Don’t Rely On Early Or Unscientific Surveys

Dr. Louis Perron
blog post louis

Surveys that are taken a long time before election day often mostly reflect awareness. In that sense, an early lead is in no way a prediction, but can actually be a sweet poison putting campaigns to sleep.

In a campaign setting, we use an early survey as a basis to make informed decisions and plan an effective campaign. We use it to determine a winning message, define low-hanging fruits or test the impact of various scenarios.

A typical beginners’ mistake is also to think that one can do his own survey or focus group. Campaigning is like walking in a labyrinth. If you take a wrong turn at the beginning (meaning: if you base your strategy on wrong data), everything that follows will be wrong as well. The quality of a survey also depends on much more than the sample size. For sure, a bigger sample size is generally better, but other factors include the flow of the questionnaire, the training and supervision of the field workers and most importantly, to get accurate sampling. The reason why polls in the US have been notoriously wrong, for example, is because pollsters try to outsmart each other with new weighing tactics instead of investing into proper random sampling and fieldwork.