Monday, 6 August 2012

Statistical Significance

When someone tries to sell you something based on “statistics” you can take it with a grain of salt, or you can analyze the analysis.  In other words, you can be skeptical or scientific.  Having studied behavioral science, including statistics, in university I make it a game to pick-apart useless statistics.

Good stats are ones that have poled a statistically significant number of people within a population.  Great stats are ones that pole a sufficient number and variety of the kind of population you’re interested in selling to.  If you want to know how people are going to respond to a jewellery line in Grande Prairie, Alberta and you’re given statistics accrued mainly from urban United States buyers, then you may not learn what you need to know.

Advertising reps throw around statistics like they are an instant roadmap to success.  The truth is that there are so many measurements taken by the ad industry (and the new car industry too) that every media outlet can pick and choose the most compelling statistics to make themselves look good.  Who do you think pays for that research and why?

Knowing what I know about standard deviation, mode, mean, median and statistical significance, I think the best policy is to take stats with a grain of salt.  When it comes to advertising, hold your ad-rep accountable for results.  When he or she says that an advertising plan will generate increased business for you whithin a certain segment of buyers, make your decision, run the ads, compile your own statistics and then report back the results.  Don’t let them use objective measurements to sell you something with subjective results.

No comments:

Post a Comment

Note: only a member of this blog may post a comment.