Dataset details: CRDA/35/DS/1

This dataset was transferred from the Essex Data Archive on a CD-ROM which was received by NDAD on 29 February 2000.

The dataset contains data collected from interviews and questionnaires. The data was collected during the period 1986 - 1987. Waves 1-4 were conducted with adults aged 18-64 who were interviewed in their own homes throughout England and Wales and homosexuals who visited gay pubs and clubs in England. Waves 3 and 4 were additionally conducted with young people aged 13-17 and 18-21.

The dataset was transferred to NDAD as four column binary data files with each one containing information for each wave of the evaluation. The data was originally recorded on punched cards. Punch cards have 80 columns each of which have 12 rows. The rows are labelled from the top to the bottom V, X, 0, 1-9 with the V punch sometimes called Y or 12 or & and the X punch - or 11. The data was stored so that each question corresponds to one or more columns, and depending on the question the punch in the column will represent either a number or response codes. This allows for multipunch data, where a question is coded but a respondent may have multiple answers.

An octal dump, done on unix, was performed on each of the data files. This outputted the data in hexadecimal format, where each four digit hexadecimal number represented one column of a card, the first pair encodes rows V, X, and 0-3 and the second pair encodes rows 4-9. Two 8-bit bytes are needed to store each 12 row column so there are two unused bits in each of the bytes. Using perl programs the first column of the octal dump, the offset into the file, was eliminated and the actual data was rearranged so that every 10 rows was put onto one line, with a comma used to separate each four digit hexadecimal number. This meant that each line now contained data that was originally stored on a single card. Each of the four wave files were then separated into different data files so that each file contained information for a different version of the questionnaire. The file relating to Wave 1 of the evaluation contained information relating to two versions of the questionnaire, Version 1 and Version 2. In Wave 2 of the evaluation Versions 3 and Versions 4 of the questionnaire were used, Wave 3 contained Version 5, 6, 7 and 8 and Wave 4 contained Versions 12, 13, 14, 15 and 16.

The first 15 columns of each card used in each version contained details of the J.N. (NDAD think this was an internal numbering system. We have no further information), Respondent Number, Questionnaire number, the Version and the Card number. These columns were therefore all singly punched so a perl program was used to translate each four digit hexadecimal number to a single decimal number, according to which row was punched in that column, using a simple search and replace method written in perl. This made it easier to identify how many cards were used for each respondent and the next step was to ensure that each respondent occupied only one line of the file. Each respondent used different numbers of cards, a core set of 2,3 or 4 cards that contained answers to questions that were asked of every respondent, 1 or 2 cards that contained details of weights applied to each respondent and an additional number of cards that represented the coloured booklets that only some respondents had to answer. The table below shows the card sequence for each version of questionnaire:

Perl programs were written for each of the version files so that each respondent was counted as one record and therefore only occupied one line of the file, with blank fields used to pad out the data since not all respondents answered the additional booklets. Now each of the version files were ready for the main transformation. The questionnaires were used to identify which rows of which columns were used for each question. These are marked on the questionnaires either beside the question or on the sorting boards. Some columns were transformed from the four digit hexadecimal number using a search and replace method: i) Where the column is singly punched and each punch represents a number, the four digit hexadecimal number was translated into this number thus a single field was made from this column. ii) Where the column is singly punched and each punch represents a coded response, the four digit hexadecimal number was translated into the response code thus a single field was made from this column with a number of coded responses.

Otherwise each column i.e. each four digit hexadecimal number was taken in turn, and turned into a binary string, giving 1's in rows that were punched and 0's in rows that weren't. From this the main types of transformation were; iii) Where the column is multi-punched, the question is coded but allows multiple responses, the bits needed were pulled from the binary string and were rearranged in the appropriate order and thus were translated into the appropriate number of fields with the binary representing yes/no answers. iv) Where a group of multi-punched columns were taken together to code a set of mutually exclusive answers to a group of questions, these columns typically used sorting boards and were grouped on the columns by answer rather than by the question. All the rows needed were pulled out of the set of binary numbers and a transpose was performed and codes were applied to the data, so that it resulted in a set of fields, one for each question, with codes used to define the answers.

• There are also cases where the transformed data needed to be concatenated eg respondent number uses three columns, one for each digit of the number so these columns were concatenated to make one field.
• It was necessary to use codes in ascending powers of 2, i.e. 1, 2, 4, 8, 16, 32 etc for the majority of coded fields, as, even though the set of answers were mutually exclusive there were cases where a respondent had more than one answer. So in these cases the data will show a number which is the sum of all the codes used, eg, if a respondent answers both the response coded 4 and the response coded 16 then this will be shown as 20. All cases where multiple answers were given are noted in Content Validation.
• In cases where a column for a field wasn't punched then all resulting fields from this column will be defined as missing and will be blank. There are also a number of columns on the cards that were not allocated to a question, so these haven't been included in the transformation.

Documents relating to the AIDS Advertising Evaluation dataset have been transferred to NDAD. See the catalogue of dataset documentation for further details. In particular, users are advised to consult the survey questionnaires which can be seen via the Dataset Documentation Catalogue.

The dataset is also held at the Data Archive at the University of Essex where it has been given the reference number 2308

Publications were produced by the principal investigators, these were: AIDS: monitoring response to the public education campaign, February 1986 - February, 1987 (HMSO, 1987) AIDS advertising campaign : report on four surveys during the first year of advertising 1986-1987 prepared for the Central Office of Information by the British Market Research Bureau Limited, London, July 1987.

The dataset comprises the following table(s):

Version 1

• The questionnaire stated that question 25 and 26 were not asked of respondents and therefore columns 27-32 of card 2 should not have been punched. Although a number of respondents had punches on row 2 of column 27 so this column was made into field V198, although the meaning of the field is unknown.
• Q.44 asked respondents how many children under the age of 16 lived in their household. This information was held on column 56 of card 2 and this was made into field V271. There was one record where the column was punched on both rows 1 and 4, although all other records were singly punched. So this record has been translated to 14 where 14 does not represent a count but the fact that row 1 and row 4 were both punched.
• No explanation was given as to the purpose of columns 16-21 on card 4. These columns were treated as singly punched columns and were concatenated to make one field, V288.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V296-V299, one for each column. Any punches on the top row of the card in these columns were translated to 12.
• No explanation was given as to the purpose of column 20 on card 6. This column was treated as a singly punched column and was made into one field, V306.
• No explanation was given as to the purpose of columns 16-20 on card 7. These columns were treated as singly punched columns and were concatenated to make one field, V365.
• No explanation was given as to the purpose of columns 16-20 on card 8. These columns were treated as singly punched columns and were concatenated to make one field, V398.
• Three extra columns at the end of cards 6 (columns 74-76), 7 (columns 47-49) and 8 (columns 49-51) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.
• One record had a punch in row 3 of column 65 card 2 but there was no explanation on the questionnaire as to its meaning. Column 65 was made into the coded field V280 and the punch in row 3 was translated to 3.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 2

• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V294-V297 one for each column. Any punches on the top row of the card in these columns were translated to 12.
• No explanation was given as to the purpose of columns 16-20 on card 8. These columns were treated as singly punched columns and were concatenated to make one field, V304.
• No explanation was given as to the purpose of columns 16-20 on card 9. These columns were treated as singly punched columns and were concatenated to make one field, V339.
• Three extra columns at the end of cards 8 (columns 49-51) and 9 (columns 76-78) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.
• Row 8 of column 53, which was made into field V84, was coded on the questionnaire as both 'Don't know' and 'Injections/Injecting drugs'. Also row 8 of column 54, which was made into field V93, was coded on the questionnaire as both 'Don't know' and 'Bisexuals'.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 3

• It was discovered that card 6, which held information from the blue booklet, had been wrongly labelled card 9. Card 9 was meant to have held information from the grey booklet, but respondents weren't asked to answer this booklet in this version of the questionnaire.
• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There are 5 records where this column was multi punched although all other records were singly punched. One case has both row 5 and row 9 punched, this has been translated as 59, three cases where row 4 and row 9 was punched, this has been translated to 49 and one case where row 2 and row 4 was punched which has been translated to 24.
• Column 35 of card 2, which was made into field V196, should only have been punched on row 1. However there is one case, on record 57, where row 0 was punched. This column was made into a logical field, so it will appear that this record has a negative response to this field, which is not the case, since an non-punch indicates a negative response. No explanation was given as to why this record had row 0 punched.
• No explanation was given as to the purpose of columns 16-21 on card 4. These columns were treated as singly punched columns and were concatenated to make one field, V322.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V330-V333 one for each column. Any punches on the top row of the card in these columns were translated to 12.
• No explanation was given as to the purpose of columns 16-20 on card 7. These columns were treated as singly punched columns and were concatenated to make one field, V340.
• No explanation was given as to the purpose of columns 16-20 on card 8. These columns were treated as singly punched columns and were concatenated to make one field, V373.
• No explanation was given as to the purpose of columns 16-20 on card 6(labelled card 9). These columns were treated as singly punched columns and were concatenated to make one field, V408.
• Three extra columns at the end of card 7 (columns 47-49), card 8 (columns 49-51) and card 6 (columns 74-76) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 4

• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There is 1 record where this column was multi punched although all other records were singly punched. This case has both row 2 and row 3 punched and has been translated as 23.
• On record 98 it was discovered that the data held on card 1 on columns 57-80 may have been punched on the card out of synch. In order to perform the transformation it was necessary to move the data in column 57 up to column 58, column 58 to 59 and so on until the end of the card. Column 57 was considered to be not punched and one of the unused columns at the end of the card was removed.
• There was one record where column 72 of card 1 had row 6 punched, although this column should only have had one or more of rows 1-5, if any, punched. No explanation was given as to the meaning of this extra punch, and it may just be an anomaly that occurred during the punching process. This punch in row 6 has been dealt with as though it was the 16th option on the sorting board so the same codes were applied to it. So this sorting board has an extra field V133 made from it, and only one record will hold data in that field.
• In column 74 of card 2 any punches in the second row have been translated as 11.
• No explanation was given as to the meaning of the punches on columns 21-24 on card 3. Each column was translated as one field, column 21 to field V318, column 22 to field V319, column 23 to field V320 and column 24 to field V321. On column 22 some records were multipunched, where both row 1 and row 4 were punched this has been translated to 14 and where both row 1 and row 2 were punched were translated to 12. In column 24 any punches on the top row have been translated to V and any punches on the second row have been translated as X.
• No explanation was given as to the purpose of column 21 on card 4. This column was singly punched and translated into field V327.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V335-V338 one for each column. Any punches on the top row of the card in these columns were translated to 12.
• No explanation was given as to the purpose of columns 16-20 on card 8. These columns were treated as singly punched columns and were concatenated to make one field, V345.
• No explanation was given as to the purpose of columns 16-20 on card 9. These columns were treated as singly punched columns and were concatenated to make one field, V380.
• Three extra columns at the end of card 8 (columns 49-51) and card 9 (columns 76-78) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 5

• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There are 18 records where this column was multi punched although all other records were singly punched. One case has both row 2 and row 9 punched, this has been translated as 29, one case where row 2 and row 4 was punched, this has been translated to 24, two cases where row 0 and row 9 were punched, this has been translated to 09 and fourteen cases where row 0 and row 7 was punched which has been translated to 07.
• The explanation of codes used for Q.26, fields V225-V230 were not included in the wave 3 questionnaire, these codes were taken from the questionnaire asked of young people aged 13-15 used in wave 4.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V391-V394 one for each column. Any punches on the top row of the card in these columns were translated to 12.
• Columns 1-5 of card 5 make up the JN No for that card, now contained in field V385. Columns 3 and 4, which are the 3rd and 4th digit in the JN No, were not punched for any records. In order to highlight this blank spaces were inserted in place of the missing 3rd and 4th digit.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 6

• There was two records where the data seemed to be punched on the cards slightly out of synch. In order to perform the transformation it was necessary to move the data on these records up one column on the particular card affected. On record 124 the data held on card 1 column 73 was moved to column 74, column 74 to column 75 and so on until the end of the card. Column 73 was considered to be not punched and one of the unused columns at the end of the card was removed. On record 378 the data held on card 2 column 50 was moved to column 51, column 51 to 52 and so on until column 56 which was moved to overwrite column 57, which was previously unpunched. Column 50 was then considered to be not punched.
• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There are 33 records where this column was multi punched although all other records were singly punched. Two cases have both row 2 and row 9 punched, these have been translated as 29, one case where row 2 and row 5 was punched, this has been translated to 25, two cases where row 0 and row 9 were punched, this has been translated to 09, 27 cases where row 0 and row 7 was punched which has been translated to 07 and 1 case where both row 0 and 2 were punched which has been translated to 02.
• There was one record where column 69 of card 1 had row 9 punched, although this column should only have had one or more of rows 1-8, if any, punched. No explanation was given as to the meaning of this extra punch, and it may just be an anomaly that occurred during the punching process. This punch in row 9 has been dealt with as though it is the 9th option on the sorting board so the same codes were applied to them. So this sorting board has an extra field V111 made from it, and only one record will hold data in that field.
• One record had row X of column 46 card 2 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V220.
• There was one record where column 67 of card 2 had row 9 punched, although this column should only have had one or more of rows 1-7, if any, punched. No explanation was given as to the meaning of this extra punch, and it may just be an anomaly that occurred during the punching process. This punch in row 9 has been dealt with as though it is the 18th option on the sorting board so the same codes were applied to them. So this sorting board has an extra field V275 made from it, and only one record will hold data in that field.
• One record had row 0 of column 79 card 2 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V302.
• Column 46 of card 3 should have been a singly punched column, although one respondent (record number 277) has rows 1, 3, 5, 6 and 8 punched. This may have been a mistake in the punching process. This record has been translated to 13568.
• Columns 1-5 of card 5 make up the JN No for that card, now contained in field V499. Columns 3 and 4, which are the 3rd and 4th digit in the JN No, were not punched for any records. In order to highlight this blank spaces were inserted in place of the missing 3rd and 4th digit.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V505-V508 one for each column. Any punches on the top row of the card in these columns were translated to 12.
• Some respondents had two occurrences of card 6, so fields V510-V516 hold data for the first occurrence and V517-V523 hold data for the second.
• No explanation was given as to the purpose of columns 16-19 on each card 6. These columns were singly punched and were concatenated to make one field, V515 on the first occurrence of card 6 and V522 on the second occurrence of card 6.
• Three extra columns at the end of card 8 (columns 49-51) and card 9 (columns 47-49) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.
• No information was given in the booklet as to the meaning of data in columns 16-72 on card 0. Each of these columns, which were singly punched, were translated into one field. Any punches in the second row were translated to 11.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 7

• There was a number of records where the data seemed to be punched on the cards slightly out of synch. In order to perform the transformation it was necessary to move the data in these records up or down one column on the particular card affected. On record number 392 the data on card 1 column 47 was moved to column 48, column 48 to column 49 and so on until the end of the card. Column 47 was then considered to be unpunched and one of the unpunched columns at the end of the card was removed. On record numbers 69 and 380 the data on card 3 column 16 was moved to column 17, column 17 to column 18 and so on until the end of the card. Column 16 was then considered to be unpunched and one of the unpunched column at the end of the card was removed. On record number 549 the data on card 3 column 34 which was unpunched was removed column 35 was then moved to column 34, column 36 to column 35 and so on until the end of the card. Column 80 was then considered to be unpunched.
• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There are 12 records where this column was multi punched although all other records were singly punched. One case has both row 3 and row 5 punched, this has been translated as 35, ten cases where row 0 and row 7 were punched, this has been translated to 07 and one case where row 0 and row 3 was punched which has been translated to 03.
• One record had row V of column 27 card 1 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V24.
• One record had row 4 of column 55 card 1 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V95.
• One record had row 0 of column 76 card 1 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V161.
• One record had a punch in row 7 of column 74 card 1 but there was no explanation on the questionnaire as to its meaning. Column 74 was made into the coded field V146 and the punch in row 7 was translated to 7.
• One record had row 7 of column 44 card 2 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V231.
• On card 2 column 52 only rows 1-2 if any should have been punched. One respondent had punches in rows 5, 8 and X and another had a punch in row 0. No explanation was given on the questionnaire as to the meaning of these punches, it may have been an error from the punching process. Row 5 was made into field V259, row 8 into V260, row 0 into V261 and row X into V262.
• One record had a punch in row V of column 54 card 2 but there was no explanation on the questionnaire as to its meaning. Column 54 was made into the coded field V267 and the punch in row V was translated to 12.
• One respondent had a punch in row 4 of column 73 card 2 whereas only rows 1-3 if any should have been punched, in accordance with the questionnaire. This column was made into field V328, a coded field. But since the respondent with the punch in row 4 also had a punch in row 1, then it was necessary to use codes in ascending powers of 2 (1,2,4,8). So a punch in row 4 would be represented by the code 8, and where there is more than one punch for a respondent the sum of codes will be used.
• Column 79 on card 2 was made into the coded field V338. As well as having a punch in one of rows 1, 2 or 3, 226 respondents had a punch in row 4. Another respondent had a punch in both row 2 and row X. No explanation was given as to the meaning of punches in row 4 or row X on the questionnaire. Again it was necessary to use codes in ascending powers of 2, where a punch in row 4 is represented by the code 8 and a punch in row X is represented by the code 16, and where there is more than one punch for a respondent the sum of codes will be used. The data in this column may in fact be more suited to question 42, which the questionnaire says is stored on column 80 of this card, where a respondent can answer yes, no or don't know (punches in rows 1-3) to knowing someone who injects drugs, and also volunteer that they themselves injects drugs (a punch in row 4).
• One record had row 0 of column 21 card 3 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V375.
• One record had row V of column 23 card 3 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V392.
• The data on column 34 card 3 contains information on the number people aged 16 or over in a household. One respondent had a punch in both row 1 and row 2, this has been translated to 12, to represent a punch in row 1 and 2, rather than a count of 12 people.
• One record had a punch in row 4 of column 38 card 3 but there was no explanation on the questionnaire as to its meaning. Column 74 was made into the coded field V407 and the punch in row 4 was translated to 4.
• Columns 1-5 of card 5 make up the JN No for that card, now contained in field V414. Columns 3 and 4, which are the 3rd and 4th digit in the JN No, were not punched for any records. In order to highlight this blank spaces were inserted in place of the missing 3rd and 4th digit.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V420-V423 one for each column. Any punches on the top row of the card in these columns were translated to 12.
• Some respondents had two occurrences of card 6, so fields V425-V431 hold data for the first occurrence and V432-V438 hold data for the second.
• No explanation was given as to the purpose of columns 16-19 on each card 6. These columns were singly punched and were concatenated to make one field, V430 on the first occurrence of card 6 and V437 on the second occurrence of card 6.
• Three extra columns at the end of card 8 (columns 49-51) and card 9 (columns 47-49) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.
• No information was given in the booklet as to the meaning of data in columns 16-71 on card 0. Each of these columns, which were singly punched, were translated into one field. Any punches in the second row were translated to 11.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 8

• There was three records where the data seemed to be punched on the cards slightly out of synch. In order to perform the transformation it was necessary to move the data on all three of these records up one column on the particular card affected. On record 178 the data held on card 2 column 41 was moved to column 42, column 42 to 43 and so on until the end of the card. Column 41 was considered to be not punched and one of the unused columns at the end of the card was removed. On record 180 the data held on card 1 column 35 was moved to column 36, column 37 to 38 and so on until the end of the card. Column 35 was considered to be not punched and one of the unused columns at the end of the card was removed. On record 243 the data held on card 1 column 52 was moved to column 53, column 53 to 54 and so on until the end of the card. Column 52 was considered to be not punched and one of the unused columns at the end of the card was removed.
• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There are 17 records where this column was multi punched although all other records were singly punched. One case has both row 2 and row 5 punched, this has been translated as 25, one case where rows 2, 3 and 5 was punched, this has been translated to 235, two cases where row 0 and row 5 were punched, this has been translated to 05 and thirteen cases where row 0 and row 7 was punched which has been translated to 07.
• One record had row 5 of column 55 card 1 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V94.
• One record had row 3 of column 59 card 1 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process as otherwise this column was not punched at all. This row of this column was made into field V112.
• One record had row 0 of column 74 card 1 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V159.
• Q.52 asked respondents how many children under the age of 16 lived in their household. This information was held on column 37 of card 3 and this was made into field V379. There was one record where the column was punched on both rows 1 and 7, although all other records were singly punched. So this record has been translated to 17 where 17 does not represent a count but the fact that row 1 and row 7 were both punched.
• No explanation was given as to the meaning of the punches on columns 47-50 on card 3. Each column was translated as one field, column 47 to field V389, column 48 to field V390, column 49 to field V391 and column 50 to field V392. In column 50 any punches on the top row have been translated to V and any punches on the second row have been translated as X.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V399-V402 one for each column. Any punches on the top row of the card in these columns were translated to 12.
• Three extra columns at the end of card 7 (columns 77-79) and card 9 (columns 49-51) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.
• One record had a punch in row 0 of column 25 card 3 but there was no explanation on the questionnaire as to its meaning. Column 54 was made into the coded field V373 and the punch in row 0 was translated to 0.
• Columns 1-5 of card 5 make up the JN No for that card, now contained in field V393. Columns 3 and 4, which are the 3rd and 4th digit in the JN No, were not punched for any records. In order to highlight this blank spaces were inserted in place of the missing 3rd and 4th digit.
• Columns 19-22 on card 1 are concatenated to make up field V7. On record number 55 column 22 wasn't punched although columns 19-21 were punched '015'.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 12

• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There are 2 records where this column was multi punched although all other records were singly punched. Both cases have both row 5 and 9 punched, this has been translated as 59.
• There was two records where column 70 of card 1 had row 3 punched, although this column should only have had one or more of rows 1-2, if any, punched. No explanation was given as to the meaning of this extra punch, and it may just be an anomaly that occurred during the punching process. These punches in row 3 have been dealt with as though they are the 23rd option on the sorting board so the same codes were applied to them. So this sorting board has an extra field V99 made from it, and only two records will hold data in that field.
• One record had row 9 of column 64, card 2 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V211.
• One record had row 0 of column 21, card 3 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V300.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V412-V415, one for each column. Any punches on the top row of the card in these columns were translated to 12.
• On the questionnaire column 22 (along with columns 20 and 21) of card 3 is given as the code list for Q.32. The questionnaire does not give Q.33 as having any code lists, although it seems that the code list in column 22 would be more suited to Q.33.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 13

• One record had row 4 of column 80, card 2 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V300.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V520-V523, one for each column. Any punches on the top row of the card in these columns were translated to 12.
• On the questionnaire column 30 (along with columns 28 and 29) of card 3 is given as the code list for Q.36. The questionnaire does not give Q.37 as having any code lists, although it seems that the code list in column 30 would be more suited to Q.37.
• There was one record where some of the columns of data seemed to be punched on the card slightly out of synch. In order to perform the transformation it was necessary to move the data on this record up one column on the particular card affected. On record 102 the data held on card 2 column 50 was moved to column 51, column 51 to 52 and so on to column 55. Column 50 was considered to be not punched and one of the unused columns after column 55 was removed.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 14

• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There are 3 records where this column was multi punched although all other records were singly punched. One case has both row 4 and row 9 punched, this has been translated as 49, one case where both rows 2 and 5 are punched, this has been translated to 25 and one case where row 2 and row 3 were punched, this has been translated to 23.
• One record had row 9 of column 29, card 2 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V171.
• One record had row 3 of column 46, card 2 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column was made into field V221.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V523-V526, one for each column. Any punches on the top row of the card in these columns were translated to 12.
• On the questionnaire column 30 (along with columns 28 and 29) of card 3 is given as the code list for Q.36. The questionnaire does not give Q.37 as having any code lists, although it seems that the code list in column 30 would be more suited to Q.37.
• No explanation was given as to the purpose of columns 16-19 on card 6. These columns were singly punched and were concatenated to make one field, V533.
• Three extra columns at the end of card 8 (columns 49-51) and card 9 (columns 47-49) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 15

• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There are 10 records where this column was multi punched although all other records were singly punched. Two cases has both row 7 and row 9 punched, this has been translated as 79, one case where both rows 5 and 9 are punched, this has been translated to 59, one case where row 3 and 9 were punched, translated to 39, two cases where row 3 and 5 were punched, translated to 35, one case where row 2 and 9 were punched, translated to 29, one case where row 2 and 5 were punched, translated to 25 and two cases where row 2 and row 4 were punched, this has been translated to 24.
• The singly punched columns 23 and 24 were concatenated to make up V8, the interview length in minutes. One record had column 23 punched on row 3 but column 24 was not punched.
• In column 72 of card 1, one record has row 3 punched and another has row 4 and 5 punched. No explanation was given on the questionnaire as to the meaning of these rows, it may have been an error from the punching process. These rows of this column have been made into fields V144-V146.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V434-V437, one for each column. Any punches on the top row of the card in these columns were translated to 12.
• On the questionnaire column 57, card 2 (along with columns 55 and 56) is given as the code list for Q.30. The questionnaire does not give Q.31 as having any code lists, although it seems that the code list in column 57 would be more suited to Q.31.
• No explanation was given as to the purpose of columns 16-19 on card 6. These columns were singly punched and were concatenated to make one field, V444.
• Three extra columns at the end of card 8 (columns 49-51) and card 9 (columns 47-49) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.
• Version 15 of the questionnaire did not contain a copy of the light green cards used in question 35. It has been presumed that the contents will be the same as the light green cards used in version 14.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

Version 16

• Field V9, which is described as for 'office use only' was held on column 25 of card 1. There one case where both row 3 and 5 were punched, this has been translated to 35.
• In column 70 of card 1, 12 records have row 3 punched. No explanation was given on the questionnaire as to the meaning of this row, it may have been an error from the punching process. This row of this column have been made into fields V142.
• According to the questionnaire column 26 of card 2 should not have been punched, although one respondent has rows 3, 4 and 6 all punched. This column was made into field V185, although only one respondent will have data in it. This is 346 to show that rows 3, 4 and 6 were punched.
• No explanation was given as to the meaning of the punches on columns 56-59 on card 3. Each column was translated as one field, column 56 to field V422, column 57 to field V423, column 58 to field V424 and column 59 to field V425. In column 59 any punches on the top row have been translated to V and any punches on the second row have been translated as X.
• No explanation was given as to the purpose of columns 17-20 on card 5. These columns were singly punched and were made into 4 fields, V422-V425, one for each column. Any punches on the top row of the card in these columns were translated to 12.
• On the questionnaire column 57, card 2 (along with columns 55 and 56) is given as the code list for Q.28. The questionnaire does not give Q.29 as having any code lists, although it seems that the code list in column 57 would be more suited to Q.29.
• Three extra columns at the end of card 7 (columns 77-79) and card 8 (columns 49-51) were punched but no explanation was given as to the meaning of them. The third last column was punched on the rows labelled V and 5 so this was made into one field and translated as V5. The second last column was punched on the rows labelled X and 5 so this was made into one field and translated as X5 and the last column was punched on the rows labelled V and 4 so this was made into one field and translated as V4.

The following lists, and gives the number of occurrences, of those fields that were coded and where a respondent has more than one answer to a set of mutually exclusive answers. The record will show the sum of codes for the answers given.

All tables

Users should be careful with any queries performed on coded fields where either some codes, or all of the codes used have integers coding for integers, as the browsing software may produce conflicting results. An example of this is the confusing coding scheme used in the field V488 in the version 16 table where the code '01' stands for '10' and code '02' stands for '9' and so on. Another example would be the field V410 in the version 8 table where the code '01' stands for '0', code '02' stands for '1' and code ' 03' stands for '2'. These types of coding may also be used in other fields within other tables.

At the start of each of the four wave books (see the Dataset Documentation Catalogue, references CRDA/35/DD/1/1-4), details are given for the sample structure for the campaign. There are a number of discrepencies between sample sizes given in this sample structure and the actual numbers of respondents in the tables. The wave 2 book states that there were 689 interviews with adults aged 18-64. Although the version 3 table, asked of adults aged 18-64 has 702 respondents. It also states that it interviewed 284 homosexuals but the version 4 table has 299 respondents. The wave 3 book states that 713 interviews were asked of adults aged 18-64. In version 7, asked to adults aged 22-64, there were 592 respondents and in version 6, asked to young people aged 16-21 there were 194 of these respondents aged 18-21, which gives a total of 786 respondents aged 18-64. In the wave 4 book the sample structure states that there were 708 interviews asked of adults aged 18-64. In version 14 asked to young people aged 18-21 there were 194 respondents, and in version 15, asked to adults aged 22-64 there were 593 respondents, which gives 787 respondents aged 18-64.

CRDA/35/DD/2/1, which provides information on card numbers and number of cases within each wave, reports that 75 respondents in the Wave 1 (referred to as Wave I in document CRDA/35/DD/2/1) data had a card 8 (Green Booklet). In Version 1 of the questionnaire 64 respondents have a card 8 and in version 2, 12 respondents have a card 8. This gives a total of 76 repondents in Wave 1 with a card 8. Also in this document the information held on Wave 3 (referred to as Wave III in document CRDA/35/DD/2/1) fails to mention that there was 264 cases of Card 87 (264 respondents with a card 7 in the version 8 table). Version 8 has 264 cases where respondents have a Card 7, this has not been reported in the document CRDA/35/DD/2/1.

There are no related datasets in this series.