What surveys won\'t tell you
Mahmudur Rahman | Published:
April 05, 2016 19:54:12
October 18, 2017 11:19:11
It may have taken a while to comprehend Mark Twain’s beef with statistics. True to his nature he was short and dismissive in his oft quoted comment, ‘There are lies, damn lies and statistics’. And it came much before the US allowed its President to play around with statistics ‘in state interest’. In short, the chief executive can use falsehood to assure the citizens.
Surveys were few and far between during Twain’s days and certainly not as pervasive as they are today. Apparently surveys are required at the drop of a hat whether to determine public opinion or a business decision. The craft of surveys and opinion polls ranging from methodology to framing of questions has become highly specialised. And with any form of specialisation the question of quality jumps out of the mass of figures and conclusions.
Pre-poll surveys before the re-election campaign of India’s Bharatiya Janata Party (BJP) had indicated a smooth return albeit with a reduced majority. Exit polls went along similar lines. And red-faced pollsters had only one escape route when results went the opposite way; pointing to the ‘undecided’ voters that invariably make up a third of any electorate.
A major multinational faced with the difficult decision of raising prices of its flagship product were informed that it would result in an 18% drop in sales. This was steep but the arithmetic made sense. When to management consternation sales volumes plummeted to 38%, the smooth marketing head pointed to advertising and promotion material not being in place in time. The cost was two-fold; implementing the change was expensive and the sales figures led to the profit and loss (P&L) sheet looking rather embarrassing. It didn't help when the research agency responsible for gauging customer pliability pointed out that their estimate had been 40% sales drop. It was apparent the figures had been manipulated to meet approbation.
In the same way the recent survey by the International Republican Institute suggests majority of those surveyed believe Bangladesh is heading in the right direction and that the law and order situation is in better shape than ever. Given the spate of horrible crimes against children and women, this is a hard one to digest. While the agencies can't be held responsible for crimes, it has been the inability to unearth the crimes and provide exemplary punishment to those responsible that has raised ire. Matters have come to a head where the devastated father of a murdered blogger says he doesn't want justice. And friends of a photographer whose body was found near Narayanganj a day after vanishing from Dhanmondi a couple of days ago say unabashedly that they don't expect justice.
Broadly, surveys are either qualitative or quantitative and serve different perspectives. Quantitative or ‘Quanti’ surveys focus more on numbers and straight forward statements in the Yes or No category. This is probably the ‘statistics’ that Twain despised so much. Qualitative or ‘Quali’ research bores deeper into the ‘why’ of the response, beyond the ‘Yes’ or ‘No’. Unless the questions themselves are sharp enough ‘Quali’ surveys can be extremely misleading and therefore open to ‘interpretation’ (read manipulation).
There have been two incidents where Agriculture Minister Motia Chowdhury and even Finance Minister AMA Muhith were dismissive of facts and figures produced by the Bangladesh Bureau of Statistics (BBS). Lest it be forgotten, the Statistical Year Book is a major informational input into the budget and Annual Development Programme (ADP) exercise. Few would admit as such, but many research projects depend on responses from individuals and there are times the field-level enumerators are guilty of filling in the questionnaires themselves rather than actual interviews. This is an outcome of too many surveys being run and the flip side of the enumerators’ action is an increasing apathy by those being surveyed to spare the time.
The Centre for Policy Dialogue (CPD) runs an annual survey among businesses on the ease of doing business, bottlenecks, corruption and way forward. Few organisations welcome this simply from the perspective of time and effort and outcome. The bulky questionnaire is thus a hot potato that changes hands frequently and is more often than not reflective of a single individual’s view. On paper this is an important one. But businesses aren't think-tanks especially for top management who squeeze their moments pretty much the way a youngster scrapes the last bit of ice cream from the container.
And now we are hearing of a UN report that isn't being viewed favourably. Obviously another report based on research findings. That isn't a surprise. As the clock ticks down to Britain’s referendum on Brexit (whether or not the country should stay in the European Union or EU), it appears that the government, which favours staying in, will only succeed if the ‘undecided’ and habitual ‘non voters’ turn up at the hustings. Apparently without them Britain will say ‘out’ and the government might well have to follow suit. But this is just one poll. The tide does tend to change and this is one survey that David Cameron will hope does prove to be wrong.