In general, variables in the CLSA dataset reflect the interview process. In some cases, follow-up questions were only asked if specific answers were given to preceding questions.

Blank values in the Baseline data can represent multiple types of missing data, including:

Valid skip patterns. For example, number of daughters and sons are only asked if the participant answered that they have at least one child. In the CLSA dataset, participants with no children will have blank values for both.

Missing data due to non-completion. There are some participants who skipped entire sections of baseline interview, and therefore have blanks for all the questions in those sections. Indicator variables such as ADM_COMPLETE_MCQ are provided in the documentation accompanying data release and should be consulted when there are large number of missing data to determine if it is due to a participant not completing a section.

In the Follow-up 1 and Follow-up 2 datasets, missing data have been assigned various codes according to the reason the data are missing. Details of the different types of missing data are provided in the data dictionaries accompanying datasets.

Within the CLSA dataset, derived variables (DVs) are variables that are created from other variables. DVs are derived by re-grouping or re-classifying the original variables, to glean information otherwise not available. Some DVs are based on published measures or scales. You will find documentation related to DVs under Researcher Resources.

Final drafts of all manuscripts, reports, reviews, pre-prints and other proposed primary publications describing research using CLSA data and/or biospecimens must be sent to the CLSA, by the primary applicant, for review at least 15 working days prior to the anticipated submission. Abstracts, posters and presentations do not need to be submitted for review but should include appropriate acknowledgements. Please review our Publication and Promotion Policy for CLSA Data Users for additional information.

As a publicly funded research platform, the CLSA encourages the dissemination of research findings from approved projects. The CLSA expects users to publish their findings in peer-reviewed journals. Multiple publications may be prepared based on a single approved project as long as the publications are directly linked to the objectives of the approved project.

To obtain all the variables contained in a questionnaire, type the two or three letter prefix (e.g. SDC for Socio-demographic Variables) into the full-text search box in the Variables listing, under “Variable properties > Name”. You can also use more general terms such as ‘food’, ‘work’, etc. (under “Variable properties > Label”) to find variables related to those terms, however, search terms are not exhaustive. For more information on the variables included in a questionnaire, please visit  Researcher Resources.

Multiple-choice questions are represented by either a single variable or multiple variables, depending on what the question allows:

– A question allowing only one response is represented by a single variable that can take on multiple values. Open-text responses are permitted in many questions; common and distinct responses are recoded to create new categories within the variable itself.

– For a question allowing multiple responses, each possible response category is assigned its own binary variable. Open-text responses are also permitted in many of these questions; common and distinct options are also recoded to create additional variables within the question scope. The number of variables corresponding to that question matches the number of response options.