Limited test should remove doubt on voters’ list authenticity

Apr 10, 2015 News

---

– EAB statistician  

…will indicate deficiencies where they exist

By Zena Henry
In the general view of things, there is always room to accommodate “illegal” occurrences with the voters’ list when especially; there is no easy way of removing unwanted names; such as deceased persons.

This negative view is growing into a greater monster; birthed by this year’s voters’ list which has grown by a colossal 75,000 from the 2011 period to now. It has thus been of grave concern to several persons, particularly the major contending parties, despite GECOM stating categorically that it has measures in place to ensure that there is no ‘hanky panky’ involving voters and the list.
GECOM’s guarantee has however failed to reassure some persons who continue to urge that the current list’s growth leaves a lot of space for undesirable occurrences and are calling for something to be done to remedy the situation.
The Electoral Assistance Bureau (EAB) has posited however, that following a ‘limited test’ on the Revised List of Electors, all spheres of concern should be quelled since the activity seeks to give credence to the authenticity of the list. While the test cannot bring 100 percent verification to the issue, the probability of its accuracy is high and the test will also indicate any issue of concern in the respective regions and will afford GECOM the opportunity of clarifying whatever problems that may exist within a specific area tested.
Troy Thomas, the professional statistician providing his service pro bono to the election watchdog agency, told Kaieteur News yesterday that the limited test is very important. He said that should results be close, “no side should be able to blame the list.” He said everyone must be satisfied with the list and the only way to address that is to go looking for these persons.
EAB is supposed to meet with GECOM officials today and will be seeking from the election agency a copy of the list. EAB will be working with 80,000 names and will be testing 1200 of those names which will be chosen randomly using among other statistical methods, stratification which deals with proportionality. This will encourage 95 percent confidence, a typical level for the activity, while ensuring a 3 percent error margin, which is also an acceptable rate for this type of activity.
The acceptable statistical sample size for this testing can be some 1054 participants, but again seeking a 95 percent accuracy rate, EAB will be working with 1200 samples.
Thomas told the publication that regardless of how it works out, the important issue is to ensure proper representation of the various areas being tested. Samples from all 10 regions will be used, but how they are selected is very relevant. Hence, to accumulate the sample size of each region, the statistician will be employing the proportional method of stratification among others, to raise his samples and will employ methods of calculation relative to the procedure. The individual samples from each region will be selected randomly.
The aim of this test is to identify and locate those 1,200 people that will be selected. When asked, Thomas reiterated that the 1,200 persons out of the 75,000 will be sufficient to provide a true reflection of what the list comprises.
“This is an adequate amount in the statistical world,” Thomas said.
The result of this test should be accepted, Thomas opined, as he expressed confidence in the mechanisms. But while the statistician said that he accepts the application and relevance of the apparatus, explaining it to the layman could be a challenge.
Former EAB Project Directors, Clairmont Lye along with Lawrence Lachmansingh are ”volunteering their time and resources to coordinate the implementation of the list test”.
Lye told the publication yesterday that their job deals specifically with the list. He said that consultations with the major parties prove that they are seriously concerned over the growth of the list.
He said given the time span, the agency will be seeking the cooperation of citizens, specifically those selected. He too explained that the test should erase doubts about the list having serious problems given the increase.
However, if discrepancies are indicated, then the necessary recommendation will be provided to GECOM which is the body tasked with ensuring that whatever discrepancies exist, are addressed.
The U.S. Embassy has committed US$300,000 for the election monitoring programmes, of both the EAB and the Organization of American States (OAS). Thomas had expressed financial concerns in relation to the body carrying out its function.
The Embassy has expressed confidence in both EAB and OAS, stating their pleasure in supporting the two organizations “in bringing their credible and critical analysis to Guyana’s 2015 electoral process.”  The significant increase in the voters’ list promoted EAB’s move to conduct the test.

---

---