You are sure — certain — that you locked the front door. You can picture your hand turning the bolt. Except the door was unlocked when you got home. So whose memory was that?
Here is the uncomfortable truth this lesson is built on: your memory is not a recording. There is no video file in your head that you press "play" on. Every time you remember something, your brain rebuilds it from fragments — and during the rebuild, it quietly fills gaps with what probably happened, what you expected, what someone told you afterward. Most of the time the reconstruction is close enough. Sometimes it invents a locked door.
This is not a bug in your brain specifically. It is how human memory works, and it has put innocent people in prison on the strength of eyewitnesses who were honestly, confidently, completely wrong. By the end of this lesson you'll know why forgetting happens, how memories get distorted, and the names — Ebbinghaus and Loftus — that the AP exam will absolutely expect you to drop.
In the 1880s, Hermann Ebbinghaus did something nobody had thought to do: he made memory measurable. Using himself as the only subject, he memorized lists of nonsense syllables (meaningless letter combinations like ZOF or BIK, chosen precisely because they carried no prior associations to contaminate the test). Then he tracked how much he forgot over time.
The result is the forgetting curve: retention drops sharply soon after learning — most of what's lost is gone within the first hour or day — and then the rate of loss levels off, so what survives the first day tends to stick around much longer. Forgetting isn't steady; it's front-loaded.
Ebbinghaus also discovered savings: even after he could no longer recall a list at all, relearning it took fewer trials than learning it fresh. That "savings" is proof the memory wasn't entirely gone — a trace remained, just below the threshold of retrieval. Savings is why a language you studied years ago and "totally forgot" comes back faster than it did the first time.
Try This. Pick something you crammed for a test last year and genuinely can't recall now. Spend five minutes relearning it. Notice how much faster it goes than it did originally. That speed-up is Ebbinghaus's savings — evidence that "forgotten" rarely means "erased."
Think of memory as a three-stage pipeline — encoding (getting information in), storage (keeping it), retrieval (getting it back out). Forgetting can strike at any of the three.
Encoding failure means the information never made it into long-term memory in the first place. You didn't forget it; you never had it. The classic demo: most people cannot draw a common coin accurately from memory, even though they've seen thousands. The details were never encoded because you never needed them. No encoding, no memory to lose.
Storage decay is the Ebbinghaus idea: physical memory traces fade over time if they're not used. The forgetting curve is decay in action. (Decay applies most cleanly to fragile, unrehearsed memories; richly encoded long-term memories are remarkably durable.)
Retrieval failure means the memory is stored — you just can't access it right now. The proof is the tip-of-the-tongue phenomenon: that maddening state where you know you know someone's name, you can feel its shape, maybe its first letter, but it won't come. If it were truly gone you wouldn't feel the gap so precisely. The memory is there; the retrieval cue isn't working. Often a cue (someone says "it starts with M…") springs it loose instantly — which proves it was stored all along.
A major cause of retrieval failure is interference — when memories compete and one blocks another. There are two directions, and the exam loves to test which is which.
Proactive interference: old learning disrupts new learning. You get a new phone number and keep "remembering" your old one. The prior memory reaches forward in time to interfere. (Pro- = forward.)
Retroactive interference: new learning disrupts old learning. You learn this year's locker combination and can no longer recall last year's. The new memory reaches backward to interfere. (Retro- = backward.)
A clean way to keep them straight: ask which memory is doing the interfering. If the old memory is the troublemaker, it's proactive. If the new memory is the troublemaker, it's retroactive.
Some forgetting may be motivated. Freud proposed repression — the idea that the mind defensively pushes anxiety-provoking memories (traumatic or shameful) into the unconscious, out of conscious reach. It's an intuitive idea and it appears on the exam, but flag it honestly: most memory researchers are skeptical of repression as Freud described it. The better-supported finding runs the opposite way — emotional and traumatic events tend to be remembered vividly, not buried. Know repression as a named concept; know that it's debated.
Amnesia is memory loss from brain damage or trauma, and it comes in two directions that students constantly flip.
Anterograde amnesia: you cannot form new long-term memories after the injury. Old memories are intact; the anterior (forward, future) direction is blocked. The famous case is H.M. (Henry Molaison), who had his hippocampus removed on both sides to stop severe seizures. The seizures stopped — but H.M. could no longer convert new experiences into lasting memories. He'd meet the same researcher hundreds of times and greet each meeting as the first. H.M. is the case that nailed down the hippocampus as essential for forming new explicit memories. Clive Wearing, a musician with brain damage from a viral infection, is a second classic example — he lives in a perpetual present, convinced each moment that he's just woken up, yet he can still play piano (procedural memory survives).
Retrograde amnesia: you lose memories formed before the injury — the retro (backward, past) direction. You can't recall the accident or the weeks before it, but you can form new memories going forward.
Mnemonic: anterograde blocks the road ahead (no new memories); retrograde wipes the road behind (lost old memories).
Here's the big idea that ties the whole lesson together: remembering is reconstruction, not playback. When you retrieve a memory, you rebuild it from stored fragments plus your expectations, beliefs, and information you've picked up since. That rebuild is vulnerable to distortion.
The most important distortion is the misinformation effect — exposure to misleading information after an event corrupts the memory of the event itself. This is Elizabeth Loftus's life work (see the Spotlight). Related distortions:
Some psychologists organize these failures as the seven "sins" of memory (Schacter): transience, absent-mindedness, blocking, misattribution, suggestibility, bias, and persistence — a handy optional framework, not required memorization. The non-negotiable takeaway: a confident memory and an accurate memory are not the same thing.
Loftus & Palmer (1974) — the car crash and the verb that changed it.
Who & when: Elizabeth Loftus and John Palmer, 1974 — the foundational demonstration of the misinformation effect.
Method: Participants watched a film of a car accident, then answered questions about it. The critical manipulation was a single verb. Different groups were asked, "About how fast were the cars going when they ___ each other?" with the blank filled by smashed, collided, bumped, hit, or contacted. The verb was the only thing that differed.
Finding: The verb shifted the speed estimates. Participants asked about cars that "smashed" estimated the highest average speed — about 40.8 mph — while those asked about cars that "hit" estimated about 31.8 mph, roughly 9 mph slower, despite everyone watching the identical film. In a follow-up a week later, participants were asked whether they had seen broken glass (there was none). Those who'd earlier heard "smashed" were more than twice as likely to falsely "remember" broken glass than those who'd heard "hit."
Why it matters: The wording of a question after the event reshaped the memory of the event — and even planted a detail (broken glass) that was never there. This is the misinformation effect in its purest form, and it's why leading questions and eyewitness testimony are treated with such caution in courtrooms. For the exam: Loftus, 1974, "smashed vs. hit," reconstructed memory.
Scenario 1. Maya switches from a QWERTY keyboard to a new ergonomic layout for work. For weeks, whenever she sits down at the new keyboard, her fingers keep flying to the old QWERTY positions, sabotaging her typing.
Which concept, and which direction? This is proactive interference: the old, well-learned QWERTY layout is disrupting the new learning. The old memory reaches forward to interfere — proactive. (If instead she'd mastered the new layout and then couldn't type on a QWERTY keyboard anymore, that would be retroactive.)
Scenario 2. After a serious bike accident, Devon can hold a normal conversation, learns his physical-therapy exercises day by day, and remembers his childhood perfectly — but he has no memory of the crash itself or the entire afternoon leading up to it.
Which type of amnesia? This is retrograde amnesia: the memories before the injury (the afternoon, the crash) are gone, while his ability to form new memories (learning the exercises) is intact. The damage wiped the road behind, not the road ahead. (Contrast H.M., who had anterograde amnesia — old memories fine, no new ones.)
Scenario 3. A witness to a robbery is interviewed twice. In the first interview, no one mentions a weapon. Before the second interview, a detective asks, "How tall was the man holding the gun?" At trial, the witness confidently testifies that the robber brandished a gun. Surveillance footage shows there was no weapon.
Which concept explains the false testimony? The misinformation effect, via a leading question. The detective's wording introduced misleading post-event information ("the gun") that got woven into the witness's reconstructed memory — exactly the Loftus & Palmer mechanism. The witness isn't lying; the memory was genuinely altered, illustrating the reconstructive nature of memory and the danger of eyewitness testimony.
Proactive vs. retroactive interference. Everyone flips these. Don't memorize definitions — ask which memory is the troublemaker. Old memory interfering = proactive (the prefix pro- = forward, the old memory pushes forward). New memory interfering = retroactive (retro- = backward, the new memory pushes back). "PRO is for the PRIOR memory."
Anterograde vs. retrograde amnesia. Anterograde = can't form new memories going forward (the road ahead is blocked; think H.M., Clive Wearing). Retrograde = lost old memories from before the injury (the road behind is wiped). Tie retro to "retro = the past."
Storage decay vs. retrieval failure. Decay means the memory faded and is gone (the trace physically weakened). Retrieval failure means the memory is still stored but inaccessible right now — proven by tip-of-the-tongue and by a cue suddenly unlocking it. Test: if the right cue would bring it back, it was retrieval failure, not decay.
Recall = reconstruction, not replay. The single biggest misconception in this unit: that memory works like a video recording you replay. It doesn't. Retrieval rebuilds the memory from fragments plus expectation and post-event information — which is exactly why the misinformation effect is possible. "Confident" is not the same as "accurate."
Four-choice MCQs in current AP format. Answers and explanations in section (h).
EBQ format. You will read three summarized peer-reviewed sources on a shared topic, then write a response that makes a defensible claim, supports it with evidence from at least two of the three sources, and adds reasoning that applies psychological concepts. Suggested time: 25 minutes.
Prompt topic: How reliable is eyewitness memory, and what factors shape its accuracy?
Using the three sources below, develop an argument that responds to the following question:
To what extent is eyewitness memory a reliable basis for legal testimony? Make a defensible claim and support it with evidence from at least two of the sources, using psychological reasoning that applies course concepts.
Source 1
Method: Participants viewed a video of a staged theft and, two days later, answered a set of questions about it. Half received a questionnaire containing a single misleading detail (a "yield" sign described as a "stop" sign); the other half received a neutral questionnaire. All participants then took an identical recognition test.
Finding: Participants exposed to the misleading detail were significantly more likely to "recognize" the false detail as something they had seen in the video (an error rate roughly three times that of the neutral group). Recognition accuracy for non-manipulated details did not differ between groups.
Source 2
Method: Researchers analyzed 250 real legal cases in which a convicted person was later exonerated by DNA evidence, coding each case for the type of evidence that had contributed to the original conviction.
Finding: Mistaken eyewitness identification was a contributing factor in about 70% of these wrongful convictions — the single most common factor. In a large share of those cases, the original eyewitness had expressed high confidence in the identification at trial.
Source 3
Method: In a controlled experiment, participants watched a simulated crime under one of two conditions: a "weapon present" condition (the perpetrator held a handgun) or a "no weapon" condition (the perpetrator held a neutral object). Afterward, participants attempted to identify the perpetrator from a lineup and to describe his face.
Finding: Participants in the weapon-present condition were significantly less accurate at identifying the perpetrator's face and at lineup identification than those in the no-weapon condition, consistent with a "weapon focus" effect in which attention narrows onto the threat at the expense of other details.
Claim. Eyewitness memory is not a consistently reliable basis for legal testimony, because it is reconstructive and easily distorted by post-event information, witnessing conditions, and even high confidence — so it should be treated with caution rather than trusted automatically. (Claim — 1 pt: a clear, defensible position that answers the question.)
Evidence. Source 1 supports this directly: participants who received a questionnaire containing a misleading detail (a "stop" sign that was actually a "yield" sign) were about three times more likely than the neutral group to falsely "recognize" that detail as part of the original video. This shows that introducing inaccurate post-event information measurably corrupts what witnesses report seeing. Source 2 strengthens the claim with real-world stakes: in an analysis of 250 DNA exonerations, mistaken eyewitness identification was a contributing factor in roughly 70% of the wrongful convictions — the most common factor — and many of those mistaken witnesses had testified with high confidence. Source 3 adds a third pathway to error: participants who saw a weapon during a simulated crime were significantly less accurate at identifying the perpetrator than those who saw no weapon, showing that the conditions of witnessing themselves degrade memory. (Evidence — 3 pts: specific findings cited from all three sources; the Evidence point requires at least two, and this answer uses three.)
Reasoning & Application. These findings cohere because human memory is reconstructive rather than a literal recording: at retrieval, the brain rebuilds an event from fragments and fills gaps with expectations and information acquired after the event. Source 1 is a textbook demonstration of the misinformation effect (Loftus & Palmer, 1974) — misleading post-event wording becomes woven into the reconstructed memory. Source 3 reflects selective attention and limited encoding: under threat, attention narrows onto the weapon, so the perpetrator's face is never encoded well enough to retrieve accurately ("weapon focus"). And Source 2's pairing of error with high confidence underscores a core principle of this unit — confidence and accuracy are largely independent, because a vividly reconstructed memory feels just as real whether or not it is true. Together these mechanisms explain why eyewitness testimony, despite feeling certain to the witness, is an unreliable foundation for conviction. (Reasoning & Application — 2 pts: explains the relationships using accurate course concepts — reconstruction, misinformation effect, encoding/attention, confidence–accuracy independence.)
1. (B). The forgetting curve drops rapidly soon after learning, then levels off — forgetting is front-loaded. (A) "steady" describes a straight line, which the curve is not; (C) and (D) reverse or distort the actual shape.
2. (C) Retrieval failure. The memory was stored all along — a cue (the textbook) unlocked it instantly. (A) encoding failure means it was never stored; (B) decay means it faded and is gone; (D) interference involves competing memories, not a cue suddenly restoring access.
3. (B). Nonsense syllables carry no prior associations, so they isolate new learning without contamination from existing knowledge. (A) is false — they're harder, not easier; (C) is false (savings still occurred); (D) is unrelated.
4. (B) Retroactive interference. New learning (this year's combination) disrupts an old memory (last year's) — the new memory reaches backward. (A) proactive is the reverse direction; (C) and (D) involve no brain damage or failed encoding here.
5. (B) Anterograde amnesia. Inability to form new memories while old ones survive is anterograde — the road ahead is blocked. (A) retrograde is the opposite (losing old memories); (C) and (D) are memory distortions, not amnesia from hippocampal removal.
6. (B). "Smashed" produced higher speed estimates (~40.8 vs. ~31.8 mph for "hit") and, a week later, more false reports of broken glass. (A) reverses the result; (C) ignores the verb's effect (the whole point); (D) didn't happen.
7. (C) Source amnesia / false memory. Priya has a vivid memory but misattributes its source — her brother's story became "her own" experience, a false memory. (A) interference involves competing real memories; (B) flashbulb memories are of shocking news events; (D) decay is loss, not a fabricated memory.
8. (B). Failing to recall a penny's details reflects encoding failure — the details were never stored because you never needed them. (A) is retrieval failure (tip-of-the-tongue); (C) is savings; (D) is retrieval (a cue restoring access).
9. (B). Repression is a Freudian concept that most memory researchers regard skeptically; evidence more often shows trauma is remembered vividly. (A) overstates badly; (C) and (D) misdefine it — repression is motivated forgetting, not encoding failure or decay.
10. (C). The Ebbinghaus curve drops steeply at first, then flattens. (A) is backwards; (B) describes constant linear loss, which the curve is not; (D) misdescribes the shape.
11. (B). Faster relearning (savings) shows a residual trace survived even after recall failed — the memory wasn't fully erased. (A) contradicts savings; (C) is false (encoding did occur); (D) overgeneralizes — savings says nothing about interference being the only cause.
12. (B) Proactive interference. The old, overlearned room number (the prior memory) intrudes on the new one — old interfering with new is proactive. (A) reverses the direction; (C) is brain-damage amnesia; (D) involves post-event misinformation, not competing self-generated memories.
13. (B). Flashbulb memories feel vivid and are held with high confidence but are not guaranteed accurate — they decay and distort like ordinary memories. (A) overstates accuracy; (C) is false and irrelevant; (D) misclassifies them.
14. (B). Differing rates of false "broken glass" reports (32% vs. 14%) tied to the verb show post-event wording shaped the reconstructed memory. (A) and (C) contradict the design (same film, no glass); (D) is the misconception the data refute.
15. (B). The misinformation effect — post-event information altering recall — most directly supports treating confident eyewitness identification cautiously. (A) savings is about relearning, not distortion; (C) interference is a real cause of forgetting but doesn't speak to distortion by post-event information the way the misinformation effect does; (D) anterograde amnesia is about a damaged brain's inability to form memories, not about a healthy witness's memory being reshaped.
---
PsyIQ · Lesson 11 of 30 · Unit 2: Cognition. Q1-style and EBQ practice modeled on the redesigned (2025+) AP Psychology exam. The three EBQ sources are instructional composites written to mirror well-established findings (misinformation effect, DNA-exoneration eyewitness data, weapon-focus effect); they are not verbatim citations. Not affiliated with the College Board. AP is a registered trademark of the College Board. Content pending external psychology QC.