The mainstream approach in analyzing public opinion survey data is mean-centric, focusing on the “average” rating (i.e. mean score). Mean score on the approval rating given by citizens is commonly used as a performance indicator of the government. A low score is usually assumed to lead to social instability while a high score is assumed to promote stability. However, mean score is not at all informative about 1) dispersion of rating and 2) skewness of rating, which I argue are crucial factors in understanding social instability. The aim of this paper is threefold. First, I highlight the importance of different parameters of the distribution of social attitudes in analyzing opinion data (e.g. variance, inter-quantile range, skewness and quantiles). Second, by analyzing the all waves of raw data about the approval rating of the Chief Executives of HKSAR (from 1997 – 2016, from the HKUPOP), I visualize the increasing cleavage in Hong Kong under the rule of our third Chief Executive and explain a paradox – why is the current average rating unresponsive to the politically catastrophic events. Third, I discuss the possibility of a new research agenda in using new methodological tools to reanalysis the existing data for new insights.
|Publication status||Published - Dec 2016|