代写ENVS222、R程序语言代写

日期：2022-11-10 10:39

ENVS222 assignment 1: root growthin

mustard seeds

You should produce a report of approximately 3 pages (including figures, but excluding your R code,

which should be added as an appendix as described below).

Background

Wewant to know whether the roots of germinating mustard seeds typically grow longer in the dark

or in the light, and whether any difference is consistent across replication attempts. We will use data

collected this year and from the 2011/12 academic year, when the experiment was set up in a

similar way.

Data

The tab-separated text file rootlengths.txt (available on the Assignment 1 page on Canvas)

contains the data. Download this file. Each row contains the data from a single dish in a single year

(there were 160 dishes in total, 48 dark and 48 light from 2011/12, 32 dark and 32 light from

2022/23). There are three variables: treatment (either light or dark), meanrootlengthmm (the

mean length of the roots produced by the 10 seeds in a dish, in millimetres) and year (the year in

which the experiment was done: either 2011/12 or 2022/23). Note that mean root length was

calculated from all seeds in each dish, including those that did not germinate (which were counted

as having root length 0 mm).

2Analysis and report

You should use R for analysis. Everything needed has been covered in the workshops from weeks 1

to 4, and you are welcome to base your R script on scripts that were provided for these workshops.

Your report should be written up individually. Because you have not yet had much practice at data

analysis, we have broken down the report into sections, and explained what we want in each.

Sections 1-5 contain material that would typically appear in the methods of a scientific paper,

sections 6-9 contain results material. Modern scientific papers will often include code in an appendix

or supporting information section, as in section 10. Within each section, where writing is involved

you should write proper sentences rather than bullet points. You do not need to write anything that

is not asked for below. The report will be marked out of 40, with marks as listed below. See the mark

scheme on Canvas for more information about how the marks are broken down within each step.

1. Explain briefly what we did to avoid bias in the design of the experiment, and why this is

important (2 marks).

2. Explain briefly why it is appropriate to work with mean root lengths from each dish rather

than the 10 individual root length measurements from the dish. You should use appropriate

terminology, for which you may find it helpful to refer to the external reading from week 4

(2 marks).

3. Based on what you saw in your own dishes, what difference might it have made if we had

included only those seeds that germinated when calculating mean root length for each dish?

How might we decide which approach is most appropriate? (3marks).

4. Suggest a statistical test to investigate the scientific hypothesis that mustard seed roots will

typically grow longer in the dark than in the light, and to investigate whether this difference

is consistent across years, with a brief (one-sentence) justification (3marks).

5. List the assumptions of this test. Check these assumptions, explain how you checked, and

what the outcome was. If necessary, choose and carry out an appropriate transformation to

address any problems. Remember that the importance of some assumptions may change as

sample size increases. Also note that violations of assumptions should suggest cautious

interpretation, which will matter less if there is very strong evidence for an effect (8marks).

6. Provide a graphical comparison of the distributions of mean root lengths (transformed as

appropriate) in dark and light dishes in each year, including appropriate axis labels, with

units where appropriate, and an informative, numbered legend underneath (4marks).

7. Carry out the test and provide a table that summarizes the results. Add a numbered legend

above the table. Remember to use an appropriate level of precision and (if relevant) choose

an appropriate way to express very small numbers (5 marks).

8. Produce an interaction plot showing how the predicted mean root length (transformed as

appropriate) depends on year and treatment. Include this figure in your report, with a

numbered legend below (4 marks).

9. Write two to three sentences summarizing what we can say about the difference in mean

root length (transformed as appropriate) between dark and light treatments, whether mean

root length depends on year, and whether the effect of treatment depends on year,

referring by number to the table from step 7 and the figures from steps 6 and 8 (8marks).

10. Include your R script as an appendix. This means I can see exactly what you did, and in the

steps above, may allow me to award partial marks even if you did not get exactly the right

answer (1 mark).

3The easiest way to do this is as follows:

a. Install the R package rmarkdownby going to Tools > Install Packages,

typing rmarkdown in the Packages box and pressing Install.

b. Go to File > Knit Document (on some versions of RStudio this will appear as

File > Compile Report), select MS Word as the output format, and click

Compile. This should produce a Word document, with a name that matches your R

script, in the same folder as your script. The document will contain all the R code you

used and the output, which makes it easy for me to check what youdid.

c. Your username will appear on the second line of the Word document, just below the

title of the script. Delete the username, so that your assignment is anonymous.

d. Paste everything else into your main Word document as an appendix.

e. If you cannot get this to work, you can just copy and paste the script directly from

RStudio to your Word document.