Tuesday, September 25, 2012

The power of statistics and the National Conversation

My survey on the Ferrari crash is a good example of the power of statistics.

The survey asked the participants to give their views on the decision of the insurance company to deny liability for the accident as the Ferrari was driven recklessly. The first 10 replies showed an equal balance between those who support and those who do not support the decision. The same results were reflected based on the first 20 replies, the first 30 replies, the first 40 replies and now a total of 50 replies. The results were split evenly based on all the progressive reports.

Clearly, this is a matter where the public is evenly split on this matter. Those who support the decision of the insurance company felt that reckless driving should be punished. Those who disagree with the decision felt that the insurance company should abide by its contractual obligations.

If a survey is carried out in an unbiased manner, the results for the first (say) 25 replies tend to be representative of the views of the population.

Perhaps the National Conversation should be assisted by survey results on several important topics affecting the people of Singapore. The survey can be based on 100 people, who are selected to represent the population. Several survey can be carried out with on different topics and can involve a different group of people - to reduce the stress on any single person answering too many questions.




No comments:

Post a Comment