To project the outcomes of General Election 15 (GE15) in Peninsular Malaysia, EMIR Research adopted multi-technique approach that is rigorously holistic:
• Historical trend analysis by looking up at least four general election results back, including general by-elections whenever data was made available;
• Quarter on quarter Emir Research sentiment analysis through a nation-wide (Kangar to Kota Kinabalu) survey that encompasses quantitative and qualitative methodologies and multi-stage sampling (cluster-sampling and quota sampling) which helps to get the most representative sample of the Malaysian population and at the same time allowing randomness.
• Modelling GE14 outcomes as a function of ethnicity and urbanisation under logit framework in a similar approach to that deployed by Ng, Rangel and Phung in their empirical work “Malaysia’s 14th General Election: dissecting the ‘Malaysian tsunami’—measuring the impacts of ethnicity and urban development on electoral outcomes” published in Asian Journal of Political Science in 2021. Ng with the colleagues have shown that the key predictive factors for the probability of BN winning are proportion of Malays in constituency and level of urbanization.
• Poll aggregation — a meta-analytical study of the poll results reported by various Malaysia research houses over the period from mid-October 2022 to mid-November 2022. EMIR Research team has paid attention to the distribution of poll-reported popular votes across few important demographic variables such as gender, ethnicity, and domicile. Then these results were combined with the actual proportions of those demographics among the voters in various states in Malaysia and averaged using meta-analytical software.
To read the full content, please click the button below: