In: Statistics and Probability
For 45 cities in the U.S, we found a correlation of 0.86 between internet use and
online shopping. The regression equation predicting online shopping is ŷ =15+ 0.8 *Internet use
a. Based on the correlation value, is the slope positive or negative? Why?
b. Honolulu had an internet use of 91% and online shopping of 85%. Find its predicted online
shopping based on the regression equation.
c. Find the residual for Honolulu. Interpret the result
Solution:
a) The slope of the regression equation is positive, because the value of the correlation coefficient between internet use and online shopping is 0.86. Means that as internet use increases then online shopping increases . There is strong positive correlation between internet use and online shopping.
b) The estimated regression equation for predicting online shopping is
When internet use of 91% then the observed or actual online shopping is y = 85%
To predict online shopping when internet use = 91% is
%
c) The residuals can be calculated as = observed value - estimated value
Residuals = 85 - 87.8
Residuals = -2.8%
Interpretation:
The residuals should be small and unstructured this means that the regression equation has been successful in explaining the variation of the dependent variable online shopping.