Reference no: EM133181692
Quiz 1:
Question 1: Which ONE of the following is NOT a requirement for the chi square test?
The observed data consists of counts (whole numbers)
The observed cell (count) values must all be greater than 5
For a 2x2 table, two variables are present or absent in the same individuals
For NxM tables, two variables are present or absent in the same individuals in each of M samples
Question 2: The null hypothesis for a chi square test when you have a 2x2 table
(i.e. two variables) is .....
The two variables are associated
The two variables are correlated
The two variables are independent
Π1= Π2 = Π3= Π4
Question 3: Apart from the chi square test of independence and the chi square test of homogeneity, which ONE of the following is NOT an additional kind of chi square test?
The chi square test for goodness-of-fit
The chi square test for trend
The chi square test for equality of medians
McNemar's chi square test
Question 4: Which ONE of the following variables demonstrates ordinality of its
categories (and would, thus, be relatively unsuitable for a chi square or Fisher's exact test)?
School quintiles
Mining commodities (e.g. coal, gold etc...)
Health districts
Drugs used for treating high blood pressure
Question 5: If the null hypothesis is true, then calculate the correct expected cell value for the cell marked as "A" in the following 2x2 table.
|
|
VARIABLE 2
|
|
ROW
TOTALS:
|
|
|
PRESENT
|
ABSENT
|
|
VARIABLE 1
|
PRESENT
|
"A"
|
|
100
|
ABSENT
|
|
|
200
|
|
COLUMN TOTALS:
|
120
|
180
|
300
|
Quiz 2:
Question 1: Which ONE of the following IS a requirement for BOTH the chi square AND the Fisher's exact tests?
The variables may consist of quantitative (measured) data
Data must be normally distributed
Categories may be nominal
For both tests expected cell counts must be > 5 in at least 90% of the cells
Question 2: For analyzing the data in a 2x2 table, the Fisher's exact test is preferred (over the chi square test) for which ONE of the following data sets? (order is a;c;b;d)
7, 12, 14, 9
9, 12, 13, 4
7, 32, 10, 5
6, 20, 8, 7
Question 3: For which ONE of the following research questions would a chi square test of homogeneity (e.g. nxm where either n>2 and/or m>2) be the most appropriate?
Is a past smoking history associated with breast cancer risk among female miners?
Does the tuberculosis incidence rate vary by altitude above sea level in quarry workers?
Is a person's ethnicity associated with their choice of general practitioner in atown with 5 GPs?
Is the proportion of workers who smoke similar by diseased/ not diseased status?
Question 4: Which ONE of the following statements about the chi square test and the Fisher's exact test is NOT correct?
The word "exact" in the Fisher's exact test is meant to imply that the test result is unbiased
The Fisher's exact test may be performed for all cell counts if computer memory is adequate
The chi square test may yield biased results if sample size is small
The chi square test for independence always yields unbiased results if observed cell counts >5
Question 5:
If the null hypothesis is true, then calculate the correct expected cell
value for the cell marked as "C" in the following 2x2 table.
|
|
VARIABLE 2
|
|
ROW
TOTALS:
|
|
|
PRESENT
|
ABSENT
|
VARIABLE 1
|
PRESENT
|
|
|
16
|
ABSENT
|
"C"
|
|
24
|
|
COLUMN TOTALS:
|
10
|
30
|
40
|
Quiz 3:
This self-assessment exercise gives you an opportunity to practice the manual performance of a chi sqOuare test for a 2X2 table.
Read the following mini-scenario and then perform the chi square test manually.
Question 1: In a cross-sectional study, 120 adults are asked if they have ever had their tonsils removed as a child (yes/no; 1/0) and, also, if they currently suffer from hay fever (yes/no; 1/0). You find that 50 of the 120 experienced a tonsillectomy as a child (the other 70 did not). Eleven of the people with prior tonsillectomy had current hay fever. Nine of those without prior tonsillectomy had current hay fever. Use a chi square test to test the null hypothesis H0: Childhood tonsillectomy and adult hay fever are independent.
Question 2: Post your final total chi square value for the 2x2 table as your answer. DO THIS CORRECT TO ONE DECIMAL PLACE. IF YOU DO IT TO TWO PLACES OR MORE YOUR ANSWER WILL BE MARKED INCORRECT; SO, TAKE CARE TO ANSWER CORRECT TO ONE DECIMAL PLACE
(e.g. 6.1 or 13.7 or 0.5 or 1.1 etc...). Answer.........
Question 3: In the above (question 1) chi square test, which ONE of the following statements is NOT correct?
The critical value for the total chi square is 3.84
The chi square test conclusion was "fail to reject"
The Degrees of freedom = 1
One expected cell value was <5
Question 4: Read the following mini-scenario and then perform the chi square test manually.
A case-control study was carried out to assess whether adults diagnosed with lymphoma (a cancer of the lymphocytes) have raised odds of tonsillectomy as a child. There are 40 cases of lymphoma and 50 people without lymphoma. Of those with lymphoma, 20 had undergone tonsillectomy as a child. Of the 50 controls without lymphoma, 10 had undergone tonsillectomy as children.
Post your final total chi square value for the 2x2 table as your answer.
You may (or may not) need to use the Fisher's exact test for one or both of the questions. Be very certain of your answers before selecting the "submit" option.
Quiz 4:
Question 1:
Examine the following 3x2 table of cell counts and then perform the hypothesis test using Stata. Use the chi square test to obtain your answers unless use of the chi square test proves less appropriate due to too many expected cell values being <5; in which case use the Fisher's exact test.
|
Measurement options
|
Management options
|
Environmental options
|
Females
|
24
|
16
|
7
|
Males
|
10
|
16
|
6
|
Table: Numbers of MPH students selecting between three "concentration options
Submit your answer. Give the answer correct to THREE decimal places. You have only ONE attempt available. The p-value is Pr = ...?
Question 2:
Examine the following 3x2 table of cell counts and then perform the hypothesis test using Stata. Use the chi square test to obtain your answers unless use of the chi square test proves less appropriate due to too many expected cell values being <5; in which case use the Fisher's exact test.
|
Measurement options
|
Management options
|
Environmental options
|
Females
|
24
|
16
|
7
|
Males
|
10
|
16
|
6
|
Table: Numbers of MPH students selecting between three "concentration options
Submit the correct answer. Give this answer correct to ONE decimal place. You have only ONE attempt available. The smallest expected cell count was .....?
Question 3:
Examine the following 3x2 table of cell counts and then perform the hypothesis test using Stata. Use the chi square test to obtain your answers unless use of the chi square test proves less appropriate due to too many expected cell values being <5; in which case use the Fisher's exact test.
|
Measurement options
|
Management options
|
Environmental options
|
Females
|
24
|
16
|
7
|
Males
|
10
|
16
|
6
|
Table: Numbers of MPH students selecting between three "concentration options
Submit the correct answer. Which ONE of the following statements about the null hypothesis is correct? [Note: heterogeneous means "diverse, or variable, in nature"]
The proportion of females does not vary across the three options
The proportion of females does vary across the three options
The proportions of men and women, by option, are heterogeneous
The proportions of men, by option, are heterogeneous
Question 4:
Examine the following 3x3 table of cell counts and then perform the hypothesis test using Stata. Use the chi square test to obtain your answers unless use of the chi square test proves less appropriate due to too many expected cell values being <5; in which case use the Fisher's exact test.
|
Region A
|
Region B
|
Region C
|
Rural
|
5
|
2
|
5
|
Peri-urban
|
40
|
30
|
20
|
Urban
|
60
|
80
|
70
|
Table: Numbers of TB patients notified by Region and rural/peri-urban/urban residence
Submit your answer. Give the answer correct to THREE decimal places. You have only ONE attempt available. The p-value is Pr = ...?
Question 5:
Examine the following 3x3 table of cell counts and then perform the hypothesis test using Stata. Use the chi square test to obtain your answers unless use of the chi square test proves less appropriate due to too many expected cell values being <5; in which case use the Fisher's exact test.
|
Region A
|
Region B
|
Region C
|
Rural
|
5
|
2
|
5
|
Peri-urban
|
40
|
30
|
20
|
Urban
|
60
|
80
|
70
|
Table: Numbers of TB patients notified by Region and rural/peri-urban/urban residence
Submit the correct answer. Give this answer correct to ONE decimal place. You have only ONE attempt available. The smallest expected cell count was .....?
Question 6:
Examine the following 3x3 table of cell counts and then perform the hypothesis test using Stata. Use the chi square test to obtain your answers unless use of the chi square test proves less appropriate due to too many expected cell values being <5; in which case use the Fisher's exact test.
|
Region A
|
Region B
|
Region C
|
Rural
|
5
|
2
|
5
|
Peri-urban
|
40
|
30
|
20
|
Urban
|
60
|
80
|
70
|
Table: Numbers of TB patients notified by Region and rural/peri-urban/urban residence
Submit the correct answer. Which ONE of the following statements is correct? [Note: heterogeneous means "diverse, or variable, in nature"]
The proportion in each residential category does not vary across the three regions
The proportion in each residential category does vary across the three regions
The observed count exceeds the expected count for urban in region A
The p-value for the Fisher's eact test is higher than the p-value for the Pearson's chis2 test
Quiz 5:
Question 1: F or a cohort study of zika virus vaccine, offered to 2000 Brazilian women of child-bearing age, assume the following table of results (only 1000 women took up the offer): ....
Zika virus-affected baby
|
|
|
Yes
|
No
|
|
Vaccine offer taken up
|
Yes
|
5
|
995
|
1000
|
No
|
40
|
960
|
1000
|
Which is the best name for the measure of association of choice in this cohort study?
The prevalence odds ratio
The relative risk
The relative proportion
The McNemar's odds ratio
Question 2:
For a cohort study of zika virus vaccine, offered to 2000 Brazilian women of child-bearing age, assume the following table of results (only 1000 women took up the offer): ....
Zika virus-affected baby
|
|
|
Yes
|
No
|
|
Vaccine offer taken up
|
Yes
|
5
|
995
|
1000
|
No
|
40
|
960
|
1000
|
Submit your answer. What is the value (correct to THREE decimal place) of the measure of association obtained for this set of data?
Question 3:
For a cohort study of zika virus vaccine, offered to 2000 Brazilian women of child-bearing age, assume the following table of results (only 1000 women took up the offer): ....
Zika virus-affected baby
|
|
|
Yes
|
No
|
|
Vaccine offer taken up
|
Yes
|
5
|
995
|
1000
|
No
|
40
|
960
|
1000
|
Submit your answer. What is the value of the upper (higher) limit of the 95% CI? Give your answer correct to THREE places of decimals.
Question 4:
For a case-control study of the association between basal cell carcinoma (a skin cancer) and the taking of thiazide diuretics among hypertensive patients who had been on treatment for hypertension for at least ten years, assume the following table of results: ....
Basal cell carcinoma
|
|
|
Yes
|
No
|
Treated with thiazide diuretics
|
Yes
|
50
|
150
|
No
|
50
|
250
|
|
|
100
|
400
|
Which is the best name for the measure of association of choice in this cohort study?
The prevalence odds ratio
The relative risk
The relative proportion
The odds ratio
Question 5:
For a case-control study of the association between basal cell carcinoma (a skin cancer) and the taking of thiazide diuretics among hypertensive patients who had been on treatment for hypertension for at least ten years, assume the following table of results: ....
Basal cell carcinoma
|
|
|
Yes
|
No
|
Treated with thiazide diuretics
|
Yes
|
50
|
150
|
No
|
50
|
250
|
|
|
100
|
400
|
Submit your answer. What is the value (correct to ONE decimal place) of the measure of association obtained for this set of data?
Question 6:
For a case-control study of the association between basal cell carcinoma (a skin cancer) and the taking of thiazide diuretics among hypertensive patients who had been on treatment for hypertension for at least ten years, assume the following table of results: ....
Basal cell carcinoma
|
|
|
Yes
|
No
|
Treated with thiazide diuretics
|
Yes
|
50
|
150
|
No
|
50
|
250
|
|
|
100
|
400
|
Submit your answer. What is the value of the lower limit of the 95% CI? Give your answer correct to THREE places of decimals.
Quiz 6:
Question 1:
A cross sectional study was carried out at a busy tuberculosis clinic in a high prevalence district. The aim of the study was to see if older people with tuberculosis were as likely as younger people to be tested for HIV. The study revealed the following situation:
Was HIV testing done?
|
|
|
Yes
|
No
|
|
Aged over 50?
|
Yes
|
80
|
20
|
100
|
No
|
270
|
30
|
300
|
Which ONE of the following four statements is correct?
Only a p-value is produced with prtest
Both a p-value and a 95% CI are produced with the chi square test
Both a p-value and a 95% CI are produced with prtest
Only a 95% CI is produced with a chi square test
Question 2:
A cross sectional study was carried out at a busy tuberculosis clinic in a high prevalence district. The aim of the study was to see if older people with tuberculosis were as likely as younger people to be tested for HIV. The study revealed the following situation:
Was HIV testing done?
|
|
|
Yes
|
No
|
|
Aged over 50?
|
Yes
|
80
|
20
|
100
|
No
|
270
|
30
|
300
|
Submit your answer. With the HIV testing data in this scenario, what is the two-tail p-value following the prtesti command?
0.0044
0.0088
0.0090
0.9956
Question 3:
A cross sectional study was carried out at a busy tuberculosis clinic in a high prevalence district. The aim of the study was to see if older people with tuberculosis were as likely as younger people to be tested for HIV. The study revealed the following situation:
Was HIV testing done?
|
|
|
Yes
|
No
|
|
Aged over 50?
|
Yes
|
80
|
20
|
100
|
No
|
270
|
30
|
300
|
Submit your answer. Which ONE of the following statements about the 95% CI for the difference between the two proportions is correct?
The difference between the population proportions is unlikely to be of much concern
The CI illustrates that there may actually be no difference other than due to sampling error
The difference between the proportions is 2.62%
The difference between the population proportions may be as large as 18%