This post is based on a presentation I made at the 52nd Actuarial Research Conference in Atlanta, Georgia in July 2017. The slides for the presentation are attached.

I was inspired to give this talk by a few conversations I had with other actuarial educators at last year's conference. I mentioned my usual approach to testing both basic and high level thinking skills, along with communication, in the same question on my exams. More than one person was extremely surprised by my approach and exclaimed "I wish I could write questions like that!"

I'm here to tell you that you can. No matter what level, what area, what class size, or what background your course is from, there are always ways of incorporating higher level learning into your assessments.

But let's back up a bit. What do I mean by "higher levels of learning"? You may be familiar with Bloom's Taxonomy of Learning, and the work done by Marzano building on it, but if you are not I encourage you to learn more about them.

The key thing to remember is that assessment is curriculum. The way you assess your students is what they will learn to do. So if you want your students to have particular skills (critical evaluation of models, clear communication of ideas, understanding of relationships between concepts, etc), you need to test them on that. You can talk about the importance of (say) good communication skills until you're blue in the face, but if every single question on every assignment and test is calculations only - actions speak louder than words - they will just focus on learning the calculations.

I'm not saying you should give up on testing the basic calculation skills. Those are essential, and should still be tested. But you can incorporate parts of questions that go deeper and allow students to really show you what they know.

Here are some ways I include higher level questions:

• list similarities and differences between two products/distributions/etc
• evaluate the effectiveness of a model
• graph some function
• explain in one sentence the logical meaning of a mathematical result
• determine whether a particular calculation is correct (and if not, explain why not)
• justify whether or not the assumptions needed for a particular model apply in a specific situation
• apply some known techniques to a completely new and unseen scenario

When you design a question, you can also do it in a way that makes the marking reasonable. I like to start with a basic calculation, giving the answer to less accuracy so students know they are on the right track, then some more complex calculations, and then a conceptual question extending the material.

There are lots of strategies for incorporating higher level questions into your assessments, and I hope you'll give them a try. The more you do it, the easier it gets to think of questions, and the more you ask this type of question, the more you will encourage your students to learn at a deeper level.