This assignment consists of two parts, both of which have to be completed. The work
you submit should be approximately 1000 words in length (specifically, no more than
10% above this length). Where requested to attach your computer output, do so in the
form. of Appendices and indicate at relevant locations in the main text where the detailed
output can be found; the appendices do not count towards the length of your submitted
work. To complete this assignment you have to make use of datasets that are available on
Moodle and which you will have used in various seminars.
The deadline for submission is 3.00pm 15th January 2018. Procedure for submission
will be posted separately on Moodle. Please note this coursework is only to be submitted
electronically.
Part1: StatisticalInferenceinPublishedArticles. Select one article from the list below
(PDFs available on Moodle).
Matthew K. Buttice and Caitlin Milazzo (2011). “Candidate positioning in Britain”.
In: Electoral Studies 30.4, pp. 848–857 Buttice and Milazzo 2011
Sin Yi Cheung and Anthony F. Heath (2007). Nice work if you can get it: ethnic
minority disadvantage in Great Britain. Oxford University Press/British Academy
Nabil Khattab (2009). “Ethno-religious Background as a Determinant of Educa-
tional and Occupational Attainment in Britain”. In: Sociology 43.2, pp. 304–322
Mary S. Senter, Spalter-Roth, Roberta, and Nicole Van Vooren (2015). “Jobs, Ca-
reers, and Sociological Skills. The Early Employment Experiences of 2012 Soci-
ology Majors”. In: Washington, DC: American Sociological Association. Retrieved
August 11, p. 2016
1
2 M11157 COURSEWORK
For the article you selected, describe the article and summarize how the article de-
scribes the relationship between two variables. As you write, try to incorporate answers
to all the following questions:
What is the research question the research(s) are investigating?
What is the dependent variable and what is it measuring?
– Is this variable categorical or quantitative? On what level of measurement is
this variable measured? Be as specific as you can.
Select one main independent variable of interest:
– Is this variable categorical or quantitative? On what level of measurement is
this variable measured? Be as specific as you can.
What are the cases?
How were the data collected (or can you not tell this from the article)?
Do the data come from an observational study or an experiment?
Does the article imply a causal association? If so, is such a conclusion appropriate?
Part 2: Performing Quantitative Analysis in R:. All components in this part pertain to
variables in the dataset USw3_completeCases.RData (available on Moodle, in the folder
Data-sets -> Understanding Society) and has several sub-questions. For all questions
below, be sure to report the result of any estimation or hypothesis test in your report, and
attach the full output and your R script. in an appendix. Use R: or Rcmdr to create graphs,
examine summary statistics, and conduct the analysis.
A geographer is interested in household expenditures on utilities and whether this
differs by type of neighborhood (e.g. ‘city’, ‘rural’, or ‘town” areas). She uses the variable
c_hood3 to measure the area a respondent lives in. She uses the variable utilSpend
M11157 COURSEWORK 3
to measure spending on utilities (gas, oil, electric expenditures for the past year, at the
house-hold level).
(1) Briefly discuss these two variables. Your discussion should include: the scale on
which each variable is measured; an appropriate measure of central tendency for
each; the number of observations of each.
(2) Construct and include an appropriate (uni-variate) visualization of the variable
c_hood3.
(3) Graphically compare spending on utilities in the different areas, city, town, and
rural. Specifically:
(a) Provide a graph showing the empirical distribution of spending on utilities
for those living in city (for those with the value city on the variable c_hood3).
(b) Provide a graph showing the empirical distribution of spending on utilities
for those living in rural areas (for those with the value rural on the variable
c_hood3).
(c) Provide a graph showing the empirical distribution of spending on utilities
for those living in towns (for those with the value town on the variable c_hood3).
(d) Compare the above three distributions (either in one graph, or three separate).
Briefly comment on their qualitative features (e.g. does it appear average
spending on utilities is different between the 3 areas?)
(4) Provide and interpret descriptive statistics of the variable utilSpend. Specifi-
cally, report and interpret the sample mean, median and standard deviation and
skewness. By investigating the mean and median, is the distribution of utilSpend
skewed? If so, in what way?
(5) Estimate the true population average spending on utilities. Report the standard
error of this estimate as well as the 95% confidence interval. Interpret both the
standard error and the confidence interval.
4 M11157 COURSEWORK
(6) Assess the claim: “the true (population) average spending on utilities (per annum)
is greater than 1,368.” Specifically, formulate this statement as an alternative
hypothesis.
(a) How would you test this hypothesis? In other words, what statistical test
would you use to test it? Why is it an appropriate test for the question? Be
sure to mention if a one-sided or two-sided test is appropriate.
(b) Conduct the test you suggested in 6a. Include the results (i.e. output) of your
statistical test and interpret your findings. (i.e. Can the researcher reject the
null-hypothesis? If so, at what level?)
(7) Assess the claim: “the true (population) average spending on utilities in cities
is different from 1,368.” Specifically, formulate this statement as an alternative
hypothesis.
(a) What is the corresponding null-hypothesis?
(b) How would you test this hypothesis? In other words, what statistical test
would you use to test it? Why is it an appropriate test for the question? Be
sure to mention if a one-sided or two-sided test is appropriate.
(c) Conduct the test you suggested in 7b. Include the results (i.e. output) of your
statistical test and interpret your findings. (i.e. Can the researcher reject the
null-hypothesis? If so, at what level?)
References.
Cheung, Sin Yi and Anthony F. Heath (2007). Nice work if you can get it: ethnic minority
disadvantage in Great Britain. Oxford University Press/British Academy.
Khattab, Nabil (2009). “Ethno-religious Background as a Determinant of Educational and
Occupational Attainment in Britain”. In: Sociology 43.2, pp. 304–322.
Buttice, Matthew K. and Caitlin Milazzo (2011). “Candidate positioning in Britain”. In:
Electoral Studies 30.4, pp. 848–857.
REFERENCES 5
Senter, Mary S., Spalter-Roth, Roberta, and Nicole Van Vooren (2015). “Jobs, Careers, and
Sociological Skills. The Early Employment Experiences of 2012 Sociology Majors”. In:
Washington, DC: American Sociological Association. Retrieved August 11, p. 2016.