In one sentence
Final-year students who reviewed staff-written digital flashcards on a 1–3–7–14–28-day schedule scored far higher on the same 20 paediatric MCQs than classmates who stuck to notes and textbooks after the same lecture.
What the researchers did
Vagha and colleagues ran a quasi-experiment during paediatric clinical postings at Jawaharlal Nehru Medical College (Maharashtra, India). 90 final-year MBBS students in two batches were split 45 / 45 by order in the batch, not by random draw.
Everyone took a 20-question pre-test on developmental milestones and immunisation, then the same didactic lecture. The intervention group then used custom digital flashcards (recall and scenario items) in supervised review sessions at 1, 3, 7, 14, and 28 days. The control group used clinic notes, textbooks, and ordinary faculty contact. The post-test reused the pre-test items (12 easy, 6 hard, 2 very hard; 17 recall, 3 applied). Attendance was recorded as complete. Control students received the cards after the study.
What they found
- Pre-test means were close: 11.42 vs 11.58 out of 20 (p = 0.057).
- Post-test: intervention 16.24 vs control 11.89 (p < 0.0001). Control’s own change (11.58 → 11.89) was not statistically significant (p = 0.11).
- Intervention SDs were reported as about 2.4 vs 2.9 on the post-test.
- In feedback, over 90% of card users agreed the method helped retention, engagement, or confidence (Likert items; e.g. about 64–71% agreed on several “effectiveness” statements).
What this means for learners and educators
- A shared, curriculum-mapped deck plus calendar intervals can beat “read the posting notes again” when both groups heard the same lecture.
- Supervised slots probably raised adherence — this is closer to a scheduled lab than to hoping students open Anki at midnight.
- If you copy the idea, write cards to outcomes, and do not treat a repeated 20-item quiz as proof of clinic-ready skill.
Limitations and what we don't know yet
Assignment was not randomised. The post-test was the same as the pre-test, so item memory can inflate the gap. Reviews were faculty-supervised during posting hours. Two topics, one college, two months. Abstract says six months; methods say September–October 2024. No delayed exam after the cards stopped. Control was traditional self-study, not another active method.
FAQ
Is this the same as using Anki on my phone?
Same idea (expanding gaps). Here the deck and timetable were staff-run, not self-scheduled software.
Why did the control group not improve after a lecture?
On this short MCQ, notes-and-textbooks did not move the mean. That does not prove lectures are useless — only that this comparison favoured scheduled cards.
Should every posting ship a 28-day card cycle?
Only if someone can write, supervise, and retire the cards. This is one paediatric block, not a universal roster.