Means Polish of the Emotional Responses to Candidates Data
In summary, exploratory data analysis is both a philosophy of and a strategy for the investigation of samples of data. The techniques emphasize resistance and robustness--insensitivity to outliers and to contamination of data, and indifference to the nature of the distribution from which the sample observations originate. Our brief and incomplete overview of the techniques has focused on those that are especially suited to the types of data that psychologists study. We have sought both to convey the potential value of these procedures for illuminating the structure of a set of data and to provide enough operational information to encourage their use.
As instructors of graduate-level courses in statistics, we have found that our students appreciate the techniques of EDA in principle but often fail to put them into practice. The major obstacle seems to be the time and effort required to carry out these procedures by hand. On these grounds, there is cause for optimism: EDA techniques are appearing with increasing frequency in standard statistical packages, making them easy and painless to use. (Indeed, stem-and-leaf diagrams and boxplots are now available in both mainframe--SAS, SPSS, BMDP--and PC packages--Systat, Datadesk; additional techniques are available in selected PC packages.) We hope this chapter contributes to the application of these procedures to psychological data.
Abelson, R. P., Kinder, D. R., Peters, M. D., & Fiske, S. T. ( 1982). "Affective and semantic components in political person perception". Journal of Personality and Social Psychology, 42, 619-630.