Article Guide

By: Sahana Srinivasan and Alison Hu

In this tutorial, you'll learn the types of articles you can write for HODP, the goal of our work, and what our publishing process looks like. We'll also learn general concepts about what data journalism is and how to write a data-driven article.

At the end, you'll be a data journalist, ready to publish your work with HODP!

Types of Articles

1. Traditional Analysis

Most of our articles! Ask a question about Harvard or its students, find data that could illuminate your questions, analyze the data rigorously, visualize it, and publish it!

2. Survey Write-up

HODP has its own survey group, and we occasionally send surveys to the student body at large. After working with our Survey Group PM to craft and send your survey, you'll analyze your results in an article. Examples include our pieces on the 2020 Presidential Primaries and on students' COVID-19 situations. You can write up your results in a traditional text-then-graph-then-text model, as the projects above are, or you can create an interactive, scrollytelling story, as described below.

3. Scrollytelling

Scrollytelling stories look like this piece from NYT on how the Iowa caucuses work. Either static graphs or interactive visualizations will occupy the bulk of the screen. As you scroll, these graphics will change or be replaced, accompanied by text blurbs that scroll up the screen as you do.

You can use our scrollytelling template to develop a project like this. You'll have to plug in your scrollable text and graphics/React components. React knowledge will be helpful (and d3 is necessary if you want to make it interactive), but if you don't know React, reach out to Sahana and she can help you out!

Check our our Fall 2020 bootcamp presentation on how to write an effective data journalism article.

How to Write a HODP Article

What's supposed to go in your article? Usually graphs and cool insights. Regardless of the type of article, we recommend including the following information. We strongly encourage coming up with your own section titles and reorganizing this info (particularly data/methods) based on what makes sense for your own piece.

Introduction and the Question

What questions are you trying to answer? Why did you send out this survey? What interesting info might this dataset contain? Are you trying to investigate or shed light on some campus phenomenon?

This section should make clear what you were trying to find out via this project. It should give the reader an idea for what they're in for, and why it's important or relevant to them.

The Data and Your Methods

Briefly, what's your data and where did you get it from? If your data collection methods were somewhat involved, briefly described how you acquired your data. If there was something special about your methods (e.g. you employed some complex ML, you scraped something live), talk about it.

We don't want this section to be so long it distracts from getting to the insights. But we want to briefly describe the data so readers can trust your analysis. And if your methods are important to or somehow qualify the validity of your insights, you should acknowledge that upfront.

If you have a whole lot of info for this section, summarize it quickly and move the rest to an appendix.

The Insights

Start answering the questions you initially proposed by presenting the results of your analysis (statistical tests, averages, garphs, trends) both visually and numerically.

Explain what each result means, rather than stating a number or the result of a test. Explain and caption graphs, and if explain (briefly) how you got results, if it's something more involved than aggregation or summary statistics.

Feel free to suggeest hypotheses behind your results, but don't speculate too much, and make sure everything you say is related to and supported by data. If you're speculating, say so.

Be sure to note any unexpected results and key trends.

This section should be the bulk of your article.

Unsanswered Questions

This can happen in the insights section or at the end of it. It's ok to talk about how the data doesn't answer all your questions, or how your results are inconclusive.

Conclusion, Significance, Future Work

So what was the point of this project? Overall, what did the data show us, and where does it leave us with respect to your original question? Is there any further work you might want to do? How was your work interesting or important?

Overall, this section should synthesize your key results and connect them back to what you initially set out to learn.

Publishing with HODP

See the style guide for information on what to turn in to get published!