Teaching Statistics Using Wild, Local and Fresh Data

Damjan Vukcevic & Dianne Cook

ICOTS 12 (Brisbane, Australia)

17 July 2026

Setting the scene

Context

Monash Business School
Business Analytics (undergrad & postgrad)

Meet (some of) the team…

Di Cook	Rob Hyndman	Damjan Vukcevic	Kate Saunders	Michael Lydeamore	Shanika Wickramasuriya
Jack Jewson	Erik Sverdrup	Ole Maneesoonthorn	Jess Leung	Klaus Ackermann	Patrick Wong

Program structure

Data analysis

Statistical thinking

Modelling / Prediction

Data examples

Cricket – Women’s Indian Premier League

Australian tourism

Road traffic and flooding

Melbourne’s Urban Forest

Wild

Real data are never delivered in a clean CSV file.

Data “in the wild”:

Inconsistent formatting
Missing values
Unexpected anomalies
Need to be collected and cleaned

Pedagogical goals

Develop skills to work with real data
Develop confidence to tackle real questions

Local

Data about our surroundings and things that impact us.

Examples:

Local transport/traffic info
Popular sports
Regional economic indicators.

Pedagogical goals

Relevance \Rightarrow engagement

Fresh

Current and timely.

Ideally: live or regularly updated.

Pedagogical goals

Relevance \Rightarrow engagement
Show that our teaching is relevant to contemporary questions

Walktober

Electricity usage

GUESTimation

Packaged up data

🐾 rwalkr

Our program

Program structure (undergrad)

flowchart TB
    node1["Data analysis (ETC1010)"]
    node2["Statistical thinking (ETC2420)"]
    node3["Predictive analytics (ETC3250)"]
    node9["More advanced units..."]

    node1 --> node2
    node2 --> node3
    node3 --> node9
    node2 --> node9

    classDef mint fill:#A2E2C6,stroke:#333,stroke-width:1px,color:#000;
    classDef periwinkle fill:#C8D3FC,stroke:#333,stroke-width:1px,color:#000;

    class node1,node2,node3 mint;
    class node9 periwinkle;

Program structure (postgrad)

flowchart TB
    node1["Data analysis (ETC5510)"]
    node2["Statistical thinking (ETC5242)"]
    node3["Predictive analytics (ETC5250)"]
    node4["Wild data (ETC5512)"]
    node5["Collaborate reproducibly (ETC5513)"]
    node6["Deeper with data (ETC5521)"]
    node7["Communicate with data (ETC5523)"]
    node8["Forecasting (ETC5550)"]
    node9["More advanced units..."]

    node1 --> node2
    node2 --> node3
    node3 --> node9
    node2 --> node8
    node8 --> node9
    node2 --> node9
    node1 --> node6
    node6 ---> node9
    node4 ----> node9
    node5 --> node7
    node7 ---> node9

    classDef mint fill:#A2E2C6,stroke:#333,stroke-width:1px,color:#000;
    classDef sage fill:#C9DBAA,stroke:#333,stroke-width:1px,color:#000;
    classDef periwinkle fill:#C8D3FC,stroke:#333,stroke-width:1px,color:#000;

    class node1,node2,node3 mint;
    class node4,node5,node6,node7,node8 sage;
    class node9 periwinkle;

Reproducibility

The ingredients:

Open-source tools
Documented workflows
Version control
Collaboration

Modern toolkit

Student perspectives

Nelson Lee:

“You’ll love the program if you’re really into data. Not because of the hype, money or skills, but genuine love for the data.”

Stephanie Kobakian:

“The technical skills I’ve developed at Monash make me feel really confident to step in to the workforce and really tackle a problem right from the start to finish.”

Key ideas

Data first
Wild, local and fresh
Reproducibility is multi-faceted

Questions?

Resources

Slides made using Quarto.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Appendix

Staff culture

Open-source collaboration
Specialised teaching tools (e.g., moodlequiz)
Lead by example: use the Git/Quarto workflows we expect from students.

Wild-caught data (ETC5512)

Concepts & skills

Open data: definitions, sources and examples
Data collection methods (incl. web scraping)
Data ethics & privacy
LLMs for data analysis
Maintenance of wild-caught data

Case studies

Natural hazards and impact warnings
Australian census
Australian election data
Credit risk and probability of default (Fannie Mae loan data)