SOAN 2120 Lecture Notes - Lecture 5: Scantron Corporation, List Of Statistical Packages, Counterargument
Document Summary
Quantitative variables: continuous, numerical (rank) ex; education (yrs of schooling) and income (in dollars) Scatterplots: play important role when exploring relationship between 2 quantitative variables. They allow us to identify problem cases, form of relationship (if its linear) and have a preliminary estimate of strength of relationship. Only when relationship is linear, it makes sense to use correlation coefficient. Correlation coefficient is a good summary of measure of strength and direction (positive or negative?) of linear relationship. Ranges from -1 (perfect negative linear relationship) to +1 (perfect positive linear relationship) and 0 means no linear relationship. Does not tell us slope (how steep line is) Outliers: something situated outside the normal body of data. If it is a part of the data, but unusual: leave it in. We have to determine if they are included or not b/c it can influence policy. We infer a causal relationship based on our understanding of the temporal relationship between variables.