The fact is that life may not fit in a straight-line formula. The 0.88 value here confirms that the prediction model is pretty good but not perfect.Īdditional Details: Regression models try to force actual results into a straight-line formula. It ranges from 0 to 1, where 1 is a perfect match, and 0 is a horrible match. I only somewhat paid attention in statistics class, but I know that a key statistical indicator is the R-squared value. Excel performs the regression and provides statistics.Choose five rows and several columns before entering the formula.Be sure to enter the function in a five-row range. To get the additional statistics that LINEST can return to show how well the results match reality, add a fourth argument: TRUE.The predictions in D11 and D12 are off by $20 each-an error of 10%. The prediction in D6 is right on the mark. Enter a prediction formula in column D to see how well the regression calculation describes sales.
Strategy: You need to do a multiple regression. I set up the table below, which shows each day's sales, temperature, and whether it rained.īased on the data I've collected, how can I determine the relationship between sales, temperature, and rainfall? On rainy or cool days, fewer people buy gelato. As I analyzed sales, I began to feel that temperature and rain might be two important determining factors in how much money I make.
After 10 days of sales, I discovered that each day, I would either make a lot of money or nearly go broke.
0 kommentar(er)
