I am currently writing a theoretical article for school, and I am not sure I am interpreting R squared correctly. In an article, the authors write the following: "The variances accounted for by all six predictors were: women same-sex: R2 = 0.086, men same-sex: R2 = 0.110, women opposite-sex: R2 = 0.110, and men opposite-sex: R2 = 0.115."
Could this be interpreted as saying that all six predictor variables (in this study) explained only between 8,6% and 11,5% of the variance in sexual harassment among the sample?
Yes, this would be my understanding as well. It sounds like the authors ran the same regression model (with the same 6 predictors) in 4 different groups. The models (with all 6 predictors taken together) accounted for between 8.6% and 11.5% of the variation in the dependent variable. The rest was unexplained/due to other predictors not in the model and/or measurement error.
What would be of interest here also is whether the authors used "conventional" (unadjusted) R^2 or some sort of adjusted/corrected R^2. With so many predictors, adjusted/corrected R^2 would likely be more realistic (typically smaller than conventional R^2). See, for example: https://www.statisticshowto.com/probability-and-statistics/statistics-definitions/adjusted-r2/
Yes it is, it means that 8.6% of that variance is explained by that factor. Cheers.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Writing