Hi,
I'm not sure if I got the question right, but the p-value only tells you if an effect is statistically significant (i.e., the assumption of the nullhypothesis that the effect appeared randomly is sufficiently unlikely). However, effect sizes tell you how large an effect is, which is important additional information. You might find an effect that is significant, but its size is so small that it's not meaningful in terms of content. This is especially important in the case of very large sample sizes (or even sample sizes that are larger than those calculated in an a priori sample size calculation), because then also very small effects that you are not interested in can become significant. As the sample size increases, even smaller effects become statistically significant, because the standard error decreases.
Best,
Brian
Hi,
I'm not sure if I got the question right, but the p-value only tells you if an effect is statistically significant (i.e., the assumption of the nullhypothesis that the effect appeared randomly is sufficiently unlikely). However, effect sizes tell you how large an effect is, which is important additional information. You might find an effect that is significant, but its size is so small that it's not meaningful in terms of content. This is especially important in the case of very large sample sizes (or even sample sizes that are larger than those calculated in an a priori sample size calculation), because then also very small effects that you are not interested in can become significant. As the sample size increases, even smaller effects become statistically significant, because the standard error decreases.
Best,
Brian
The kind of statistical techniques yyou use will give you the effect size aside the P value.
Correlation will give you the levels of relationship along side the p value
Regression will give you the contribution of the IV (Adjusted R) along the P value and even the equation line to consider possible effect size if you want to forecast.
The p value eill give you insight to what the effect size will be in terms of its level and grade.
Best of luck!!
In that of correlation, the coefficients will give you all you red fo r an effect size
- Badges
- Science topic
- Similar topics
- Cognitive Science
- Thinking