Jason Smith

I'm Jason. This is my home on the web where I write. Stick around. It's going to get interesting. 

Why you cannot trust survey results

The thing about survey results is that they are so appealing to read. The headline for an article sprouting a survey result speaks as though it were declaring pure gospel, it's facts not debatable for in them lies a foundation as stable as a rock; a scientific survey.

But a survey, by definition, is not the truth, but merely a supposed representation of the truth based on a sampling of people's answers. Which on the surface should be good enough to build an accurate picture in the minds of the information seekers.

But what needs to be taken into account when conducting a survey, and also when reading someone's survey results, is the fact that people are usually under no compulsion to complete a survey, and thus a survey is going to attract a certain type of person. 

The other day a polling company rang my house wanting to speak with a woman in the house between the ages of 18 and 34. My wife happens to be 34, so I went and asked her if she felt like participating in the survey. She was chilling in front of the TV spending time with our younger children, and quite enjoying herself. She declined, and that was that.

Now, I have no idea what the survey was going to be about, but whatever it was it's safe to assume that the results are likely to be skewed given that women in that age bracket with young children and a good friend network and with nothing to get off their chest and no craving to share their opinion about the world with a random phone surveyor, are not likely to participate.

I used to work for a company conducting surveys. One survey we conducted was at Perth Airport, and it was all about how people felt about the parking rates and services provided by the airport.

Problem again was, people that were too successful and thus too busy to spend 5 minutes answering a survey questionnaire were not participating in the results. So whatever answers that demographic would have given, were not included in the final results.

I have noticed that people with the most money, most successful careers, and busiest lives are the people companies most want to understand. Yet these are the LEAST likely people to give their precious time to a random telemarketer wanting to complete a survey.

That's not natural

How do you measure who's winning?