This is an old revision of the document!


There is no variable label on the annotated questionnaire. What does this mean?

Within the NSHD meta-data repository, annotated questionnaires are available. For the majority of the questions, a variable label is located next to the question indicating that these questions have been coded. These variables will have been coded either directly from the questionnaire or to a coding frame if open-ended responses were recorded. However some questions dont have a variable label. One reason is that the question is open-ended and has not been coded. The other exception occurs in the early questionnaires, when a number of questions were incorporated together to produce a single derived variable. As these were never coded separately, no variable label has ever been produced. For example, the breastfeeding questions on the Maternity Surveys have no variable labels but have been used with a couple of other questions to produce derived variables about breastfeeding [BRE] and weaning [WEA]. Looking at the topic summaries of the data collected will give you a good idea of the variables available for analysis.

I have made my final selection of variables. What do I need to consider before submitting the request?

When building a data basket using SWIFT it is recommended that users spend time ensuring their basket contains a) the most appropriate variables and b) a complete set of the variables required for the current stage of their research project. When identifying and selecting variables suitable for your research, please use the meta-data repository carefully and seek advice from the designated NSHD co-ordinators and the Data Librarian.

Once a basket has been built, the list of variables will be reviewed by the NSHD co-ordinator assigned to your project. Some revisions may be made to the basket at this stage; additional variables may be recommended and unsuitable variables pointed out. Once the basket has been finalised the basket will be converted into a 'scrambled' data set, encrypted and password protected and sent to you via email.

If you realise, once you have received your dataset, that you require additional variables, the above process will need to be repeated. Requesting additional variables is a time consuming process, for you and for us, and may delay your project.

Prior to submitting your request, please think carefully about which variables you require and ask for help or recommendations, if needed.