Faculty of Sciences
SCIE4401 Data Use in Science
Assignment Three
Data sets
You have to download the data for this assignment from the Meat and Livestock Australia website: see below.
Submitting Assignment THree
Your assignment is to be uploaded on the LMS. Please make sure you include an assignment cover sheet as the first page of the Assignment. Upload it as a pdf.
Assignment Due Date
The assignment is due by Sunday 8th September, 11:55pm.
Assignment content
All figures and tables must be appropriately formatted with axis labels etc. An example of an appropriate figure was provided with Assignment 2. An example of an appropriately formatted table is provided below, but with generic information included.
Table 1: Pairwise t-test results (p-values)
Location 1 Location 2 Location 3
Location 1 Value 1 - -
Location 2 Value 2 Value 4 -
Location 3 Value 3 Value 5 Value 6
Note: p-values have been adjusted to control the familywise Type ## error rate using the ##### method.
General Marking matters
The question will be marked out of 100. All figures and tables must not only be appropriately formatted, but must also be set at an appropriate scale.
Alpha level
Use an alpha value of 0.10 for all hypothesis testing.
Question 1
The research question
There is economic theory that suggests that prices in different markets in different regions should be related to each other, otherwise someone can make money by trading between markets. But there are other factors at play as well, including transport costs, and local market conditions. You have been asked to undertake some analysis about whether, on average, the prices being paid for live cattle is different in 4 different markets in Australia, and, if they are statistically different, which markets you would assess as having statistically significant higher prices, and which tend to be lower.
Because your client has some training in statistics, they have asked for the analysis to be conducted with some specific requirements. The alpha level that you have to use is 0.1. You also have to report the results of comparisons between markets using a Holm correction for multiple comparisons and also, not using any adjustment for multiple corrections.
Data set
There is no data set provided. You are required to obtain the data directly from the Meat and Livestock Australia website.
To get to the relevant website you can follow the link: https://www.mla.com.au/prices-markets/market-reports-prices/.
There is a very clear requirement of which data you have to use.
You should conduct the analysis using weekly cattle prices for the years 2018 for the following four markets:
Armidale
Dubbo
Wagga
Finley
There are several different types of cattle traded: you one need is
yearling steers.
There are different weights of steers: You want the
330-400kg
There are also several different prices reported: you need
$/head
There are some criteria which should be switched to All: Sales prefix, Muscle score, Fat score
A brief video on how to download the data is included in LMS.
You need to watch this as it’s an important guide to showing you how to download the right data.
You need to download the relevant data into excel format for each of the towns. Then edit the data so you get the right time period. Then combine the results for each town into a single file it so that it can be used in the analysis, and then saved as a CSV file.
Structure and marks
20 marks out of 100 will be based on your generating the correct data set.
Your research report should provide an analysis about how average prices differ across the 4 markets.
The format of your report should be as follows (marks out of 100 reported):
Introduction (marks 10)
One or two sentences that state the purpose of the report. Your objective for this section is to leave the reader with a clear impression of what is to follow. The section should be sharp and focused. Original research is not required, you just need to restate the key information provided in the question.
Methods (marks 20)
This section is a little different to what a methods section would look in a traditional paper. For this question you can state the data source; the testing approach that you have used; the alpha level used for each test; the method of adjustment used for multiple comparisons etc. Your objective for this section is to ensure the reader is convinced you have followed a logical approach to testing.
Results (marks 40)
This will be the main section of your response. Report details in a table of the mean and standard deviation and number of observations of the prices, for each of the markets, in a well formatted table. Here you will also include an exploratory figure.
You should report the results of the steps you have taken in the analysis. You should also report summary tables of results. As the data comes from 4 markets (i.e. you have 4 samples) you have been asked to provide analysis for multiple comparisons using the Holm adjustment, as well as one not undertaking any adjustment for multiple corrections. You should make it clear what the implications are of using these two approaches, and which set of results you have based your final conclusions on and why. You will make clear short statements about your research findings. Your objective at the end of this section is to make sure the reader understands whether or not there are any differences in the group means that are statistically significant.
Discussion/ Conclusion (marks 10)
This will be perhaps three or four sentences. In this section your objective is to take the detail from the results section and translate this detail into a form that matches the research objective stated in the introduction.
In the sections on Methods and Results, 50% of the available results will be allocated to the clarity of your discussion and presentation. Brief dot points, and simple reporting of outputs from R are not sufficient here. You have to show that you can communicate what you have done, and the interpretation of results.
Structure of report
The general structure of the report should follow the same style as the “mock answer” provided for Assessment 2. It should not be more than 2 pages long.
In addition to the 2 pages you should include an Appendix that contains the R script that can exactly reproduce the figure/results reported in the text.
You should also include the results in the Appendix from running a summary() command on the data frame.
Failure to include the appendix will mean it’s difficult to give discretionary marks if you make an error somewhere. You should also include the standard assignment cover sheet. The entire report should be submitted online as a single pdf document. A hard copy can be submitted the following day, but the online version is the official version.