Week 8 Blog Statistics and Surveys By Cynthia Dempster

When my mother buys a lottery ticket, she only buys one set of numbers. She doesn't believe in or consider statistics or odds. She says one ticket is all God needs if he wants her to win. She says chance has nothing to do with it.

I have a friend who reads statistical reports diligently and makes major decisions based on them. At a recent election, she liked two political leaders equally. She read the statistical predictions and chose the leader that was most likely to win. Predicted results influence her behaviour. She thinks this is common sense. Is it?

If we read that there is a 60% chance of rain and we take an umbrella, is this common sense, or, are we guided by a sense of authority and/or knowledge outside ourselves? Do we trust and/or want to trust statistical predictions? Are statistical predictions our modern oracles? Statistics can be like the zodiac guiding people's lives.

Looking at these samples, I see that people can have an attitude or relationship with statistics that is more mythological or religious than factual. This approach to statistics is also apparent in the rejection of statistics as in my mother's case. God is more powerful than numbers. Are numbers really factual in their nature? Are myths and religions less or more factual(or real) than numbers?

Surveys, also, are far from factual. D.L McFadden says, "when respondents cannot recall or construct the exact answer to a question, they can interpret the task as one of constructing a 'correct' response, much like responding to a teacher's questions in school."(McFadden,2005,p.3)

Human imagination and beliefs are sometimes active and present in the operation and non-operation of statistics and surveys.

I have been disheartened to hear statistics about employment. I deal with this though by reassuring myself that there are deeper causes and effects influencing outcomes. It may be true that a certain number get work and a certain number don't. However, the causes and the reasons for employment success are not necessarily numerical.

McFadden, D. (2005). Statistical analysis of choice experiments and surveys. Marketing Letters(vol. 16), p. 3.