STATS 240 Mid-semester Test PART B 2020
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
STATS 240 Mid-semester Test
PART B: Analysis of survey data (Barry Milne)
2020
An international survey was conducted on people’s opinions on how well governments have handled the global covid- 19 pandemic. The world was divided into continents and n=1000 adults from each country in each continent were contacted to complete the survey. (a) Was stratified sampling used? If so what were the strata? YES – COUNTRIES (OR COUNTRIES WITHIN CONTINENTS) (b) Was cluster sampling used? If so, what were the clusters? (If cluster sampling was used more than once then describe for each stage.) NO (c) What are the PSUs (primary sampling units) in this survey? PEOPLE/ADULTS |
3 |
|
List two reasons why a researcher might consider cluster sampling? • IT’S CHEAPER • DON’T NEED A COMPLETE SAMPLE FRAME • IF INTERVENTIONS ARE PLANNED THESE CAN BE DONE AT THE LEVEL OF THE CLUSTER |
2 |
|
Page Total |
5 |
|
Page 1
B3 If we had the data for the whole population of interest and calculated a mean, what would the associated standard error be? Why? ZERO BECAUSE THERE IS NO VARIABILITY/ERROR DUE TO SAMPLING |
2 |
|
|
B4 (a) What is a bubble plot and how is it used for plotting survey data? A SCATTERPLOT WHERE THE SIZE OF THE PLOTTING SYMBOL IS PROPORTIONAL TO THE SAMPLING WEIGHT. (b) Give one reason to add smoothers to bubble plots TO OVERCOME THE MISLEADING APPEARANCE OF DENSITY TO REVEALS TRENDS
B5 Consider the output below:
Summary of var1: ---------------- Population estimates: Est. Pop. Size 584627952.1 Is var1 a numeric variable or categorical variable? NUMERIC (A CATEGORICAL VARIABLE WILL SHOW % IN GROUPS RATHER THAN MEAN., MEDIAN ETC.) |
|||
2 |
|
||
1 |
|
||
|
|
|
|
B6 The output below contains summary statistics and an ANOVA test for the association between ‘agecat ’ (a categorization of age in years into groups) and ‘HI_CHOL’ (ranging from 0- 1; higher scores indicate higher cholesterol).
Summary of HI_CHOL by agecat: ----------------------------- Population estimates: 25% 0 0 0 0 Wald test for agecat (ANOVA equivalent for survey design) F = 72.812, df = 3 and 13, p-value = 2.2067e-08 Null Hypothesis: true group means are all equal Alternative Hypothesis: true group means are not all equal
Interpret the output, commenting on whether you think the null hypothesis should be accepted or rejected, and whether (and if so, how), age group is associated with high cholesterol.
NULL HYPOTHESIS IS REJECTED AND ALTERNATIVE HYPOTHESIS IS ACCEPTED (AS P IS VERY LOW: P~0.00000002) AGE IS SIGNIFICANTLY ASSOCIATED WITH HIGH CHOLESTEROL, WITH INCREASING HIGH CHOLESTEROL WITH INCREASING AGE TO 40-59, AND A SLIGHT DIP IN THE OLDEST AGE GROUP (60+) |
2022-08-24