Second, my analytical skills, as well as general infrastructure in terms of analytical possibilities have expanded exponentially. Since that first multivariate regression, I have learned how to use multilevel and growth models to investigate spatial and temporal variations, made use of latent classes spatial lags, as well as delved into more south africa rcs data exotic subjects such as social networks and sequence analysis. In terms of infrastructure, I have migrated to software such as Stata and Mplus, and want to embark on a steep learning curve into R. Statistical analysis software has become much faster, datasets much larger, and access more open. of possibilities to do analysis, so that it becomes more and more important to find out which questions really matter to you as a researcher, and society at large.
Crucial data skills specific to sociology in my opinion can be narrowed down to conceptual and heterogeneity issues:
In contrast to qualitative scholars, quantitative social scientists often conduct secondary analysis: they try to answer questions for which the questionnaire was not originally developed. This means it is not always straightforward to conceptually map data onto theory. Concepts like social position, social capital and health can be measured in different ways, and this affects how strong relationships examined will be. How well different pieces of information fit together can be examined both theoretically and statistically, making use of (confirmatory) factor analysis, a technique I have been teaching for a number of years in short courses at the Cathie Marsh Institute. Thinking through how we measure things in surveys as such is a key skill for a social scientist.