1. Introduction
This part of the assessment for INF6027 Introduction to Data Science comprises a piece of individual coursework to assess your
ability to analyse data using R/RStudio and to then communicate your findings. Given a specific topic and dataset (see Section 2),
you should identify a specific problem or topic you would like to investigate. You will then need to pre-process and analyse the
dataset to identify patterns and relationships that address your selected problem/topic. This should involve using techniques learned
throughout the practical sessions that will help you to demonstrate your R skills, such as summarising datasets, statistical modelling
or data visualisation, to highlight and illustrate particular aspects of the data you want to communicate (e.g., particular patterns or
trends).

This coursework aims to follow the stages involved in a ‘typical’ data science process: (i) define the question(s) to address (note,
sometimes this does not come at the start of the process, but after initial exploration of the data); (ii) gather data; (iii) transform, clean
and structure the data; (iv) explore and analyse the data; and (v) communicate the findings of the data analysis. This often occurs in
an iterative manner and centred on one or multiple questions you are seeking to address. For example, the data discovery process
in Figure 1 presents an example of the stages involved in data discovery as an iterative process1 and you can find more details in
Section 3. This is also similar to the data science process we have been using in class from the “Doing Data Science” book (O’Neil &
Schutt, 2013).

Fig. 1 Example data discovery process (Jones, 2014: p.2)


You should write a 3,000 word structured report (see Section 4) that describes the approach you have taken to explore and
analyse the data for the selected problem/topic. You report should clearly communicate the results of your data analysis and be
written in a way that helps the reader interpret your findings. Note: charts, tables, and appendices are not included in the word count.

This assessment is worth 100% of the overall module mark for INF6027. A pass mark of 50 is required to pass the module as a
whole. Submission deadline: 10am Monday 17th January 2022 via Turnitin. See Section 5 for more general information about
Coursework Submission Requirements within the Information School.


2. Our World in Data COVID-19 Dataset
There has been a lot of recent interest in analysing publicly available datasets to identify patterns and gain insights into the impact of
COVID-19, see for example the Coronavirus Resource Center by John Hopkins University2. There is also an increasing use of
COVID-19 data used in the media to highlight aspects of the pandemic and related activities (see, e.g.,
https://www.bbc.co.uk/news/uk-51768274). The dataset to be used in this assessment is the Our World in Data COVID-19 dataset, a
collection of public global COVID-19 data (this is an example of Open Data). A description of the data is available here:
https://ourworldindata.org/coronavirus. The data is provided as CSV files and can be downloaded from Github.3



1 You can find out more about this process in (Jones, 2014: p.2): https://tanthiamhuat.files.wordpress.com/2015/07/communicating-data-with-tableau.pdf
2 https://coronavirus.jhu.edu/
3 https://github.com/owid/covid-19-data/tree/master/public/data
The dataset is a collection of COVID-19 data and includes the following data: vaccinations, tests & positivity, hospital & ICU,
confirmed cases & deaths, reproduction rates, policy responses, and other variables of interest.
You can select any data from the Our World in Data COVID-19 dataset. (This may require multiple downloads.) You can also
aggregate the dataset with other open data sources if you want (e.g., census data), which would demonstrate your ability to join
datasets (although you don’t have to do this to pass the coursework as the emphasis of the coursework is on how you carry out your
analysis in R/RStudio and communicate your findings on the Our World in Data COVID-19 Dataset).


3. What you need to do

The following sections describe what you need to do in order to carry out the coursework. This roughly follows the steps shown in
Fig. 1, but you don’t have to be constrained by this or follow them in this particular order; it is just a suggestion. Also, all the R we
have done in the practical sessions (and the final sessions) should be enough to conduct the coursework, although you may need to
investigate certain areas further that relate specifically to the problem you tackle in your investigation.


3.1. Review the literature and identify research question(s)

As mentioned previously, you should select a specific problem/topic related to the data (the ‘question’ stage in Fig. 1). To decide
what area to focus on you could start by undertaking a brief review of the relevant literature around areas, such as analysis of
infection data, geographical analysis of infections, predictive modelling, analysis of vaccinations statistics, etc. For example, these
articles may be a useful starting point:

Latif, S. et al. (2020). Leveraging Data Science to Combat COVID-19: A Comprehensive Review. IEEE Transactions on
Artificial Intelligence, Volume 1, Issue 1, pp. 85-103, IEEE. (Available online: https://doi.org/10.1109/TAI.2020.3020521)

Callaghan, S. (2020). COVID-19 is a Data Science Issue. Patterns, Volume 1, Issue 2, pp. 100022. (Available online:
https://dx.doi.org/10.1016%2Fj.patter.2020.100022)


Reviewing past literature will help you understand what kinds of analyses are undertaken using COVID-19 data and provide a
possible source of ideas for what you could do with the dataset mentioned in Section 2. Examples of possible topics include, but are
not restricted to, the following:

• Evolution of COVID-19 infections in an area over time;
• Models and predictions of infection rates;
• Analysis of infections and vaccinations;
• Comparisons of the spread of variants in various regions;
• Clustering and classification of data
• Normalisation and integration with other datasets (e.g., LSOA census statistics);
• Focus on a certain census dimension (e.g., demographics in the area);
• Visualisation of the data (e.g., on maps).


3.2. Download, pre-process and explore the data

As well as reviewing relevant academic literature you should also download some data as clarified above and perform an exploratory
analysis (i.e. ‘play’ with the data), to better understand the dataset and also help you to identify a particular problem or topic you
might want to focus on.

This part of your investigation will include steps to pre-process and transform the data, such as cleaning up the data, dealing with
missing values, standardising numeric values, etc. This may also include combining or joining the data with further datasets, e.g.
census or deprivation data. This reflects the ‘gather’ and ‘structure’ stages in Fig. 1. (Note: this part of the analysis could take a lot of
time so don’t underestimate how much time you will need to spend on this part of the coursework.)


3.3. Analyse and explore the data

As you identify a topic of interest for your analysis then you should identify the most appropriate techniques (using R and associated
packages) for carrying out your analysis and exploring the data, e.g. you might want to predict infection rates using regression or
compare levels of recovery rates using statistical tests. This might also be an iterative process whereby you perform some analysis
and then gather (or remove) more data. Where possible relate you analysis to the relevant literature. This relates to the ‘exploring
data’ stage in Fig. 3.

Note that this is often an iterative process: as you explore the data you may end up re-designing your research questions, having to
gather more data or having to perform further cleaning as more data quality issues arise. Again, this is all a part of the data discovery
process.


3.4. Write up your findings

Once you have performed analysis on the data and have some results then you need to write up your investigation into a report (this
is the ‘communicate’ stage of Fig. 1). The report should be structured as outlined in Section 4. You will be evaluated on your ability to
plan and undertake data analysis and exploration of the pandemic based on named dataset, your ability to engage with the relevant
literature, your use of R (and appropriate packages) and RStudio to process and analyse the data, and the way in which you
communicate your findings within the report for your given problem/topic.

You should also provide your R code as an appendix and marks will be awarded for your clarity, consistency and way in which you
comment your R code (see, e.g. http://stat405.had.co.nz/r-style.html). The specific style you use is not as important as how well you
comment your code so that someone else can follow what you have done and being consistent in whichever style you adopt.

The minimum requirement to pass is to perform at least one type of data analysis (e.g., clustering, prediction, time-series analysis,
etc.) and include at least two visualisations (e.g., charts, maps, etc.) in the report. To obtain a higher mark and more effectively
communicate your findings, you may decide to use more than one dataset or present more than one type of data analysis and/or use
multiple visualisations. Again, you should also engage as much as possible with the appropriate literature.


4. Report structure
You are required to produce a structured report that includes the sections detailed in Table 1. You must state the word count on the
first page of the report. As there is a word count limit (3,000 words) you should aim to make your writing as concise and informative
as possible. Also note that your work will be assessed taking into account the word limit; therefore, we are not expecting detailed
multiple analyses in the report; rather the emphasis should be on the clarity, accuracy and quality in communicating your findings.
Note that words within tables and appendices are not included in the word count.


Table 1: Required content of the structured report.


Section

Description

Examples of what we will be
looking for and mark allocation

Maximum allocated
marks
Structured
abstract
This should provide a summary of your report in a
structured manner, e.g. objective, methods, results,
conclusions. This is not included in the word count.

• Brief but informative abstract that is
clearly structured.
Required, but 0 marks
Table of
contents
This should include section titles and page numbers.
This is not included in the word count.

• Clearly structured Table of Contents
with use of numbering for sections.
Required, but 0 marks
Introduction and
aim(s)
This section should describe your selected problem or
topic addressed in the report and that forms the focus
for your data analysis. This should include a (brief)
summary of the literature around analysis of COVID-19
data relevant to your selected topic that helps to provide
the background to your chosen topic. You should also
state why you chose this problem/ topic and why you
think it is an important topic to consider in this dataset
(ideally support by the relevant literature)

• Clear statement regarding the overall
goal of your investigation.
• Brief literature review of data and
crime analysis.
• More marks for engagement with the
relevant literature.

10 marks
Methodology This section should describe the process you have
used to gather the data, pre-process and clean the
data, conduct your analyses and visualise the data
(note, you could follow the stages in Fig. 1). This will
include ways in which you gathered, pre-processed,
transformed, and sampled/ filtered the data.

You should try to justify your choices and include
references to relevant literature where appropriate. This
should also include details of the experimental setup,
e.g. which R packages you have used etc. Think of it
like this, if someone else had to replicate your
methodology have you provided enough details (and
clearly enough) for them to reproduce your results.

• Expect to see a clear description of
methodology used in your analyses.
• Clear list of the datasets used (and
links to sources) and variables in the
dataset(s).
• Clear discussion of methods for pre-
processing data (and appropriate use
of R packages).
• More marks for examples of the data.
• More marks for multiple data sources
used.
• More marks for the range of
techniques used, appropriateness,
links to supporting literature etc. (e.g.,
methods for trend prediction, spatial
20 marks
As well as describing the methodology used to generate
your results, you should list all datasets used (e.g., data
covering different regions or time periods). You should
also list any additional external datasets used (e.g.,
shape files or census statistics for LSOA areas).
Describe all datasets used, any pre-processing and
how they were joined together (e.g., over LSOA area
identifiers).

data analysis etc.). Techniques can
include types of visualisation and
references to which R libraries have
been used
• More marks for the detail of the
description provided, e.g., could
include use of group_by(),
aggregate() etc.
• More marks for use of methods to
deal with data quality issues, such as
missing values.
• More marks for discussing use of
appropriate techniques for different
types of data, e.g. categorical data.

Results and
discussion
In this section you should present the results of your
data analysis and exploration (e.g., statistics, maps,
trends, predictions). You should use the results to
address the selected problem by presenting and
discussing tables and charts as appropriate.

You should present your findings in a way that helps the
reader interpret the results. You should focus on
effectively communicating the results of the analysis to
the reader by highlighting the trends or patterns you
have observed during your data analysis.
• More marks for correct use of
statistics and visualisations.
• More marks for packaging results etc.
into tables rather than simply using R
output or command line code.
• More marks for a clear narrative and
structure (e.g., adding sections and
sub-sections and guiding the reader
through the analysis).
• More marks for clearly explaining the
results and graphics used (e.g., use
of legends etc.).
• More marks for using graphics that
convey information (e.g., combine
results) and help identify insights
(e.g., use of log scales to dampen
effects of high values etc.).
• More marks for bringing out insights
rather than leaving the reader to
interpret the findings.
• More marks for not over-interpreting
the results and recognising biases.
• More marks for re-labelling the
variable names in graphs and tables
(rather than using default names).
• More marks for how well the data is
summarised and made accessible for
comparison.

50 marks
Conclusion In this section you should summarise the main findings
of your analysis and lessons learned. You should state
the main message the reader should come away with
from your analysis.

You should also highlight any weaknesses of your
analysis and state what you would do to improve your
analysis if you had more time.

• Summary of the main findings of the
analysis with respect to the original
aim(s) of the investigation.
• More marks for highlighting
limitations/ weaknesses of your
methodology and analysis.
• More marks for a clear set of take-
away messages.

10 marks
R code You should include the full R code as an appendix.

• More marks for well-commented
code.
• More marks for clarity of presentation.
• More marks for consistent style.

5 marks
Presentation The overall presentation of the report will be given a
separate mark, including how well you have presented
your results, clarity of writing and use of literature.
• More marks for use of appropriate
references.
• More marks for clarity of writing
• More marks for use of appropriate
charts and tables and their
presentation quality.

5 marks