## 28 Apr Complete the table, accounting for regression to the mean. Do you understand why adding 10% to

Complete the table, accounting for regression to the mean. Do you understand why adding 10% to each store’s sales is __not__ the correct forecast? If not, re-read chapter 17. This site can also help. Repeat this step until you arrive at the correct answers, and you can explain why they are correct

Part 1

1. Recruit five friends, family members, or acquaintances.

2. Ask them to complete the table, using the instructions provided above the table only (in other words, do not provide any other guidance, instructions, or hints).

3. Ask them to explain how they arrived at their answers. In other words, ask them for a commentary or a play-by-play of how they went from “read the instructions” to “filled out numbers in the table.”

4. Compile their data into one single spreadsheet, and include their responses (how they arrived at their answers). This spreadsheet will be submitted (see Part II, question b, below). See template here.

5. Interpret each respondent’s data. Make a new column and enter into this column a dichotomous categorical variable labeled *Regression*. A value of ‘0’ indicates *respondent did **not** consider regression to the mean*. A value of ‘1’ indicates *respondent **did** consider regression to the mean*. Tally the percentage that considered regression to the mean and hold on to this value (you’ll be reporting it later

Part 2

a. A definition of regression to the mean.

b. A copy of the spreadsheet containing your complete results for all five participants (data + responses). You should copy and paste the spreadsheet directly into your .doc or .pdf file.

c. The percentage value you derived in question #5 from Part I. From this value, extrapolate and report a conclusion about whether regression to the mean tends to be ignored or not when the average consumer predicts future performance from past behavior.

d. In this class, we have assumed that System 1 does not by default consider regression to the mean (i.e., we assume that if regression to the mean *is* considered, then System 2 must have jumped in). Using respondents’ verbal reports as evidence, discuss the merits of this assumption. For example, did anybody’s reports show evidence of System 2 overriding System 1? Did those respondents who failed to consider regression to the mean show evidence of System 2 thinking? Did those respondents who considered regression to the mean show evidence of arriving at these answers by default? Be sure to clearly define System 1 and System 2 in your discussion. We are looking for a reasoned and thoughtful response here; there are no right or wrong answers (except for the definitions of System 1 and System 2).