Monday, March 08, 2010

Political Surveys - How Reliable Are They?


“I gather, young man, that you wish to be a Member of Parliament. The first lesson that you must learn is, when I call for statistics about the rate of infant mortality, what I want is proof that fewer babies died when I was Prime Minister than when anyone else was Prime Minister. That is a political statistic.” (Winston Spencer Churchill)



The use of research whether applied to marketing or scientific endeavors that better the human condition is a vital and noble occupation. Advancements in most of the things that people now enjoy have been, in one way or the other, helped out by research. It is a very powerful tool in the pursuit of new knowledge and the improvement of existing wisdom.

Like most effective scientific methods with wide practical utility, research is easily subjected to misuse though, sometimes, unwittingly by one or two of the parties involved in the research.

My own personal view is that it becomes suspect when used in politics, especially in the local context, where the motive of the fact finding is often ill intended
Is it research when the intention is to corroborate a preconception that puts in a good light the advocate? Is it research if all that it wanted was to bring out a set of results which would be beneficial to specific interest groups? Is it research if the goal was to discredit and besmirch personalities and institutions? Is it research if it falsely tries to create impressions of irreversible trends to sway constituent preferences? Is it research if it is used to help in the deception and in the cover up the ugly truth of personalities and institutions?

Of course there are legitimate political researches. You can include those researches whose results have been used to improve strategies whether logistical and or image related (although a debatable ethical area). Other legitimate uses of research are exit poll surveys (if done properly) who serve a beneficial purpose in that it establishes results in the soonest possible time so that manipulations may be preempted and the early settlement of the outcomes of the election eases the tension and allays the fears of the voting public. Political research is useful to political scientists in the study of electoral choices of different groups and the underlying reasons for their preferences. Sociologists are able to trace the evolution of social issues by studying their shifts. And for the citizen, it allows his voice to be heard and it gives him an idea of how he differs with other individuals and against sectoral aggrupations other than his own.

Good research, bad research. It all depends on the motives of those commissioning the research and the willingness of the researcher to go along with them. There is also the media who may abet the ill intentions of the research by completing the disgraceful cycle when they make the results public.
There is no arguing the right of the research to be published as the constitution guarantees press freedom and freedom of speech, but as in any of the freedoms, it carries with it the responsibility of publicizing only those that are not inimical to the public interest. The harm of malicious research cannot be underestimated as it has the power to sway people’s opinions and create strong emotions that could be injurious to a person or to a state.

Research is too complicated to be fully appreciated by laymen or even from those a rung higher. Statistics has the unfortunate association to lies. Great thinkers and statesmen have remarked about their discomfort with statistics because it has been often used by their detractors and by themselves to make vague the issues that are indefensible in plain terms.

The obvious solution to the problems of political polling is to educate the masses about research so that they may understand the full extent of the information as well as the limitations of the data being shown. They should be able to understand the limits of the inference that can be made of the data and to know the difference between data that may be confidently considered as reliable against those that show a tendency towards an outcome but cannot be relied on. This, of course, presupposes that the research is truly objective and that the methodology is the most appropriate to arrive at the research objectives and that it is free from bias in the design, the sampling, wordings of the questions, the order of questions, the fieldwork’s operations and personnel. The approach to the analyses of the data should ensure that it is exhaustive and no omissions or commissions in the analysis should favor a particular result. It should also make sure that the statistical tests and the advanced statistical methods, when employed, are the appropriate ones.

The education of the electorate is a good thing but it may take a long time in progressing. I would favor the creation of a group made up of professional researchers, political scientists and those in media responsible for the publication of the results of surveys who would scrutinize all research projects that are to be commissioned and those non-commissioned but subscriber supported surveys(syndicated researches) before they are allowd to go on field. This is not censorship of the content of the survey. It endeavors to make the surveys fair to affected parties and personalities by ensuring that it has undergone a rigorous check that will result in findings that are with integrity.

No comments: