Alexis Bushnell

Asking Fedi cos my tutor doesn't know the answer to this:

It surprises me that a correlation coefficient of +/-0.5 or more is considered strong correlation in psychology. That seems like an awfully low bar for "strong" correlation. I would have assumed that +/-0.7 or more would be considered strong and +/-0.4 to +/-0.6 moderate.

Is there some reason that 0.5 was chosen as the bottom end of "strong"?

#Psychology #StudentsInWales #AskFedi

Jan 18 at 10:40amWeb
Show threadrojunJan 18
@alexisbushnell Sounds like something I'd liked to dive in to figure out. My guess is that it has to do with sample sizes and variance. I also found something about effect size statistics in case that results complementary https://www.theanalysisfactor.com/effect-size/
A Comparison of Effect Size Statistics - The Analysis Factor

You have surely heard that p-values don't measure the size of an effect. You also need to report effect size statistics. What are they?

The Analysis Factor
Show threadπŸ’‘πš‚π—†π–Ίπ—‹π—π—†π–Ίπ—‡ π™°π—‰π—‰π—ŒπŸ“±Jan 19
@alexisbushnell
Sounds more like a Psychology person has forgotten the definition from when they were taught it in Statistics and made up their own definition. If your tutor is a Maths tutor then they should know better. I can tutor Maths online
Show threadAlexis BushnellJan 19

@SmartmanApps thank you so much! He's a psychology tutor, not a maths one.

I'll send this over to him to clarify further.