As we conclude this election cycle, pundits and clients alike are questioning what went wrong with the polls. While there are some possible explanations, though perhaps the final error may be on par with historical performance, this scrutiny of polling and results illustrate three best practices we should adopt so that we are able to better design, analyze, and present primary research to clients and stakeholders.


First, we need to define our sample audience to best match the population of interest.

Election polls take a sample of likely, registered, and non-voters to extrapolate the vote choice of those most likely…

Even before the Covid-19 pandemic shifted the majority of our social lives online, social media has been the arena where many Americans expressed their political activism, advocacy, and anger. But does social media politicking translate into the ultimate political outcome of voting? My analysis of data from the 2018 Cooperative Congressional Election Study (CCES) reveals that many Americans engage in cheap talk, following, sharing, and discussing politics on social media, but then do not vote. …

Though a small portion of the electorate, Asian-Americans could play a critical role in the 2020 presidential election. According to the Pew Research Center, Asian-Americans make up 5% of the eligible voting population and are the fastest growing ethnic or racial group of voters. However, Asian-Americans are historically politically apathetic, with low interest in politics, political participation, and turnout. But as the findings below show, Asian-Americans have significant potential political power if they can overcome historical apathy and could significantly impact the outcome of 10 states with a total of 123 electoral votes.

1. Nearly 7 in 10 Asian-Americans don’t…

Smart use of comparison, conversion, and contextualization when sharing research findings are essential to effective data storytelling. Instead of listing results like a lab report or a phone book of numbers, using principles of shock and awe data storytelling help our data come to live. Doing so helps us be more effective in explaining data findings to diverse audiences, help our audiences better understand the scale and scope of our findings, and highlight actionable imperatives and implications.

1. Contextualize large numbers by comparing to known places.

When reporting aggregate counts of people, events, engagements, or likes, totals often sum in the millions. To contextualize this, use references to known…

Dueling, as we are reminded in the cultural phenomenon that is Hamilton, was a critical part of America’s colonial and early national society. As we are reminded in song 15 of Act 1, even this most uncivilized way to resolve conflict had rules. Below is my re-imagining of these commandments when we do data analyses. While we may not always get to be in the room where the analytics happens, we can be satisfied knowing that our right hand man and obedient servants will have these rules on our side when we tell our story.

Photo credit

In the tune of Ten…

When we do research, we are almost always working in data deficit environment. To do reasonable research, we need to balance we can say with the data we have with what we cannot say because we lack better data.

photo credit

In an ideal world, there would be perfect match between research questions and available data to conduct empirical analyses. In the real world, we simply do what we can with best available data. Optimizing the confidence with which we answer the questions the data lets us answer with proper recognition of limitations in our analysis is critical. …

While police initiate contact with Black and White Americans at equal rates, Black Americans are over twice as likely to be arrested, over twice as likely to be threatened or experience use of force, and nearly three times more likely to be killed by police. Black Americans are also nearly six times more likely to be incarcerated after a police encounter.

Image source

In previous posts, I discussed how Black Americans are arrested and incarcerated at higher rates than White Americans. One potential counterargument is that Black Americans more likely to be arrested and incarcerated because they are more likely to commit…

In my last post, I show Black Americans were over twice as likely to be arrested than White Americans. But racial disparities persist beyond arrest to incarceration.

Image sourced here

Combining data from the FBI Uniform Crime Reporting program with the Bureau of Justice Statistics National Prisoners Statistics program, I find arrested Black Americans are nearly three times as likely to be incarcerated than white Americans. And as a percent of population, Black Americans have an incarceration rate more than five times higher than that of White Americans.

While reasonable people can disagree about the causes and remedies to these disparities, the data…

Is recent violence against George Floyd, Ahmad Arbery, and Breonna Taylor an outlier amplified by social media or emblematic of systematic racism?

Picture sourced here

This question should not only be answered with polemics, however passionate, but smart analysis of best available data. Bringing empirical analysis into the public square is essential for reasonable discourse; it guards against recency bias, flawed extrapolation from anecdotal evidence, and provides perspective to put individual events in national and historical context.

To that end, I analyze data from the 2018 FBI Uniform Crime Reporting program (the most recent year for which data is publicly available) and identify…

Excel is the American cheese of the analytics world.

Like American cheese, Excel is underappreciated and derided as an inferior product to more sophisticated data management and statistical analyses solutions. Excel is a limited software platform that should not be used for purposes for which it is not well suited. But it is useful and can play important roles in your analytics workflow. Here are the three ways I incorporate Excel in my day-to-day statistical analyses.

Store and share results from statistical software. Excel is a useful resource to store results of statistical analyses from your data science platform. Sharing…

Joshua Wu, PhD

Market research, analytics, PhD in Political Science.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store