1 Overview
As the course is about teaching you how to analyse quantitative data, the as- signment will test you about how to analyse quantitative data. You will choose suitable data, derive a research question, select dependent and independent vari- able, and apply the skills you have acquired in the course. After this, you will write up your results into a paper.
2 Format
The paper should broadly follow the style of a quantitative journal article, with the exception that you should focus on the analysis and explaining your analysis. It is not necessary to include a literature review, although you may choose to cite papers to support some of your choices e.g., your research question, your choice of variable, the assumptions you make, and so on. However, note that the learning ourcomes are focussed on analysis and the quality of the explanation and interpretation.
You will be required to submit a paper which is produced in R Studio and compiled using R Markdown. You will submit both the source file (with an .Rnw extension) and the output (which can be Word, html, or pdf). Any code included in the output will not count towards the word limit. You can find out more about R Markdown here: Using R Markdown for Class Reports and/or R Markdown for R Studio1. There is also a Datacamp course on using R. If you already know LATEXthen note that you can produce the paper in R Studio using that language. Markdown. Having an R Markdown cheat sheet to hand will also be helpful. We will be using R Markdown during most of the tutorials. Writing the document in this way follows best practice for quantitative research (and also data science). It will make it easier to manage your analysis and write-up.
3 Content
The paper should outline what your research question is and what data you will use to address it. You will then go on to analyse your data. Your analysis should include:
• Summary statistics
• Data visualisations
• Bivariate associations
• Hypothesis tests
• Regression model
• A discussion of the regression assumptions and whether they are met • A summary of your findings
Note that the assignment is not about mindlessly churning out R output. It is about your ability to select appropriate data and methods to address your research question, and to critically interpret the output of the tests/methods. Reading journal articles which employ quantitative methods will help you to see how your work should be written up. Strong emphasis should be put on the regression model. It is mentioned explicitly in the course’s learning outcomes, and will therefore carry significant weight in the marking. A regression model will also allow you to address other statistical concepts e.g., hypothsis testing, and bivarate associations.
4 Marking
As usual, the purpose of the assignment is to assess the extent of your attainment of the course’s intended learning outcomes. The intended learning outcomes of this course are:
• Have the skills to manage, visualise, summarise, and present univariate and bivariate data.
• Understand the concept, logic and assumptions of linear regression anal- ysis.
• Understand the theory of hypothesis testing for interval, binary and cat- egorical variables.
• Be familiar with the use of the statistical software R for quantitative anal- ysis of univariate variables, bivariate associations and linear regression analysis.
• Be able to critically evaluate theories, test hypotheses, and answer sub- stantive research questions using quantitative approaches with available data from a variety of social science perspectives beyond the examples given in this course.
• Be able to describe quantitative methods, and to interpret and write up the results of quantitative analyses, clearly and concisely.
For more information about assessment at the University of Glasgow, please consult the Code of Assessment. Please take time to familiarise yourself with the university’s policy on plagiarism.
5 Data
As there are so many programmes on the course, I have decided to give you the option choosing which data you want to work with. However, be careful with your choice as a poor choice may make it difficult to complete the assignment to a high standard. Here are some points to consider:
• Is the data a match for your research question?
• Do you have a mix of types of variable i.e. interval/scale, ordinal, cate-
gorical?
• See whether you have the data to run a regression model e.g., you will need a continuous dependent variable
• Do you have enough observations? Probably at least 200 will be needed but this will be discussed during the course.
• Is the quality good enough? Is the data well documented?
• Is the data cross-sectional? We only consider cross-sectional analysis in
this course, so that’s what you should use for the assignment.
I would recommend selecting a dataset from the UK Data Service. You may choose to work with the health data used in the lectures, however you will not be able to do the same analyses with the same variables. There are, however, many variables in the dataset so there is scope for other work to be done.
Choosing something you are interested in will help you maintain motivation during the process. Note that you will not be able to submit the same work for another assessment. Bear this in mind when selecting a research question and data. Your work may, however, be used to inform other work.
6 Getting help
There is time available during the tutorial sessions for you to work on your assignment. This will mean you will have access to your instructor. They will not tell you exactly what to do, but will help to point you in the right direction. Questions about the formal requirements for the assignment should be directed to the convenor, David McArthur. You should post any questions on the assignment forum on Moodle. This ensures that everyone has access to the same information. Note that emails sent directly to the convenor may go unanswered if they should have been posted on the forum.
Reviews
There are no reviews yet.