I Finally Got Published in Psychological Methods!
It feels almost like a miracle, but I finally succeeded in publishing a paper in Psychological Methods—one of the top journals in psychological methodology.
Why a miracle? If you’ve ever submitted papers for publication, you might relate. My work is typically rejected multiple times before it gets accepted somewhere. I always aim for the highest-tier journals first, gradually working my way down. This process can be time-consuming and, at times, frustrating. However, this time, I managed to get into one of the best journals.
What Is the Paper About?
The core argument of my paper is straightforward: Pearson correlations should never be used for studying dimensionality.
For example, if you conduct a factor analysis using Pearson correlations and then interpret the factor loadings to determine whether your data has five or six or ten dimensions, you’re likely overestimating the number of dimensions. Pearson correlations tend to yield an upper bound rather than an accurate estimate of dimensionality.
Why does this happen? Simply put, Pearson correlations are designed for making predictions between variables—not for assessing dimensionality. This distinction has important practical implications: if you analyze dimensionality using a better-suited measure, such as the H-index, you will likely find fewer dimensions than you previously assumed. In fact, I can mathematically guarantee that you will never find more dimensions using the H-index than using Pearson correlations.
Want to Read More?
If you’re interested in the full details, you can find the official publication here (APA subscription required):
https://doi.org/10.1037/met0000736
For those without a subscription, I have made my personal copy available here:
Link
Additionally, I was invited by the APA to create a Kudos page summarizing my research for a general audience. You can check it out here:
https://www.growkudos.com/publications/10.1037%25252Fmet0000736
Let me know your thoughts, and if you decide to test your own questionnaire or model with the H-index, I’d love to hear about your findings!