Assignment 1: Reliability and tools for reproducible data science

The purpose of this assignment is to present estimates of reliability of measures collected in the physiology lab. A second purpose is to use tools for reproducible data science. The report that you are expected to hand in therefore has some strict requirements in its format (see below). The assignment is a group assignment and at least three students are expected to contribute to each report.

Elements of the report

The report should describe one test that you have performed in the physiology-lab. Select the test that is most interesting to you. The test should be described with a detailed protocol, including preparations of the participant (that is being tested), standardization of the test, and post-test data preparation. Post-test data preparation refers to steps needed to get data from e.g. equipment used during the test. This section should take into account and reference (Halperin, Pyne, and Martin 2015; Tanner and Gore 2012)

The next section should contain descriptive data from the test. This could be measures of central tendency and variability in the measures you have collected. If possible, try to find similar estimates in the scientific literature.

Finally, we are interested in reliability. Here you need to calculate an estimate of reliability of the test. Use (and reference) (Hopkins 2000). Try to be clear with what measure of reliability you are using and what it is telling you.

Importantly, the report should contain:

• At least one table (created from your data)
• At least one figure (created from your data), and
• data presented in the text.
• The report should use a bibliography file to manage references.

The format of the report

The report should be uploaded to github with both a source file (.Rmd or .qmd) and report file as output (html, pdf or docx-file). The github folder should also contain the dataset being used in the calculations. Work on your assignment as a R project. Contributions from members of the group can be made directly to the github repository.

Resources

• See Chapter 8 in the course notes for a step by step approach to getting started with git.

• An even more comprehensive guide to using github can be found here (Happy git with R)

• See Chapters 5-7 for graphs, tables and report writing in R in the course notes

Follow this checklist on how to get started:

• Create a github account. All members of the group should have their own account.

• Be sure to download and install git on your computer.

• Once signed in to github.com, create a new repository with an informative name. Make sure that the repository is public. One member of the group can create this repository in their own account. Alternatively, fork the portfolio exam template repository https://github.com/dhammarstrom/idr4000-student-assignments. Rename the repository and hand in your version of the repository.

• Start up RStudio, start a new project, select the Version Control option and copy the address to the repository.

• Add or modify files to your project and upload to github. You need to add files, commit them and push using git. Use the command line in your terminal:

git add -A

git commit -m "A short message to describe your changes"

git push

• When you want to download the latest version of your project, type in the terminal:

git pull

• Alternatively, use GitHub desktop to commit/push/pull your changes.

• You may encounter conflicts if pushing changes that overwrites changes made by other group members. These may be tricky but can be resolved… Have patience!

If you are contributing to the project owned by another person in your group it is recommended that you create a pull request with your changes.

Data

If you want more data, you may “borrow” data from previous students:

VO_2max tests (Group 1) VO_2max tests (Group 2) Lactate threshold tests

How to hand in the report.

Copy the link to your github folder into canvas.

References

Halperin, I., D. B. Pyne, and D. T. Martin. 2015. “Threats to Internal Validity in Exercise Science: A Review of Overlooked Confounding Variables.” Journal Article. Int J Sports Physiol Perform 10 (7): 823–29. https://doi.org/10.1123/ijspp.2014-0566.

Hopkins, W. G. 2000. “Measures of Reliability in Sports Medicine and Science.” Journal Article. Sports Med 30 (1): 1–15. http://www.ncbi.nlm.nih.gov/pubmed/10907753.

Tanner, R. K., and C. J. Gore. 2012. Physiological Tests for Elite Athletes 2nd Edition. Book. Human Kinetics. https://books.google.no/books?id=0OPIiMks58MC.