A sample t-
test for the difference between group means sufficient is
:
Here' the truth: The answer is a big" NO."
- You cannot have statistical test determine a causal
effect.
- Please read on to get an idea of what you can and can't acheive
with a t-test.
- Causal relationships are established by experimental design,
not a particular statistical test.
- You could use a correlation as your statistical test and
demonstrate that the high quality true experiment you conducted
strongly implies causation.
- You could perform a t-test as your statistic and show a
relationship in your quasi or observational study.
- But, that statistic does not, in and of itself, justify a
causal explanation.
- Causal effect just gives data with respect to whether the two
means are equivalent or essentially unique