I support Dr. Cheddi (Joey) Jagan Jr.’s view (KN Dec 3) that opinion surveys (called polls) must be scientific.
In Guyana, I use the scientific method for the NACTA polls and have been quite successful in capturing the mood of the electorate as reflected in election results measured against the findings.
If properly conducted, polls can be a predictor of elections. But a poll can be off target if there is a last minute swing by the electorate. The important point about polling is it must be scientific and not “straw polling” as done over the radio or TV or internet. A scientific poll has nothing to do with political commentaries.
In the U.S, where I studied polling and social science survey techniques, polls relating to the political establishment are conducted with regularity (every month) by dozens of different polling organizations to measure the President’s approval ratings, support for the war in Iraq, Afghanistan, etc. and other issues. Polling is an ongoing process in which support is not static but dynamic. For example, when Obama became President, he had an approval rating in the 70s and now it has slipped to below 50% but it has see-sawed during the last 11 months.
All of the polls that seek to capture the representative views of the population are based on scientific techniques, which is a very complex process that can be described briefly in the following steps: defining the universe (whole voting population), constructing a sample (representative slice of the universe), preparing valid questions (not loaded towards an answer), selecting and controlling the means by which the poll will be taken (conducted at random); and reporting the findings (publication).
In scientific polling, pollsters usually draw a random sample or probability sample from among the universe. A pollster devises an appropriate methodology for a poll depending on the location. Ideally, opinions are taken from a systematically randomized sampling of the population. The poll-taker interviews people from among the universe. In the U.S, it is made easy by computer selection.
In Guyana, NACTA randomly interviews (face to face) voters. In the U.S, pollsters generally interview about 1500 people to represent the universe of about 250 million voters.
In Guyana, about 600 is a good number to represent a voting population of about four hundred thousand.
In the U.S, the sample is constituted to proportionally represent the geographical (regional), ethnic, class, age, educational, religious and gender diversity, etc., of the population. In Guyana, NACTA makes every effort to include the diversity of the population in the sample.
A scientific statistical sampling of the population to include the varied social categories would in theory give very accurate results of a poll. In the U.S, pollsters use a complex mathematical equation from a (SAS or SPSS) program that generates a sample from the registered voters’ list. Then the names are telephoned with their opinions recorded by the computer or interviewer.
The program then calculates the results and errors using a standard formula.
Pollsters usually analyze the results at a 95% significance level which means that in 19 out of 20 cases the results are accurate. In Guyana, the lack of telephones in every home makes it very difficult to conduct polling over the phone. Therefore, polling must inevitably be conducted by interviewers in what is known as “intercept contact” done at random.
American pollsters recently used this intercept contact methodology in their third world surveys. Pollsters in India, Mexico and other third world countries also use this method.
I conducted an opinion survey on current issues last week using the scientific method described above. I used six interviewers who have worked for me over the last several years — including four educators and two former UG students who are engaged in ongoing survey activities for institutions in Guyana. Three intellectuals offered their comments about the questionnaire sheet before it was sent out to the fields for responses. The findings will be released soon.
Feb 22, 2020Regal All Stars sink Ariel Speedboat at Providence By Sean Devers A disappointing crowd saw Regal All Stars beat Ariel Speedboat by 14 runs in the Softball encounter before on-target bowling from...
Feb 22, 2020
Feb 22, 2020
Feb 22, 2020
Feb 22, 2020
Feb 22, 2020
Oil revenues will not save this country. A sound economic model will. Unless the problems in the economy are fixed, no... more
Editor’s Note, If your sent letter was not published and you felt its contents were valid and devoid of libel or personal attacks, please contact us by phone or email.
Feel free to send us your comments and/or criticisms.
Contact: 624-6456; 225-8452; 225-8458; 225-8463; 225-8465; 225-8473 or 225-8491.
Or by Email: [email protected] / [email protected]