1. What percentage of the entire sample is made up of Indian respondents? Does the percentage change if you only consider respondents who gave a valid response to the race question?
Commands:
tab race, missing
tab race
Answer:
2.25% of the sample is made up of Indian respondents. Yes, the percentage does change to 2.31%.
2. Which of the racial groups makes up the largest proportion of urban residents?
Commands:
tab race if metro ~=1
Answer:
Africans have the highest proportion of urban residents. Of the 2042 urban residents, 55.24% are Africans.
3. What do respondents think will happen to the quality of their lives under a new government? How many think a new government will improve the situation for their household? What percent of those surveyed think a new government will not change their household’s quality of life?
Commands:
lookfor government
Answer:
572 households (58.73%) believe a new government will make their lives better.
169 households (17.35%) believe a new government will not change their household’s
quality of life.
4. What percentage of the sample is made up of grandchildren of the household head?
Commands:
lookfor grandchildren
lookfor relationship
Answer: 14.50%
5. How many households located in rural areas report a total monthly income of R300? Use _n in finding your answer.
Commands:
lookfor total monthly income
lookfor income
Answer: 5 households
6. How many observations do not have a valid hhid value, meaning the observation’s hhid value is missing? Provide three different methods of obtaining the answer to this question, using at least one of the following commands in each – count, codebook, list
Commands:
count if hhid==.
codebook hhid
Answer: 0 households
7. What percent of females live in the old
Commands:
tab province if gender_n==2
tab province if gender_n==2 & age<50
Answer:
18.79% of females in the sample data set live in Kwa Zulu Natal.
47 women in Kwa Zulu are under the age of 50.
EXTRA CREDIT - THIS IS VERY HARD
8. Using the system variable <_n>, create a measure that reflects the number of people in the household (i.e. a household size variable) called “size”. Label the new “size” variable with the following description: “household size”. You may not use the existing household size variables in constructing your new measure, although you may compare the frequency distribution of “size” with the frequency distribution of the existing household size variables in the data, are they the same?Commands:
gen temp = 1
replace temp = temp + temp[_n-1] if hhid == hhid[_n-1 ]