In the figures below you will see several plots for data comparing a university student population and quarterly sales for a nearby pizza place. Instructions: Investigate any data point, especially any suspicious ones, by hovering with the mouse over the point in any graph. The corresponding observation is automatically highlighted in all the other graphs. Scatter diagram (upper left): There should be a scatter of points along the regression line. Systematic departures such as points aligned along a parabola would indicate an assumption violation. • Residual versus Predictor (upper right): There should be a random scatter of points in this graph. A common violation would be for the residuals (errors) to increase as x increases producing a funnel pattern. Standardized Residuals versus Prediction (lower left): Again, there should be a random scatter of points in this graph. And again the most common violation is a funnel pattern with residuals becoming larger as predicted values increase. Also, the range of the standardized residuals (y axis) should be between about -2 and +2 for small samples, between -3 and +3 for larger samples, and almost always between -4 and +4 regardless of sample size. • Normal Quantile-Quantile Plot (lower right): If the residuals are from a normal distribution, as assumed, then the points in this graph should fall along the diagonal line. . examine the residual plot (upper-right) of the armand's pizza data. which best describes the pattern of the residuals? model form not adequate (curvilinear, etc.) nonconstant variance (funnel) good pattern (dispersion) -select- 2. examine the normal probability plot (lower-right) for the armand's pizza data. is it reasonable to conclude that the assumption that the error term has a normal probability distribution is not violated? yes no -select-