定性数据统计分析的对偶标度

Dual Scaling for Statistics of Qualitative Data

定性数据已被广泛地应用在社会科学、自然科学的许多分枝上，对偶标度是处理定性数据的一种非常有效的方法。对原始定性数据运用方差分析方法计算变量权，根据相关比最大原则计算样品最优得分，就是使之有最大的离差，进而可以看出变量和样品的关系。对偶标度具有一个重要的性质就是对偶性，掌握这个性质在实际应用中可以节省大量的计算机运算时间。

Qualitative data have been applied in the many branches of social science and natural science. Dual scaling is very effective method that deals with qualitative data. By using variance analysis variable weight valus for qualitative data have been calculated, and then sample score has been calculated based upon maximizing correlation ratio, maximum variance dispersion. From the calculations the relation between variables and sample can be obtained. Duality is an important characteristic of dual scaling, so run time of computer is reduced in practice application by mastering this characteristic. Hence the work efficiency is enhanced.