PSY 250 Lecture Notes - Lecture 45: Type I And Type Ii Errors, Confidence Interval, Statistical Parameter
Document Summary
5 drawbacks of hypothesis testing: over-emphasis on p-values, association does not imply causation, multiple comparisons, high type ii error (low statistical power, fallacy of comparing statistical significance, over-emphasis on p-values. We end up losing part of the story. Clinically unimportant effects may be statistically significant if a study is large. Focus on effect size (how much of the behavior we are looking at is attributable to the situation) and confidence intervals. Interval around point estimates to indicate actual population value different from point estimate. Range of numbers inferred from sample with a probability of capturing the population parameter. Probability relates to the area in a normal curve. Approximately 95% of individual measurements lie within 2 standard deviations of the mean. Regardless of shape of the population distribution, sampling distribution of the mean is normally distributed (as long as the sample size is large enough) We know the probability of the true population parameter.
