Poor survey question caught by the media

I love it when the non-statisticians in the media catch a statistical error. I spotted this piece in the Washington Post by the guy who does the traffic reporting – Dr. Gridlock. The piece was actually on upcoming local hearings on the proposed fare increases for the Washington metro area Metro.

The problem was in a survey being conducted online by Metro. As Dr. Gridlock says

Some questions encourage respondents toward a certain view of the transit authority: “Metro’s six-year Metro Forward rebuilding program has reached its halfway point. Riders are already seeing improvements in overall reliability of trains and buses, as well as escalator availability, continued investments over the next three years will build on those improvements.”

That is a big no-no in survey questionnaire design. The preamble to any question should not put forth a particular position.

Thank you Dr. Gridlock, and please Metro don’t do that again.

2 Comments

  1. Kara says:

    Theres no question in that quote? What was the survey question?

  2. Larry says:

    I should have included the next paragraph of what Dr. Gridlock had to say. What I quoted was the introduction to a series of questions. The problem is that the intro can bias the questions that follow. What he added was:

    Then it goes on to ask, “How important would you say each of the following initiatives are for Metro?”

(Comments are closed)
  • Subscribe to Blog via Email

    Enter your email address to subscribe to this blog and receive notifications of new posts by email.

  • November 2017
    S M T W T F S
    « Oct    
     1234
    567891011
    12131415161718
    19202122232425
    2627282930  
  • Recent Posts