Dear Editor,
Stabroek News and Kaieteur News are the nation’s only independent newspapers in a country dominated by state media. What I have always stressed with Mr Bisram and his NACTA organization is how he presents information, his selective use of information and lack of verification. Why does Stabroek News have an uncommon duty to carefully screen Bisram?
Because many politicians and parties often parade these NACTA findings as evidence for their parties or against their opponents in the heat of election season. This will be a very close election. Mr Bisram’s latest submission is his mid-November 2011 poll. His sample is already a running joke. He used 44% Indians in his sample when academic, Rishee Thakur, recently conducted a revealing analysis using mortality, fertility and migration trends to show that Mr Bisram’s estimate of the Indian population in his sample is 8% higher than what it actually is. There is no way Bisram as a veteran pollster does not know this.
Mr Bisram’s modus operandi is simple: increasing the number of Indians in the sample increases favourable findings for the PPP given that Indians overwhelmingly vote PPP. Mr Bisram’s statistical manipulation does not end there. His sample changes, albeit slightly, without reasonable explanation or support. His July, August and October 2011 polls pegged Indians at 45% in his sample. The figure then dropped to 44% a few weeks later while the Amerindian population pegged at 8% in the July, August and October 2011 polls increased by 1% in mid-November 2011. He had Amerindians at 8% for three polls since July 2011, notwithstanding the fact that the 2002 Census confirmed the Amerindian population at 9.16% and it is along with the Mixed Races the fastest growing population.
Bisram’s sample size goes from 780 persons in July 2011 to 420 in August to 500 in October to 560 in mid-November 2011.
Even the primary school maths student knows that the bigger the size of the sample, the better indication of statistically reliable results. The PPP was 46% in July with a sample of 780 and increased to 50% in mid-November 2011 with a sample of 560, a whopping 220 persons less. A decreasing sample as the election gets closer, coupled with what appears to be an inflation of the size of the Indian population in the sample, unequivocally favours the PPP.
Another problem I have noticed with Mr Bisram is what I call the voter turnout debacle. He missed it completely in 2006. A smaller sample size, inflated Indian population size in the sample and reduced number of undecided voters clearly favour the PPP in his findings. Undecideds went from 22% in July to 20% in August to 17% in late October and finally to 9% in mid-November 2011. To put that into perspective, a 91% turnout in 2011 would match the highest ever turnout in any election in Guyana. Why is Bisram not telling us how many and what percentage allegedly told him they will not vote and from which ethnic, age, class and regional segments? Is it because it hurts the PPP?
Mr Bisram has been supposedly conducting polls for a long time now. We have never seen any statistics of how various ethnic groups intend to vote in any of these polls.
Mr Bisram engages in the game of trying to create a self-fulfilling prophecy that favours the PPP. This is a proven tactic that works with unsophisticated voting populations whereby PPP supporters will see the majority findings and those thinking of voting for another party will simply fold into the PPP tent. Stabroek News should beware. Same for the voters out there.
Yours faithfully,
M Maxwell
