1. Introduction
We hope that a shared goal of educators is to help students learn material in ways that are both durable and efficient (Rawson and Dunlosky 2011). One finding that has gained recent attention is the utility of active, effortful retrieval practice in facilitating this type of learning (e.g., Brown et al. 2014, Rowland 2014). Essentially, humans tend to learn best when they are asked to actively generate or recall knowledge for themselves, rather than receiving knowledge through passive techniques such as (re-)reading, highlighting, copying, etc. One interesting consequence of this is that the act of administering a test or an exam to students can itself help them learn the material—a phenomenon sometimes referred to as the “testing effect” (see Rowland 2014 for review and meta-analysis). As Roediger and Karpicke (2006: 181) explain, “testing not only measures knowledge, but also changes it, often greatly improving retention of the tested knowledge.”
How, then, can instructors leverage this effect for better learning in their courses? In this paper, our goals are to (a) provide a framework for both practice and assessment within which students can organically develop active study habits, (b) share resources we have built to help implement such a framework in the linguistics classroom, and (c) provide some examples and evaluation of their success in the context of an introductory phonetics / phonology course.
Our approach combines several pre-existing pedagogical ideas into a novel form, facilitated by a purpose-built piece of software. We use sets of open-ended questions made available to students after each class session, which are then combined into individualized, highly customizable random-sampled exams by our open-source software. While many of the individual components of our approach have been used before, this software in particular is novel in that it allows for much more user-specified cross-categorization of topics and other randomization criteria (see section 4) than is typical of other similar open-source or proprietary applications. In turn, this allows for the other especially novel component of this approach, which is that the ‘exams’ in our courses are quite short (2–6 questions). The actual ‘testing effect’ component comes more from students’ preparation for these randomized exams, where they are encouraged to test themselves using the full range of possible exam questions.
We present these ideas in the specific context of teaching linguistics. Although the strategies we describe are based on general principles of learning and should be applicable to all disciplines, we think there are a number of reasons why it is useful to describe them for linguists. First, linguistics is a field that lends itself to an open-ended, explanatory approach to learning because most ‘real-life’ applications of linguistics involve slower, analytical tasks rather than rapid exact recall. At the same time, this may lead instructors away