Question 1

In the recent COVID-19 pandemic, many disruptions to various aspects of our daily life have been created and show up visibly in many areas when data analytics is applied. You are to identify and investigate one such area that exemplifies the disruptions caused by the virus. Some examples include the economy, society, climate, energy consumption, internet usage, and so on. In this ECA, you are required to formulate a research question and find two or more publicly available datasets that can be used to address the question. Apply the techniques of data processing, visualization and analytics to support your claims and conclusions.

Marks will be awarded according to the following parts:

(a) Write an abstract to describe a project which applies Python programing to analyse some data in the given context. Use up to formulate the research question, and provide the scope and depth of the question.

The research question should require a reasonable breadth and depth of investigation that necessitates the use of multiple datasets (at least two) and various perspectives of the data in order to be addressed adequately. For example, a suitable research question could be: Can the COVID-19 pandemic lead to the downfall of retail REITs in Singapore?

To address this question, you can use datasets that may include consumer spending in malls and online retail, profitability/annual reports of shops/malls, tenant occupancy rates and rents of shop spaces, number of persons employed in the retail sector and REIT stock prices, etc. You could then address the research question from different perspectives provided by the trends and observations generated from the various datasets.

(Refer to Appendix 1 for examples suitable for this assignment.)

(b) Provide a description of the datasets, and include details such as data quality analysis and data preparation.

A description of what the datasets contain; how and why they are employed to address the research questions. They should be inspected for data quality issues such as missing values, errors, duplicate values, outliers or extreme values, etc. As different datasets could come in different formats, preparation and pre-processing work such as data cleaning, imputation, transformation and/or merging operations should also be presented. As far as possible, data preparation should be done using Python programming. However, if you face a lot of difficulty in using Python, you can also prepare the data by other means (e.g., using MS Excel), but make sure you give your reasons.

(Refer to Appendix 2 for some possible sources of datasets, but not limited to those sources.)

(c) Perform data visualization and analyses.

For analyses of the datasets, summary statistics and visualizations should be generated to illustrate trends, patterns or anomalies to provide an answer to the research question. Logical and thoughtful analyses linking the observations to explanations that address the research question should then be provided. The discussion should lead to sound predictions or practical recommendations. Extra marks are awarded for elucidating important insights that are not obvious or counter-intuitive but have far-reaching consequences.

(d) Provide Conclusions and References.

The main findings are to be summarized clearly leading to important conclusions of the investigation. The research question is then addressed with final concluding statements. The report makes use of a good range of references and is cited properly in a recognized format.

(e) Provide Python code in Jupyter notebook.

Develop your code in "eca.ipynb" and submit it along with the datasets used in a single zip folder. The program should also have sufficient comments to describe the code steps, explain any functions used, and analyse any control flow logics. Include your full name and PI number at the start of your code.

Present your work for parts (a)-(d) above in your report using the template provided (Appendix 3).

Appendix 1 Proposing Research Questions

Example 1: Does the COVID-19 pandemic increase unemployment in Singapore?

Too Narrow. A single summary statistic on unemployment rate will answer the question.

Example 2: How will the COVID-19 pandemic affect globalisation?

Too Broad. The range of data required for a comprehensive analysis is too wide. There are too many countries to analyze for the scope of this assignment.

Example 3: Can the COVID-19 pandemic lead to the downfall of retail REITs in Singapore?"

Suitable for this assignment. To address this, the multiple datasets that could be analyzed include consumer spending in malls and online retail, profitability/annual reports of shops/malls, tenant occupancy rates and rents of shop spaces, number of persons employed in the retail sector and REIT stock prices etc. One could then address the research question from different perspectives provided by the trends and observations generated from the various datasets.

Appendix 2 Sources to look for datasets

Singapore's Public Data, https://data.gov.sg/

Smart Nation Singapore, https://www.smartnation.sg/resources/open-data

The United Nations, http://data.un.org/

The World Bank, http://data.worldbank.org/

The Global Open Data Index http://index.okfn.org/place/

US Government Data, https://www.data.gov/

UK Government Data, https://data.gov.uk/

Canada's Open Data Exchange, https://codx.ca/

World Health Organization, http://www.who.int/gho/en/

Wikipedia, https://en.wikipedia.org/wiki/Main_Page

Appendix 3 Report Template

Title

Abstract - Briefly describe the datasets used, research question, methods, and main findings.

1. Research Question

Document your work for part (a) here.

2. Data Preparation and Pre-Processing

Document your work for part (b) here.

Provide screenshots of the relevant Python code and its output wherever appropriate.

3. Data Visualization and Analyses

Document your work for part (c) here.

Provide screenshots of the relevant visualizations to aid your explanations wherever appropriate. For ease of referencing, insert the figures/charts directly before/after the explanatory paragraphs. Do not consolidate your figures/charts in the Appendix.

4. Conclusions

Document your work for part (d) here.

Summarize your findings and identify possible limitations.

References