Python Tools for Management Research

3c: Presentations and Other Research Outputs

Jason T. Kiley

From manuscript to research outputs

Quarto is one tool for many kinds of output we use in research, teaching, and life in general.

We can get a lot of mileage from a relatively small set of tools.

More outputs

More mileage from the same skillset

Quarto is versatile.

Manuscripts

Write prose, citations, tables, figures, and code-generated results in one source file.

Presentations

Turn the same project into slides for workshops, conferences, and coauthor meetings.

Websites

Share project pages, documentation, teaching materials, and supplementary resources.

Why slides from the project?

Presentations have the same copy/paste risks as manuscripts:

  • sample sizes drift;
  • figures get updated in one place but not another;
  • regression results change after the slide deck is built;
  • tables and text quietly stop matching.

If the talk draws from the project, we reduce those failure points.

Quarto presentations

Quarto revealjs slides are still .qmd files.

---
title: "A Short Research Talk"
format: revealjs
---

## Research question

Why do some firms attract more stakeholder attention?

Headings become slides, and the same Markdown basics still work.

Reuse a result

The sample includes `{python} n` firm-year observations.

Attention is associated with performance
(b = `{python} round(attention_coef, 2)`,
p = `{python} f"{attention_p:.3f}"`).

This is the same idea from the paper: values come from the analysis, not from retyping.

Reuse a figure

axes = scatter_matrix(results_data[analysis_vars])
fig = axes[0, 0].figure
fig

A figure can appear in the paper and in the slides without saving and inserting it by hand.

Other outputs

The same project can grow into other useful artifacts:

  • a project website;
  • an appendix or supplement;
  • a polished PDF;
  • a Word document for coauthor track changes;
  • a slide deck for a research talk.

Easy things easy. Hard things possible.

Hands-on

Open the 3c activity page.

Course summary

What we learned: foundations

Python

Names, objects, functions, control flow, notebooks, and enough syntax to read and adapt useful code.

Data

Polars for reading, inspecting, transforming, joining, aggregating, and saving research datasets.

Projects

Git, GitHub, devcontainers, Codespaces, and pyproject.toml for more reproducible research work.

What we learned: outputs

Documents

Quarto manuscripts with citations, references, HTML, PDF, and Word output.

Results

Generated sample sizes, tables, figures, regression results, and in-text values.

Outputs

Slides and other research products that can draw from the same project.

Start small

You do not have to change your whole workflow at once.

Start with one part:

  • clean one dataset in Python;
  • generate one table from code;
  • put one paper draft in Quarto;
  • use Git for one project.

Build toward a more integrated workflow over time. It took me years to figure out what I thought the best combination was for me, and I did it incrementally.

Tools can meet at the boundaries

Reading and writing data gives you a lot of flexibility.

spreadsheet -> Python -> CSV or Parquet -> Stata, R, Python, or Quarto

That means you can adopt one tool where it helps and get comfortable before adding more.

Watch for updates

Many of these tools are under active development:

  • Python data tools are getting faster and more capable;
  • Quarto keeps adding output features;
  • VS Code and Codespaces keep improving;
  • package ecosystems keep filling practical gaps.

Staying current helps get more mileage from the tools. I like to star things on GitHub and occasionally look at that feed.

Join the community

There is now a solid network of computationally minded researchers in our field and adjacent fields.

The AOM Content Analysis PDW is one good outlet, both for presentations and for proposal expert roundtables.

It’s nice to talk shop or to simply hang out and not talk shop with like-minded folks.

Keep in touch

I am always interested to hear what you build, what works, and what gets in your way.

Thank you for spending the week learning these tools together.

Thank you!