5

Cohen's d is a measure of effect size calculated as:

$d = (x_1-x_2) / \sigma_{\text{pooled}}$

where $x_1$ is the mean of one group, $x_2$ is the mean of a second group, and $\sigma_{\text{pooled}}$ is the pooled standard deviation.

Let us say that one treatment has an effect size of 0.6 and another treatment has an affect size of 1.2. How can I tell whether this difference is statistically significant? I.e. does one treatment provide a significantly different outcome from another?

EDIT: The samples are not independent; each sample consists of the same 7 participants.

I've briefly searched around but didn't find an answer for this question.

Andy
  • 18,070
  • 20
  • 77
  • 100
Angry_at_Linux
  • 101
  • 1
  • 1
  • 3
  • 3
    Part of the reason effect sizes were suggested in the first place is to get away from the practice of testing every single statistic for "significance". See [Cohen (1994): The Earth is round (p < .05)](http://scholar.google.com.au/scholar?cluster=10018019703886757098&hl=en&as_sdt=0,5) – Marius Mar 21 '12 at 01:10
  • 7 people in the first group and 7 people in the second group or 7 people altogether? – Wolfgang Apr 04 '14 at 09:32

0 Answers0