Use The “Bloom Test” Before Making Anki Cards


So you’ve been told to “learn before you memorize,” perhaps as part of the 20 rules of formulating knowledge. (aka the SRS prophet’s commandment tablets)

And I can tell you that it’s good advice for making good cards that help you retain knowledge…until you realize the fact that not everyone has the same level of intuition and awareness to know whether they’ve fully understood something.

Thus, asking:

How do you know whether you’ve understood something well enough that it’s time to put it into Anki?

Is there a signal?

An a-ha moment, perhaps?

Let’s find out…from your teachers.

Clues from educational research

If you were to ask teachers in the early 1950’s “what’s the most important thing they want students to gain after each class”, they’ll say:

I want students to “really understand” the content.

Yet when you ask them what they meant about “understanding the content”, they didn’t have a clear answer either. So it’s not a surprise that they, too, were finding it hard to create questions that test what the student has “truly understood.” Familiar problem, huh?

Then this guy named Benjamin Bloom came up with the solution — a comprehensive set of criteria (some kind of meta-prompts) that, when met, tells you whether a student could “really understand.”

As Krathwohl (2002) put it, Bloom aimed to “reduce the labor of preparing annual comprehensive examinations”. Short for “turning question-making into a less dreaded task,” I suppose.

And in 1956, the now-famous, widely-used, “Bloom’s Taxonomy” was born.

So what does this have to do with us?

If you’re studying a textbook, a lecture, or basically if you’re following some kind of academic curriculum…

Know that the learning objectives of that entire thing is structured mostly based on an instructional design principle similar to Bloom’s idea.

We can therefore say that:

the purpose of learning objectives is to test whether the student has “really understood” the contents or not.

And therefore, we can take advantage of that property by coming to a nice conclusion:

If you can meet the existing learning objectives (from chapter objectives, syllabi, or curricula), then it’s safe to say that you’ve understood enough of the contents.

“Enough” here meaning “at the level the instructional designer wants you to be.”

I call this the “Bloom Test,” and if you apply it, you’ll be able to:

  • set a strong foundation of knowledge — especially of the “big picture”
  • make your cards useful in helping you retain knowledge in the long-term
  • get most of your cards correct during the first review
  • avoid the dreaded “ease hell” without obsessing about settings; and most importantly,
  • know exactly when you’ve studied enough before moving on to your revision phase

How to do the “Bloom Test” before making your flashcards

For easier implementation, I broke down The Bloom Test into three steps:

  1. Find the objectives for the topic
  2. Turn those objectives into questions
  3. Answer them one-by-one during/after you study

1. Find the objectives for the topic

For most topics, it’s simple — just open up a primary textbook you’re using and go to the beginning of the chapter.

Sometimes, though, the book won’t have chapter objectives. That’s not unusual.

If you’re reading this, you’re probably learning well-established academic subjects like Anatomy, Physiology, Physics, etc. so the best thing to do here is to extract objectives from other textbooks.

Textbooks on the same subject won’t vary much in the underlying ideas, anyway.

This is especially true if you’re in undergrad and not in some “frontiers of human knowledge” kind of place.

Remember, we’re doing this simply to take advantage of an expert instructional designer’s opinion about “whether we’ve understood or not.”

If you don’t want that, you can always go the DIY route — create learning objectives yourself by using the cognitive process dimension of the Revised Bloom’s Taxonomy. Here’s a good overview.

2. Turn those objectives into questions

You can consider this as the “border” between educational research and flashcard formulation skills.

All you have to do is turn your objectives into questions/prompts. This should be easy enough, because you can use the objectives themselves as a cognitive scaffold.

If it’s not, then treat this as a weakness in your flashcard formulation skill — which is great, because now you know that you can still improve!

Here’s an excerpt from the study system course, taken from the objectives above:

The beauty in this approach is that:

  • You’re training yourself to make GOOD questions — because let’s face it: we’re trained to answer questions, not to make them. We need this training!
  • If you do this right before you consume a topic, you’re basically biasing your attention to these important pieces while reading — you now know exactly what’s important from a textbook! Throw that highlighter away, will ya?
  • Or if you do it right after the topic, then you’ve just given yourself “big questions” to pull your atomic questions from.

Once you’re done, you can start to answer them one-by-one.

3. Answer them one-by-one during/after you study

To know whether you’ve understood enough or not, you simply need to test one thing:

Have I met the objectives or not?

If yes, make flashcards that test you of the ideas. Else: read again, you may have been memorizing the words instead of processing the contents.

And just to be clear:

Just because you can’t explain something simply doesn’t mean you don’t understand it.

There’s no need to “explain it simply” to say that you’ve understood it. Just meeting the learning objectives is enough.

In fact, if I were you, I won’t force Feynman Technique-ing the hell out of your 5-year old sister, or even an imaginary set of ghosts.

In my experience, doing so only makes your study sessions unnecessarily long, not to mention unnecessarily difficult.

Save the simplification for when someone asked you to explain!

For yourself, just meet the objectives.

C’mon, do you honestly think your Ph.D professors are dumb as heck that they can’t explain things without using godlike celestial lifeform language all the time? Of course not.

So, for the n-th time:

If you can answer the questions (derived from the objectives) correctly, then you can start to make flashcards for the contents.

Gotta pass that Bloom test.

Some more thoughts for implementation

  • You don’t need to make a flashcard for every fact. Formulate prompts from your insights! Think, as a general rule, that 1 flashcard should represent 1 insight. (See: Using Anki for Complex Information)
  • Getting to “understanding” is a different story. It’s in the realm of encoding — that is, manipulating/altering/connecting new info to what’s already stored in your memory. This is a crucial skill if you want to make the most out of Anki.
  • Struggling to write good cards is NOT always a sign of poor understanding. Sometimes it simply means you need practice. Like any other, the entire process of making great Anki cards requires many skills working together — like machinery! Keep working on your weak links and you’ll get faster over time 🙂