I confess to being a data geek. I’m okay with it. I accept it. I am calling on all my fellow data geeks to unite around our love of good data.  Regardless of how any of us use the data, how we interpret it, or what actions we think should be taken as a result of data, we can all agree that we need reliable, valid data to form the basis of our arguments, which we then revel in debating over with other data geeks.
 
Census forms are being delivered as we speak. The short-form is still mandatory but the long-form has been replaced with a voluntary National Household Survey (NHS), in a controversial decision by the Federal Government that was based on rhetoric rather than facts. Statistics Canada is doing everything they can to ensure that the NHS data are going to be useable, but the data geek in me is anxious - and I am not the only data geek who is feeling this anxiety. We all have unanswered questions: How will Statistics Canada be able to ensure that the NHS is representative of the Canadian population? How will we be able to track trends with data from a new survey? How can we be sure that this survey will integrate with other data from other Stats Canada surveys?
 
I am particularly anxious about making sure Canadians living on low-income are accounted for and that we are reporting accurate rates of poverty, in addition to other valuable data about poverty in the country. If not, then come 2012 we will not have an accurate picture of poverty in this country and therefore will have limited knowledge about the best policies and programs to reduce it.  Poverty is now figuratively a hidden problem; without good data, it will soon be literally hidden.

We know that people living in poverty are less likely to voluntarily complete a survey of this nature, for various perfectly valid reasons. As such, I am quite troubled that we may see a trend indicating declining poverty rates, when in fact, all we have seen is a decline in response rate from people living in poverty.  I’d be overjoyed if we significantly reduced poverty but I don’t want to celebrate a false victory arising from faulty data.
 
Statistics Canada is doing everything they can to make the NHS a valid and reliable form of data collection but the reality is that even with all of their efforts this will not be enough. Indeed, most researchers, policy analysts, and users of census data agree that the NHS is not likely to produce a source of data that inspires the same kind of confidence as the long-form of the census. Unfortunately, our collective uproar failed to make even the smallest dent on the decision to scrap the mandatory long-form.
 
For now, please complete the census, complete the NHS, and encourage everyone you know to do so. The NHS is being sent to 1/3 households and completing it is very important to getting responses that are truly reflective of Canadian society. If you are feeling particularly feisty, please use the comments to let Statistics Canada know you are a proud data geek and you want the long-form census back.