Project Brief: Analyse a spatial network

Assignment description

You need to identify and analyse a spatial network you are interested in. This involves:

  • Selecting a spatial network, which is of interest for you; finding, using and analysing relevant network data – more on this below.

  • Setting up research questions. You will need to set up clear research questions that your analysis will answer. Although I don’t expect a full literature review section in your report, you need to link these research questions to some previous literature.

  • Analysing your network. You will need to employ relevant methods and tools that we discussed during the course (from data acquisition to network analysis and from diversity measures to spatial interaction models) to answer your research questions. The methods you will use should be appropriate for the research questions you need to answer. I will encourage you (and give extra points) if you utilise relevant methods beyond those discussed during the classes. Your methodological choices should be justified in the text.

  • Writing up the report, which needs to be adequately structured (i.e. introduce the problem, link it to some previous literature, describe the methods and the data, present and discuss the results, indicate some key outcomes). The report should be written as an RMarkdown document and knitted as an .html document that allows the reader to see both the text, the code you will develop and the output of this code. Your report should include tables and figures where appropriate – these form an important element of the assessment. Be critical to what outputs will be included in the report. The report needs to be appropriately referenced.

Data sets

In order to address the different interests the cohort might have, the data choice is up to you. I offer below a few ideas, but if you want to analyse other data, please get in touch with me directly to discuss. You will need to do this early on during the term, so please start thinking about the report early.

  • US internal migration flows. You can use data from the American Community Survey (ACS). There are quite a few R packages, which enable you to directly access these data through an API.

  • Africa migration flows from WorldPop.

  • Data for Good by Facebook. There are quite a few interesting datasets there. Some of them are ready to download, some require an application. If you are interested in the latter, apply sooner than later.

  • Internet submarine cables

  • NYC Taxi trips

  • Flight data

Word limit and marking criteria

The report should be 3,500 long excluding references and code. It is worth 100% of the unit’s mark and on top of the programme marking criteria, I will also consider:

  • The reproducibility of the analysis.

  • The quality of visualisations and tables and the appearance of the knitted .html document.

  • The relevance and the novelty of setting up specific research questions.

  • The use of relevant tools beyond the ones directly demonstrated during this unit.

  • The robustness of the analysis and modelling.

  • The complexity of the data that you will use.