You could use the Cochran-Mantel-Henzel statistic, i.e. chis-square test of the relationship between 'whether or not likely to depend on video games' & interval rartings (low, medium low, medium, medium high, high) controlled for the 5 traits. See the R script in attachment

The scoring from low to high is actually ordinal not interval, but I suspect that these are multi-item scales, so they would yield the equivalent of interval-level scores.

To select between regression and MANOVA, you need to decide if you are treating one or the other of your key concepts as a dependent variable. If you do want to do regression with a dichotomous dependent variable, then you need to do logistic regression.

Binuri Ranathunga appears to be describing the so-called Big 5 Personality Test. IIRC, the most common version has 44 items, as shown here:

• https://fetzer.org/sites/default/files/images/stories/pdf/selfmeasures/Personality-BigFiveInventory.pdf

Although it does not say so in the scoring instructions on that page, I think the 5 scores are obtained by computing means of the relevant items (after reverse-coding where it is required).

Finally, I suspect that the 5 personality traits are explanatory variables, and that video game dependency is the (dichotomous) outcome variable. If so, then binary logistic regression is indeed one way to go. However, with 5 explanatory variables, by Frank Harrell's 20:1 rule of thumb (for reducing the likelihood of overfitting), you would need to have at least 100 people in each of the two dependency groups. More info about that can be found here:

• https://discourse.datamethods.org/t/author-checklist/3407
• https://discourse.datamethods.org/t/author-checklist/3407#arrow_rightoverfitting-and-lack-of-model-validation-26

HTH.

PS- By Babyak's somewhat less onerous rule of thumb (15 events-per-variable rather than 20), you would need at least 75 people in each of the dependency groups.

• https://people.duke.edu/~mababyak/papers/babyakregression.pdf