- You survey high school students on their academic habits and performance, using the data to regress high school GPA (measured from 0-4) on the number of hours spent studying per week. You create a variable that takes the value of 1 if a student studies for at least five hours per week and 0 if not. You create the regression shown below and find that β1 equals 0.01, and that this difference is statistically significant. GPA=α0+β1(5 or more hours). Based on the information from the prompt, select all statements that you think are in concordance with the evidence presented here:
- More studying causes improved academic performance.
- The regression results are economically significant.
- The estimates have a bias because students could study for more than 5 hours.
- We cannot tell if this effect is causal or not.
- All of the above
‣
- Select all statements that are FALSE related to the pros and cons of expressing policy outcomes with standard deviations. More studying causes improved academic performance.
- Standard deviations allow for unit-free comparisons across different variables, but they can be difficult for non-technical audiences to interpret.
- Using standard deviation eliminates selection bias, eliminating the need to conduct randomized control trials.
- A pro of standard deviations is that it assures us of having statistical significance if the effect is at least 2 standard deviations away from the mean.
- A change in 2 standard deviations is usually considered large.
‣
- A city’s arrest rate in 2017 was 7,127 arrests per 100,000 residents, but after a policy intervention in 2018, that rate had fallen to 6,947 arrests per 100,000 residents. What is the percent decrease in arrests with which the policy was associated?
- 0.18 percent
- 2.5 percentage points
- 0.0018 percentage points
- 2.5 percent
‣
- There is a promising new white noise machine that has been shown to help increase the hours people sleep. The developers of this product want to grow sales, but an investor is concerned that people who bought this machine may also be more likely to use other sleep aids, like melatonin, also influencing their amount of sleep. What is the the equation below represent? E[Yi1∣Di=1]−E[Yi0∣Di=1]
- This represents selection bias
- This represents the naïve difference
- This represents the causal effect
‣
- A reporter is doing a piece on big tech CEOs (Microsoft, Meta, etc.) and noticed a pattern among the people she interviewed – most of them were college dropouts. Which of the following expressions represents the exercise that the reporter did?
- E(Being a CEO | College dropout=1)
- E(College Dropout | Being a CEO=1)
- E(Being a CEO | College dropout=1) - E(Being a CEO | College dropout=0)
- None of the above
‣