Group Quiz 2 PDF

Preview

Summary

Steinbuch Group Quiz 2 ...

Description

QM222 Team Quiz 2 Instructions: This quiz is closed notes and closed book. Please first complete the quiz individually and in silence. The professor will announce when to begin the team quiz portion.

1. You would like to estimate the proportion of the US population who think the legal drinking age should be lowered to 18. Which of the following is most likely to produce selection bias in a survey? a. Using a random sample of 5,000 people from the US population. b. Asking a follow-up survey question about the appropriate penalties for driving under the influence of alcohol. c. Conducting interviews by telephone or text message rather than in person. d. Using a random sample of 20,000 college students. 2. Which of the following is the best example of selection bias? a. More than two-thirds of studies of antidepressants given to depressed children found the medications were no better than sugar pills, but companies published only the studies that found a significantly positive effect. As a result, published evidence convinced MDs to give antidepressants to depressed children. b. George W. Bush promoted a tax cut plan that would give 92 million Americans an average reduction of over $1000. But he was later criticized because the median tax cut was less than $100. c. One year, New York had 329 homicides, whereas Boston had only 40. Therefore Boston was a safer city to live in that year d. In 2008, Baylor University offered financial rewards to admitted students to retake the SAT in hopes of increasing its average score and thus improving the school’s US News ranking. //only published studies got read 3. Which of the following statements make a conclusion not supported by the facts. (Hint: The incorrect statement uses a misleading baseline to support its conclusion.) a. At iCampus Inc., employees with flexible hours account for just 10% of the total population of workers, and yet, they make up 80% of workers reporting “very high” levels of job satisfaction. Therefore, flexible work hours are associated with higher job satisfaction. b. At Imaginary State University, 60% of all hazing incidents involved fraternity or sorority members. Therefore, hazing is more common in the Greek system than among unaffiliated dorm residents. //what percent of the school is Greek? c. At iCampus Inc., workers with an MBA earn 30% higher salaries than workers without an MBA on average. Therefore, at iCampus, Inc., MBAs are associated with higher earnings.

d. At Imaginary State University, 95% of work-study participants were employed within 3 months of graduation, compared to 80% of all other students. Therefore, work-study participants were more likely to be employed after graduation. //EXPLAIN

You run a regression analysis to answer the question of whether high school athletes receive higher test scores on average. ATHLETE is a dummy variable that equals 1 if the student is an athlete. TEST SCORE reports student performance on a standardized test taken by all students. You get this regression: PREDICTED TEST SCORE = 70 + 15*ATHLETE (10) (5) (Standard errors in parenthesis) Use this regression to answer questions 4 – 6. 4. Which of the following statements is true? a. Non-athletes score 70 points higher than athletes on average. b. Students with higher scores are 15% more likely to be athletes. c. Athletes score 15 points higher than non-athletes on average. d. Athletes received an average score of 15. 5. Which of the following statements is false?

a. The coefficient on Athlete is statistically significant. b. We cannot reject (with 95% certainty) that athletes’ test scores are just 6 points higher than for non-athletes. c. We can reject (with 95% certainty) that athletes’ average test score is 20 points higher than non-athletes’ average score. d. We can reject (with 95% certainty) that the average test score of non-athletes is 95. // how do you find confidence interval 6. Suppose you instead had run the regression with a dummy variable for NON-ATHLETE instead, which equals 1 if the student is not an athlete. TEST SCORE = b0 + b1 *NON-ATHLETE What would the intercept b0 be in this new regression? a. 85 b. 15 c. 55 d. 70 //70 + 15, because athletes earn 15 pts higher than non-athletes on average 7. After a local 5K race, you ask all runners about their race times (in minutes) and the number of hours they spent training for this event over the past 3 months. You run a regression and get the following results: RACE TIME = 45.42 – 1.22*(TRAINING TIME) Which of the following is the best interpretation of this regression? a. Each additional minute of race time was associated with 1.22 hours less training time on average. b. Those who spent no time training will run the race in 44.20 minutes on average. c. Each additional hour of training was associated with 1.22 minute shorter race time on average. d. Every runner who trained for 10 hours finished the race in exactly 33.22 minutes.

8. You collect data on the annual sales of organic foods, and the autism prevalence per 1000 children.

Autism rate per 1000 children

Organic Food and Autism 16 15 14 13 12 11 10 9 8 7 6 5

10

15

20

25

30

Organic food sales (in $ billions)

Which of the following numbers is most likely to be the correlation coefficient for these two data series? a. -0.010 b. 0.956 c. 1.115 d. -0.812 9. Which of the following is NOT true about the correlation coefficient: a. The correlation coefficient is always between -1 and 1. b. A correlation coefficient of 0 means there is no linear relationship between X and Y. c. A correlation of 1 means that all dots lie exactly on an upward sloping line. d. The steeper the slope of the best fit line, the larger the absolute value of the correlation coefficient. //correlation coefficient is always between 1 and -1 , 1 is positive -1 is negative Zero is no relationship.

10. You run a regression and estimate that b1 = 5 with a standard error of 10. A textbook argues that the true estimate is 8. Calculate a t-statistic to test the null hypothesis that the textbook is correct. a. b. c. d.

-0.15 -0.3 0.5 0.3

//the null hypothesis is that there is no positive relationship! Saying the slope is 0 Estimated – true / sd of estimate = z  (5-8)/10...