(Problem Statement as given to the students)

Expectations

First, watch this talk for inspiration – https://www.youtube.com/watch?v=hVimVzgtD6w

Also, poke around the videos and case studies linked here — https://www.callingbullshit.org/.

Can also check https://www.data-wto-viz.com/ for exploring different types of graphs suitable for the data.

You will start from a (mostly) raw dataset, it might have various imperfections e.g. missing data. Also, you might get multiple datasets representing different aspects of a sector. Your task is to write a data-driven summary that draws certain conclusions and presents arguments for the conclusions either visually or via summarizing tables or statistical tests. You will be judged on the kind of insights, how well you present your argument (e.g. visual illustrations, any number crunching e.g. correlations derived, statistical tests used to support your argument, etc). Note that this will be mentored by the Writing Studio, so will be an iterative process.

Length – roughly 1000 words (excluding code).

Examples of such reports (much longer) – https://www.education.gov.in/sites/upload_files/mhrd/files/statistics-new/udise_201920.pdf https://www.education.gov.in/sites/upload_files/mhrd/files/statistics-new/aishe_eng.pdf

Analysis Themes

The following are the analysis themes to choose from.

  • Understanding our campus life
    • Food Wastage Dataset 2019-2020 Dataset Link: (Dataset Link removed for Confidentiality)
      The dataset is from Jaiswal Dining Hall for the year 2019-2020. For each day, there are four values, the amount of food (in kgs) wasted in Breakfast, Lunch, Snacks and Dinner. You can choose a specific time period for evaluation (whole dataset, one year, few months, etc.) This dataset can be combined with the monthly mess menu and different events in the calendar (summer holidays, exams, etc.) to get insights about food habits. Further suggestions can be made for strategies/insights from the data to reduce food wastage.
    • Campus energy consumption data – 2020+2021 (monthly aggregate).
      For reference and possible use, here (Dataset Link removed for Confidentiality) is the energy consumption data for all Indian states.

  • Telling the story of India
    A lot of official datasets are available at – Open Government Data (OGD) Platform India You need to decide on a sector and choose a dataset from there. As an example, for the education sector, you can navigate to the following page for the datasets.
    https://data.gov.in/dataset-group-name/school-education-statistics

  • Charting the growth of AI in different countries
    Can we take the top ML conferences in the last few years, get the author institutions and countries, and summarize the leaders / rising institutions/trends? E.g. this report – The race to the top among the world’s leaders in artificial intelligence
    Dataset for NeurIPS – All NeurIPS (NIPS) Papers | Kaggle Other conferences – https://github.com/martenlienen/icml-nips-iclr-dataset

  • World inequality and how it has changed

  • World inequality dataset
    https://wid.world/

  • How are different countries developing?
    Various country-level statistics collected by UN. You can ask questions e.g. how well are the different countries doing on the various developmental goals and how are they making progress. https://unstats.un.org/home/ https://unstats-undesa.opendata.arcgis.com/

  • Looking at Primary Education across multiple countries
    PISA is a well-known test for measuring how well a particular country is doing in providing quality education. There are data available across years at – https://www.oecd.org/pisa/data/ You can talk about how the indicators for specific countries have changed over time. Participating countries are available at https://www.oecd.org/pisa/aboutpisa/pisa-participants.htm

  • How are vaccination drives going + the inequalities underneath
    Dataset on worldwide covid19 vaccinations https://www.nature.com/articles/s41562-021-01122-8 A broader dataset for covid-19 (for current vaccination theme) can be found at: https://github.com/owid/covid-19-data/tree/master/public/data https://www.kaggle.com/datasets/gpreda/covid-world-vaccination-progress

  • Effects of Economy, natural resources distribution, government policies on CO2 emissions by countries
    CO2 emissions dataset: https://github.com/owid/co2-data

  • Prevalence and reasons of childhood stunting in India (malnutrition, child diseases, sanitation, etc.)
    World Bank datasets: https://data.worldbank.org/indicator/SH.STA.STNT.ZS?locations=IN, https://data.worldbank.org/indicator/SH.STA.ORTH?locations=IN WHO datasets: https://www.who.int/data/gho/data/themes/theme-details/GHO/gho-nutrition

  • Neglected Tropical Diseases in Africa and their eradication
    WHO datasets: https://www.who.int/data/gho/data/themes/neglected-tropical-diseases
  • Maternal and Reproductive Health
    WHO datasets: https://www.who.int/data/gho/data/themes/maternal-and-reproductive-health
  • Road Safety
    WHO datasets: https://www.who.int/data/gho/data/themes/road-safety