Tips and tricks in “broadband” data analysis

Stats Can analyst Ben Veenhof weighed in recently with further comments about how to interpret – and not interpret – broadband-related data. His email comes in response to a question I had about drawing parallels between a) CIUS respondents who report they’ve never used the Internet and b) respondents (from a U.S. Dept of Commerce/NTIA study ) who say they do not have broadband in their homes. The results of the 2009 Canadian Internet Use Survey were released in The Daily, May 10, 2010 – here.

I noted in my Nov 27 post that the Stats Can Internet analysts had shown a gratifying willingness to field my questions, and not spare the details. It’s terrific that the staff of such a highly respected institution are able to offer outreach of this kind. Certain other federal bodies could learn a thing or two about conveying quality information to the taxpaying public. It would also be very gratifying if the CRTC took note of how Stats Can handles the naming issue, i.e. sorting out broadband and high-speed when talking to those taxpaying and much confused Canadians.

Ben’s email follows verbatim, in blue (emphasis original). None of Ben’s comments, nor those of any of his colleagues cited elsewhere, is to be construed as approval or endorsement in any way of my opinions.


Hi David,

Thank you for sharing your articles with me.  I’ve had a few moments to go through them, and can confirm that all of the information you have sourced to the CIUS is accurate.  Just a couple of small points regarding information in your 2nd posting:

In the section ‘Non-adopters’, you note:

“But if we look at the whole adult population, use of the Internet at home with a high-speed connection is cited by only 70% of all lindividuals 16 and over. While this figure is up from 59% in 2007, it means 30% of Canadian adults did not have broadband at home as of one year ago.”

Based on the terminology used in the survey, one could write ‘30% of Canadian adults did not have high-speed at home as of one year ago.” (I realize that many people use the terms ‘broadband’ and ‘high-speed’ interchangeably, but you may recall from an earlier exchange (if my memory is correct), we discussed how the term ‘broadband’ was less familiar to survey respondents (it did not perform reliably in survey testing) and so the term ‘high-speed’ was used on the CIUS questionnaire.

Another comment:  I noticed that in the list of ‘Top reasons cited by Canadians 16+ for never having used the Internet’, you rolled up the reasons least-often mentioned, in to one category called ‘All other (10.1%)’.  As the denominator you’re using speaks to the % of people (never users), as opposed to % of reasons, I re-ran the combined categories, which produced a figure of 9.8% (since some people may report more than 1 reason).  A small difference, I know, but my suggestion would be to report the figure as ‘Other reasons (9.8%)’.  ie., 9.8% of those who never used the Internet cited reasons in addition to those appearing on the list.

Concerning the fact that the data from the U.S. report cover broadband non-adoption and the CIUS data relate to Internet non-users, my comment would be that indeed, a direct comparison of the reasons for non-use of broadband vs. non-use of the Internet itself is not possible.  However, your article notes clearly the main differences in how the questions are framed, which is important, while highlighting a number of parallels.