Skip to main content

Colloquy Podcast: Testing and the Origin of Big Data

You’re being tested. You don’t know the criteria used to determine your score—or even your results. The test is being administered not by a human teacher or moderator, but by machines. And it’s going on 24 hours a day, every day of your life. Harvard Griffin GSAS historian Juhee Kang traces the emergence of the obsession with mass data collection in the early 20th century.

This transcript has been edited for clarity and correctness.

In April 1920 at the Ministry of Education in Tokyo, Japan, a group of high officials asked psychologist Ueno Yoichi: What is a test? Ueno, who had just come back from the United States and Europe to learn the latest in human resource management, replied: That's a difficult question. If anything, a test is a scientific experiment. This was the moment when tests emerged as a solution to a pressing social problem. 

As a historian, my favorite part of research is illuminating how people of the past tackled problems––problems that are similar to those we face today. And I found the problems of 1920s Japan all too familiar. Here, the Japanese government was wrestling with the problem of how to distribute resources in a fair and efficient way. By the 1920s, Japan had become a multiracial empire with more than six ethno-races under its influence. 

This meant that people who differed not only in skin color but also in languages, customs, and ways of living now had to live side by side and follow the same rules––rules that they would learn at school. Starting in the 1890s, the empire made four years of education mandatory and free for all its subjects. Less than half a century later, however, it needed another system. 

As rapid industrialization drew people to its urban centers, the changing demographic disrupted the existing order. To maintain control, the imperial state first had to ensure that all its subjects knew the rules of being a Japanese citizen and, second, enable them to be both happy and useful. Scientists proposed a test to address this challenge. 

As a technology to ascertain whether a person possessed certain qualities, tests were considered scientific experiments. And by being a scientific experiment, they were also believed to be fair and efficient. Fair because if you had received the mandatory education, regardless of your background, you now had the knowledge to be tested; efficient because your test result would place you in relation to millions of other Japanese subjects who had also taken the test.

And in this experiment, the hypothesis was that you are a good Japanese Imperial citizen. And then the variable was your intelligence. When I started my doctoral research in 2018, I went on to collect historical artifacts that allowed me to explore my research questions. How did people outside the West define and measure human capabilities? And how did those measurements change as the world became more interconnected? 

These questions led me to seven archives in Japan, South Korea, and Germany, where I collected historical tests. I also read a lot of historical newspapers, magazines, laws, congressional meeting notes, and scientific articles to contextualize how those tests came about and were implemented. Let me give you an example of what my research discovered. This is one of the more unique tests that I found. This test works like this. 

First, you will receive a piece of paper like this. Starting from the top, you would add two adjacent numbers and note the sum in the units place. So here we have five and seven. What do we get? Thank you. 12. So you just dropped one and noted the two. Continue this process for a minute, then move on to the next row. After 15 rounds, you would take a 5-minute break and then resume for another 10. 

When you are done, the administrator will connect the last numbers to draw a curve. Different parts of this curve would tell your general intelligence propensities to develop certain mental conditions and personality at work. This test had three components as a scientific experiment: first, numbers and simple math. Numbers were considered universally understandable and required no linguistic or cultural pre-knowledge. The performance of simple math does function as a unit of measurement. 

Second, the mechanics of the test and the curve. The stages of repeat, rest, and repeat were to mimic the process we require new knowledge. In this way, the shape of the curve was to visualize our learning pattern. And finally, the tedious procedure. Scientists who invented this test knew that repetition is boring. 

But our modern life, once industry developed in a relatively safe nation-state, was also thought to be just full of repetition. So the artificially induced state of boredom stimulated the human condition in modern life. And one's intelligence was then measured in their ability to learn and improve while enduring that boredom. In 1926, this test, although briefly, did replace the standardized aptitude test to enter some of the best schools in the empire. 

Now, as a historian, I'm not here to argue how we should handle our problems. My work is to tell the stories. History offers an array of comparable stories where people, under all kinds of different conditions, tackle problems––problems that we still and will forever face. So as our society grapples with questions like how to distribute resources in a fair and efficient manner, or how to judge a person's merit, we will continue to test things out. After all, as the Japanese psychologist said, all tests are experiments. And even when experiments fail, we learn something for the better. 

Harvard Griffin GSAS Newsletter and Podcast

Get the Latest Updates

Subscribe to Colloquy Podcast

Conversations with scholars and thinkers from Harvard's PhD community
Apple Podcasts Spotify
Simplecast Stitcher

Connect with us