A later retrieval failure can depend on competing information rather than simply time passing.
proactive interference
Earlier learning disrupts a later target.
retroactive interference
Later learning disrupts an earlier target.
intrusion
A response from outside the intended target set.
01 · TRY
When an old password intrudes during recall of a new one, which direction is the interference?
Draw timelines with the tested target circled and interference arrows labeled. Use the historical study card to explain why interpolated activity matters. Write 200 words proposing a cue-overlap manipulation that distinguishes response competition from unequal initial learning.
Complete the labeled table or trace with the supplied records. Explain each step in words before using the technical term.
Make a provisional claim
Loading saved evidence…
Inspect evidence and test an explanation
The investigation controls appear when your local evidence is ready.
04 · UNDERSTAND
Connect evidence and mechanism
Interference: which learning disrupts which memory?
Distinguish proactive from retroactive interference and test similarity separately from elapsed time.
Question and prediction
When an old password intrudes during recall of a new one, which direction is the interference?
Proactive interference occurs when earlier learning disrupts later learning or retrieval; retroactive interference occurs when later learning disrupts earlier material. Identify the target memory before naming the direction. If an old code intrudes when a new code is requested, the influence runs forward from old to new. If learning the new code makes the old one harder to recall, it runs backward toward the older target.
Interference can arise when cues activate competing responses or when materials overlap. It need not mean the original trace has been erased. McGeoch and McDonald’s 1931 study varied intervening material and found retention differences associated with its meaningful relation to the original material. That bounded design helps distinguish activity during an interval from elapsed time alone; it does not establish that every kind of forgetting has the same cause.
The fictional groups learn A, spend equal time either learning similar B or doing an unrelated activity, then recall A. Lower A recall after B supports examining retroactive interference. The conclusion depends on comparable initial learning, equal intervals and attention. Naming an effect is only the start: classify intrusion errors and vary cue overlap to investigate competition.
Time-based decay, encoding differences and retrieval conditions remain distinct explanations. An equal-delay comparison with different intervening material challenges a time-only account, but does not rule out all time-related processes. The unfamiliar case reverses the target. Draw old/new/target arrows before choosing a term, and propose a cue-specific test rather than assuming the memory has vanished.
Inspect the relationships
Sequence
Target tested
Candidate direction
Learn A → learn B → test A
Earlier A
Retroactive
Learn A → learn B → test B; A intrudes
Later B
Proactive
Worked example
A-target retention differs by 8−5=3 items with equal delays; later B learning is the candidate interfering event, so the direction is retroactive. In the separate B-target case, old A intrudes on new B, making the direction proactive.
Authentic relearning evidence
Savings in relearning is not the percentage recalled
A headline says “only 21.1% remembered after a month.” Audit the measured quantity before accepting that interpretation. What would you need to know about the task, denominator and population?
Read the source and method boundaries
Ebbinghaus, Memory (1885), Chapter VII, sections 27–29, in Ruger and Bussenius’s 1913 translation. The seven summaries below concern relearning syllable series, not counts of items recalled. Initial learning and relearning used the same two-errorless-recitation criterion. Reported savings excluded estimated recitation time; some short-interval results involved time-of-day corrections. The summary table rounds the earliest interval to 0.33 hours; the detailed table labels it 19 minutes.
Chapter III: methods and sources of error describes the author's own learning procedure and concerns about varying materials, conditions and expectation. Repeated tests on the same investigator do not create a representative sample of students. This is historical documentary analysis, not a request to conduct the procedure yourself.
Early points crowd together in the full linear view. The explicitly labeled early window separates them; the vertical 0–100% scale stays fixed. No line supplies unobserved values.
All seven reported summary points, retained when the graph window changes
Source ID
Reported hours
Published savings %
In graph window?
Percent of items recalled
EB-01
0.33
58.2
Yes
Not supplied by this savings measure
EB-02
1
44.2
Yes
Not supplied by this savings measure
EB-03
8.8
35.8
Yes
Not supplied by this savings measure
EB-04
24
33.7
Yes
Not supplied by this savings measure
EB-05
48
27.8
Yes
Not supplied by this savings measure
EB-06
144
25.4
Yes
Not supplied by this savings measure
EB-07
744
21.1
Yes
Not supplied by this savings measure
Compare unequal intervals
Savings change: 44.2 − 58.2 = -14.00 percentage points over 0.67 hours. Interval-average change: -20.895522 percentage points per hour. This divides an endpoint difference by elapsed hours; it is not a measured instantaneous rate.
Compare the first interval with the final one. Explain why equal horizontal spacing for all seven rows would obscure elapsed time. Then critique the headline: what does the vertical value measure, and which claims about present-day students or mechanisms remain unsupported?
Transfer: a separate invented relearning record
These invented effort values already exclude recitation time and use an unchanged success criterion. They are not reconstructed historical rows. Savings = 100 × (initial effort − relearning effort) ÷ initial effort. Here effort is seconds spent learning, not elapsed delay between sessions.
Predict which claims survive changing the denominator while holding saved seconds constant. Does a zero recall result force zero savings? Does negative savings need to be hidden, or does it describe more work on relearning under the stipulated record?
Compare explanations and revise
The invented first record has 40% savings and zero measured recall. These are different outcomes. The second saves the same 40 seconds but gives 25%, because its initial effort is 160 seconds. The third has −20% savings; clamping it to zero would erase the extra relearning effort. Unknown recall is not zero recall.
The observed interval averages describe changes in a particular measure. A declining sequence alone cannot isolate elapsed-time decay from intervening activity, changes in conditions or retrieval demands. Specify what existing comparative evidence would discriminate those explanations. A fitted curve, if added, would also require explicit assumptions and validation; none is used here.
Revise the headline into a source-specific claim, use two source IDs, distinguish savings from recall, and propose one missing comparison. Do not prescribe a universal study schedule or infer a person's ability from this packet.
Copy your relearning report
Feedback opened this visit: no. Reset clears writing and restores the full window, first interval and zero-recall case, retaining feedback exposure. Reload clears temporary work. Copy before leaving; capture does not certify unseen work or mastery.
The source-based relearning packet and the original fictional A/B recall cases use different outcomes. Keep savings, recall and inferred mechanism distinct when comparing them.
Your investigation
Draw timelines with the tested target circled and interference arrows labeled. Use the historical study card to explain why interpolated activity matters. Write 200 words proposing a cue-overlap manipulation that distinguishes response competition from unequal initial learning.
Unfamiliar transfer
Two equal delays yield different recall after different activities. Which claim is weakened? Explain your reasoning before checking feedback.
Fictional retroactive comparison Both groups initially recall 9/10 items from list A. After equal 10-minute intervals, the similar-list-B group recalls A at 5/10; an unrelated-activity group recalls A at 8/10.
Fictional reversed target A separate task asks for new code B. Participants repeatedly give old code A. No A-recall test follows.
Authentic evidence and its limits
Original intervening-material comparison — John A. McGeoch and William T. McDonald (1931). The study compared retention after different intervening materials, including semantically related and less-related material. Recall differed with the relation between original and intervening learning.
Bounded paraphrase of the original experiment, not a claim that similarity explains every memory failure. The A/B lists and scores in the exercise are invented.
Read the original source. Original LearnAI paraphrase and attributed factual results; no reproduced figure or substantial verbatim passage. Original publication retains its rights.
Worked example
A-target retention differs by 8−5=3 items with equal delays; later B learning is the candidate interfering event, so the direction is retroactive. In the separate B-target case, old A intrudes on new B, making the direction proactive.
