Best Practices for Legal Education

It’s time to write final exams again. It’s also time to struggle with what role test-taking speed should play in our assessments.

William Henderson’s Study

As Professor William Henderson’s ground-breaking study demonstrates, test-taking speed – how fast students can read and answer test questions– is often an independent variable when students take in-class timed law school exams.

As Professor Henderson cautions, given the high stakes nature of law school exams, we need to be cognizant of the test-taking speed variable and consciously decide whether it is one we think is important in terms of the substance or skills being assessed. As he notes, test-taking speed may have limited relationship to how lawyers use doctrine and legal analysis and it also may have a discriminatory impact.

Potential Solutions

Some faculty members address the test-taking speed issue by giving take-home exams or papers. Others do not want to give take home exams for a host of reasons, including the belief that the material does not lend itself to a take-home exam or concerns about policing academic honesty.   For those of us giving timed, in-class exams, are there ways to decrease the impact of test taking speed? Below I share a few ideas. I also invite those of you who have grappled with this issue to share what you are doing.

A. Pre-release exam instructions

One way to help students at least prepare for time constraints is to release exam instructions ahead of time. In many classes, the exam  instructions identify how many questions and provide suggested time allocations. This information, shared in advance, can help students plan their time before they take the test.

B. Pre-release a couple of questions

In addition to pre-releasing instructions, 24 hours before an exam, I pre-release a couple of short answer questions [worth 10-15% of the total raw score points]. This allows students to prepare answers to those questions ahead of time.  For those questions, I minimize the test-taking speed issue.

I allow collaboration on the pre-released questions so I do not need to police students. I also warn students about the dangers of collaboration and letting others lead you down the wrong path.

C. Time yourself

I also take the test myself, timing how long it takes me to answer a question or set of questions. I double or triple the time allotted based upon how long it took me, or in some cases how long it took a colleague who gave my exam a test run. For example, if it takes a professor about two minutes per question for a set of multiple choice questions, I allot four to five minutes per question for my students.

Some may argue that allowing five minutes per multiple choice question does not prepare students for the bar exam – an exam in which students have under two minutes to answer often fairly complex multiple choice questions. That point is correct and it raises the interesting question of whether one’s course grade should be capturing bar exam taking skills.  It also raises the bigger question about whether test-taking speed is a variable that we, and bar examiners, should be assessing.

Reliability and Validity concerns

Significant time to answer a question necessarily means a test with fewer questions. For some, this prompts reliability and validity issues [fewer questions may reduce the reliability and thus the validity of an exam]. On the other hand,  test-taking speed as an independent variable also raises questions about test validity.

Making Sure We Understand the Issues

These are complicated issues and different faculty may have different answers to the test-taking speed questions. The key is not necessarily how we answer these questions – it is that we are asking ourselves the questions as we draft our exams.

Filed under: Uncategorized |