01Research Talk Deck

Spaced retrieval may improve retention in introductory science courses.

An illustrative pilot study of review schedules and delayed quiz scores.

Northfield Learning Lab | Research seminar
Research Talk Deck01 / 09
02Research Talk Deck

We will follow the question from study design to interpretation.

  1. Why delayed retention matters
  2. How the two review schedules differ
  3. What the illustrative scores show
  4. What this pilot cannot establish
Research Talk Deck02 / 09
03Research Talk Deck

A single review session may not support durable recall.

Study question: Does distributing the same review time across three sessions improve recall after three weeks? Hypothesis only; not an observed finding.

Research Talk Deck03 / 09
04Research Talk Deck

Both groups receive equal review time on different schedules.

Day 0

All 48 illustrative participants complete the same lesson.

Days 1-7

One group reviews once; the other splits equal time across three sessions.

Day 21

Both groups take the same delayed recall quiz.

Research Talk Deck04 / 09
05Research Talk Deck

The comparison isolates the timing of review.

Single session

  • One 30-minute review on Day 1
  • Same lesson and practice items
  • Delayed quiz on Day 21

Spaced sessions

  • Three 10-minute reviews on Days 1, 4, and 7
  • Same lesson and practice items
  • Delayed quiz on Day 21
Research Talk Deck05 / 09
06Research Talk Deck

Spaced review shows higher delayed recall in the illustrative data.

Mean score (%)
78Single D179Spaced D166Single D773Spaced D754Single D2168Spaced D21
TakeawayIllustrative values only; the Day 21 difference is 14 percentage points, not a verified study result.
Research Talk Deck · Sample figures, replace with sourced data06 / 09
07Research Talk Deck

The pilot design supports a focused comparison, not a broad claim.

48
Illustrative participants
30 min
Review time per participant
21 days
Time to delayed quiz
Research Talk Deck · Sample figures, replace with sourced data07 / 09
08Research Talk Deck

Interpretation depends on assignment and measurement quality.

These are design considerations for the fictional example, not findings from a real study.

  • A small, single-course sample limits generalization.
  • Prior knowledge and review outside the study may affect scores.
  • Quiz reliability and group assignment must be checked.
Research Talk Deck08 / 09
09Research Talk Deck

A larger study should test whether the pattern persists.

  1. Verify the result with preregistered group assignment.
  2. Report uncertainty and item-level quiz performance.
  3. Discuss which course topics might benefit most.
Research Talk Deck09 / 09