Question

Please answer the following two questioins.

Explain how to determine the effect size of an outcome based on the limits stated for a confidence interval.

How does estimation differ from hypothesis testing in terms of teh decisions researchers make?

Answer 1

(a)

The Effct Size (d) of an outcome based on the limits stated for a Confidence Interval:

Cohen's d is given by:

,

where

M₁ = Mean of Sample 1

M₂ = Mean of Sample 2

SD_pooled is given by simpller equation given by Cohen as follows:

(b)

Interpretation of Effect Size d:

When a difference is concluded by the researchers based on hypothesis testing as statistically significant, it simply indicates that there is a significant difference. Instead of reporting the difference in terms of , such as number of points earned on a test, all Effect Sizes are calculated on a common scale. This standardisation of Effect Size allows the researchers to compare the effectiveness of different programs on the same outcome.

The guidelines devoloped by Cohen are as follows:

If d < 0.1, Trivial effect

If 0.1 < d < 0.3 : Small effect

If 0.3 < d < 0.5: Modrate effect

If t > 0.5: Large difference effect