Statistical Problems on Epidemiology

This article covers statistical problems on epidemiology including internal validity, research bias, confidence level, case-control design, cross-sectional study, and more. It also includes a sample ANOVA table and a study on the effects of chronic smoking on LV MPI after acute MI in successfully revascularized patients with primary percutaneous coronary intervention (PCI).

Statistical Problems on Epidemiology
Student Name: Student ID:
Unit Name: Unit ID:
Date Due: Professor Name:

Answer 1: (F)
Reason: Internal validity refers to the success of an experiment. It also checks if there is more
than one possible variable acting at the same time.
Answer 2: (T)
Reason: Research bias, also called experimenter bias, is a process where the scientists
performing the research influence the results, in order to portray a certain outcome. Some bias in
research arises from experimental error, and failure to take into accounts all of the
possible variables. Other bias arises when researchers select subjects that are more likely to
generate the desired results, a reversal of the normal processes governing science.
Answer 3: (T)
As the confidence level increases, the width of the confidence interval also increases. This
increases the precision of the experiment. However, the accuracy decreases.
Answer 4: (T)
In the case control, design initiates with part of the results, that are cases and bases or controls or
referents (coincidental or not), and try to study what was the exposure.
Answer 5: (T)
The result rate is usually lower than the prevalence of exposure; Cohort studies generally require
larger samples to have the same potency as a case-control study.

Answer 6: (T)
The design of the case-control study is frequently used in the learning of rare diseases or as a
prelude study where little is known about the alliance between the risk factor and the disease of
interest.
Answer 7: (T)
Cross-sectional study investigates the information from a population or a representative subset,
at a specific point in time.
Answer 8: (T)
In the AB / BA crossover study, the sample group is divided into two subgroups. The first group
receives first treatment "A" and then treatment "B". The second group receives treatment "B"
first and then treatment "A". Here each patient serves as his own control
Answer 9: (T)
As each patient fulfills requirement of his or her own control in a cross-over study, variation in
response between patients is eliminated when comparing treatments.
Answer 10: (F)
Cross-design would not be effective in comparison with treatments, because the effects of
training in the cross design result in a substantial loss of efficiency in estimating the relative
effect of treatment (Dutra & Glantz, 2014).

Answer 11: (F)
The sampling error decreases as the sample size increases, although it depends on the sampling
method and the characteristics of the population.
Answer 12: (T)
Cross-sectional studies are used to establish prevalence. They are comparatively rapid and easy,
but do not allow the distinction between the cause and the effect of the disease.
Answer 13:
a. Response variable: Anxiety levels
Explanatory variables: Music type and Shyness
b.
i. At 5% level of significance, three music types were statistically different. As,
statistical significance (F = 6.98, p < 0.05) was observed in the ANOVA table,
three music types have statistically different effect on anxiety level.
ii. At 5% level of significance, two shyness types were statistically different. As,
statistical significance (F = 11.03, p < 0.05) was observed in the ANOVA table,
two shyness types have statistically different effect on anxiety level.
c. From post hoc analysis, it was observed that the three music levels have mixed level of
interaction. Difference of effect of Classical music and Pop music was statistically
significant (MD = 41.00, p < 0.05). Jazz music had significant difference with Pop music

(MD = 46.25, p < 0.05) for effect on anxiety. But, difference in effect between Jazz and
Classical music (MD = 5.25, p = 0.92) was statistically not significant.
d.
i. Interaction between Music and Shyness (F = 4.36, p < 0.05) was statistically
significant for the effect on anxiety.
ii. From the interaction plot, Classical and Jazz had interactive effect on anxiety
level for shy people. For non shy people, all the three music levels had low level
of effect on anxiety level. The three music level had higher effect on anxiety for
shy people.
Answer 14:
a. The case-control study has the effect of moving the cohort study from one cause to
another. The case study begins with the disease, while the cohort study begins with
healthy people (Peacock et al., 2018).
b. Cohort studies endow with the best information on the causality of the disease, as they
result from a person's exposure at the onset of the disease. With the data from the cohort
study, accumulated incidences can be calculated. These are the most direct measure of
the risk of developing an illness. An added benefit is that it can examine a range of results
caused by an exposure (Garcia-Aymerich et al., 2006).

Answer 15:
a. This study was to examine the effects of chronic smoking on LV MPI after acute MI in
successfully revascularized patients with primary percutaneous coronary intervention
(PCI), particularly its association with the smoker’s paradox.
b. A total of 429 middle-aged (40–65 years old) male patients presenting with acute ST
elevation MI were enrolled between September 2006 and January 2012 from 2 centers.
c. Prospectively means, the baseline data has been collected at enrollment in a prospective
study. As this data included age, risk factors for CAD and other information, the data can
be used for predicting new future events.
d.
i.The smokers and nonsmokers were compared to show their demographical difference in
table 1.
ii.The values 54.6 and 53.5 were the average age of nonsmokers and smokers. The values
6.7 and 6.8 were the standard deviations of the age data.
iii.Unpaired t-test was used to compare age of the two groups. The t-test was used to
compare between the groups.
iv.NS indicated non-significant, the NS entry in p column for age indicated that there was
no statistically significant difference in age in smokers and non-smokers.
v.In the hypertension row 16, 51 were the numbers of non-smoker and smoker suffering
from hypertension. The bracketed terms were the percentages of the above categories.
vi.The p-value was NS in the hyper tension row. It indicated no significant difference
between the proportion of age of smoker and non-smokers.

vii.To compare the proportion (percentage) of people suffering from hypertension due to
smoking and non-smoking, chi-square test was the best option. The p-value was reported
from chi-square test. The hypertension was dichotomous in nature; hence choice of chi-
square was obvious (Bacaksiz et al., 2013).
e. Under Two-dimensional echocardiography, smokers and non-smokers differ statistically
on Wall-motion score index as the p-value was less than 0.05. The difference in
magnitude between the two groups was evident. The non-smokers Wall-motion score
index was between [19.4, 29] whereas for smokers the index was between [20.9, 34.7].
References
Bacaksiz, A., Kayrak, M., Vatankulu, M. A., Ayhan, S. S., Sonmez, O., Akilli, H., ... & Ozdemir,
K. (2013). The Effect of Smoking on Myocardial Performance Index in Middle‐Aged
Males after First Acute Myocardial Infarction. Echocardiography, 30(2), 155-163.
Dutra, L. M., & Glantz, S. A. (2014). Electronic cigarettes and conventional cigarette use among
US adolescents: a cross-sectional study. JAMA pediatrics, 168(7), 610-617.
Garcia-Aymerich, J., Lange, P., Benet, M., Schnohr, P., & Antó, J. M. (2006). Regular physical
activity reduces hospital admission and mortality in chronic obstructive pulmonary
disease: a population based cohort study. Thorax, 61(9), 772-778.
Peacock, A., Hutchinson, D., Wilson, J., McCormack, C., Bruno, R., Olsson, C. A., ... &
Mattick, R. P. (2018). Adherence to the Caffeine Intake Guideline during Pregnancy and
Birth Outcomes: A Prospective Cohort Study. Nutrients, 10(3), 319.