What the Numbers Can (and Can't) Tell Us
A plain guide to reading a correlation honestly.
This last post is about reading results honestly: what a correlation can show, and what it cannot.
How to read a correlation
A Pearson correlation gives a number between −1 and +1. Near +1 means the two measures rise together, near −1 means one rises as the other falls, and near 0 means no clear linear relationship. A "significant" result means the pattern is unlikely to be chance alone, not that it is large or important.
What it cannot say
- It cannot say that gratitude causes duty, or the reverse.
- It cannot speak for people outside the sample.
- It depends on honest and thoughtful answers, which the researchers cannot fully control.
Even a strong result is a beginning
Work such as Ma and colleagues, linking gratitude to filial responsibility through mediating factors [Ma et al. (2026)], suggests the relationship may involve other variables. Those would be the next questions, for the team or for others.
Thank you for reading
If this series made you think about your own family, the comment section is open. And if you have ever said "retirement plan, yarn?" with a laugh and a small pang, you are exactly who we wrote this for.
Sources used in this post
Pass this on
Know someone who carries this at home? Send it their way.
Comments
What would a result need to look like to change your mind about how gratitude and duty relate?
Be kind, don’t name relatives, and please don’t give mental-health advice. Comment policy
Loading comments…