Abstract
The paper argues that it is necessary to teach social scientists `data analysis' rather than `statistics'. Instruction in statistical principles needs to be closely integrated with the use of appropriate software packages and secondary data sets. It is also suggested that a discussion of interval level techniques such as correlation and regression should precede the introduction of the supposedly simpler technique of cross tabulation. The merits of intensive short course module teaching are also briefly examined.
Get full access to this article
View all access options for this article.