PSY30100 Lecture Notes - Lecture 8: Null Hypothesis, Sampling Distribution, Statistical Hypothesis Testing
Document Summary
A hypothesis test is a statistical method that uses sample data to evaluate the integrity of a hypothesis about the population. General goal to show that results obtained in research are not caused by sampling error i. e. chance. If the sample mean is noticeably different from the population mean, then the difference cannot be explained by just sampling error. Hence, the independent variable manipulated in the study had a measurable effect on the sample. If the difference is not significant enough, the default cause for the difference in sample statistics and population parameters is always sampling error. In order to model the default difference that is expected, we use standard error from the sampling distribution of sample means. If the difference is greater than expected (> m) it is attributed to other causes. Step 1 state the hypotheses for the population. Null hypothesis (h0) difference due to sampling error.
