METHODOLOGY OF POLL CONDUCTED EXPLAINED

March 21, 2007

There have been a number of questions raised over a poll conducted by the University of Belize and the Belizean Studies Resource Center of Saint John’s College. The poll released yesterday indicates that Belizeans are leaning toward a new Prime Minister and a new Government. Today Love News spoke with Yasmin Andrews from the Belizean Studies Resource Center of Saint John’s College who said they methodology used in this survey was similar to that used in the previous four polls. She said most of their polls have proven to be precise and accurate.

Yasmin Andrews:

“We selected the sample from the latest statistic we looked at rural population and urban population. We looked at each district and from each district when we selected a proportionate sample from the entire private population. Most like Belize will have the biggest sample because it comes from the biggest district and then Cayo the second to the largest and so on and so fort. The type of sampling technique that was used is called “Systematic Random Sampling” and there is two ways you can do this there is one use the phone book and the second method is using a computerized system but at the end of the day the important thing is that we try to eliminate the bias that would exist if we use something like “Convenient Sampling” so that will explain the sampling methodology. The statistics that we use we have been using the same programme that we have been using “Statistical Package for Social Scientists”. The method of data collection the University of Belize selected four students four so Belizean Studies Resource did not have input in the selection of the students; however we trained those students how to conduct the polling so the training was done by myself  and the Registrar’s Office at the University, it was done during the weekend. When we start a poll the timing is not very important however it has to be consistent saying that you cant do one Saturday and do another one on Monday it has to be right after the other, ,to eliminate again any bias that may occur where people may change their minds between Saturday and Monday. The actual collection of data was done by these students over phone, it was done over the period of two days and the numbers are just phone numbers so therefore there is no way we could connect a phone number to any name and the students are all trained with one questionnaire and they should ask the same question uniformly that means it cant be changed it unless they have to interpret it into Spanish and again we have the Spanish format and again we have the Spanish format that they just read each phone number each questionnaire. At the end what they are entering are just codes to these questionnaires and we show them how to do that. After the collection we have the data reentered into our computer programme coding everything so again absolute confidentiality we have no idea where these numbers are the numbers are randomized in order to represent a generalized population and then we do the analysis.” 

Andrews said the questions were compiled by both institutions.

Yasmin Andrews:

“If you recall the last poll that was done it was a generic question “Would you vote for a third party?” however at that time there weren’t the name of the third party. What we did this time was detailed what should this third party consist of and we know it consist of four names. So what we did was asked each question individually like “Would you vote for Cornelius Dueck?” “Yes or no” next question would you vote for this person yes or no. the questions were again in collaboration of both institutions where we looked at it we both tested it, we tested the questionnaire for reliability and validity so  what we are asking is exactly what we are getting and the validity is on the samples that we collected is exactly representative of the entire population of Belize. One of the questions about the leadership we know the type of system we have in Belize however that question about “Would you vote for a Prime Minister?” more relates to the level of leadership of any party. So we asked “Who would you vote for a Prime Minister?” relates to not system per say but it relates to the leadership capability, so we used that nationwide instead if using it for just one area or one target population and the same thing for the other questions.” 

Andrews said any member of the public who would like to find out more about the poll can visit their offices.

Yasmin Andrews:

“We have all of our data all the hard paper available we have all the statistical output and the methodology of the sampling the phone numbers which they could feel free, ,now the public can go back and check where those numbers came from and any other doubt that any person has out there kindly come in I will personally sit down and I think even Dr. Young will be willing to sit down with anyone and because we also have research and statistics back ground and try to eliminate any questions and doubt that could be out there and that’s reasonable to assume. But that’s what we are here for I think that at the end of the day what we are trying to do is one to encourage more research, two to try and engage our students in conducting research and thirdly to try and gather opinions based on a scientific manner and I guess that is the ultimate objective. The margin of error might have been plus or minus five percent and for each result you could either add five percent or subtract five percent.” 

Yasmin Andrews from the Belizean Studies Resource Center of Saint John’s College.

|

|