Download Example question 1

Example question 1
A study is conducted to learn about the relation between regular fish
consumption (>2 times a week) and colon polyps. In a group of 500 regular
fish consumers, there were 6 cases with colon polyps. There were 15 cases
with colon polyps in 650 people that did not consume fish regularly.
What is the proper effect measure of polyp risk as associated with fish
consumtion and how do you calculate it?
a. the odds ratio = 6/15 = 0.4;
b. the prevalence ratio = 6/15 = 0.4;
c. the odds ratio = (6/500)/(15/650) = 0.52;
d. the prevalence ratio = (6/500)/(15/650) = 0.52.
Example question 2
The authors conducted a study in the state of Iowa to determine whether
increased consumption of nitrate from drinking water and other sources was
associated with pancreatic cancer. After exclusions for insufficient data, 189
cases of pancreatic cancer were compared to 1,244 non-cases. No
association was observed between pancreatic cancer and increasing levels of
nitrates in community water supplies (American Journal of Epidemiology,
2004; 159:693 - 701).
What was the design of the study?
a. Case-control study
b. Cross-sectional study
c. Parallel intervention study
d. Prospective cohort study
Example question 3
Suppose that 93 children are selected at random from the general population.
Their C-reactive protein (mmol/L) are evaluated. The mean C-reactive protein
was 3.2 mmol/L with standard deviation 6.9 mmol/l. The 95% confidence
interval for C-reactive protein in the population is therefore calculated as
a. 3.2  1.96 
b. 3.2  1.96 
6.9
93
6.9 2
93
c. 3.2  1.96  6.9
d. None of these answers.
Example question 4
Which of the following statements is true?
I. Non-parametric tests and log-transformations can be used for non-normally
distributed variables.
II. Three ways of visual inspection of whether a variable is Normally
distributed are the histogram, the scatterplot and the Q-Q-plot
a.
b.
c.
d.
I is true, II is false;
I is false II is true;
I and II are true;
I and II are false.
Output
Researchers wanted to compare BMI at baseline between subjects involved in
an intensive 6-week weight reduction programme with subjects that
participated in a one-day course on healthy eating.
These were the results from their SPSS analysis:
Group Statistics
BMI
diet
6-week programma
N
One-day course
Mean
Std. Deviation
Std. Error
Mean
20
34.3830
2.25188
.50353
21
36.2806
1.34923
.29443
Independent Samples Test
Levene's Test for
Equality of
Variances
F
BMI
Equal
variances
assumed
Equal
variances
not
assumed
7.473
Sig.
.009
t-test for Equality of Means
t
df
Sig. (2tailed)
Mean
Differenc
e
95% Confidence
Interval of the
Difference
Std. Error
Difference
Lower
Upper
-3.292
39
.002
-1.89762
.57645
-3.06360
-.73164
-3.253
30.793
.003
-1.89762
.58330
-3.08758
-.70766
Example question 5
What is the result of the statistical comparison?
a. The p-value for the difference between the two groups is p=0.009;
b. The p-value for the difference between the two groups is p=0.002;
c. The p-value for the difference between the two groups is p=0.003;
d. The p-value for the difference between the two groups is p=0.504.
Example question 6
What conclusion can you draw from this SPSS output?
a. Subjects in the intensive programme have a 1.90 kg/m2 significantly lower
BMI at baseline compared to subjects in the one-day course, with a 95%
confidence interval of -3.06 to -0.73;
b. Subjects in the intensive programme have a 1.90 kg/m2 significantly lower
BMI at baseline compared to subjects in the one-day course, with a 95%
confidence interval of -3.09 to -0.71;
c. Subjects in the intensive programme have a 1.90 kg/m2 non-significantly
lower BMI at baseline compared to subjects in the one-day course, with a
95% confidence interval of -3.06 to -0.73;
d. Subjects in the intensive programme have a 1.90 kg/m2 non-significantly
lower BMI at baseline compared to subjects in the one-day course, with a
95% confidence interval of -3.09 to -0.71.