Statistics > Correlation and Regression > Linear Two Variables Statistics > Correlation and Regression > Linear Two Variables This utility performs computations that determine the correlation and regression between two variables. To use this utility, you must provide the paired x (independent/predictor) and y (dependent/response) values in separate columns. You must also provide the significance level for the hypothesis test for correlation. For correlation, the following values are computed: Linear correlation coefficient: Test statistic: Critical values , where t is the inverse probability distribution of (1 - significance / 2) in the Student's t distribution with n - 2 degress of freedom p-Value of the test statistic r For regression equation Y = b0 + b1 x, the y-intercept b0 and the slope b1 are calculated as follows: The null hypothesis H0 for a hypothesis test for linear correlation is that there is no linear correlation (ρ = 0), and the alternative hypothesis H1 is that there is a linear correlation (ρ ≠ 0). For variation, the following values are computed ( is the mean of the y variable values, is the y value predicted by the regression equation): Explained variation: Unexplained variation: Total variation: Coefficient of determination: Standard error of estimate: Dialog Inputs In the Independent/dependent variable series list, select the columns containing data values for the input variables. The x variable is the independent variable, and the y variable is the dependent variable. Click Clear Input List to clear the input columns list. Significance level: Enter the significance level of the hypothesis test for correlation. Select the Show a scatterplot for all pairs of data values check box to display a scatter plot showing data values in each pair of input variables. Use the provided options to customize the scatterplot.