Real material
A classroom replication with an actual course chapter.
Master’s thesis defenseExample deck · fictional study
Does retrieval practice beat rereading?
A classroom test of the testing effect with real course material.
Overview
Four parts, twenty minutes
- The question 3 min
- Prior work and the gap 4 min
- Method 5 min
- Results and limits 8 min
01
The question
Retrieval practice
noun · also “the testing effect”
Learning by pulling information out of memory, rather than putting it back in by rereading.
“Closing the book and writing down what you remember is the practice.”
Prior work
A century of evidence, mostly in the lab
- 1917
Gates
Recitation outperforms rereading for children.
- 2006
Roediger & Karpicke
Testing beats restudy on a delayed test.
- 2011
Karpicke & Blunt
Retrieval beats concept mapping.
- 2013
Dunlosky et al.
Rates practice testing a high-utility technique.
The gap: few studies use real course material under classroom conditions.
02
The method
Design
Two conditions, one week apart
LearnRead a 1,200-word chapter
PracticeReread or recall
WaitOne week
Final testShort-answer questions
Participants and measures
A between-subjects design
| Group | Practice phase | Students | Final measure |
|---|---|---|---|
| Reread | Read the chapter a second time | 60 | Short-answer test, % correct |
| Retrieve | Free recall, then check against the text | 60 | Short-answer test, % correct |
Illustrative design and sample size for this fictional example study.
03
The results
Final test after one week
Retrieval practice scored higher
The retrieval group answered more questions correctly a week later.
Illustrative data (fictional study).
Effect size
Cohen’s d
The difference between group means in units of their pooled standard deviation.
Effect size
0.84
A large effect by Cohen’s conventional benchmarks (0.8 and above).
Illustrative value (fictional study).
Hypothesis: practising recall improves long-term retention of course material.
Finding: supported, with a large effect after one week.
Limitations
What this study cannot tell us
- One course, one chapter. Other subjects may respond differently.
- One delay. We tested at one week, not across a term.
- Self-checked recall. Feedback quality varied between students.
Contribution
What this adds
A simple protocol
Ten minutes of recall that any instructor can adopt.
Open materials
Test items and scoring rubric shared for reuse.
Thank you.
- QuestionsBackup slides follow
- MaterialsAppendix B, test items and rubric
- NoteExample deck: the study and data are fictional