iX6Iazv1Rfq-biostatistics Mine

Summative Quiz 2 (Continuous Data Measures) Solutions:
1. Which of the following is true about an observational prospective cohort study being
designed to study the association between cigarette smoking (smoking ≥ 1 pack of
cigarettes per day, smoking < 1 pack of cigarettes per day, or no smoking) to be assessed
on 3/1/19, and catching a cold between 3/2/19 and 3/31/19?
Answer: Subjects self-select to be in the exposure groups, so potential confounders will

need to be considered when assessing the outcome/exposure relationship.
2. Which of the following statements defines the 97.5th percentile value for a data
sample, regardless of the sample data distribution?
Answer: The value such that 2.5% of the sample observations are larger than this value,
and 97.5% of the sample observations are less than (or equal to) this value.
3. 500 randomly selected university students were selected for a study on smoking status
and stress levels. At the time of the study, students were asked to rate their stress levels
on a scale of 0 – 100 with 0 indicating no stress. Students were also asked to disclose
whether they were a non-smoker, a light smoker, or a heavy smoker.
Of the 500 students, 75 identified themselves as heavy smokers, 150 identified

themselves as light smokers, and 275 identified themselves as non-smokers. The
average stress level reported for heavy smokers was 60.4. The average stress level
reported for light smokers was 45.7. The average stress level reported by non-smokers
was 56.6. Based on these results, which of the following statements is true?
Answer: On average, heavy smokers have stress values of 3.8 greater than non-smokers
Reasoning: 60.4 – 56.6 = 3.8.
4. Which of the following sample statistics is least sensitive to the influence of outliers for
a sample continuous data measures?
Answer: The sample median (50th percentile)
Reasoning: The sample median is the “middle” value in the data set, ie: the value that is
greater than or equal to 50% of the data values, and less than 50% of the data values.
Outliers, small or large, do not affect the position of the median (or most other
percentiles) among all the data points, but will affect quantities that depend on the
actual data values and not just the order of the data point.
5. This graph shows the total payment amounts in 2011 (in thousands of US$) for
a random sample of 400 US physicians who received payments from any of 8 major
pharmaceutical companies in 2011. The vertical axis represents the percent of
physicians. (Source: Propublica “Dollars for Docs” Online Database)
Which of the following statements best tells the story of these data?
Answer: The majority of physicians received relatively “small” payments (<$25,000) and
a smaller percentage received payments greater than this majority.
6. This graph shows the payment amounts (in thousands of US$) for a random sample of
400 US physicians who received payments from any of 8 major pharmaceutical
companies in 2011. The vertical axis represents the percent of physicians. (Source:
Propublica “Dollars for Docs” Online Database)
Given the distribution of the sample data, what is the most likely distribution of the
payment data in the population of all US physicians in 2011?
Answer: right (positively) skewed
Reasoning: The distribution of values in a sample from a larger population set should
mimic (imperfectly) the distribution of values in the population. These 400 sample
values have a right-skewed distribution.
7. This graph shows the payment amounts (in thousands of US$) for a random sample of
400 US physicians who received payments from any of 8 major pharmaceutical
companies in 2011. The vertical axis represents the percent of physicians. (Source:
Propublica “Dollars for Docs” Online Database)
Which of the following boxplots corresponds to the physician payment distribution in

this sample?
Answer: Boxplot C
Reasoning: Both boxplots A and C evidence of a left skew, and B shows evidence of
symmetry in the data distribution. Only boxplot C shows a right-skewed distribution,
which is consistent with histogram. (Also, you can note that the empirical percentiles
given in the table accompanying the histogram correspond to those show in boxplot C).
8. Suppose another researcher takes a random sample of 1000 subjects from the same
population (US physicians who received payments from any of 8 major pharmaceutical
companies in 2011). You have not yet seen these data. Likely, how will the sample
standard deviation of these 1000 values (s1000) compare to the sample standard
deviation of individual physician payments in the sample of 400 (s400 =17 )
Answer: s1000 will likely be similar in value to 17 (s400), but the exact relationship
between these two sample standard deviation values cannot be predicted based on the
information given.
Reasoning: There is no systematic link between the expected change in the value of a
sample standard deviation and the size of the sample on which the standard deviation is
computed. While these estimates should be similar in value, there is no way to predict
how the two estimates will exactly compare.
9. This graph shows the payment amounts (in thousands of US$), and selected sample
summary statistics, for a random sample of 400 US physicians who received payments
from any of 8 major pharmaceutical companies in 2011. The vertical axis represents the
percent of physicians. (Source: Propublica “Dollars for Docs” Online Database)
Using these data from this sample of 400 physicians, estimate the percentage of US
physicians who received payments of greater than $10,000 in 2011. This estimated
percentage is:
Answer: 25
Reasoning: The 75th percentile estimated from these 400 observations is 10 ($10,000).
The 75th percentile is the value that is greater than or equal to 75% of the sample
observations and less than the remaining 25%. The 75th percentile estimated from this
sample is the best estimate of the corresponding 75th percentile in the entire
population of US physicians receiving payments from the 8 pharmaceutical companies in
2011
10. Which of the following is true about x̄, the sample mean?
Answer: The sample mean, x̄, is the best estimate of the underlying population mean,
based on the sample data.

iX6Iazv1Rfq-biostatistics Mine

Uploaded by

Copyright:

Available Formats

iX6Iazv1Rfq-biostatistics Mine

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

iX6Iazv1Rfq-biostatistics Mine

Uploaded by

Copyright:

Available Formats

Summative Quiz 2 (Continuous Data Measures) Solutions:

Answer: Subjects self-select to be in the exposure groups, so potential confounders will

Of the 500 students, 75 identified themselves as heavy smokers, 150 identified

Reasoning: 60.4 – 56.6 = 3.8.

Answer: The sample median (50th percentile)

Answer: right (positively) skewed

Which of the following boxplots corresponds to the physician payment distribution in

You might also like