This is an abridged Engilish version of the original article in Malay language written earlier today in this blog.
This prediction model is based on multiple regression analysis of data in all of the 222 parliamentary constituencies in previous elections. Several factors have been chosen and the data are tested as independent variables.
These factors are found to be highly significant, namely number of Malay, Chinese and Indian voters as well as those of the natives of Sabah and Sarawak. Other factors include the competing partys’ traditional influence in the constituency and the number of times the incumbent stood in previous elections. All the above factors are found to be highly significant and have substantial impact on the outcome of the election.
I have done 12 by-elections and the Sarawak State Election forecast from 2009 to 2011. All my past by-election forecasts turned out to be accurate, except one.
My forecast for the outcome of the Sarawak State election was very accurate up to the number of seats that the opposition obtained. You may access all these past forecasts in this in this blog.
In this prediction, for Peninsular Malaysia, I estimate that 61% of the Malay voters will vote for Barisan Nasional, 25% of the Chinese and 60% of the Indians . These figures provide input to the model along side with the data on the traditional influence of the party in each constituency. In Sarawak and Sabah a slight variation of the model is used in view of the different racial composition in these states.
The prediction derived from the model is shown in the table below:
|Sabah & Sarawak|
The above table shows that Barisasn Nasional will, on worst case scenario, win a toal of 159 seat, DAP 35, PKR 19 and PAS 8 seats. In a best case scenario BN will win 170 seats while DAP will take 33, PKR 14 and PAS 5 seats.
In order to have 2/3 seat majority, Barisan Nasional needs to only win 148 seats.END