EC 348 Winter 2024
Assignment #1
Objectives
. Get familiar with a large micro data set
. Properly account for the survey weights
. Create publication-quality tables
Questions
(1) Many health related statistics come from the Canadian Community Health Surveys in Canada. We have access to the public-use microdata files through our library. Using one of the surveys, let’s get familiar with the microdata set.
(a) Using the Canadian Community Health Survey 2017-2018, Annual Component (Public- use file), produce a table that shows the age distribution in Canada. In the data, the age variable is labeled as DHHGAGE .
The table will have two columns:
(1) without survey weights, and (2) with survey weights, and each cell should be %.Do not copy and paste your Stata output directly. Your table should be publication quality.
(b) Explain briefly why we should use the ones with the sampling weights.
(c) Confirm your result in (a) with the Statistics Canada official population statistics. You should be able to find it from Statistics Canada website. For the purpose of this assign- ment, just cite the website (no need to copy-and-paste the official statistics).Note also that the numbers do not necessarily match exactly each other partly because the CCHS only targets population above 12 years old who lives in a private dwellings (that is, excluding people in institutions etc.).
(2) Now turn into some health outcome that is measured in the CCHS 2017-18. One of theleading causes of death in Canada is heart disease, which is recorded as CCC_085 .
(a) Create the similar table as (1a) that shows the proportions of people who have heart disease and who don’t have heart disease.
(b) Based on this data, what is the proportion of the Canadians who have heart disease?
(3) Age is a good predictor of health. Let’s confirm this notion with this data.
(a) Create a cross-tabulation table that shows the heart disease incidence among each age group. The table should properly takes into account of survey weights (no need to show the unweighted version here).
(b) Comment on this table.
Table for (1a) looks like:
|
Age |
(1) no weights % |
(2) weighted % |
|
Age between 12 and 14 |
— |
— |
|
Age between 15 and 17 |
— |
— |
|
··· |
··· |
··· |
|
Age 80 and older |
— |
— |
|
# Obs. |
— |
— |
Table for (3) looks like:
|
Age |
Heart Disease Yes No % % |
|
|
Age between 12 and 14 |
— |
— |
|
Age between 15 and 17 |
— |
— |
|
··· |
··· |
··· |
|
Age 80 and older |
— |
— |
|
# Obs. |
— |
— |