# 2.3 study types, measures of association, statistical tests, & | biostatistics

Directions: Write   the information requested in each column. Please refer to the Variables Supplement found in   the Resources. Each scenario is 14 points each, total 70 points. Question F   is worth 5 points for a total of 75   points.

Select from   these study designs, using each only once:

Case   control study

Prospective   cohort study

Retrospective   cohort study

Cross-sectional   study

Intervention   trial

Select from   these measures of association and tests of significance, some of which are   used more than once:

Chi-square

Chi-square   w/o Yate’s Correction

Difference   between the means

Chi-square   Fishers Exact Test

Odds   ratio

p-value

Relative   risk

t-test

Scenario   A: A team of researchers set out to study the risk   factors for pancreatic cancer. From   hospital records, they assembled a group of 50 patients who were either still   living or who had a next of kin identified and willing to answer questions   about medical history, lifestyle, and prior experiences of the patient. They assembled another group of 50 patients   from the same hospital who were admitted for other causes and did not have   cancer of any type. The investigators   conducted a review of all available records of these patients, then   interviewed them or their next of kin using a standardized format, either in   person or by telephone. They focused   on ten possible risk factors which could either be answered as yes/no or   rated on an ordinal scale. In this   study, 22 of the 50 pancreatic cancer patients were recorded as having at   least a 20 pack-year history of cigarette smoking, while 7 of the 50 in the   other group of patients had that history.

1. What   type of study is this   and what attributes make it so? (See   Week 2 PPT):

Case-control   study

2. What is the independent variable in this scenario, and what type is it? What is the dependent   variable in this scenario, and what   type is it?

Independent= 20 pack a year   smoking

type=

Dependent=pancreatic cancer

type=

3. A null hypothesis states that   there is no effect, no relationship, or no difference between the variables   being tested. What might be a null hypothesis   for this study?

4. Odds   Ratio is the   most appropriate measure of association to evaluate the relationship of each   risk factor for the occurrence of pancreatic cancer and measures 4.47. Use the table and link below to show the data. What do the results   mean?

The   Odds Ratio is calculated as (a/c) / (b/d).

In   this scenario, which group   is ___ times more likely to develop pancreatic CA?

What   types of variables can be tested by OR?

Odd Ratio   Calculator

# of Ills

# of Wells

Total

# of Exposed

A =

B =

A + B = 50

# of Unexposed

C =

D =

C + D = 50

Total

A + C = 50

B + D = 50

A+B+C+D = 100

§ OR = 1 suggests that there   is no difference between the groups; i.e. there is no association between the   suggested exposure and the outcome (being ill)

§ OR > 1 suggests that the   odds of exposure are positively associated with the adverse outcome compared   to the odds of not being exposed

§ OR < 1 suggests that the   odds of exposure are negatively associated with the adverse outcomes compared   to the odds of not being exposed.  Potentially, there could be a protective   effect

5. What is the most appropriate test   of significance for this scenario?   (See Week 2 PPT)

Chi-square   test for Independence w/o Yate’s correction

6. Calculate   the chi-square test result using this on-line calculator. What is the result?

7. What   is the interpretation of the test result and what do the results mean?

The chi-squared statistic is a   single number that tells you how much difference exists between your observed   counts and the counts you would expect if there were no relationship at all   in the population.

Scenario   B: Another team of researchers set out to test a new   medication developed to prevent the onset of diabetes in patients with a   strong family history of diabetes. They assembled a group of 200 patients and randomized them into two   equal groups, one to receive a placebo and the other to receive the   medication. They followed them each   year for five years and ascertained the occurrence of diabetes using a   glucose tolerance test. At the end of   the trial, 11 of 100 patients in the placebo groups had developed diabetes   mellitus, and 8 of 100 patients in the medication group had developed   diabetes mellitus.

1. What type of study is this and   what attributes make it so? (See week   2 PPT)

Intervention   trial

2. What is the independent variable   in this scenario, and what type is it? What is the dependent variable in this   scenario, and what type is it?

3. Please   state a null hypothesis.

4. What is the most appropriate   measure of association for this scenario? (see week 2 PPT)

Relative   risk

5. What is the most appropriate test   of significance for this scenario? (See week 2 PPT.) What do the results mean?

Chi   square

6. Fill in the online Chi-square   contingency table.    Set it up just as you did earlier. What   is the result? Chi-square is again the most appropriate test for   significance. Why?

7. What   do the results mean in terms of the null?

Relative Risk is very similar to Odds Ratio,   however, RR is calculated by using percentages, whereas Odds Ratio is   calculated by using the ratio of odds.  Relative Risk values are greater   than or equal to zero.  A value of 1 indicates a neutral result meaning   the probability of a variable being present in one group is the same   probability of a variable being present in the other group.  However, a   value of zero indicates that none of the cases in group 1 had the event occur   while x number of cases in group 2 had the event occur; or   in other words, the numerator was a zero (A = 0) and the denominator was any   number greater than zero (B = x, where x >   0).

Scenario   C: A team of health sciences students decided on a   project that would give them experience with their statistical analysis   skills as well as some practice in taking vital signs. They assembled a group of 40 undergraduate   students, 20 science majors and 20 majors in other fields of study, brought   them to a makeshift clinic in the science building and measured their blood   pressures with a hand-held BP cuff. The results were as follows:

The science majors’ SBPs were: 100, 110, 116, 120, 122, 128, 128, 128, 130, 132, 132, 134, 140, 144,   146, 146, 150, 152, 154, 158.

The other majors’ SBPs were: 112, 118, 120, 126, 130, 130, 132, 134,   140, 142, 144, 144, 148, 150, 158, 160, 170, 172, 176, 190

1. What type of study is this and   what attributes make it so? (See Week   2 PPT)

Cross-sectional   study

2. What is the independent variable   in this scenario, and what type is it? What is the dependent variable in this   scenario, and what type is it?

3. Please   state a null hypothesis.

4. What is the most appropriate   measure of association for this scenario (see week 2 PPT)? Calculate that measure of association.

Significance of the difference   between the means

5. What is the most appropriate   statistical test of significance of this measure of association. Why?

Independent   t-test

6. Calculate the independent t-test result using the on-line calculator (Formula) What are the results?

7. What   do the results   mean and how   do we state our conclusion regarding the null hypothesis?

Scenario D: Another team of   investigators reviewed medical records to determine which military personnel   among several basic training units had received a previous type of meningitis   vaccine during training two decades ago. Two study groups were formed, those who had not received the vaccine   and those who had. These were followed   forward by record review throughout the duration of their active military   service, and if they had a record of care through the VA system, these   records were searched as well. The   outcome of interest was the occurrence of any of a number of autoimmune   diseases that were thought to have a possible association with the   vaccine. In this study, of a vaccine   group of 70, 4 individuals were found to have been diagnosed with lupus at   some subsequent point in time. In the   non-vaccine group of 105, only 2 individuals had that diagnosis

1. What type of study is this and   what attributes makes it so? (See week   2 PPT):

Retrospective   cohort study

2. What is the independent variable   in this scenario, and what type is it? What is the dependent variable in this   scenario, and what type is it?

3. State a null hypothesis.

4.  What is the most appropriate measure   of association to compare the rates of occurrence of autoimmune disease   between the two groups and Why?   Calculate that measure of association.

Relative risk

5. What is the most appropriate   statistical test of significance of this measure of association and Why? Calculate the test result using   the on-line calculator.  Which test option should you use? What is the result?

Chi square, fisher’s   exact test

6. Using the same on-line calculator as for problems #1 and #2,   perform this test. What are the   2-tailed p-value results?

7. What do the results mean   regarding the null hypothesis?

Scenario   E: Yet another team of investigators embarked on a   long-term study to assess the impact of low versus high levels of blood   cholesterol in Native Americans on subsequent occurrence of myocardial   infarction. They recruited 100   patients from three Indian Health Service hospital systems, and divided them   into two groups based on cholesterol LDL/HDL ratio (high ratio is bad) – 42   in the high group and 58 in the low group. By the end of the fifteen-year follow-up period, 16 in the high group   and 7 in the low group had experienced a myocardial infarction.

1. What type of study is this and   what makes it so?

Prospective   cohort study

2. What is the independent variable   in this scenario, and what type is it? What is the dependent variable in this   scenario, and what type is it?

3. State   a null hypothesis.

4. What is the most   appropriate measure of association to compare the rates of myocardial   infarction in the two groups and Why?   (See Week 2 PPT) Calculate that measure of association. What are the results?

Relative   risk

5. Chi square w/o   Yates correction   is the most appropriate statistical test to test the significance of this   measure of association. Why and what   are the 2-tailed p-results?

6. What do the results mean in terms   of the null hypothesis?

1.   How the study design is selected.

2.   How the measure of association is determined.

3.   How the measure of significance is determined.

4.   How interpretation of the results is accomplished.

5.   What you would like to understand better.

Total

/75