I have u survey where respondents answer with I use/I don't use on 30 questions (30 dependent variables). One independent variable is age (16-74). I have a large sample N=1138 (unequal groups).

The question is whether Point-biserial correlation is the right test to find a correlation between each pair of variables (scale-dichotomous). As stated "point-biserial correlation, is used to determine the strength of a linear relationship between one continuous variable and one nominal variable with two categories", bus scale variable is usually dependent.

Therefore answers like this case, where gender is dichotomous independent sound OK: "There was a statistically significant correlation between gender and engagement score, with males more engaged with the advertisement than females (5.56 ± .29 versus 5.30 ± .39)."

In my case, dichotomous dependent (usage) does sound weird: "Results show a statistically significant correlation between age and laptop use, with participants using laptop computer being at younger average age than non-users (40.55 ± 14.14 versus 45.13 ± 15.15). Laptop usage accounted for 2.1% of the variability in age."

Thanks for the suggestion.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.