How many options in branching scenario decisions?

How many options do you need in a branching scenario for each decision point? What number gives the best balance of realism and manageable complexity?

When you create branching scenarios, each decision point provides learners with multiple choices. How many options do you need to provide in a branching scenario? Reducing the number of options can help manage the complexity of branching scenarios, but reducing it too far may provide less realistic decision-making practice.

How many options in branching scenario decisions?

Drawbacks of 4 options

By default, we usually see multiple choice questions with 4 choices or options. That’s the traditional number, used to reduce the chance of getting it right just by guessing.

I sometimes see branching scenarios where the instructional designer has created 4 choices for each decision. I suspect that’s mostly out of habit rather than really a good practice for branching scenarios.

First of all, it’s hard to write good distractors (wrong answers). The more of them you have to write, the harder it is. Providing 4 options means you have to come up with 3 choices for every decision point that are realistic, relevant, and reflect mistakes people actually make.

Second, in a branching scenario, each of those choices leads to more choices. The complexity can grow exponentially (4 choices to 16 choices to 64 choices). Even if you reuse options, let people correct mistakes, or use other techniques to manage the complexity, 4 choices at each decision point quickly become unwieldy.

2 options works sometimes

Sometimes providing 2 options works for branching. It’s much easier to manage, and the number of slides and links doesn’t get out of control so fast.

However, if every decision point can be reduced to a simple “good or bad” choice, is this really a skill that requires a branching scenario? Branching scenarios are great if the process or skill has some “gray area” where decisions might be OK but not ideal, or they might be partially correct. If the situation doesn’t have any shades of gray, maybe you should use another approach to storytelling.

Two options might work for a branching scenario if you have a very long process to show, where providing even 3 options would be unmanageable. If you have over 50 total decision points, for example, I would strongly consider reducing the number of options to 2. The classic Choose Your Own Adventure books mostly offer 2 choices, partly because they are lengthy stories.

Two options might also work if you’re trying to focus on a particular mistake, and that decision really only has two paths forward. I’d rather see a scenario use 2 realistic choices than 2 realistic choices plus 1 obviously made-up, wrong option.

Two options can also work in combination with more choices. There’s no absolute rule that every choice in a branching scenario has to have the same number of choices. I sometimes narrow the options to 2 near the end of a scenario because that fits the actual process better.

3 options is usually best

Most of the time, I use 3 options in my branching scenarios. I aim for one good choice, one bad, and one OK. The OK choice is often partially correct or a minor error that would be easy to recover from.

I find that 3 choices is more manageable. It doesn’t expand as quickly as 4 choices, but it has enough complexity to provide a rich scenario with some shades of gray.

Research support for 3 options

At the beginning of this post, I noted how multiple choice questions often have 4 choices in an attempt to reduce guessing. However, 4 choices isn’t the best option for traditional multiple choice questions either. In fact, research shows that 3 options is the best for multiple choice. Why?

But the number of answer choices is not the only thing that determines the chance of guessing correctly. Quality of answer choices makes all of the difference.

Patti Shank in How many answer choices is best for a multiple-choice question? Probably not what you think.

Instead of spending time and effort creating additional wrong answers, reducing the number of choices to 3 means you can work on making those 3 options higher quality.

In one 2011 study, researchers randomly removed wrong answers and compared questions with 3, 4, and 5 options.

So here’s the main finding: no significant differences were found in terms of item difficulty. There were also no differences found in terms of test reliability. Thus, Baghaei and Amrahi (2011) concluded that three answer options are all you need. If the test characteristics are essentially the same, there doesn’t seem to be any reason to spend our time developing additional answer options.

Christine Harrington in Three Answer Options Are All You Need on Multiple-Choice Tests!

Patti cites Rodriguez (2005). Harrington summarizes findings from both Rodrigues and Baghaei and Amrahi (2011). Read their posts for summaries of the research, or check the full citations at the bottom.

However, this research was about more traditional multiple choice questions, not branching scenarios. Research findings don’t always apply consistently when you change the context, so some caution is warranted. I’m not aware of any research directly assessing the number of options in branching scenarios though, and the explanations for the effect seem to apply to branching scenarios as well. I think it’s reasonable to stretch the research to assume 4 options wouldn’t provide additional benefit for branching scenarios.

Citations

Baghaei, P., & Amrahi, N. (2011). The effects of the number of options on the psychometric characteristics of multiple choice items. Psychological Test and Assessment Modeling, 53(2), 192-211.

Rodriguez, M. C. (2005). Three Options Are Optimal for Multiple-Choice Items: A Meta-Analysis of 80 Years of Research. Educational Measurement: Issues and Practice, 24(2), 3-13.

Leave a Reply

Archives