Dear Editor,
I have just completed an assessment of the “preliminary” Region Four results, as declared on two separate occasions by the Returning Officer for Region Four, and the PPP/C’s published Statements of Polls. The following are my conclusions:
We conclude that for all the “preliminary” results, we can reject the null hypothesis, at even the 1% significance level, that there is no difference between the distributions of votes cast for the two major political parties in 2020 and the average votes cast in their favour in the elections held over the period 2001 – 2015. This is an important result because it means that one cannot use the results of the elections held over the period 2001 – 2015 to contend that any of the preliminary results for 2020 is more “reasonable” or more “valid” than the other. Similarly, it also suggests that one may not be able to use these earlier elections to make predictions about the relative electoral fortunes of the two major political parties in the 2020 elections.
Additionally, at the 1% significance level, the two declarations made by the Returning Officer for Region Four were completely unrelated to each other, even though they were arrived at using the same process for the same population of votes. This latter is a very serious finding, as it calls into serious question both the declarations made by the Returning Officer for Region Four for the two major political parties and compounds the more obvious problem created by the clearly incorrect tallies for the third parties that contested the 2020 General Elections.
Finally, we note that if the PPP’s SoPs are correct, then we can say with 99% confidence that (only) between 52.64 – 53.01% of the valid votes would have been cast for the APNU+AFC in Region Four. In contrast, the two declarations by the Returning Officer for Region Four that either 62.9% or 62.8% of the valid votes were cast in favour of the APNU+AFC are outside the realm of what is probable if the PPP’s SoPs are correct. We note moreover, there has been at least one potentially independent corroboration of the PPP’s SoPs.
It is safe to conclude therefore that there is no logical reason why the declaration of the Region Four results should be based on either of the GECOM declarations. Additionally, if the process employed by GECOM as the basis for the two declarations made for Region Four produced two declarations that were wholly independent of each other, and if the unofficial (PPP SOP-based) process is wholly incompatible with the results inherent in GECOM’s declarations without being completely independent of them, there may be some significant merit in making a simple comparison of the SoPs for Region Four that were already verified with the SoPs of the PPP/C, to determine if there is sufficient correspondence between the two to justify either the completion of the process of verifying the SoPs for Region Four or proceeding with a recount. Statisticians may even be able to calculate the size of a random sample of these verified SoPs, to save GECOM the cost of doing a full check.
Anyone desirous of reading the full assessment might do so at https://drive.google.com/open?id=1KbEdQoT3Wu96h1XJW7wV0l4UUsZ1CY1t.
Yours faithfully,
Thomas B. Singh
