This post walks through an exploratory k-means clustering analysis as applied to numerical data from a dot probe task. Steps to preprocess and analyze the data are discussed and the results are presented through visualizations, as well as a discussion of the results. All code used in the analysis and in the production of this blog post is available at https://github.com/kdestasio/hc_analysis. For additional resources on clustering, see the references section at the bottom of the post.

When learning to write code for research projects, it can be overwhelming to figure out how to set-up a project and the novice programmer may not yet have the experience necessary to forsee potential pitfalls of a given, seemingly inconsequential decision. This post provides a discussion of best practices1 for developing code-based projects and for writing R code in a research setting with an eye toward proactively avoiding common pitfalls.