The Centre for Governance, Economic and Election Research-Ghana has predicted a win for the presidential candidate for the NDC, former President John Mahama ahead of Saturday’s polls.
The survey which was conducted in November 2024 in some selected swing constituencies portrayed a clear victory for John Mahama.
The survey was spearheaded by Dr. Gbensuglo Alidu Bukari, Head of the Department of Political Science for the
University for Development Studies and
Mr. Maclean Ninsaw Gbati of the Department of Political Science at the University of Ghana (Legon).
According to Dr Bukari, the survey was based on the rational choice model, which recognized that most voters, particularly the swing voters, are not identified with any political party.
Below are the findings;
Summary of findings
In this study, six hundred (600) respondents were randomly selected in a cross sectional survey in six swing constituencies.
This sample was selected at 95%
confidence level of significance, and the margin of error was 5%. Based on the
response rate, we expect the total voter turnout on 7 December 2024 to be 80% (±5).
In the study, 57.4% of our sample respondents said they would vote for the NDC
presidential candidate.
Also, 27.5% of the respondents said they would vote for the NPP presidential candidate, while 15.1% indicated their support and vote for other candidates (key issues for consideration were performance, policies, personality, party unity, economy, education, health, jobs, agriculture, and vice-presidential running mates). Based on the findings, we expect the NDC candidate to win the presidential election with a victory margin of 52.4 or 57.4, all things being equal.
Methodology
The study design was cross-sectional. We adopted a multi-stage cluster sampling
approach. For the purpose of this study, the whole country was clustered into three (3) electoral zones (i.e., the northern electoral zone, the middle electoral zone, and the coastal electoral belt). The study was then conducted in six (6) randomly selected constituencies in the electoral zones from 1 November 2014 to November 30, 2024. In the study, two constituencies were randomly selected from each of the three (3) electorally clustered zones.
The study questionnaires were then administered to two hundred (200) randomly selected respondents in each of the six (6) selected study constituencies in the three (3) electoral zones of Ghana (for the purpose of our study).
The sample frame for our survey includes all citizens of voting age (18 years) and
above within the country who are registered voters at the time of the study. The
determination of the sample size was based on the following assumptions.
i. The study assumed confidence interval of 95 percent. It is assumed that at 95 percent confidence level of statistical significance, the right decision was made in choosing the sample size for the study. This statistically significant level means that the probability of obtaining the findings by chance only will
be less than 5 percent. In other words, there are about 5 chances in 100 that the study would reject the null hypothesis, when in fact it should be accepted.
ii. Since we do not have control over the respondents in the survey, the study assumed the response rate to be 90 percent. The study assumed a split decision of 50-50 chance of responding to our questionnaire;
iii. Only Ghanaians who are 18 years and above, and are registered voters were
qualified to respond to the questionnaire designed for the study;
iv. The random selection of the respondents was based on registered voters, with a significance level of 0.05 percent at 95 percent confidence level for the study.
The sampling frame was the Electoral Commission of Ghana 2024 voters register
which captured all eligible voters of 18 years and above for the 2024 elections.
The minimum statistically acceptable sample size was determined by employing a
simplified formula to calculate sample size at 95% confidence level and P = 0.05 are
assumed. It states that; 𝒏 =𝑵(𝟏+𝑵(𝜶𝟐)
, Where; 𝒏 = sample size; 𝑵 = sample frame
or total registered voters of 18 years (18,774,159); α = confidence interval or level
of precision; and 𝟏 = constant. By substitution, we have 399.76 (n = 400).
So, to cater for the non-response, we added 200. Therefore, the sample size of 600
respondents was considered for the study.
Source: Myxyzonline.com
