Before this course: Learning How to Learn (course 9.6). You already know the forgetting curve, the testing effect, spaced practice, interleaving, the learning-styles myth and a simple weekly plan. Here we use those ideas on a real subject, and we look at the numbers and the limits more closely.
Before this course, and how to read learning numbers
Two runners try a new shoe. One says "I got a bit faster." The other says "I got much faster." To judge the shoe you need to know how much faster, and for who. In learning research, an effect size is the "how much".
- An effect size says how big a difference between two groups is. Cohen's d and Hedges' g are two common versions. Both measure the gap in standard deviations. A common rule of thumb is about 0.2 = small, 0.5 = medium, 0.8 = large. It is only a rule of thumb: what counts as big depends on the setting and the cost.
- Numbers you will meet in this course:
- Classroom quizzing, 222 studies and 48,478 students: g = 0.499 (medium), Yang et al. (2021).
- Interleaved maths practice, one large classroom trial: d = 0.83, Rohrer et al. (2020).
- Interleaving across 59 studies: g about 0.42 (moderate), Brunmair and Richter (2019).
- Lab studies and classrooms differ. Many memory results come from short lab tasks (word lists, facts). Real subjects are messier: more time, more material, more motivation problems. Always ask what the people learned, and how long after they were tested.
- An average hides variation. A medium average effect can include some studies with no effect or a negative one.
- Tip (not a research result): when you read "research shows," ask four questions. How big? For whom? What material? Tested how long later?
| Question | Why it matters |
|---|---|
| How big is the effect? | Small, medium and large change what you should expect. |
| What material was used? | Word lists do not always act like maths or essays. |
| How long until the test? | A test in 5 minutes and a test in a month can give opposite winners. |
| Who took part? | University students are not every learner. |
- Pick one subject you will study for the next three weeks.
- Write the test or goal and its date, for example "20-question quiz on 3 December."
- Write how you study it now, in one sentence.
- Write one study claim you have heard, for example "highlighting works."
- Ask the four questions in the table and look for a number in this course.
- Rewrite the claim with a size and a limit.
- Check: your rewritten claim has a number (or says "no number") and one limit.
Videos by other creators on YouTube, for reference. They are not made by or affiliated with The Future Market.
Read the video script
[Host] "Study tips are everywhere. The question is: how much do they really help?"
[Text on screen: Effect size = how big the difference is] "An effect size tells you how big a difference is, measured in standard deviations."
[Screen: bar for quizzing, g about 0.5] "In a large review of classroom quizzing, the average effect was medium: about 0.5."
[Host] "Medium, not magic. Ask: how big, for whom, and tested how long later?"
Write one study claim you have heard. Rewrite it using the four questions (how big, what material, how long later, who). Keep both versions.
ClaudeStuck? Copy a ready-made prompt for Claude
Copy this, open Claude, paste it, and Claude will walk you through the lesson step by step.
For a grown-up: Claude is for adults (18+). Copy this, paste it into Claude, and use the answer to explain the lesson to your child.
I'm learning "Learning Strategies: memory and practice" on The Future Market. Lesson 1: "Before this course, and how to read learning numbers". What the lesson teaches: - An effect size says how big a difference between two groups is. Cohen's d and Hedges' g are two common versions. Both measure the gap in standard deviations. A common rule of thumb is about 0.2 = small, 0.5 = medium, 0.8 = large. It is only a rule of thumb: what counts as big depends on the setting and the cost. - Numbers you will meet in this course: - Classroom quizzing, 222 studies and 48,478 students: g = 0.499 (medium), Yang et al. (2021). - Interleaved maths practice, one large classroom trial: d = 0.83, Rohrer et al. (2020). - Interleaving across 59 studies: g about 0.42 (moderate), Brunmair and Richter (2019). - Lab studies and classrooms differ. Many memory results come from short lab tasks (word lists, facts). Real subjects are messier: more time, more material, more motivation problems. Always ask what the people learned, and how long after they were tested. - An average hides variation. A medium average effect can include some studies with no effect or a negative one. The exercise I'm working on: Write one study claim you have heard. Rewrite it using the four questions (how big, what material, how long later, who). Keep both versions. I'm stuck. Please: 1. Explain the main idea in plain, simple words with one everyday example. 2. Walk me through the exercise one small step at a time. Ask me what I've done so far before giving the next step. 3. Don't just give me the final answer. Help me do it myself, and check my work when I share it.
I'm a parent helping my child (about 8 years old) with a lesson called "Before this course, and how to read learning numbers" from the course "Learning Strategies: memory and practice". Here's what the lesson says: Two runners try a new shoe. One says "I got a bit faster." The other says "I got much faster." To judge the shoe you need to know how much faster, and for who. In learning research, an effect size is the "how much". The activity: Write one study claim you have heard. Rewrite it using the four questions (how big, what material, how long later, who). Keep both versions. Please explain this so an 8-year-old understands: short sentences, a fun everyday example, and no hard words. Then give me 3 simple questions I can ask my child to check they understood, and a way to do the activity together, one tiny step at a time.
Sources
- Yang, Luo, Vadillo, Yu & Shanks (2021), "Testing (quizzing) boosts classroom learning," Psychological Bulletin: doi.org/10.1037/bul0000309
- Rohrer, Dedrick, Hartwig & Cheung (2020), "A randomized controlled trial of interleaved mathematics practice," Journal of Educational Psychology: doi.org/10.1037/edu0000367
- Brunmair & Richter (2019), "Similarity matters: A meta-analysis of interleaved learning and its moderators," Psychological Bulletin (PubMed): pubmed.ncbi.nlm.nih.gov/31556629/
Retrieval practice in depth
This lesson is part of full access: every course and every level for one payment.
Spacing: choose the gap
This lesson is part of full access: every course and every level for one payment.
Interleaving: when mixing helps
This lesson is part of full access: every course and every level for one payment.
Why good methods feel bad
This lesson is part of full access: every course and every level for one payment.
Put it together on a real subject
This lesson is part of full access: every course and every level for one payment.
Measure, adjust and know the limits
This lesson is part of full access: every course and every level for one payment.
Final project
This lesson is part of full access: every course and every level for one payment.
Quiz
What does an effect size tell you?
How big a difference is, in standard deviations (for example Cohen's d or Hedges' g).
What was the average effect of classroom quizzing in Yang et al. (2021)?
About g = 0.499, a medium effect, across 222 studies and 48,478 students.
What should you do right after a retrieval attempt?
Check your answers and fix the gaps.
What did Cepeda et al. (2008) find about the best gap between study sessions?
It gets longer as the time until the test gets longer, but it gets smaller as a share of that time.
Roughly what share of the test delay was the best gap for a 1-week delay, and for a 1-year delay?
About 20% to 40% for 1 week and about 5% to 10% for 1 year.
Which hurts more, a gap that is too short or a bit too long?
Too short.
What did the interleaved group score in Rohrer et al. (2020) compared with the blocked group?
61% versus 38% on a test about a month later.
Give one limit of interleaving.
It took more time in the trial, it helps less for some material (such as word pairs), and it is not meant for first learning of a brand-new idea.
What did Kornell and Bjork (2008) find about how learners judged methods?
Most thought massed study was better, even after results showed mixed study was better.
What should a personal test of a study method include?
One method, one measure, a baseline, a time period, one other possible explanation and a plan to change if there is no gain.
Final exam
This lesson is part of full access: every course and every level for one payment.
Certificate
This lesson is part of full access: every course and every level for one payment.