Abstract:
The objectives of this research were (1) to develop the methods of forecasting election results by applying Pavias method, (2) to compare the deviation of the forecasting results of the election among Pavia's method, Pavia's method application and the actual election results and (3) to forecast the election results by applying Pavia's method which used data of the election of the House of Representatives in B.E. 2562 (2019) in Bangkok. The data used in the research consisted of 1) secondary data from the Suan Dusit Poll in opinion polls about the elections of House of Representatives in B.E. 2554 (2011), the total of 19,130 samples and 2) primary data from opinion polls about the elections of House of Representatives in B.E. 2562 (2019) in Bangkok, the total of 3,600 samples collected by the researcher. The research results were as follows: 1. Development of methods of forecasting election results by applying Pavias method 2. The result of comparison the deviation of the forecasting results of the election between Pavia's method and Pavia's method application with the actual election results found that the error of Pavia's method compared with the actual election results was percentages of 21.21 and the error of Pavia's method application compared with the actual election results was percentages of 3.03. These represented that Pavias method application which developed by the researcher was able to predict the forecasting election results more accurately than Pavia's method and also reduced the error of forecasting election results. 3. The forecasting election results by applying Pavia's method in House of Representatives elections in B.E. 2562 (2019) found that this method can actually forecast the election results by percentages of 82.28. When comparing the error of the forecasting results from Pavia's method application to the actual election results, there was percentages of 17.72. Therefore, election forecasting by Pavia's method application can predict the election results more accurately and less error in the forecasting results