STAT670 - Introductory Statistics

Semester 2, 2017Statistics Report

Due Friday 3rd November, 5.00pm

Submit only via the Turnitin link in the ASSESSMENTS area of iLearn.

Instructions:

  1. The answers toall questionsmust be word processed and presented in report format.
  2. This assessment task should be created in Word or a similar package and then saved as a PDF file (created from a Word processed document). Other formats, including a PDF created from an image, will not be accepted by the system. This cover sheet must be attached to your document (i.e. you can edit your report into this file).
  3. Please note that iLearn is fairly reliable, but, like any website, it is occasionally unavailable for short periods. Therefore, do not wait until the due date/time to start trying to submit your work. If you have technical difficulties in submitting this assessment task, it is your responsibility to ensure these difficulties are resolved by the submission deadline. You can also contact Petra in case of difficulties but do not do this at 5pm Friday!
  4. No submissions will be accepted by email or other means.
  5. When you think you have submitted your assessment, please check to make sure: In the iLearnturnitin area you should be able to click through to see your uploaded files. You should also receive an email confirming the successful submission.
  6. This assessment task is worth 20% of your total grade for STAT670. You must complete the analysis and write up your reports on your own, questions may not be discussed on the forum or elsewhere. You may email DrPetra Grahamwith any questions you have. Submitting your assignment via the iLearn turnitin link is a declaration that the work is your own.

Information:

This assessment task is worth 20% and is marked out of 80. Penalties apply as follows:

Submitted on Friday after 5pm or Saturday:Deduct 4 marks

Submitted onSunday:Deduct 8 marks

No assessments will be accepted after Sunday.

Q1 / Q2 / Total
40 / 40 / 80

Report 1 (40 marks)

A study was undertaken to investigate whether death rates could be predicted from factors related to air pollution, weather and/or socioeconomic variables. The Minitab file, Mortality.mtw, which is available in the Assessments area of iLearn, contains data recorded on a random sample of 60 metropolitan areas in the USA. Each metropolitan area contains information from the late 1950s-early 1960s recorded on the each of the variables described below.

Column VariableDescription

C1IDMetropolitan area identifier

C2PrecipAverage annual precipitation (inches), 1960

C3JanAverage January maximum temperature (°F), 1960

C4JulyAverage July maximum temperature (°F), 1960

C5HhsizeAverage household size (number of people in a household), 1960

C6WhitePercentagewhite population in urbanized areas, 1960

C7SO2Relative sulphur dioxide pollution potential in 1960

C8MortTotal age-adjusted death rate per 100,000 in 1960

C9High_NOx0=Low nitrous oxide pollution potential in 1960

1=High nitrous oxide pollution potential in 1960

C10July_2000Average July maximum temperature (°F) in 2000

Adapted from: McDonald, G.C. and Schwing, R.C. (1973) 'Instabilities of regressionestimates relating air pollution to mortality', Technometrics, vol.15, 463-482.

  1. Use the Minitab file, Mortality.mtw, to fit a simple linear regression model for predicting death rate from each of the 6potential predictors described in C2-C7 (ie. six separate models).
  2. Write up a statistical report on your analyses (ie. one report that addresses all six models). Your report should include a comparison of the models, a determination of the best predictor and interpretations of any significant predictors. Your report should include an introduction, a methods section, a results section, a conclusion and an appendix which contains all relevant Minitab output. You will be marked on both the content and the presentation of your report.

Report2 (40 marks)

The data set above was also used to answer these additional research questions:

  1. Was the average household size in 1960 equal to 3.67, as claimed by the website Statistica.com?
  2. Was there a difference between average death rates in metropolitan areas with low and high nitrous oxide potential in 1960?
  3. Was there a change in mean July (maximum) temperatures between the data collections in 1960 and 2000? Use an appropriate 95% confidence interval to answer this question.

Use the Minitab file, Mortality.mtw, to carry out appropriate analyses to address the research questions in a to c of Report 2 and write a second statistical report on your analyses (ie. one report that addresses all threequestions a to c of Report 2). Again, your report should include an introduction, a methods section, a results section, a conclusion and an appendix containing all relevant Minitab output. You will be marked on both the content and the presentation of your report.

1

STAT670 Statistical Report Session 2 2017