1. Context A New Survey Measure of Disability: the Disability Screening Questions (DSQ)
1. Context

The Disability Screening Questions (DSQ) were developed in the context of the New Data and Information Strategy on persons with disabilities (PWD), which is being developed at Employment and Social Development Canada (ESDC) in order to have more efficient, more relevant and more accessible data on PWD. The three main pillars of the Strategy are: survey data, administrative data, and an open data platform to maximize accessibility and usefulness of this information. As part of Pillar 1 (survey data), the new DSQ were developed in partnership with Statistics Canada (STC) and a Technical Advisory Group (TAG) made up of experts in the field of disability. The goal is to include the DSQ in various existing general population surveys, thereby enabling direct comparisons of PWD with the overall population on a more frequent basis.

Until recently, several general population surveys at STC used the “Census filter questions”^{Note 1} to identify PWD. These questions, which first appeared on the 1986 Census of Population long form, served as filters to create a survey frame for a post-censal survey^{Note 2} on disability. However, these questions do not provide a good measure of disability. In fact, early quantitative tests of these questions revealed that although most PWD were concentrated among people who reported at least one positive answer to these filters (true positives), a relatively large number could be found among people who reported “No” to all the filters (false negatives).

The Census filter questions were modified and improved over time, but even those on the latest version (used on the 2011 National Household Survey which replaced the long-form Census) are considered to be a misleading measure of disability because they tend to miss people with non-physical disabilities (false negatives),^{Note 3} yet overestimate the rate of disability (false positives).^{Note 4} Nonetheless, these questions being the only ones^{Note 5} available as a proxy for disability, they were still being used to estimate the prevalence of disability, for example, for subprovincial estimates. Some general population surveys also incorporated these filter questions to obtain an approximate identifier of PWD. However, in 2010, Statistics Canada decided that these questions could no longer be used by surveys to identify disability, and a new set of questions to replace them on general population surveys became a priority.

The objectives of the DSQ were to improve the measure based on the Census filter questions by having: 1) a more inclusive measure for persons with non-physical disabilities (learning and mental health-related disabilities in particular); 2) a more precise measure including type and level of severity; and 3) a consistent way of measuring disability based on the social model^{Note 6} that defines disability as the relationship between body function and structure, daily activities, and social participation. However, to be implemented in general population surveys, the DSQ could take no more than 2 minutes to administer. This constraint was a main driver in the format of the DSQ, and also, in the decision to have both a long and a short version.

This document describes the creation of the new DSQ, explains why a long and a short version were developed, and shows how these two versions work, how they differ, and how to use each version.

The DSQ continue to undergo changes. The version described in this document was used in the Canadian Survey on Disability (CSD) in 2012. A section at the end of this report describes subsequent changes and future plans for the DSQ.

This report is structured as follows. Section 2 is an overview of how disability has been measured at Statistics Canada over the past 30 years. Section 3 covers the development of the DSQ, starting with a description of the social model of disability on which the new DSQ are based, followed by a general overview of the questions that worked well during qualitative testing. Section 4 describes the goals in creating the DSQ, the disability types covered, and the reasoning behind the use of filter and screener questions, and presents the wording of the questions for each of the 10 disability types. It also presents information about the two major quantitative tests conducted to assess the reliability of the DSQ, definitions of disability that were considered and retained, and the length of time needed to administer the DSQ. Section 5 describes the severity score. Section 6 explains the need for a short version and how it differs from the long version. Section 7 explains how the long and short versions of the DSQ are currently used at Statistics Canada. Section 8 compares DSQ results with those of other health measures used by Statistics Canada surveys. Section 9 contains information on the latest developments of the DSQ, and Section 10 summarizes the discussion.

• Previous
• Next