I want to test the relationship between age (continues variable) and deppresion level which range from 1 to 6 (1=normal, 2=borderlin, 3=mild, 4=moderat, 5=sever, 6=extremely sever). I confused which test (Pearsons's or Spearman's ) is better in this situation?
Since depression level is probably ordinal in nature (ordered categorical), Spearman might be a better choice.
NEITHER PEARSON NOR SPEARMAN because Pearson is used only for (quantitative x quantitative); it is not the case here. Spearman rho is used only for (quantitative x ordinal), and this is also NOT the case here.
AGE is always nominal whether it is grouped or individually selected. Nominal data by defined as data ghag is used for "identification" and age is used for identification, nkt ordi al. We identify age as a person with a number of years living, not ranking people who is older or younger.
LEVEL od any kind is quantitative. A level is a series of measurable value belonged to a quantitative number set. The "verbal" description after the numerial 1, 2, 3, etc. Does not change its quantitative nature. The level used here is similar to the Likert scale. The levels 1-6 here is an e en scale.
Quantitative x Nominal = Point biserial
LINK:
https://en.m.wikipedia.org/wiki/Point-biserial_correlation_coefficient
I would like to state my disagreement with Paul Louangrath 's response. Age doesn't need to be treated as a nominal variable. And it is usually better to treat it as ordinal or interval or ratio, as the method of measurement permits.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress