OPINION POLLS IN KENYA. ARE THEY RELIABLE?

Kenya can be credited with being one of the few countries in Africa that has stable and consistent pollsters who are able to operate freely and without control from government and other interested parties. The opinion polls however do come with varied findings, although the pollsters have attempted to explain the varied findings. Most explain the differences to be as a result of the variance in their research design.

Estimates based on the Standard Newspaper Online Report, See here

While the trend of opinion polling is encouraging and vital in a democratic system, such as Kenya, pollsters and the public needs to watch out for particular concerns regarding validity and reliability of such polls. In a society as diverse as Kenya, and one in which election is characterized and influenced by ethnicity, errors are almost inevitable. The acceptable scale of errors becomes the major question.

In the past, certain opinion polls have resulted in large inaccuracies all over the globe. The USA is a good case in study. Labovitz and Hagedorn (1971) in their book “Introduction to Social Research”, give several cases in which opinion polls misled the public with regard to elections. I will use two cases here to explain my case.  In one case, the writers indicate that, in the Literary Digest Poll of over two million persons that predicted, in 1936, a landslide victory for Alf Landon over Franklin Roosevelt. Of course, the 1936 election overwhelmingly favoured Roosevelt. The poll had erred by approximately 20 percentage points.

A second instance occurred when in 1948, when polls by Gallup and others forecasted the wrong winner of a presidential election. Some polls showed that Harry Truman would receive less than 45% of the popular vote, when on the contrary; he received just about 50% of the votes and, of course won the election.

These cases show how polls may mislead the public and worse, contestants into false confidence. In the Kenyan case, in 2007, despite several polls tipping Raila Odinga as the winner of the election by comfortable margins, the results, however controversial, give us an indication of just how misguiding opinion polls can be.

I am a strong believer of research, as a matter of fact, I am a researcher. However, research is faced by a common problem of generalizability. In this case therefore, however genuine polls in Kenya might be, the ability of such findings to be generalized as a representation of the opinion of Kenyans remains questionable. The accuracy of the polls, as any person would say, remains to be seen come the March, 2013 results. Contestants in the 2013 elections should beware lest they have false perception in their popularity or their lack of it. Until then, however, Prime Minister Raila Odinga remains the man to beat.