Bisram’s sample was manipulated

Dear Editor,

Vishnu Bisram of NACTA fame claims he did a recent poll in early July 2011. Mr Bisram published some findings of those polls in two letters in Stabroek News titled ‘Poll finds 42% approval rating for government’ (SN, July 16) and ‘NACTA poll finds PPP in lead’ (SN, July 17). The second letter stated that the PPP had 46% support. These polls are statistical subterfuge.  The problem starts in the formation of the sample. Bisram states, “The poll randomly interviewed 780 voters to yield a demographically representative sample (45% Indians, 30% Africans, 16% Mixed, 8% Amerindians, and 1% others) of the population.” That is not a demographically representative sample. The sample is based on race. So why does Mr Bisram have to interview people if he is simply selecting them based on race? You either appear African or Indian or Amerindian or Mixed Race or you don’t. I cannot see the need for an interview when the pollster is solely selecting on race. An interview process can taint the entire poll. Mr Bisram’s poll is inconsistent with the population’s statistical trend since 1980. The sample and resulting poll results are structured to give the PPP an advantage. It is to generate a self-fulfilling prophecy.

In 1991, which is just before the PPP returned to power, the ethnic percentages of the major ethnic groups out of the total population were as follows: Indians 48.63%, Africans 32.26%, Mixed Races 12.14% and Amerindians 6.46%. In the 2002 census, these percentages were as follows: Indians 43.45%, Africans 30.20%, Mixed Races 16.73% and Amerindians 9.16%. The truth is that between 1991 and 2002, the ethnic populations changed as follows with respect to the total population: Indians decreased by 5.18%, Africans decreased by 2.06% while Mixed Races increased by 4.33% and Amerindians increased by 2.7%. If we venture back to the 1980 census, the Indian population was 51.93%. Only two ethnic populations have grown in 20 years of census taking.

The Indian population has steadily declined since the 1980 census. It has not grown. It did not grow when the party it votes for overwhelmingly in the form of the PPP returned to power in a wave of euphoria. It actually fell. Is Mr Bisram saying that after the 2002 census when crime, cost of living and corruption under the PPP escalated to their highest levels under the wonderful watch of President Bharrat Jagdeo, that the Indian population grew in Guyana? Is Mr Bisram saying that a population that fled this land left, right and centre actually bucked and reversed the trend of the 1980, 1991 and 2002 censuses and actually grew? Yes, Mr Bisram is telling us this statistical Anansi story.

I am calling on Vishnu Bisram to prove his 45% Indian demographic representation in his sample. I am calling on Mr Bisram and NACTA to prove that the Amerindian population fell from 9.16% to 8% between 2002 and July 2011. I am calling on Mr Bisram and NACTA to prove that the Mixed Race population fell by 0.73% between 2002 and July 2011. Mr Bisram has reduced the Amerindian and Mixed Race populations in this tricky representative sample he masterminded. It is another act of statistical manipulation. For the Amerindian and Mixed Race populations do not overwhelmingly vote PPP. To reduce their representation in the sample at the expense of increasing the Indian representation is Machiavellian. A proper demographic representative sample using statistical evidence gleaned from censuses and statistical trending should put the PPP’s current support in July 2011 somewhere around 41% using Mr Bisram’s own results. It will put the current PPP government’s support in July 2011 at around 37% to 38% again, using Mr Bisram’s own numbers. The two-card trick games played by Mr Bisram and others is at its end. Let change and truth prevail.

Yours faithfully,
M Maxwell

Around the Web