Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 60 of the 65 references it cites were found in a registry or an index; 10 of the 49 sources its claims cite were checked for support.

Replication and Analysis of Ebbinghaus’ Forgetting Curve

Open access
Murre, Jaap M. J., Dros, Joeri · 2015 · PLOS ONE

Abstract

We present a successful replication of Ebbinghaus' classic forgetting curve from 1880 based on the method of savings. One subject spent 70 hours learning lists and relearning them after 20 min, 1 hour, 9 hours, 1 day, 2 days, or 31 days. The results are similar to Ebbinghaus' original data. We analyze the effects of serial position on forgetting and investigate what mathematical equations present a good fit to the Ebbinghaus forgetting curve and its replications. We conclude that the Ebbinghaus forgetting curve has indeed been replicated and that it is not completely smooth but most probably shows a jump upwards starting at the 24 hour data point.

Overview

CoChat Summary

## TL;DR

The paper reports a successful single-subject replication of Ebbinghaus’ forgetting curve using the savings method, with results broadly similar to Ebbinghaus’ original data. It also argues that the curve is not perfectly smooth and likely shows a jump beginning at the 24-hour/1-day point.

## Key findings

  • The replication used 69 learned-and-relearned lists in the experimental phase (plus 14 practice lists relearned after 20 minutes), with 70 hours of data collection over 75 days.
  • The main forgetting curve was similar to earlier data: the authors say the four curves (Ebbinghaus, Heller et al., Mack, and Dros) share many characteristics, though the 31-day savings score for Dros was much lower than the others.
  • Learning time increased over the course of the experiment by 2.67 s per day per list (linear regression explained 56.18% of variance); correcting for this would raise the 31-day savings from 0.0410 to 0.137.
  • For curve fitting, Ebbinghaus’ original functions fit his own data very well: the 1880 power function and 1885 logarithmic function explained 98.8% of variance; adding a “boost” improved fit modestly (average AIC -22 to -23.6, SSD 0.00932 to 0.00628, variance explained 98.7% to 99.1%).

## Methods

  • Design: A replication of Ebbinghaus’ savings-based forgetting experiment with one subject; retention intervals were 20 min, 1 hour, 9 hours, 1 day, 2 days, 6 days, and 31 days.
  • Sample/data: The only subject was the second author, J. Dros, age 22, male, native Dutch speaker; the materials were 70 lists of nonsense syllables, each list containing 104 syllables arranged in 8 rows of 13.
  • Analysis: They computed savings from learning/relearning times, tested interval effects with one-way ANOVAs, examined serial position curves, and compared several fitted functions using nonlinear fitting in Mathematica 9 and model-comparison metrics including AIC, SSD, and R².

## Limitations

  • The paper states that this was a single-subject design, so generality to other people remains unclear.
  • The authors note they could not reproduce Ebbinghaus’ exact schedule: they had only 75 days instead of seven months, used 10 replications per interval rather than Ebbinghaus’ 12 to 45, and could not keep a fixed time of day.
  • It is evident from the text that the 31-day condition may be confounded by massed early learning, intervening-list differences, and a strong time/order confound; the authors say they could not disentangle number of intervening lists from time.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0120644

Cites

60 of 60 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Learning schedule during 2011–2012 for all lists, where labels in bold indicate when each of the lists 1 to 10 was first learned for each retention interval. Relearning times are not shown but can be derived by adding the retention interval (e.g., 6 days).
Fig 1 Learning schedule during 2011–2012 for all lists, where labels in bold indicate when each of the lists 1 to 10 was first learned for each retention interval. Relearning times are not shown but can be derived by adding the retention interval (e.g., 6 days).CC BY · PubMed Central
Fig 2 Four loglog graphs with savings as a function of retention interval with fitted power function curves and curves with best fitting power functions with boost at 1 day (see text for an explanation).
Fig 2 Four loglog graphs with savings as a function of retention interval with fitted power function curves and curves with best fitting power functions with boost at 1 day (see text for an explanation).CC BY · PubMed Central
Fig 3 Four log graphs with savings as a function of retention interval with best-fitting Memory Chain Model retention functions (see text for an explanation).
Fig 3 Four log graphs with savings as a function of retention interval with best-fitting Memory Chain Model retention functions (see text for an explanation).CC BY · PubMed Central
Fig 4 Normalized savings scores as a function of retention interval on a logarithmic scale, rescaled so the first data point is 1.0 for all curves.
Fig 4 Normalized savings scores as a function of retention interval on a logarithmic scale, rescaled so the first data point is 1.0 for all curves.CC BY · PubMed Central
Fig 5 Learning time per list as a function of day of experiment with a fitted straight line.
Fig 5 Learning time per list as a function of day of experiment with a fitted straight line.CC BY · PubMed Central
Fig 6 Serial position for correct relearning scores for each retention interval and for the average of all retention intervals (see text for an explanation).
Fig 6 Serial position for correct relearning scores for each retention interval and for the average of all retention intervals (see text for an explanation).CC BY · PubMed Central
Fig 7 Proportion correct as a function of retention interval on a logarithmic scale. (a) Proportion correct, averaged over all serial positions, shown with Dros’ savings scores for comparison. (b) Proportion correct curves for different groups of serial positions and for the average over all 13 positions.
Fig 7 Proportion correct as a function of retention interval on a logarithmic scale. (a) Proportion correct, averaged over all serial positions, shown with Dros’ savings scores for comparison. (b) Proportion correct curves for different groups of serial positions and for the average over all 13 positions.CC BY · PubMed Central

Verification

Claims the paper makes

6 Valid · 2 Not valid · 22 Not checked · 60 Own results and interpretations

32 more claims

Create a free account to see all 90 claims and to verify the 1 reference still open.

See all claims — free account
  • It may state two separate claims as one. 47
  • Probably not a substantial claim. 17
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Not confirmedF(6, 512) = 0.753reported p > .608 · recomputed p = .607

    “F(6, 512) = 0.753, p > .608”

  • ConsistentF(6, 69) = 0.691reported p = .658 · recomputed p = .658

    “F(6, 69) = 0.691, p = .658”

  • ConsistentF(6, 512) = 3.85reported p < .01 · recomputed p < .001

    “F(6, 512) = 3.85, p < .01”

3 of 3

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Self-check results appear here once an independent check has confirmed them.

Citations

Cites

Show all 65 references

Research works printed but not linked: 1

  1. WA Wagenaar, 1986, My memory: A study of autobiographical memory over six years, Cognitive psychology, 18, 225, 10.1016/0010-0285(86)90013-7 · DOI · Paper page

Cited by

See all 777

Metadata

Authors
Murre, J. M. J., & Dros, J.
Affiliations
University of Amsterdam (NL)
Corresponding author
Murre, Jaap M. J.
Venue
PLOS ONE
Published
2015-07-06
Volume
10
Issue
7
Pages
e0120644
ISSN
1932-6203 (ISSN-L)
OpenAlex
W1933687796
Cited by
777
References
65
Citations by year
2026: 89 · 2025: 125 · 2024: 114 · 2023: 102 · 2022: 92 · 2021: 92 · 2020: 58 · 2019: 44 · 2018: 36 · 2017: 18 · 2016: 6 · 2015: 1
FWCI
12.21
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • UvA-DARE (University of Amsterdam) (publishedVersion, cc-by)
  • UvA-DARE (University of Amsterdam) (acceptedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Data Archiving and Networked Services (DANS) (submittedVersion, other-oa)
Funders
Universiteit van Amsterdam
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Memory Processes and Influences
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Memory and Neural Mechanisms
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Cognitive Functions and Memory
Keywords
Ebbinghaus forgetting curve, forgetting, relearning, serial position, word list learning, replication
MeSH terms
Humans, Learning, Memory

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-09-28, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-09-28 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Replication and Analysis of Ebbinghaus’ Forgetting Curve. https://app.cochat.ai/papers/10.1371/journal.pone.0120644
The paper
Murre, J. M. J., & Dros, J. (2015). Replication and Analysis of Ebbinghaus’ Forgetting Curve. PLOS ONE. https://doi.org/10.1371/journal.pone.0120644

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.