Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit



MET212

Statistical Methodology


Question 1

Malaria is a major public health concern in developing countries. Inadequate access to information, healthcare and anti-malarial resources results in inability to implement preventive measures. To assess current disparities in access to Insecticide-Treated Bed-Nets (ITN) between rural and urban locations in Cambodia, data from 16,823 survey respondents to the Demographic Health Survey (DHS) 2005 were used. The percentage of households in each area that had at least one ITN was estimated. Table 1 shows the main characteristics of respondents and the number of households that had at least one ITN.

Table 1.1

Variable

Mean

Std. Dev.

Highest year of education

3.27

1.664

Current age of respondent (years)

29.67

10.289

Variable

Frequency

Percent

Area of residence

Urban

4152

24.7

Rural

12671

75.3

Household possession of ITN

No

6293

37.4

Yes

10530

62.6

Source: adapted from Welch & Fuster, J Vector Borne Dis 2012

a. Specify the type of variable for each of those shown in Table 1.1 (4 marks)


b. Calculate the 95% reference range for the age of respondents and interpret (4 marks)


c. What is the assumption when calculating a reference range? (1 mark)


d. Calculate and interpret the 95% confidence interval for the mean age of respondents

(5 marks)


e. Explain the difference between a reference range and 95% confidence interval (3 marks)


f. The odds ratio of having an ITN at home and living in an urban area was 1.3 (95%CI: 0.9-1.4). Interpret this result (4 marks)


g. Which statistical test could you use to confirm this result? What would be the null hypothesis for this test? (4 marks)



Question 2

a. Explain what a correlation coefficient (denoted by r) is used for, and the range of possible values it can take. (6 marks)

Figure 2.1

b. For each of the scatter plots in Figure 2.1, explain whether it is appropriate to calculate r and if so approximate its value. Justify your answers (12 marks)

c. If the aim was to predict how y changes when x changes:

i. What technique would you use? (2 marks)

ii. Give any assumptions for this technique (3 marks)

iii. Write down the standard regression equation (2 marks)



Question 3


An intervention study was conducted in a central region of Mexico to assess the impact of exposure to second hand smoke in young subjects who spend time in a discotheque, by comparing within-subject baseline and post-exposure urinary cotinine levels. A total of 100 non-smoking volunteers provided a urine sample before entering a discotheque and another sample an average of 6 hours after the end of exposure. Concentrations of cotinine (ng/ml) were measured in the urine laboratory and results of the first 20 volunteers are shown in table 3.1.


Table 3.1 Level of cotinine in urine samples of non-smoking volunteers collected before exposure


Cotinine levels (ng/ml)

Cotinine levels (ng/ml)

ID

Gender

Age

Before

After

ID

Gender

Age

Before

After

1

Male

20

1.7

52.1

11

Female

18

1.9

10.7

2

Male

22

2.5

55.1

12

Male

22

2.7

32.3

3

Female

19

1.8

13.7

13

Female

17

0.9

22.1

4

Male

25

3.9

49.6

14

Male

25

4.2

37.9

5

Female

23

1.9

12.3

15

Male

23

1.5

30.5

6

Male

17

3.6

37.5

16

Female

16

3.1

21.8

7

Female

21

2.5

16.3

17

Female

23

3.0

15.7

8

Female

22

2.6

20.0

18

Female

23

2.4

27.8

9

Male

24

5.8

38.8

19

Male

25

5.5

256.0

10

Male

20

4.3

42.1

20

Female

20

0.5

10.1


a. What type are the following variables and how would you summarise them:

a. Gender

b. Cotinine level after exposure (4 marks)

b. Which statistical technique could you use to compare the mean cotinine levels before and after exposure? (2 marks)


c. Give 4 statements that describe the main results shown in Table 3.2. (4 marks)