Correlation & Regression Correlation does not specify which variable is the IV & which is the DV. Simply states that two variables are correlated. Hr:There is a correlation between education and income. Regression specifies an IV & DV. States that the IV influences an outcome on the DV. Hr:People with higher levels of education will earn higher levels of income.
Level of Measurement IV & DV are interval/Ratio or a scale Exception: Dummy variables Examples: Education (measured in years) and income ($) Education & TV hours Income and number of children
Education and Income Hr: 0 Ho: 0 (rho) = correlation in the population Hr:There is an association between years of education and income. Ho:There is association between years of education and income. See Scatter plot on the board X axis = years of education Y axis = income in dollars
Pearson’s r tells us 2 things: 1. The direction of the relationship between the X & Y 2. The strength of the relationship between the X & Y
Direction: Range: r can range from –1 to +1 -r = inverse (negative) relationship. As values of X increase, values of Y decrease (see example on board) +r = positive (direct) relationship. As values of X increase, values of Y increase
Strength: -1 is perfect negative correlation. +1 is perfect positive correlation. 0 indicates no correlation Rule of thumb: r =.60 (+or-) and above is a strong correlation. r =.30 (+or-) is a moderate correlation. r =.10 (+or-) is a weak correlation.
Calculating Pearson’s r See formula on the board (example height & weight in children) r = SP (sum of the products) / (divided by) square root of SSx (sum of squares x) * SSy Like t and ANOVA, r determines the extent to which variation in Y can be explained by X.
Significance of r SPSS will give the actual probability of error associated with r. Calculated Pearson’s r Pearson’s r for height & weight example =.61 Table r Alpha.05 df N-2 (two variables) = 8 –2 = 6 Critical (table) r =.7067 Calculated r must be = to or larger than.7067