The data trail begins with questions and ends with interpretation
Have you ever responded to a phone call and been asked to participate in a survey and wondered, after the fact, what happens to the information collected? Sometimes it seems like so much is asked that no one could possibly analyze it all. Or maybe the questions seem so irrelevant that you aren't sure how the information collected would be useful. We are a group of researchers who are analyzing the data collected from these types of surveys as part of a broad-based series of studies Ð studies that are mapping literacy and disability data in Canada.
We thought it might be helpful to take you on a journey along the data trail to show what happens to the information that is collected and why it is important. As an example, we have chosen to use environmental barriers to long-distance travel as measured for the Participation and Activity Limitation Survey (PALS) database collected in 2001. This piece of data is important because many people in Canada, including many readers of Abilities, may face such barriers.
CHOOSING A TRAIL
Of course, the first step in any data journey is deciding which trail to take. This is determined in large part by the structure of the survey itself. A great deal of time and energy goes into survey design because the data collected will be useful for a wide range of applications. This is particularly the case with large government-sponsored surveys, such as PALS. It is important to note that these surveys are often modified from year to year as theories change, models of disability and rehabilitation evolve, concepts of human rights emerge, political and economic priorities shift, and technical problems with previous versions are uncovered. While this does usually lead to improved information gathering, it can also make comparing information over time, from one version of the survey to another, difficult.
The illustration on the next page starts at Step 1 with a question that asks participants what prevents them from travelling long distances. This is just the beginning. The real journey is what follows-Steps 2 to 4. Our goal is to illustrate the environmental barriers to long-distance travel on a map in our atlas, so we need to tally up the responses to the question by geographic location. In other words, we want to know if there are more environmental barriers for people with disabilities in Quebec, Ontario or Manitoba. Each response to this particular question is a separate piece of data, and in this case, each option represents one barrier to long-distance travel (Step 1). Some respondents to the survey may have more than one barrier and we want to be sure this is reflected in our analysis.
Closer examination of the possible options included in the survey leads us to conclude that some of these options belong in other categories (Step 2). For instance,
'unsupportive staff" isn't an environmental barrier, but rather an attitudinal barrier, and "too costly" is clearly a financial barrier.
Step 3 in the journey is to classify the data into logical categories according to the goals of the study. This can be a tedious process since it requires careful interpretation of the questions asked along with a vision of the big picture. In this case, we consider the means of overcoming environmental barriers to be quite different from attitudinal or financial barriers and therefore it is important to remove the latter from this analysis so that our measure of environmental barriers is as accurate as possible.
TRAIL GUIDE
We've now gotten to the point on the data trail where it would be easy to get lost and difficult to explain to others where we are. We are developing an understanding of where the data can take us and we have a vision of the big picture, but we need a guide to help us navigate and communicate the journey so that others may follow.
First, we created a detailed guide, from the bottom up, by examining each question to determine whether it is useful for our analysis and, if so, how it should be categorized. This process results in a guide that is limited by the questions that in turn determine the possible routes we might follow in our analysis. Undoubtedly there are things we would like to know but don't because of the survey design (as examples, this particular question does not ask about access to medical personnel, and the PALS survey was not administered in the Territories).
The next step is to re-examine this detailed guide from the top down (Step 3). The resulting simplified guide is useful for communicating the purpose and character of the journey we've taken without getting into the distracting details required to follow in our footsteps.
CONCLUSIONS FROM OUR JOURNEY
Our numbers are weighted so that they represent the actual population and, finally, maps are made (Step 4). Often the data gathered by one question in the survey can lead to many different places, as in our case. Map 1 shows the total number of barriers encountered by people with disabilities who are prevented from travelling. This map reflects, more or less, the total population of the province, though curiously British Columbia ranks second in barriers, yet it is third, behind Quebec, in population.
Map 2 shows the percentage of the provincial population with disabilities prevented from long-distance travel regardless of the number of barriers they encounter. Map 3 shows the number of people with disabilities prevented from travelling categorized by the number of barriers they face. In other words, the colour of the bar refers to the number of barriers encountered, and the size of the bar reflects the number of people (so provinces with larger populations generally have higher bars). This map shows that more people have multiple barriers to travel in British Columbia than in Quebec, which explains why British Columbia ranks higher than Quebec in the number of barriers encountered, even though more people live in Quebec. It also shows that a very large number of people prevented from travelling face only one barrier and they live in Ontario.
A possible next step might be to explore the different types of environmental barriers to determine if there is one barrier that is particularly problematic for a large number of people and, if so, communicate this to policy-makers so they may help remedy the situation.
We have travelled a long way, from question to analysis to visualization to interpretation. The trip was simple in some ways and hard in others. It has been simple in that we have travelled along a main line from a question regarding difficulty of travel to an interpretation of barriers. However, along the way we have had to make choices between routes. There are different understandings of the questions asked, different ways of analyzing the data collected, different methods of visualizing the data, and different interpretations of the results. In the end, hopefully the choices we make will lead to a better life for people with disabilities.
Two roads diverged in a yellow wood...
(We) took the one less travelled by
And that has made all the difference.
-From Robert Frost, Mountain Interval, 1920, "The Road Not Taken"