# Effect Size, Probability & Hypothesis testing

• Effect Size, Probability & Hypothesis testing
• Effect Size
• A general term for describing the thing we are interested in finding
• Cohen's D
• An effect size that expressed difference between two means in standard deviation units
• d = (M1 -M2)/?((SD12+SD22)/2)
• How meaningful, impactful, influential is our effect
• Ranges on a numerical continuum; often has cut-offs for “small,” “medium,” and “large”
• Not affected by sample size
• Probability
• Probability distribution
• A curve describing an idealised frequency distribution of a particular variable from which you can get the probability
• Null Hypothesis Significance Testing
• an approach to evaluating our findings
• Experimental/ alternative hypothesis = effect.
• Evaluating the degree to which the data are incompatible with a model summarised by “the null hypothesis”
• alpha value
• the cut-off at which we decide whether we can reject the null hypothesis
• If p ?, then we typically cannot reject the null hypothesis
• Statistical power
• the probability that we will reject the null hypothesis when it is appropriate to do so
• affected by sample size
• Type I and Type II error
• Type I = False positive
• Type II = False negative