Correlation Home Work Review

JohnM

TS Contributor
#2
Just a few observations:

- adjust/shrink the x-axis range on your graph so that the dots take up more of the area

- use the Excel chart feature "Add a Trend Line" so that you can see the regression line, as well as r^2 and the equation on the chart

- the question asks whether the relationship is linear, non-linear, or nothing, so in addition to the linear, so you'll need to check other types of trend lines to see if it "improves" the fit (increases r^2)

- you may want to add commentary that talks about the regression equation (i.e., for every $xxx increase in retail sales, we'll see an approx $xxx increase in online sales, etc.)

- your last comment about "causation" is incorrect --> since this is an observational study and not a controlled experiment, you cannot make any statements about cause-and-effect, only that there is "an apparent relationship or association" between the variables
 
#3
One More Question

- you may want to add commentary that talks about the regression equation (i.e., for every $xxx increase in retail sales, we'll see an approx $xxx increase in online sales, etc.)

I know how to do regression in excel but what am I looking for I dont see it
 
#5
Follow on

tell me if Im thinking right
if the regression equation computes
Intercept -30795.79304
Total Retail Sales 0.050928154 (Slope)

then the following statement could be made

Regression equation: y = -30796 + 0.0509x

So if I am thinking right for every $1,000,000 Retail sales goes up Online sales go up $50.928.15