mcfarland, 21sep2000 


UCLA Soc. 210a, Assignment 3:  

Univariate Frequency Distributions and Summary Statistics


1. In the Moore and McCabe book, do the following exercises
beginning on page 23, referring back to Section 1.1 as needed:
	1.1, 1.7, 1.11, 1.13, 1.17, 1.19, 1.25, 1.31, 1.39	

2. Do the following exercises beginning on page 58, referring
back to Section 1.2 as needed:
	1.43, 1.47, 1.57, 1.61, 1.65

3. If you have not already done so, run the hamilton.exe file, to
extract the datasets it contains. (Make sure you have it in its
own \hamilton directory, not in the same directory with any of
your other files.)

Load stata, and invoke the following stata commands:
        use e:\hamilton\vttown.dta (or replace "e:" with the
designation of your Zip drive)
        describe
        summarize
        tabulate educ
        graph educ
Write a paragraph describing the results.


4. Use your wstates.dta file created in Assignment 2, to do the
same sort of descriptive statistics on that dataset, again
describing the results in words and sentences.


5. If you have not already done so, go to the links page of the
Soc. 210a ClassWeb site, and download to your Zip disk the file
gss94.dta, which is a Stata-format version of the 1994 General
Social Survey dataset.

Go to the online GSS codebook at the icpsr site.  Look up the
variable EDUC.  Which of the numerical values represent various
forms of non-response rather than amounts of education?  Write
out stata command(s) that would replace those values with the "."
which stata uses as missing data code.  

Load the stata software.  Use the gss94.dta file that you saved
on the Zip disk.  Replace with "." the educ values that represent
missing data.  Invoke the stata commands:
        summarize educ
        tabulate educ

Using your favorite wordprocessing software, create a document in
which you will briefly describe the EDUC variable, in words and
sentences, though perhaps also copying material from the stata
tables into the document. Keep that document available, to create
and add a histogram.

Invoke the stata command:
        graph educ
How many "bins" are the cases sorted into by default, in
constructing the histogram?  Try overriding that default value,
to make the histogram more informative.  You may also override
the default axis labels.  

Copy your finished histogram into the document begun earlier, and
write a paragraph describing it. Be sure to discuss the relative
advantages of using the full distribution (frequency table or
histogram) vs. using summary statistics (mean and standard
deviation; or alternatives), vs. using both.


6. For each of the following GSS variables, list:
        valid values
        values that should be changed to missing
        level of measurement
Explain briefly.

        affrmact
        chldidel
        marital
        news
        tvhours

Based on the frequencies shown in the codebook, with any
supplementary calculations you wish, suggest how to best describe
the distribution of each of those five variables.