Quiz 1 Solutions

ECON 303 Quiz 1

Name: ____________________________________________

You have 25 minutes to complete this quiz. Please answer the following questions in the space provided.

(1) In recent weeks, there have been several contradictory polls of physicians’ opinions of health care reform. For instance, the New England Journal of Medicine published two polls of physicians. One poll found that: 78% of physicians believe that addressing societal health policy issues is part of their professional obligation as a physician 73% of physicians believe it is their professional obligation to care for the uninsured or underinsured, 67% favor limiting reimbursement for expensive drugs and procedures if that would help expand access to basic health care for those currently lacking such care. This survey was mailed to 2000 physicians under the age of 65 randomly selected from the AMA Masterfile (a list of all practicing physicians) in May of 2009 and responses were collected throughout the summer. Those not responding to the first mailing received two subsequent mailings. All of those surveyed received a book, and those who responded received $25. Of those selected, 61 could not be contacted, leaving 1939 participants. Of these, 991 (or 51%) returned completed surveys. The other survey asked physicians to indicate which option they would most strongly support Provide people under the age of 65 the choice of enrolling in a new public health insurance plan (like Medicare) or in private plans. [62.9% of respondents chose this option] Provide people with tax credits or low-income subsidies to buy private insurance coverage (without creating a public plan option). [27.3% of respondents chose this option] Elimination private insurance and cover everyone in a single public plan like Medicare. [9.6% of respondents chose this option] This survey was based on a random sample of 6000 physicians drawn in April from the AMA Masterfile. Physicians from US territories and physicians in training were excluded from the sample, leaving 5157 participants. Physicians were placed into four groups primary care, medical specialties, surgical specialties, and other

ECON 303 Quiz 1

specialties, and equal numbers of individuals were sampled from each of the groups. The mailed survey was sent in three waves. The first wave included a $2 bill and was sent in late June and responses were collected throughout the summer. After each wave physicians were called and asked to complete the survey and were given the option of returning the survey by fax or email. 2130 physicians completed the survey. Respondents were slightly less likely to be women and were slightly (1 year) older than non-respondents. Finally, Investors Business Daily conducted a poll of physicians and found that: 65% of respondents opposed the proposed health care plan 45% of respondents indicated they would consider leaving their practice or taking an early retirement if reform passes The questionnaires were sent out Aug. 28 to 25,600 doctors nationwide. The sample was purchased from a list broker, Lake Group Media of Rye, N.Y. One hundred of those responding were retired, and their answers were not included in the final results. At the time of reporting (earlier this week), 1,376 participating physicians had responded.

Each of these surveys purports to describe the distribution of physician opinion of health care reform. Yet, at first glance, the results seem inconsistent. Please briefly describe what might explain this disparity. Of the different results, which do you think are most credible. Why?

The first two surveys are far more credible than the third for several reasons. First, he first two surveys were drawn from random samples of the AMA masterfile. As such, the population surveyed was a random sample of the full population of physicians. The third survey was drawn from a purchased list which, at least in the provided documentation, does not claim to be representative of the population of physicians. Simply drawing a random sample does not guarantee that the responses actually represent the population. It is possible that certain types of people are more likely to respond to the survey. The response rates for the first two surveys are relatively high. The researchers put a lot of effort into getting people to respond. That reduces (but does not eliminate) the possibility of bias. To further rule out response bias, we would want to see that the distribution of observables are similar to the distribution in the population overall. The IBD survey, though, has an abysmal response rate. At the time these data were presented, respondents had only had a few weeks to respond. As such, only the results reflect only the most motivated respondents – hardly a representative group. Finally, the questions in each of the surveys are vastly different. The first survey asks very vague questions that are easy to agree with. The third survey asks questions that obviously frame reform in a negative way (and are likely to generate

ECON 303 Quiz 1

negative responses). Even the second survey asks somewhat misleading questions. The three choices presented are not mutually exclusive. Given that tax credits and subsidies do not replace a public option, the tax credits and subsidies should have been included in the first response option to more clearly frame what the researchers were clearly interested in – physician support of the public option. Overall, though, the second survey is likely the most credible for the reasons outlined above.

(2) Many people argue that the government should support greater investments in early childhood education (e.g., Pre-Kindergarten). Proponents argue that participation in early childhood education programs improves long-run educational outcomes (e.g., reduced drop-out rates) and lowers adolescent crime. Please describe the data you would like to have if you were going to test the hypotheses lurking in the statement above. What is an example of simple evidence you could produce that would support this position? What would be the ideal evidence to support this position?

To test the hypotheses in the statement, at a minimum, we need data that matches student educational outcomes to student attendance in a pre-K program. If students who attend pre-K have different mean outcomes than students who did not attend pre-K, that would be suggestive, simple evidence for pre-K effects. Ideally, though, we would identify a large, representative population and randomly assign some students to attend pre-K while others do not. Then, we would follow these students over time and compare their outcomes. Any differences would be attributable to attendance in the pre-K program.

(3) Recall the course evaluation and beauty data we discussed in class on Wednesday. In the data we examined, the average course evaluation score for men was 4.07 and the average course evaluation score for women was 3.90. The difference between these means is 0.17 and the standard error for this difference is 0.05. Can we assert that the difference in the mean course evaluation for men and women is statistically significant at the 5% level? Yes. The standard error for the difference in means is less than half the coefficient size (suggesting a t-stat>1.96).

ECON 303 Quiz 1