PS102 Lecture Notes - Lecture 3: Statistical Significance, Statistical Inference, Selection Bias
Document Summary
Involves measuring and determining the relation between 2 variables; (co-relation- relation of 2 things) Strength of a correlation: higher the number, stronger the correlation, the positive or negative reflects direction not strength, 0-1= positive. Why correlation studies: you can have access to 2 pieces of info but may not be able to manipulate the variables: asking pregnant woman to drink. Correlation is not causation: research has found a strong correlation between stress and clinical depression. However, this correlation does not tell us whether stress causes depression, depression causes stressful events, or other factors, such as poverty, produce etc. The experimental design: control group- not exposed to independent variable- measure change in dependent variable, experimental group- exposed to independent variable- measure change in dependent variables. Experimental research methods: examines how one variable causes another to change, hypothesis, prediction, cause-effect relationship, manipulation. Pros and cons of experimental methods: pro: can establish cause and effect, limitations to overcome, selection bias.