Retrieval Practice: Multiple-choice or short answer questions?
Saturday morning homework with the preteen – the part of the week when time normally standstill. A couple of weeks ago, biology homework involved her completing two quizzes. The first quiz was a multiple-choice quiz; the second was a short answer quiz. Both were based on content from the previous lesson. So, how did it go? My preteen flies through the multiple-choice quiz, getting all the questions correct on her first attempt. Next up, the short answer quiz. Despite covering similar content, she finds it a lot more challenging and finishes with a score of 63%.
After the homework was complete, we discussed the two approaches to quizzing. My preteen told me she preferred the multiple-choice questions because she found them easier to complete. Why? Because in many cases, she quickly identified the correct answer as seeing it triggered a memory response. When she did not immediately know the correct answer to a multiple-choice question, she worked through a process of elimination to find the correct answer. She also correctly guessed a couple of solutions. However, the short answer questions she found more challenging because she could not always recall the correct answer without a prompt being available. This is likely because she had not yet fully learned what had been taught in class. At this point, I will also add, that after looking at the multiple-choice questions, the distractors were not always plausible!
What surprised me was the difference in my daughter’s performance using multiple choice and short answer questions. Had my daughter not completed the short answer questions, both she and I might have come to the false conclusion that she knew her stuff about the topic being tested. Seeing my daughter using the two quiz formats for the first time led me to stop and reflect on how quizzing is used in learning.
A range of research suggests low-stakes testing/quizzing is beneficial to learners. There is no denying that frequent quizzing can reduce test anxiety; the work of Smith, Floerke, and Thomas in 2016 illustrates this. Students who receive an intervening test after the initial learning experience generally perform better on a later final test than subjects given only the final test. This phenomenon has come to be referred to as the testing effect. It has been demonstrated with diverse study stimuli, including word lists (Darley & Murdock, 1971), paired associates (Runquist, 1986), pictures (Wheeler & Roediger, 1992), general knowledge facts (McDaniel & Fisher, 1991), and prose passages (LaPorte & Voss, 1975; for a review, see Roediger & Karpicke, 2006a).
However, how often do we reflect on the format low-stakes quiz/test questions should take? Multiple choice and short answer formats are probably two of the most common. When completing short answer questions, students must think of and produce the correct answer. In contrast, multiple-choice questions provide several possible solutions and require the learner to choose the correct one. But which is best?
Research by Kang, McDermot and Roediger (2007)1 suggests that short-answer questions improve learning more than multiple-choice questions as they require students to produce the answer. So, does this mean we need to stop using multiple-choice quizzes in favour of short answer quizzes? Certainly not. In geography, students are often faced with multiple-choice quizzes in the exam, so they need to be well versed in completing them. Additionally, a study by Smith and Karpicke (2014)2 has indicated that students who practised retrieval (either multiple choice or short answer quizzes) performed better than those in a control group. However, the differences between the performance of the students using the two forms of retrieval (short answer and multiple-choice) were small.
Considering this, it is clear that both approaches to quizzing for retrieval practice have a place in geography. However, it is worth considering when each is most appropriate. For example, in the short term, after teaching a concept or topic, multiple-choice questioning (containing effective distractors – take a look at these tips) might be well suited to support embedding learning. Following this, the scaffolding that multiple-choice questions can provide to students can be removed, transitioning to short answer questions to check for knowledge and understanding. This way, students have to work harder to by having to retrieve their answers from memory.
To support Internet Geography Plus subscribers in developing the use of short-answer questions. We’re creating a bank of questions and answers that can be shared with students in Word documents along with a bank of self-marking Google Forms/Microsoft Forms that subscribers can copy over to their accounts. To get the ball rolling we’ve added a bank of short answer questions (and answers) covering hazards, tectonics and plate margins in Word, Google Form and Microsoft Form format that Internet Geography Plus subscribers can access now. Log in or take out a low-cost subscription to Internet Geography.
We’ve also developed an example of a hybrid quiz containing multiple choice and short answer quiz questions available in the 4Rs of Revision area in Internet Geography Plus.
If you’ve experience combining the use of multiple-choice and short answer questions as part of your retrieval strategy, please share in the comments below.
- Kang, McDermot, and Roeriger (2007), Test format and corrective feedback modify the effect of testing on long-term retention – https://www.scinapse.io/papers/1981846272#fullText.
- Smith and Karpicke (2014), Retrieval practice with short-answer, multiple-choice, and hybrid formats