Countering Espeut's assertions

Bill Johnson , GUEST COLUMNIST

Peter Espeut's column on Friday, July 8, relating to the results of two questions in the recent Gleaner poll, is an example of distortion and misrepresentation and why some people are distrustful of what they read, see or hear in the news media.

Espeut is described as a "sociologist and survey research practitioner", so he really has no excuse for his unfounded diatribe against some of the survey's findings.

Espeut starts off his column fair and balanced. He doesn't seem to dispute the fact that 60 per cent of those interviewed believe they themselves, and people like them, would be better off today if Jamaica had remained a British colony, compared to 17 per cent who feel they would be worse off. He states: "Both the poll results and history are an indictment on the quality of our political governance since Independence."

But he then claims that the next question, which asks Jamaicans if they think the present Westminster form of government should be continued or be replaced by a republic form of government, was biased because of what he claims is "question-loading".

Question-loading, whether intentional or not, is a serious charge, and I personally take great offence to it.

The fact is that there were 48 questions that preceded the one Espeut has concerns about. In theory, any one of them might have influenced, to some extent, those which came after it. But serious researchers like myself do everything possible to minimise or eliminate that possibility.

The actual wording and sequence of the two questions was:

"As you know, it was almost 50 years ago that Jamaica became independent from England ... :

48. Generally speaking, do you think Jamaicans like yourself would be better off today if Jamaica had continued to be a colony, or do you think you'd be worse off today if Jamaica was still a British colony?

Jamaica has had the so-called Westminster form of government since it became independent from England ... under the Westminster form of government, the people elect members of Parliament, who then elect the prime minister ... .

49. Some people have said that Jamaica should change its form of government to a republic ... . Under a republic, the people would directly elect a president, who would be the highest and most powerful elected official ... . They would also elect members of Parliament ... . From what you know about it, do you favour replacing the present Westminster form of government with a republic form of government, or do you think Jamaica should continue to have the Westminster form of government?"

As a Gleaner columnist, Espeut certainly had access to these two questions.

With all due respect to the survey research practitioner, what on earth is the basis for his assertion?

"Asking whether Jamaica would have been better off under British rule, and then asking whether we should retain the British monarch as our head of state, is another good example of loading the question; since many believe we might have been better off under the British, maybe they were led to answer that we should retain the British monarch as our head of state."

One could be forgiven for believing that these accusations were made out of whole cloth.

First, question 48 was not whether or not Jamaica would be better or worse off, but whether the respondent and those like (him or her) would be better or worse off.

Second, nowhere in either question is the term British monarch to be found. Although it is true that under a republic the British monarch would be replaced, much of the conversation about whether or not to retain the Westminster form of government centres around the direct election of the country's leader, and that was the emphasis of the question. A by-product of this would be the replacement of the monarch.

As much as serious researchers should avoid question-loading, they should avoid reading something into a question which just isn't there and also fancifully interpreting what the question might have meant to the respondents.

By the way, additional analysis of the results of the two questions shows there is no apparent relationship (or possible bias) between the answers to the first and second question.

Maybe it's time for the survey research practitioner to take a refresher course in survey research methodology.

Bill Johnson is a pollster and researcher. Email feedback to columns@gleanerjm.com.