OR A
PAIR OF BLANK CARDS, ACCORDING AS ONE OR TWO CARDS ARE USED FOR
EACH RESPONDENT.
THIS PROGRAM ASSUMES THAT INPUT IS FROM THE USER TERMINAL AND
OUTPUT IS TO THE DISK. IF A DIFFERENT INPUT DEVICE, SUCH AS THE
SYSTEM CARD READER, IS TO BE USED, IT SHOULD BE ASSIGNED LOGICAL
UNIT 5 PRIOR TO RUNTIME. ALSO, FOR INPUT FROM THE CARD READER,
THE USER SHOULD RESPOND TO THE PROGRAM'S REQUEST FOR DATA BY
TYPING "@CDR:". FOR INPUT FROM A FILE ON THE DISK, THE USER
SHOULD TYPE "@DSK:FILNAM.EXT", WHERE 'FILNAM.EXT' IS THE NAME
OF A CARD-IMAGE SOURCE FILE.
SPECIAL INSTRUCTIONS FOR NONSTANDARD SOURCE DECKS
DATA CARDS DO NOT NEED TO BE PREPARED IN EXACT ACCORDANCE WITH
THE ABOVE INSTRUCTIONS TO BE ACCEPTABLE, PROVIDED THAT (1) THERE
ARE NO MORE THAN TWO CARDS PER RESPONDENT, (2) EACH CARD CON-
TAINS AN IDENTIFICATION LABEL, AND (3) THE ONLY RESPONSE CODES
USED ARE DIGITS, MINUS SIGNS, AMPERSANDS, AND BLANKS. WHEN A
NONSTANDARD SOURCE DECK IS INPUT, THE PROGRAM CARD SPECIFYING
THE NUMBER OF QUESTIONS MUST BE REPLACED BY A FORMAT CARD GIVING
THE BLOCKS OF COLUMNS IN WHICH IDENTIFICATION AND RESPONSES ARE
PUNCHED ON EACH DATA CARD.
FOR EXAMPLE, IF THE SOURCE DECK CONTAINS ONE CARD PER RESPONDENT
WITH RESPONSES PUNCHED IN COLUMNS 1 THROUGH 72 AND IDENTIFICA-
TION IN COLUMNS 74 THROUGH 80, THE FORMAT CARD SHOULD READ:
74-80,1-72
FOR A SOURCE DECK CONTAINING TWO CARDS PER RESPONDENT, HAVING
IDENTIFICATION IN THE FIRST 4 COLUMNS OF EACH CARD AND RESPONSES
IN COLUMNS 5 THROUGH 80 OF THE FIRST CARD AND 5 THROUGH 64 OF
THE SECOND CARD, BUT WITH THE LAST 8 RESPONSES TO BE IGNORED,
THE FORMAT CARD SHOULD READ:
1-4,5-80/1-4,5-56
RELATED PROGRAMS
THE PROGRAM SORTER TABULATES THE RESPONSES TO EACH QUESTION OF
THE QUESTIONNAIRE BY FREQUENCIES AND PERCENTAGES. THE PROGRAM
CROSS CORRELATES RESPONSES TO SELECTED QUESTIONS TO GIVE MAR-
GINAL FREQUENCIES, CROSS-TABULATIONS, AND OTHER STATISTICS AND
PERMITS THE COLLAPSING OF TABLES AND THE COMBINING OF VARIABLES.
THE PROGRAM MERGE COMBINES UP TO 64 FILES OF RESPONSES TO THE
SAME SET OF QUESTIONS. THE PROGRAM UNITE COMBINES TWO FILES OF
RESPONSES TO DIFFERENT QUESTIONS BY THE SAME RESPONDENTS.