When you’re lost and frustrated, statistics only make you want to run in the OPPOSITE direction! If you’re looking for the basic info to consider in your publication, here are a few starting points:
- Use your methods section to tell the reader what you did and why. How did you identify confounders? What are they? How did you select your sample? Why? Who should this ideally represent? What analytic method did you choose? What assumptions were tested?
- Everybody wants to know who is in your sample, and what the sample members look like. Any journal will inevitably be resistant to your manuscript if you don’t tell them, for example, the demographic make up (age, gender, race, education level, etc), and tell them something about your main variables of interest (describe your explanatory/independent/predictor variables, then describe your outcome/dependent/criterion variables).
- A table is meant to summarize more than one bit of information. In other words, don’t include in a table what could more easily be stated in a sentence. Capitalize on the table structure to present, for example, comparable models, comparisons for demographic measures, or multiple analyses. Bring it together for the reader!
- Don’t forget the analytic assumptions. Nothing is worse than to have someone read your statistics section only to realize that you didn’t consider (or make it evident that you did) the assumptions underlying your models. If you don’t tell the reader what you did, we assume you didn’t do it.
- Results sections are usually for the presentation of your results, not your interpretations of the body of results. The discussion section is the discussion of the results, and the connections to the current literature (among other things). Let the results show the results, ideally in an objective fashion.