DESIGNING EFFECTIVE MULTIPLE CHOICE QUESTIONS
Multiple choice assessments have anecdotally been the pariah of the assessment family. But its perceived inferiority as a valid form of assessment is unfounded, as research by Smith and Karpicke (2014) attests. However, for the format to be just as effective as short answer questions, the design of the test requires careful consideration, and I shall now outline the key characteristics of an effective multiple choice test.
1 BUILD THE LEVEL OF DIFFICULTY, gradually
Understanding schema is everything, as always. An awareness that you are building your students’ schema of a topic will help shape the design of your multiple choice questions. Butler, A. C., Marsh, E. J., Goode M. K., & Roediger, H. L., III (2006) discovered that adding too many lures as distractors to the novice learner not only negatively impacted on motivation, but also inhibited later recall of that content when compared to the performance of a student with better developed schema. This makes sense because novices are not yet able to distinguish between the distractors because their knowledge is not secure enough. While it may be tempting to make the questions harder by adding in lots of other knowledge, it is not an effective strategy.
We also know that there is the possibility that a novice will learn from the ‘incorrect’ lures/distractors presented (Marsh, E.J., Roediger, H.L., Bjork, R.A. (2007)), further evidence that we need to be cautious and precise when designing multiple choice questions for novice learners.
KEY TAKE AWAY
The design of the questions should emulate the way the knowledge was taught: incrementally building in difficulty.
2 MASTER SPECIFIC KNOWLEDGE FIRST
Initially, individual pieces of knowledge that form part of a larger key concept need to be retrieved. Much of the content of multiple choice questions at this stage of the learning journey would be based on factual knowledge that simply has to be retained to help shape understanding of more complex knowledge at a later time. The advantage of using these questions to dominate the fundamental stages of your retrieval strategy is that you will be able to isolate misconceptions and gaps in learning immediately; the reality is that if a student is struggling at this stage, then they either haven’t studied or paid enough attention to the content. By approaching the design of your assessment in this way, you are ensuring that your students can walk before you expect them to run.
KEY TAKE AWAY
As a retrieval strategy, multiple choice tests should help a student master individual components of the course before they strive to test several and eventually all components of the course.
3 ACTIVELY ENGAGE RETRIEVAL
There are several design choices that strengthen the validity of a multiple choice question being able to assess learning.
- Odegard, T. N., & Koen, J. D. (2007) suggest that there are certain questions, such as ‘none of the above’ that you shouldn’t ask, as they potentially don’t encourage retrieval as none of the relevant information is being recalled. Also, of concern is that one of the wrong answers may incidentally and inadvertently be retrieved.
- Answers should include at least 2 plausible options, otherwise a student can choose an answer by elimination, which is not necessarily strengthening the retrieval of the correct answer. For example, a poor design would be: What is the capital of Australia? A) London, B) Canberra C) Paris, D) Berlin. In this question the student doesn’t have to know it is Canberra, they could just eliminate the other options that they would have heard of before. If option D) was Sydney, then they would have to think and retrieve harder.
- The number of plausible options should increase as the retrieval stretches to include multiple components of the course.
- As the course proceeds and the domain of knowledge increases, the range of questions increases to include previous learning as well as the current learning. Adding options that are wrong in the current question but correct to another question has been shown to be effective: Little, J. L., Bjork, E. L., Bjork, R. A., & Angello, G. (2012). This strategy is only useful however when a student has a well developed schema about the content, otherwise incorrect answers could be again inadvertently retrieved, but now on two occasions.
- Feedback AS RETRIEVAL – Besides automatic marking, multiple choice questions provide 2 extra bonuses: they help make feedback more precise, and a prepared discussion of why certain plausible options are not quite the right answer presents another excellent retrieval opportunity as students see the correct answer in context and how it is connected to other pieces of knowledge. Below is a good example of this:
KEY TAKE AWAY
There is a science and an art to designing multiple choice questions. Understanding the research on what works and what doesn’t will render your design an effective assessment for/of learning tool as well as an excellent retrieval activity, or simply a tokenistic waste of time.
4 MITIGATE GUESSING
By asking several questions about the same concept the tutor can safely eliminate that students have guessed their way to success.
The same can be done by ensuring there are at least 4 options as answers for questions: every extra option statistically reduces the chance of guessing correctly.
KEY TAKE AWAY
If you provide enough questions, and enough options inside those questions, statistically you’ll be in a better position to assess learning.
5 MASTERY PATHWAYS
Eventually, the multiple choice test you design will strive to assess not just individual pieces of knowledge, but more of the domain. The domain will be made of many individual components, which are in turn made up of many individual pieces of knowledge. When designing the domain tests, questions should be created with a mastery approach in mind, where there will be 3 streams of knowledge: core, developmental, and foundational.
A student who incorrectly answers a question in the core stream shouldn’t be encouraged to continue with the quiz in this ‘core’ stream of questions until they can address the error: the error produces a learning gap that can be confounded later if not fixed now.
A mastery pathway enables this by redirecting the student to a ‘developmental’ stream of questions to help strengthen and eventually secure the correct knowledge necessary to return to the core stream. The developmental stream is comprised of 3 – 4 questions that are hierarchical in difficulty, eventually building to be analogous to the original question. Students who simply made a mistake or pressed the wrong choice for example, are encouraged by this process to be more precise in the future – they are also presented further retrieval opportunity, and so still gain from the perceived waste of time, provided they are aware of the teaching strategy (more on the power of metacognition in the next post).
If a student is unsuccessful in the developmental stream they are indicating that they need further knowledge building. Such a student would be redirected to a ‘foundational’ stream, where questions take the student back to basic factual and elementary pieces of knowledge. Success in this stage provides access back to the developmental stream and then eventually back to the core stream, and crucially, possessing the required knowledge to progress in the course. The video below illustrates this process.
KEY TAKE AWAY
It may take some students longer to arrive at the required level of knowledge, but at least they will eventually arrive – that is not something that every teacher could guarantee presently.
THE CURSE OF TIME
Of course, designing a multiple choice sequence is a time consuming affair. Sometimes coming up with the ‘wrong’ distractor options is actually quite difficult. Having to then design extra questions to satisfy a mastery pathway is even more demanding. But, once created, the multiple choice test is able to be used multiple times, over many years, and will have significant benefits to students who present with learning gaps. Also, it will actually save you time in the long run as less energy will have to be spent addressing gaps further into a course.
So, in summary, the key things to consider when designing multiple choice questions are:
Butler, A. C., Marsh, E. J., Goode, M. K., & Roediger, H. L., III (2006). When additional multiple-choice lures aid versus hinder later memory. Applied Cognitive Psychology, 20, 941-956.
Little, J. L., Bjork, E. L., Bjork, R. A., & Angello, G. (2012). Multiple-choice tests exonerated, at least of some charges: Fostering test-induced learning and avoiding test-induced forgetting. Psychological Science, 23, 1337-1344.
Marsh, E.J., Roediger, H.L., Bjork, R.A. et al. The memorial consequences of multiple-choice testing. Psychonomic Bulletin & Review 14, 194–199 (2007). https://doi.org/10.3758/BF03194051
Odegard, T. N., & Koen, J. D. (2007). “None of the above” as a correct and incorrect alternative on a multiple-choice test: Implications for the testing effect. Memory, 15, 873-885.
Smith MA and Karpicke JD (2014) Retrieval practice with short-answer, multiple-choice, and hybrid formats. Memory 22: 784–802.
I’m Paul Moss. I’m a learning designer. Follow me @edmerger