top of page

Making standardised testing the standard practice

federick j kelly standardised

Who is Frederick J. Kelly? What "test" is too crude to be used? Why should I care?

In 1914, the population of the United States was growing rapidly. The country had also just introduced at least 2 years of compulsory high school education. As a result, the number of students per cohort in the United States swelled from 200,000 in 1890 to 1.5 million in 1914.


Concurrently, there was a shortage of teachers. The training of new teachers was not as fast as the growth in students, and... World War 1 was ongoing - some teachers were gone, joining in the war in some form or another. 

More students; insufficient teachers - there was hence an urgent need to examine how scarce teaching resources can be distributed to serve the students. 

Enter statistician Frederick J. Kelly who came up with his version of the standardised test — the “item-response” model. Doesn’t ring a bell? Today, it is known as the multiple-choice test.


Kelly did not expect that his item-response test would soon be used as widely as it eventually did. He initially thought it would be an easy way to assess how much students knew about a topic. But, it soon adopted a different purpose - it solved the most pressing problem of the day. Kelly's standardised test was a quick and efficient way to grade and categorise students into different bands. It can then be decided how much of the limited teaching resource to allocate for each of students.   


Kelly started to champion a different direction. First, he explained that his test was intended for “lower-order” processing, i.e. simpler questions. Later, as president of University of Idaho, he rejected standardised-testing, pushing for liberal, integrated, and problem-based learning, claiming that: “college is a place to learn how to educate oneself rather than a place in which to be educated.”

Here’s the irony. The creator of a testing metric suggested changes to how his metric should be used. But the only thing that was changed was him. Because of his views, Kelly was asked to step down as President after just 2 years, while the test he created, which so well fitted the narrative of education, became increasingly widespread. Today, multiple-choice questions feature heavily at all levels of education; you and I have probably taken many a multiple-choice test. Indeed they are a critical component for entry into all universities in the world, including the very best.

The history of Frederick J. Kelly and the standardised test reinforce something that has been repeatedly emphasised across these pages:

- Stories are the way we explain and understand the world. For example, most of us understand the story of standardised tests as a "standard practice" in education. 

- Before we formulate our stories, we are naturally influenced by many factors. But once we have made up our minds about what the story is, it becomes very difficult to change.

- Even if the original writer of the story comes out to explain: hey, you guys got the story wrong, that's not how I intended the story to be! Even then, it hardly matters. We believe the story more than the author who wrote it. 

Related links:

Planck's principle: Change occurs one funeral at a time. 

How far must one go to prove his case? Barry Marshall and stomach ulcers

How can we actually change the views of others?

Why does evidence not change our minds?

What effect dos evidence have in changing views on climate change?

bottom of page