As the first step in the decommissioning of sasCommunity.org the site has been converted to read-only mode.


Here are some tips for How to share your SAS knowledge with your professional network.


Eliminating Response Style Segments in Survey Data via Double Standardization Before Clustering

From sasCommunity
Jump to: navigation, search
Murali Krishna Pagolu and Goutam Chakraborty
Oklahoma State University, Stillwater, OK

Abstract

Segmentation is the process of dividing a market into groups so that members within the groups are very similar with respect to their needs, preferences, and behaviors but members between groups are very dissimilar. Marketers often use clustering to find segments of respondents in data collected via surveys. However, such data often exhibits response styles of respondents. For example, if some respondents use only the extreme ends of scales for answering questions in a survey, the clustering method will identify that group as a unique segment, which cannot be used for segmentation.

In this paper, we first discuss the different data transformation methods that are commonly used before clustering. We then apply these different transformations to survey data collected from 959 customers of a business-to-business company. Both hierarchical and k-means clustering are then applied to the transformed data. Our results show that double-standardization performs better than other transformations in eliminating groups that identify response styles. We show how double-standardization can be achieved on any data using SAS® programs and SAS® macros.

Online Materials