AP Psychology

Lesson 1.1 · Unit 1 · Research methods

How psychologists ask questions: hypotheses and designs

Everyone has theories about people: that opposites attract, that music helps you study. Psychology's advantage is procedure: it states a claim precisely enough that some result could prove it wrong, then collects that result. The design then fixes what you may conclude; experiments alone establish cause.

Lesson 1.2 · Unit 1 · Research methods

Operational definitions and measurement

Psychology studies things nobody can hand you: happiness, intelligence, aggression, grit. These are constructs: ideas with no ruler attached. The fix is procedural: say exactly what you will count. A study is only as good as the sentence that turns its construct into a measurement.

Lesson 1.3 · Unit 1 · Research methods

Research ethics and the APA guidelines

Every rule in this lesson exists because a study once harmed someone. Ethics is not decoration on a finished design; it is a gate the design must pass first: a committee weighs what a study could teach against what it costs the people in it, and can say no.

Lesson 1.4 · Unit 1 · CED topic 1.1

Heredity, environment, and the evolutionary perspective

Nature versus nurture is a badly posed question. Genes never act outside an environment, and environments act on organisms that already have genes. The answerable question is how much of the variation in a trait tracks genetic differences.

That distinction carries history. The eugenics movement read heritability as destiny and used it to justify forced sterilization and exclusion. It was indefensible and statistically wrong, and psychology rejects it.

Lesson 1.5 · Unit 1 · CED topic 1.2

The nervous system, the endocrine system, and the neuron

Your body runs two messaging services at once. The nervous system sends fast, targeted electrical signals down wires; the endocrine system floods the bloodstream with hormones that act slowly and everywhere. Fear needs both: a jump in a quarter second, and a jittery half hour after.

Lesson 1.6 · Unit 1 · CED topic 1.3

Neural firing and neurotransmitters

A neuron is not a dimmer switch. It waits at a negative charge, collects incoming signals, and either fires at full strength or not at all. Everything you feel is built from the timing and chemistry of those identical pulses.

Lesson 1.7 · Unit 1 · CED topic 1.4

Drugs and neural firing

Psychoactive drugs do not invent new experiences. They work by imitating, blocking, or lingering in systems the brain already runs, which is why their effects are predictable from the synapse, and why the brain pushes back against them.

Lesson 1.8 · Unit 1 · CED topic 1.5

The brain: structures, plasticity, and how we study it

The brain is not a set of tidy departments, but damage and imaging both show that specific regions do specific jobs. Learn each structure as an answer to two questions: what stops working when it is gone, and what changes when it is used hard?

Lesson 1.9 · Unit 1 · CED topic 1.6

Sleep and dreaming

Sleep is not the brain switching off. It is a scheduled sequence of states, each with its own electrical signature and its own work, and the work includes finishing what you learned that day. Cut the sequence short and you lose the part you needed.

Lesson 1.10 · Unit 1 · CED topic 1.7

Sensation: transduction and the senses

Nothing physical gets into your head. Light, air pressure, and molecules stop at a receptor, which converts them into the only currency the nervous system accepts. Everything after is the brain working from a translation.

Unit 1 review · 10 multiple-choice

Unit 1 review: Biological Bases of Behavior

Ten questions in the three styles the exam uses (concept application, data analysis, and scientific investigation) one drawn from each lesson in the unit; click an option to see why that choice is right or wrong.

Multiple choice

1. Described study

   A researcher wants to know whether background music helps students learn vocabulary. She posts a sign-up sheet offering two sessions. Students who sign up for the 4 p.m. session study a 30-word list for 20 minutes in a room with instrumental music playing; students who sign up for the 6 p.m. session study the same list for 20 minutes in silence. Both groups take the same recall test immediately afterward. The music group recalls a mean of 19.4 words and the silence group a mean of 22.1 words. The researcher concludes that background music impairs learning.

   Which of the following is the most serious threat to the researcher's causal conclusion?

   The dependent variable, words recalled, was measured immediately rather than after a delay of several days. Students chose their own session, so the two groups were not formed by random assignment and may have differed before the study began. The study did not include a third condition in which the music had lyrics. A difference of 2.7 words is too small to be a real difference between conditions.

   An immediate test narrows what the study can say about durable learning, which is a fair limitation. But the delay was identical in both conditions, and a weakness that applies equally to both groups cannot explain why they differed.

   This is self-selection, and it is the confound that matters here. Students who pick a 4 p.m. slot may differ from students who pick 6 p.m. in sleep, course load, or study habits, and every one of those differences travels with the condition. Random assignment is the design feature that would rule them out.

   A lyrics condition would tell you more about which kind of music matters, but it would leave the existing comparison just as uninterpretable. Adding conditions does not repair groups that differed before the manipulation began.

   You cannot judge whether a 2.7-word gap is real by looking at it; that needs the spread of the scores and a significance test. And even a large, clearly real difference would still have self-selection as a rival explanation.

2. Data: four measures of sleep quality given to the same 150 adults. Test-retest reliability is the correlation between scores taken two weeks apart. The validity column is the correlation between each measure and total sleep time recorded by a wrist actigraph on the same nights.

   Measure	Test-retest r	r with actigraph sleep time
   Single item: "I am a good sleeper" (yes/no)	0.88	0.15
   Nightly 1–5 rating of how rested you feel	0.61	0.24
   Nightly diary of minutes asleep	0.79	0.68
   Bed partner's report of how well you slept	0.55	0.34

   Which measure best illustrates that a measure can be highly reliable and still have poor validity?

   The nightly diary of minutes asleep The bed partner's report The single yes-or-no item, "I am a good sleeper" The nightly 1–5 rating of how rested you feel

   The diary is the measure that works: reasonably consistent across two weeks at 0.79, and by far the closest to the actigraph record at 0.68. It is the example of reliability and validity together, not of the two coming apart.

   The partner's report is the weakest entry on both counts, 0.55 across occasions and 0.34 against the actigraph. A measure that is neither consistent nor accurate cannot illustrate a point that requires one high number and one low one.

   This item is the most consistent of the four across two weeks, at 0.88, and the least related to how long the person actually slept, at 0.15. People answer it the same way every time because it reports a stable self-image, which is exactly how a measure can be perfectly reliable and still miss the construct.

   This measure is mediocre on both counts, 0.61 and 0.24. Because its reliability is the second lowest in the table, it cannot be the example of high reliability paired with weak validity.

3. Described study

   Researchers recruit 48 first-year college students for what the consent form describes as a study of group problem solving. Each participant is seated with three other people who appear to be fellow participants but are working for the researchers and who, on twelve of eighteen trials, give the same obviously wrong answer to a simple perceptual question. The measure is how often the participant gives the group's wrong answer. Participants are told they may stop at any time, and responses are stored without names. Sessions last about twenty minutes.

   Because the procedure uses deception, which additional safeguard is required?

   A debriefing after the session in which the researchers disclose the deception and explain the study's actual purpose. Random assignment of participants to a confederate condition and a no-confederate condition. Anonymity, so that no participant's responses can be linked to a name. Telling participants in advance that some of the people in the room are working for the researchers.

   Deception is permitted only when a study cannot be run without it and participants are told the truth as soon as the session ends. Debriefing restores what an incomplete consent form could not cover, and it is the step this description leaves out.

   Random assignment is a design feature that supports causal claims, not an ethical guideline. Adding a comparison group would strengthen the inference and do nothing at all about the fact that participants were misled.

   The description already says responses are stored without names, so anonymity is in place. It is a genuine guideline, but it is not the one that the use of deception specifically triggers.

   Telling participants beforehand would eliminate the deception rather than justify it, and the behavior under study would vanish with it. The rule is not that deception must be disclosed in advance but that it must be disclosed afterward.

4. A news story reports that a large twin study estimated the heritability of height in a Dutch sample at about 0.80. A reader concludes that 80% of his own height was produced by his genes and 20% by his diet. Which of the following best identifies the error?

   Heritability estimates apply only to psychological traits, not to physical ones such as height. Heritability is the correlation between identical twins, so 0.80 means identical twins match on height 80% of the time. The estimate must be wrong, because genes and environments interact and cannot be separated at all. Heritability describes the share of the differences among people in one population that track genetic differences; it never divides a single person's trait into genetic and environmental parts.

   Heritability is estimated for physical traits as routinely as for psychological ones, and height is one of the standard textbook examples. The trouble with the reader's conclusion is what the statistic refers to, not which kind of trait it describes.

   Heritability is not a twin correlation read off directly. It is calculated from how much more similar identical twins are than fraternal twins, and it describes a population rather than a match rate for any pair of people.

   Genes and environments do interact, but that does not make heritability meaningless. It stays a well-defined statistic about variation within a specific population at a specific time, which is precisely the part the reader drops.

   The statistic is about variance between people: in this Dutch sample most of the differences in height track genetic differences. It does not partition the reader's own height, and in a population where nutrition varied far more, the same trait would show lower heritability.

5. During an unexpected fire alarm, Amara's heart pounds and her hands go cold within seconds. The alarm is cancelled after two minutes, but ten minutes later she is still shaky and her heart is still fast. Which of the following best explains why her arousal outlasts the alarm?

   Her somatic nervous system keeps contracting skeletal muscle until she deliberately relaxes it. Her sympathetic division also triggered the adrenal glands to release epinephrine into her bloodstream, and circulating hormones are cleared far more slowly than neural signals stop. Her parasympathetic division cannot begin working until the pituitary gland signals that the threat has ended. Her sympathetic and parasympathetic divisions are both firing at full strength, and the conflict between them sustains the arousal.

   The somatic nervous system carries voluntary commands to skeletal muscle; it is not what raises heart rate or constricts the vessels in her hands. Those organs are under autonomic control, which is the whole reason the division exists.

   This is the two-speed design of the stress response. The sympathetic division acts in a fraction of a second through nerves, and it also tells the adrenal glands to release epinephrine into the blood, where it keeps acting on the heart until it is metabolized: minutes after the danger has passed.

   The parasympathetic division works through the vagus nerve and does not wait on a pituitary signal. The pituitary belongs to the slower cortisol pathway, not to the switch that turns arousal off.

   The two divisions are not locked in a tug-of-war at full strength; the parasympathetic division becomes dominant as arousal subsides. Sustained shakiness is explained by how long a hormone stays in the blood, not by simultaneous maximum firing.

6. Dmitri brushes a hot pan lightly and feels a mild sting; a moment later he grips it and feels intense pain. Which of the following best describes what differs at the level of the individual sensory neurons?

   Each neuron still fires at the same fixed strength, but the stronger stimulus makes each one fire more often and recruits more neurons into firing. Each action potential is larger, because the size of a neuron's spike is proportional to the strength of the stimulus. The neurons' resting potential rises above threshold, so they no longer need incoming input in order to fire. The refractory period lengthens, which lets charge build up between spikes and produce a larger response.

   This is the all-or-none principle. A neuron either reaches threshold and fires at full strength or does not fire at all, so intensity has to be coded some other way, by how often a neuron fires and by how many neurons are recruited.

   This is the most common misconception about action potentials. Once threshold is crossed the spike is the same size every time; a stronger stimulus does not produce a bigger spike, any more than pressing a doorbell harder makes it louder.

   A resting potential sitting above threshold would mean the neuron fired continuously with no input at all, which would make it useless as a signal. Resting potential stays near −70 mV, and it is depolarization from incoming input that carries the cell to roughly −55 mV.

   The refractory period is the brief window after firing during which the neuron cannot fire again, so lengthening it would lower the maximum firing rate rather than raise the signal. Charge does not accumulate across spikes to build a larger one.

7. Data: described bar graph of mean fluid consumed per day by two groups of laboratory rats.

   The horizontal axis has two groups: rats housed alone in standard cages, and rats housed together in a large social colony. Each group has two bars, one for a sweetened morphine solution and one for plain water. The vertical axis is mean millilitres consumed per day and runs from 0 to 20. For the isolated group the morphine bar reaches 8.4 mL and the water bar 12.0 mL. For the colony group the morphine bar reaches 2.1 mL and the water bar 18.3 mL.

   Which conclusion is best supported by the graph?

   The colony rats drank less fluid overall, which is why their morphine consumption was lower. Morphine is less addictive than is commonly believed, since most of the animals avoided it. Both groups drank the same total volume, 20.4 mL, but the isolated rats took about 41% of it as morphine solution against about 10% for the colony rats, so housing is related to how much of the drug was consumed. Living alone causes substance use disorder in people who live by themselves.

   Add the two bars for each group and both totals come to 20.4 mL. Total fluid intake is effectively held constant across the groups, so it cannot explain a four-fold difference in morphine solution.

   The graph says nothing about addictiveness in general; it compares two housing conditions in one species. The colony animals' low intake is a fact about those rats in that environment, not a property of the drug.

   8.4 plus 12.0 and 2.1 plus 18.3 both equal 20.4 mL, so the groups differ in what they drank rather than in how much. 8.4 of 20.4 is about 41% and 2.1 of 20.4 about 10%: the comparison the display actually licenses.

   Two problems at once: the participants were rats, and even inside the study, housing was manipulated in a laboratory rather than across human lives. A causal claim about substance use disorder in people needs evidence this design cannot supply.

8. After a stroke, a woman speaks in short, effortful bursts, "want... water... cup", and is visibly frustrated by her own speech. She follows spoken instructions accurately and points to the correct object every time she is asked. Damage to which area best accounts for this pattern?

   Wernicke's area, in the left temporal lobe The hippocampus, in the medial temporal lobe The cerebellum, behind the brainstem Broca's area, in the left frontal lobe

   Damage to Wernicke's area produces the opposite pattern: speech that flows easily but carries little meaning, paired with impaired comprehension. This woman understands everything said to her, which rules that area out.

   Hippocampal damage blocks the formation of new explicit memories; it does not disrupt the production of speech. She is not failing to remember the word: she knows what she wants to say and cannot get it out.

   The cerebellum supports balance, coordinated movement, and procedural learning. Cerebellar damage can slur articulation, but it does not produce the grammar-stripped, telegraphic output described here.

   Non-fluent aphasia (halting, effortful speech with comprehension largely intact and the person aware of the problem) is the signature of damage to Broca's area. Its position beside the motor cortex also explains why right-side weakness so often accompanies it.

9. Data: one adult's overnight sleep recording, divided into four successive cycles of about 90 minutes. The table gives the minutes of N3 (slow-wave) sleep and of REM sleep in each cycle; the remaining minutes in each cycle were N1 and N2.

   Cycle	1	2	3	4
   N3 sleep (minutes)	45	30	10	0
   REM sleep (minutes)	8	18	32	45

   A student whose sleep follows this pattern goes to bed, sleeps through cycles 1 and 2, about three hours, and then gets up to study. Which of the following is best supported by the table?

   She loses most of the night's slow-wave sleep, because N3 clusters in the final cycles. She loses 77 of the night's 103 REM minutes, about three-quarters, while keeping 75 of its 85 N3 minutes, because REM periods lengthen across the night while N3 periods shrink. She loses REM and N3 in roughly equal proportion, because every sleep cycle has the same composition. She loses nothing that matters for consolidating a skill she practised that day, because consolidation happens entirely during N3, which she kept.

   The table shows the reverse. N3 falls from 45 minutes in cycle 1 to none at all in cycle 4, so slow-wave sleep is front-loaded and a shortened night keeps most of it.

   Add the columns she skipped. Cycles 3 and 4 hold 32 plus 45, or 77, of the night's 103 REM minutes, close to 75%, while the two cycles she kept contain 75 of the 85 N3 minutes. That asymmetry is why a short night costs REM first, and why REM rebound tends to follow the next night.

   The four cycles in the table are plainly not identical: N3 runs 45, 30, 10, 0 while REM runs 8, 18, 32, 45. Treating cycles as interchangeable is the assumption these data contradict.

   Consolidation is not confined to one stage; overnight gains on motor sequence tasks are associated with sleep containing both slow-wave and REM periods. Keeping N3 while losing three-quarters of REM is a real cost, not a neutral trade.

10. In the last hour of a long shift, a baggage screener fails to flag a prohibited item that is clearly visible on the scanner image. Her supervisor then announces that inspectors are running test images through the line, and over the next hour she flags several harmless objects as suspicious. Which of the following best explains the change in her performance?

    Her absolute threshold for visual detail fell, so she could detect fainter images than before. Sensory adaptation to the scanner display first reduced and then restored the firing of her retinal receptors. Her response criterion shifted: the warning made a miss more costly than a false alarm, so she began reporting on weaker evidence, which produces both more hits and more false alarms. Weber's law predicts that the just-noticeable difference between a threat and a harmless object grows as a shift goes on.

    An absolute threshold is the weakest stimulus a person detects half the time, and it does not drop because a supervisor makes an announcement. Nothing about her eyes changed; what changed is how much evidence she demanded before saying yes.

    Sensory adaptation is a receptor-level decline in response to unchanging stimulation, and the images on her screen change every few seconds. It also cannot explain a rise in false alarms, which is a decision made about ambiguous evidence.

    Signal detection theory separates sensitivity from the criterion. Fatigue and a long uneventful stretch pushed her criterion toward "nothing there," producing a miss; the warning raised the cost of a miss and dragged the criterion down, and the predictable by-product of catching more real threats is flagging more harmless objects.

    Weber's law describes the proportional size of a just-noticeable difference between two stimuli, such as two weights or two tones. It is about discriminating magnitudes, not about the decision rule a person adopts when the incentives change.

Lesson 2.1 · Unit 2 · CED topic 2.1

Perception and attention

Your eyes deliver a flat, upside-down smear of light; you experience solid objects at measurable distances. Everything in between is perception: built partly from the signal and partly from what the brain expects. Attention is the gate on that construction, which is why something can sit in plain view and never be seen.

Lesson 2.2 · Unit 2 · CED topic 2.2

Thinking, problem solving, judgment, and decision making

Most thinking is fast, cheap, and good enough. The shortcuts that make it fast are not random errors: they fail in predictable directions, which is exactly why psychologists can study them. Learn each bias as a shortcut doing its job in the wrong situation.

Lesson 2.3 · Unit 2 · CED topic 2.3

Introduction to memory: the three-stage model

Memory is not a single container. Information passes through stores with different capacities and different lifespans, and almost everything is dropped at the first two. That is a feature: a system that kept every sound and glance would have nothing left to think with.

Lesson 2.4 · Unit 2 · CED topic 2.4

Encoding memory: depth, spacing, and testing

Studying feels like exposure: more hours in front of the page, more learning. It is not. What ends up stored depends on what you did with the material while it was in front of you, and on when you came back to it. Both of those are choices.

Lesson 2.5 · Unit 2 · CED topic 2.5

Storing memory: systems, structures, and amnesia

Long-term memory is not one archive. Different kinds of learning are supported by different structures, which is why a single injury can destroy the ability to remember yesterday and leave the ability to acquire a new skill completely intact.

Lesson 2.6 · Unit 2 · CED topic 2.6

Retrieving memory: cues, context, and state

A memory you cannot produce is not necessarily a memory you lost. Most ordinary forgetting is a failure of access: the file is intact and the search term is wrong. That makes retrieval something you can design for instead of hope for.

Lesson 2.7 · Unit 2 · CED topic 2.7

Forgetting and memory distortion

Memory does not replay; it rebuilds. Every retrieval assembles a version out of fragments, general knowledge, and whatever you have heard since, and the rebuilt version feels exactly as vivid as an accurate one. That is why confidence is such a poor guide to accuracy.

Lesson 2.8 · Unit 2 · CED topic 2.8

Intelligence and achievement: theories and testing

Intelligence is a construct, not an organ, so arguments about it are partly arguments about what a test ought to measure. Two questions run through the topic: is intelligence one thing or many, and does a score describe a person or a moment?

Lesson 2.9 · Unit 2 · Research methods

Descriptive statistics: center, spread, and the normal curve

A set of scores is not a finding until you say something about its shape. Two numbers do most of the work (one for where the scores sit, one for how far apart they are), and picking the wrong one is how honest data end up misleading people.

Lesson 2.10 · Unit 2 · Research methods

Correlation: scatterplots, r, and what it cannot tell you

Correlation is the workhorse of psychology, because most of what we want to know cannot be assigned to people. It is also the most misread statistic on the exam. The questions are almost never about computing r; they are about what r does not entitle you to say.

Unit 2 review · 10 multiple-choice

Unit 2 review: Cognition

Ten questions in the three styles the exam uses (concept application, data analysis, and scientific investigation) one drawn from each lesson in the unit; click an option to see why that choice is right or wrong.

Multiple choice

1. During a chemistry demonstration Priya is counting how many times the instructor stirs a beaker. She does not notice a classmate walk slowly across the front of the room holding a large yellow sign. Which of the following best explains what happened?

   Change blindness: she failed to detect a difference between two successive views of the same scene. Sensory adaptation: constant stimulation reduced her visual receptors' response to the classmate. Inattentional blindness: her attention was committed to the counting task, so a fully visible but unexpected event never reached awareness. Perceptual set: her expectations about the demonstration changed what she saw the sign to say.

   Change blindness needs two views with something altered between them: a photograph that differs after a flicker, a person swapped during an interruption. Nothing here changed; an object was present the whole time and simply never registered.

   Sensory adaptation happens at the receptor level and requires constant, unchanging stimulation, like the smell of a room you have been sitting in for an hour. A classmate walking past is novel and moving, and Priya's retina responded to him normally.

   This is the finding behind the gorilla-in-the-video study: a demanding counting task consumes attention, and an unexpected object in plain view goes unreported. Her eyes worked; the bottleneck was attention, which is why "she wasn't paying attention" names the outcome rather than the mechanism.

   A perceptual set biases how an ambiguous stimulus is interpreted, as when you read a smudged letter as the one you expected. Priya did not misread the sign: she never saw it at all.

2. Theo cancels a beach vacation after a week of news coverage of a shark attack, then drives eight hours to the mountains instead without a second thought, even though the drive carries the far higher risk of injury. Which of the following best explains his judgment?

   The availability heuristic: vivid, heavily covered events come to mind easily, and ease of retrieval is mistaken for frequency. The representativeness heuristic: he judged the beach by how well it matched his prototype of a dangerous place and ignored base rates. Confirmation bias: he sought out news stories that agreed with a fear he already held. Anchoring: the first risk estimate he encountered pulled all his later estimates toward it.

   Shark attacks are rare and highway crashes are common, but coverage reverses how easily each comes to mind. The availability heuristic substitutes "how quickly can I recall an example" for "how often does this actually happen," and a week of coverage is what loaded his memory.

   Representativeness judges probability by resemblance to a stereotype: deciding someone is a librarian because he is quiet and tidy. Theo is not matching the beach to a category; he is over-weighting one event he can picture vividly.

   Confirmation bias would mean he already feared sharks and then went looking for stories that supported it. In the stem the coverage arrives first and the fear follows, which is availability rather than a biased search.

   Anchoring requires a specific number to anchor on: a listed price, a first offer, an initial estimate that later judgments drift toward. No figure appears anywhere in the stem.

3. A pharmacist hears a 12-character prescription code read aloud once and, without writing it down, types it correctly ten seconds later. On the way to the keyboard she has recoded it as a four-digit year, a three-letter drug abbreviation she uses daily, and a five-digit lot number. Which of the following best explains her success?

   Iconic memory held the whole code, which is why every character was still available ten seconds later. The code was transferred into long-term memory by deep semantic processing during those ten seconds. Her short-term capacity is unusually large, closer to twelve items than to the typical seven. Chunking: she reorganized twelve characters into three meaningful units, which fits comfortably inside ordinary short-term capacity while rehearsal in the phonological loop keeps them alive.

   Iconic memory is a visual store that fades in well under a second, and she heard the code rather than seeing it. Even echoic memory, its auditory equivalent, lasts only a few seconds and could not carry her to the keyboard.

   Deep semantic processing builds durable long-term memories, which is not what happened here: ask her an hour later and the code will be gone. Recoding something for immediate use is not the same as encoding it for storage.

   Capacity limits are remarkably stable across people; experts do not get extra slots, they pack more into each slot. Assuming she has unusual capacity skips the mechanism the stem hands you.

   Chunking changes what counts as an item. Three familiar units sit well inside the roughly seven-item limit, and the phonological loop rehearses them for the few seconds she needs. This is why experienced people look like they remember more within their own field.

4. Data: described bar graph. Participants answered one question about each of 60 words (about the word's appearance, its sound, or its meaning), and then took an unexpected recognition test.

   The vertical axis is the percentage of words later recognized, running from 0 to 80%, and there are three bars. The bar for appearance questions ("Is the word printed in capital letters?") reaches 18%. The bar for sound questions ("Does the word rhyme with train?") reaches 39%. The bar for meaning questions ("Does the word fit in this sentence?") reaches 63%. A note beneath the graph gives the mean time taken to answer each type of question: 0.6 seconds for appearance, 0.8 seconds for sound, and 1.1 seconds for meaning.

   Which statement is best supported by the data as presented?

   Meaning-based encoding caused the higher recognition, and the answer times rule out time spent with each word as an alternative explanation. Recognition rose as the question required deeper processing (63% for meaning against 18% for appearance, about three and a half times as many words), but because the meaning questions also took longest to answer, time spent with each word remains an alternative explanation these data cannot rule out. The data show that how often a word is repeated is what determines whether it is recognized later. Because every participant saw the same 60 words, this design contains no possible confound.

   The answer times point the other way. Meaning questions took nearly twice as long as appearance questions, so processing time varies right alongside depth instead of being held constant, which is what a confound looks like.

   Both halves of this statement come straight off the display: 63 divided by 18 is about 3.5, and the note shows the deepest condition also consumed the most time. Naming the finding and then the alternative explanation it cannot exclude is exactly what a data-analysis item rewards.

   Repetition is held constant here: every word was presented once and only the question about it differed. A variable that does not vary cannot explain a difference in the results.

   Using the same words for every participant controls for word difficulty, which is a real strength, but it does nothing about the timing difference the note reports. Controlling one variable is not controlling all of them.

5. A man who developed anterograde amnesia after damage to both hippocampi is taught to use an unfamiliar espresso machine. Across five daily sessions his time to produce a cup falls steadily from four minutes to under one. Asked each day whether he has used the machine before, he says he has never seen it. Which of the following best accounts for this pattern?

   Retrograde amnesia prevents him from recalling events from before the damage, including the previous day's sessions. His working memory is intact, so he holds the sequence of steps in mind from one day to the next. Procedural learning is implicit and does not depend on the hippocampus, so the skill improves while the explicit, episodic memory of having practised it cannot be formed. The improvement reflects sensory adaptation to the machine rather than any form of memory.

   Retrograde amnesia erases memories formed before an injury, and these sessions happen after it. His difficulty is with the new, which is precisely what anterograde amnesia means.

   Working memory holds information for seconds, not across a night's sleep. Nothing held in working memory could survive from one daily session to the next, so it cannot explain a five-day improvement.

   This dissociation is the classic finding from the patient known as H.M.: medial temporal damage blocks new explicit memories while procedural learning, supported by the cerebellum and basal ganglia, proceeds normally. He knows how without knowing that.

   Sensory adaptation is a decline in receptor response to constant stimulation; it does not make anyone faster at a sequence of actions. A steady gain across five sessions is learning, not a change in sensitivity.

6. Described study

   Researchers tested 32 members of a university diving club. Each diver learned one list of 36 unrelated words while standing on a beach, and a second, equivalent list while submerged about fifteen feet underwater in scuba gear. Each diver was then tested for free recall twice: once in the setting where that list had been learned and once in the other setting. The order of lists and settings was counterbalanced across divers. Recall was substantially higher when the learning and testing settings matched than when they did not.

   Every diver completed all four learning-and-testing combinations. What does this feature of the design accomplish?

   It controls for individual differences: because each diver serves as his or her own comparison, the gap between matched and mismatched conditions cannot be explained by some divers simply having better memories. It makes the study correlational rather than experimental, since nobody was assigned to a separate group. It guarantees that the results will generalize to students studying in ordinary classrooms. It removes the need for an operational definition of recall, because the same people are measured more than once.

   This is a repeated-measures, or within-subjects, design. The comparison happens inside each person, so stable differences in memory ability, diving experience, and motivation are held constant instead of being left to chance.

   The match between learning and testing setting was manipulated by the researchers, and every diver met every combination, which makes this an experiment. Repeated measures is a way of assigning conditions, not a retreat from manipulating them.

   Generalization depends on who the participants are and how large the manipulation is, not on how conditions were assigned. A beach and open water differ far more than a dorm room and a classroom, so the everyday version of this effect is much smaller.

   Recall still has to be operationalized: here as the number of words from a 36-word list produced without cues. Measuring the same people repeatedly changes what the comparison controls, not whether the measure needs defining.

7. Described study

   Researchers showed 90 adults at a driving school the same 30-second video of a collision between a car and a bicycle. Participants who attended the morning sessions were asked how fast the car was going when it smashed into the bicycle; those who attended the afternoon sessions were asked how fast it was going when it hit the bicycle. Everything else about the questionnaire was identical. Mean speed estimates were 41 mph in the morning group and 34 mph in the afternoon group. The researchers concluded that the verb changed the estimates.

   Which of the following is the most serious threat to that conclusion?

   A speed estimate in miles per hour is not a valid operational definition of what a person remembers. Wording was assigned by session time rather than at random, so anything that separates morning attendees from afternoon attendees is confounded with the verb. Watching a video is not as stressful as witnessing a real collision, so the estimates are too low. The study needed a third group asked with a neutral verb such as "contacted."

   A speed estimate is a serviceable operational definition: a number, produced under stated conditions, that another researcher could collect the same way. The weakness here is not the measure but what produced the difference between the groups.

   Random assignment is what makes two groups equivalent before a manipulation. Assigning by session time lets a whole bundle of differences (alertness, driving experience, who is free in the morning) ride along with the verb, so the 7 mph gap has more than one candidate explanation.

   This is a genuine limitation on generalizing to real witnesses, and it would affect the absolute level of both groups' estimates. But it applies equally to everyone, so it cannot account for the difference between the groups.

   A neutral-verb condition would make the finding richer by showing which direction the shift runs from. It would not repair a comparison in which the groups were formed by when people happened to show up.

8. Naomi is a strong mathematics student. Just before a difficult placement test, the proctor remarks that the test is designed to measure whether students like her can handle advanced coursework. Naomi spends much of the period worrying about how her score will be read and finishes fewer items than usual. Which concept best explains her performance?

   Test bias, because the test predicts later performance less accurately for her group than for others. The Flynn effect, because average performance on tests like this one has risen across generations. A self-fulfilling prophecy, in which the proctor's low expectations changed how he scored her work. Stereotype threat: making a negative stereotype salient creates worry that consumes working memory, lowering performance on a test the person is otherwise prepared for.

   Test bias is a property of a test's predictive accuracy across groups, established with data on how scores forecast later outcomes. Nothing in the stem concerns prediction; what changed was Naomi's state in the room.

   The Flynn effect is the long-run rise in raw test scores across generations, a fact about norms over decades. It has no bearing on one student's performance on one afternoon.

   A self-fulfilling prophecy runs through someone else's expectations changing their behavior toward the person: different teaching, different grading, fewer opportunities. The proctor scored nothing differently here; the effect ran through Naomi's own attention.

   Stereotype threat is the term the exam wants for this pattern: making a negative stereotype relevant adds a monitoring load, and working memory spent on worry is unavailable for the problems. The prediction is specific: the effect shows up on difficult items and disappears when the framing is removed.

9. Data: summary statistics for the same 100-point final project in two sections of one course, each with 28 students.

   Statistic	Section A	Section B
   Mean	78	78
   Median	78	85
   Standard deviation	4	14
   Lowest score	70	31
   Highest score	86	96

   Which statement is best supported by the table?

   Section B's scores are negatively skewed, a few very low scores pull the mean below the median, so the median of 85 describes a typical Section B student better than the mean of 78 does. Section B is positively skewed, because the distance between its lowest and highest scores is much larger than Section A's. Students in Section A learned more, because their scores are far more consistent. Because both sections have a mean of 78, the mean describes a typical student equally well in both.

   When the mean sits below the median, the tail runs to the left: the extreme scores are low ones, and they drag the mean down while barely moving the middle value. A low score of 31 and a standard deviation of 14 are what produce that gap, and the median is the defensible summary.

   Positive skew means a long right tail, with the mean above the median. Section B shows the opposite ordering, mean 78 and median 85, and a wide range by itself tells you nothing about which direction the tail runs.

   Spread measures how alike the scores are, not how much anyone learned. Section B's median is seven points higher than Section A's, so the section with the tighter distribution is not the higher-scoring one.

   Equal means do not imply equally useful means. In Section A one standard deviation puts roughly two-thirds of students between 74 and 82; in Section B the same band runs from 64 to 92, which describes almost nobody.

10. Data: correlations between four self-reported study habits and final exam score, from a survey of 180 students in one course. No variable was manipulated.

    Variable	Correlation with final exam score
    Hours per week spent self-testing	+0.41
    Hours per week spent rereading notes	+0.06
    Hours of sleep the night before the exam	+0.22
    Hours per week at a part-time job	−0.35

    Which statement is best supported by these data?

    Rereading notes is negatively related to exam performance. Self-testing has the strongest linear relationship with exam score of the four variables, but because nothing was manipulated, these data cannot show that self-testing raises scores. Hours at a part-time job matter less than hours of sleep, because the correlation for the job is negative. Students who cut their work hours will raise their exam scores.

    An r of +0.06 is essentially no linear relationship, and its sign is positive in any case. Reading a near-zero coefficient as a negative relationship confuses a weak association with a reversed one.

    Strength is distance from zero, so +0.41 is the largest of the four, just ahead of −0.35. The second clause is the part the exam is really testing: in a survey with no manipulation, students who self-test more may also be the students who attend more, sleep more, or care more.

    Direction and strength are separate readings. The sign says which way the relationship runs; the distance from zero says how strong it is, and 0.35 is larger than 0.22, so the job variable is the stronger of the two.

    This is a causal claim drawn from correlational data, with a prescription stacked on top of it. Work hours may stand in for financial pressure, commuting time, or lost sleep, and cutting the hours would not remove those.

Lesson 3.1 · Unit 3 · CED topic 3.1

Themes and methods in developmental psychology

Developmental psychology asks how people change across a whole life. That creates a problem no other subfield has: the variable you care about is time, and nobody can assign it.

Three arguments run under the unit: smooth growth or stages, stability or change, body or experience.

Lesson 3.2 · Unit 3 · CED topic 3.2

Physical development across the lifespan

The sequence of physical development is close to fixed; the timing is not. Knowing which parts run on a biological clock and which wait on experience is what separates ordinary variation from a delay worth checking.

Lesson 3.3 · Unit 3 · CED topic 3.3

Gender and sexual orientation

Everyday talk bundles three things together that psychology pulls apart: what a body is, what a culture expects, and whom a person is drawn to. They vary independently, and precise, respectful terms are part of the science rather than a courtesy added afterward.

Lesson 3.4 · Unit 3 · CED topic 3.4

Cognitive development across the lifespan

Children are not small adults with less information. For a while they reason by rules that are internally consistent and simply wrong, and the errors are the evidence: they show what a mind can and cannot yet do.

Lesson 3.5 · Unit 3 · CED topic 3.5

Communication and language development

No one teaches a toddler grammar, yet the order in which it arrives is nearly identical across languages. The puzzle is how much of that schedule a child brings and how much the people talking to her supply.

Lesson 3.6 · Unit 3 · CED topic 3.6

Social-emotional development across the lifespan

A baby arrives with a reactive style already in place and starts building relationships on top of it. The question is what those relationships look like when they work, and how anyone measures something as vague-sounding as a bond.

Lesson 3.7 · Unit 3 · CED topic 3.7

Classical conditioning

Some learning takes no effort and no intention. A signal that reliably precedes something important starts producing the reaction on its own, which is why a ringtone can tighten your chest before you read the name on the screen.

Lesson 3.8 · Unit 3 · CED topic 3.8

Operant conditioning

Classical conditioning explains reactions that happen to you. Operant conditioning explains what you do on purpose: behavior is shaped by what follows it, and the timing of consequences matters more than their size.

Lesson 3.9 · Unit 3 · CED topic 3.9

Social, cognitive, and neurological factors in learning

Conditioning cannot explain learning you never performed and were never rewarded for. Watching is enough, and so is thinking: something happens between stimulus and response that a strictly behaviorist account leaves out.

Lesson 3.10 · Unit 3 · Research methods

Effect size, significance, and replication

A finding is not a fact until it survives being run again. Two questions decide how much weight a result carries: how big the difference is, and whether anyone else can produce it.

Unit 3 review · 10 multiple-choice

Unit 3 review: Development and Learning

Ten questions in the three styles the exam uses (concept application, data analysis, and scientific investigation) one drawn from each lesson in the unit; click an option to see why that choice is right or wrong.

Multiple choice

1. Described study

   To study how memory for names changes with age, researchers recruit 90 volunteers in three groups: 30 people aged 20 to 29, 30 aged 50 to 59, and 30 aged 75 to 84. All are tested on the same afternoon in the same room. Each learns twenty face-and-name pairs and is tested ten minutes later. The mean number of names recalled is 14.2 in the youngest group, 11.6 in the middle group, and 8.3 in the oldest group. The researchers conclude that memory for names declines with age.

   Which feature of this design most limits that conclusion?

   The dependent variable was measured only ten minutes after learning. It is a cross-sectional design, so age is confounded with cohort: the three groups differ in schooling, health care, and life experience as well as in age. Thirty participants per group is too few for any comparison between the groups to be meaningful. The researchers should have used random assignment to place participants into the three age groups.

   A ten-minute delay limits what the study can say about long-term retention, but every group faced the same delay. A feature shared by all three conditions cannot explain the differences among them.

   Cross-sectional designs compare different people at a single moment, so everything that separates someone born in 1945 from someone born in 2000 travels with the age difference. A longitudinal or sequential design is what pulls age and cohort apart.

   Thirty per group is a workable sample for a comparison of this kind, and the reported differences are large. Sample size affects how precise an estimate is; it is not the structural problem in this design.

   Participants cannot be randomly assigned to an age: age is a participant variable, not something a researcher can hand out. That is exactly why developmental research so often has to settle for quasi-experimental comparisons.

2. Jonas, 16, can explain in detail why texting while driving is dangerous and scores well on the written portion of his driving test. With three friends in the car on a Friday night, he texts anyway. Which of the following best explains the gap between what he knows and what he does?

   Subcortical reward and social-sensitivity systems mature earlier in adolescence than the prefrontal cortex, which keeps developing into the twenties, so impulse control lags behind reasoning: especially when peers are present. He has not yet reached the formal operational stage, so he cannot reason about hypothetical consequences. Synaptic pruning has not yet begun, so his brain holds too many connections to act efficiently. The personal fable leaves him literally unable to perceive that the behavior carries any risk.

   The developmental mismatch is the point. Reasoning about risk is fully available to him, the written test proves it, but the systems that make a reward feel urgent, particularly a social reward, come online years before the frontal machinery that brakes them.

   Formal operational reasoning is precisely what he demonstrates when he explains the danger and passes the written test. His difficulty is not thinking about hypotheticals; it is which system wins in the moment.

   Synaptic pruning begins in early childhood and continues prominently through adolescence, so it is not a process still waiting to start. The raw number of synapses is not what settles a single decision either.

   The personal fable, the sense of being uniquely protected, is a real feature of adolescent egocentrism and may well contribute here. But "literally unable to perceive" overstates it: he perceives the risk and discounts it, which is the distinction this item is testing.

3. Described study

   A gender-role inventory built in the early 1970s was scored from ratings by American undergraduates of how desirable each of hundreds of traits was for a man, or for a woman, in their society. Twenty traits rated far more desirable for men, twenty far more desirable for women, and twenty rated neutral became the scale. A researcher now administers that inventory, with its original scoring key, to 800 undergraduates at the same university and finds that far fewer students score in the masculine range than the original norms recorded.

   Which interpretation of the new result is most defensible?

   Students today are less masculine than students were in the 1970s. The inventory has low test-retest reliability, since the scores it produces have changed. Because neither sample was randomly selected from all adults, no comparison between them is possible. The scoring key records what one group of raters in one culture and one decade judged desirable for each sex, so a shift in scores may reflect changed cultural norms rather than changed people.

   This treats the scale as a fixed yardstick, when the yardstick itself was built out of one era's ratings. A trait that counted as masculine in 1974 may simply no longer be sorted that way, which would move scores without anyone changing at all.

   Test-retest reliability concerns whether the same people score similarly across a short interval, not whether two different samples fifty years apart score alike. Genuine change over decades is not evidence that a measure is inconsistent.

   Neither sample was drawn randomly from all adults, yet comparing two student samples from the same university is still informative. The caution belongs on what the scale's norms mean, not on whether any comparison can be made.

   This is the limitation the exam expects you to name for scale-construction and survey research: the norms are cultural data. The result is genuinely interesting and genuinely ambiguous between a change in people and a change in what the culture codes as masculine, and this design cannot separate the two.

4. A five-year-old watches as ten pennies in a tight row are spread out into a longer row. He counts ten pennies before the change and ten after it, then insists the longer row now has more. Which of the following best describes his reasoning?

   He lacks object permanence, so the pennies stop existing for him while they are being moved. He is accommodating his schema for number to fit the new arrangement. He fails to conserve number: he centers on the length of the row and cannot mentally reverse the spreading, which is characteristic of preoperational thought. He is showing egocentrism, because he assumes the adult sees the row exactly as he does.

   Object permanence, knowing a hidden object still exists, is typically in place before a child's second birthday, and these pennies were never hidden. He can see all ten of them the entire time.

   Accommodation means changing a schema when new information will not fit it, which is the thing he is failing to do. He is assimilating the spread-out row to a rule he already holds: longer looks like more.

   Conservation is the understanding that quantity is unchanged by a change in appearance. Centering on a single dimension and being unable to mentally undo the transformation are the two limits Piaget identified, and both show in his answer even though his counting is accurate.

   Egocentrism is difficulty taking another person's perspective, as when a child hides by covering his own eyes. Nothing in this task turns on what he believes the adult can see.

5. At two, Lena said "I ran" and "two mice." At three and a half she says "I runned" and "two mouses." Her grandfather worries that her language is going backwards. Which of the following is the best response?

   The change reflects telegraphic speech, in which function words and endings are dropped from short utterances. This is overgeneralization, and it is a sign of progress: she has extracted the regular past-tense and plural rules and is applying them to irregular words she used to produce as memorized chunks. She has entered a critical period during which errors of this kind are expected and then disappear on their own. It is a phonological error: she can hear the irregular forms but cannot yet articulate them.

   Telegraphic speech is the two-word stage, "want cookie," "Daddy go", in which endings and function words are missing altogether. Lena is adding endings, not dropping them.

   Earlier, "ran" and "mice" were whole items copied from adults. Now she has abstracted the rules (add -ed, add -s), and applies them everywhere, including where English is irregular. The error is the evidence for the rule, which is why it appears after the correct forms rather than before them.

   Critical and sensitive periods describe windows during which experience has an outsized effect, such as the early years for a first language. The label does not explain why one particular error emerges, and "they disappear on their own" is not a mechanism.

   She produces both the -ed and the -s ending accurately, so articulation is not the problem, and she says "runned" rather than mispronouncing "ran." The error is grammatical, not phonological.

6. Data: 100 one-year-olds observed in a laboratory playroom through a standard sequence of separations from and reunions with a caregiver. Trained coders classified each infant's reunion behavior and recorded two timed measures.

   Classification	Secure	Avoidant	Resistant
   Number of infants	65	21	14
   Mean seconds exploring while caregiver present	142	118	63
   Mean seconds of contact-seeking at reunion	38	9	165

   A reader concludes from the table that the avoidant infants were the least distressed by the separation. Which response is best supported by the data?

   The conclusion is supported, because seconds of contact-seeking is a direct measure of distress. The conclusion is supported, because the avoidant infants also explored more than the resistant infants did. The conclusion is contradicted, because the table shows that avoidant infants explored the least of the three groups. The conclusion goes beyond the data: the table records contact-seeking at reunion and exploration while the caregiver is present, never distress during the separation, so a low contact-seeking score describes a reunion strategy rather than how upset an infant was.

   Contact-seeking is a behavior, not a feeling, and the whole point of an operational definition is that the two are not interchangeable. An infant can be highly aroused and still not approach the caregiver.

   Exploration while the caregiver is present speaks to whether the infant uses her as a secure base, and it is measured before the separation even happens. It cannot establish what the separation did.

   The table shows resistant infants exploring the least, at 63 seconds, with avoidant infants at 118. That column does not contradict the reader's claim: it simply does not address it.

   This is a "what the data do not show" item, and the answer sits in the row labels. Neither timed measure records separation distress, so the reader has quietly swapped a coded reunion behavior for an internal state the study never measured.

7. A patient who is receiving chemotherapy eats a lemon drop in the waiting room before each infusion. The drug reliably causes nausea a few hours later. After four rounds, the smell of lemon alone makes her feel sick. In this example, what is the conditioned stimulus?

   The nausea produced by the drug The lemon flavour and smell The chemotherapy drug The queasiness she now feels at the smell of lemon

   Nausea produced directly by the drug is the unconditioned response: an automatic reaction that required no learning. A response is never a stimulus in this scheme, so it cannot be the conditioned stimulus.

   The lemon was neutral before the pairings and now triggers the reaction on its own, which is the definition of a conditioned stimulus. Taste aversions like this one form remarkably fast, often after a single pairing, because organisms are biologically prepared to link flavours with illness.

   The drug is the unconditioned stimulus: it produces nausea with no learning history at all. Confusing the unconditioned stimulus with the conditioned stimulus is the most common error on these items, and the test is always which one was neutral to begin with.

   Her queasiness at the smell of lemon is the conditioned response: the learned reaction to the conditioned stimulus. It looks much like the unconditioned response, but it is triggered by the signal rather than by the drug.

8. Priya checks her email roughly every four minutes throughout the day. Messages she cares about arrive a few times a day at unpredictable times, and checking more often does not make one arrive any sooner. Which reinforcement schedule best describes her checking?

   Fixed-ratio, because a set number of checks is required before a message appears. Variable-ratio, because the number of checks needed before a reward varies unpredictably. Variable-interval, because reinforcement becomes available after unpredictable amounts of time, regardless of how many times she checks. Fixed-interval, because messages arrive at regular, predictable times.

   A fixed-ratio schedule pays off after a set number of responses (every tenth sale, every fifth punch on a coffee card), and produces a burst-and-pause pattern. Nothing in the stem counts her checks.

   Variable-ratio is the slot-machine schedule, where responding faster genuinely earns more payoffs. The stem rules it out directly: checking more often does not make a message arrive sooner, so the reward is not tied to her response count.

   The question that separates these schedules is whether time or responses control availability. A message becomes available as time passes, and her checks matter only in that one of them eventually finds a waiting message: the signature of a variable-interval schedule, which produces a steady, moderate rate of responding.

   Fixed-interval means reinforcement becomes available after a set amount of time, which produces the scalloped pattern of a student who studies harder as a weekly quiz approaches. Her messages arrive unpredictably, so the interval is variable rather than fixed.

9. Data: described bar graph. Preschoolers who already chose to draw during free play were randomly assigned to three conditions. In the expected-reward condition each child was promised a certificate for drawing; in the unexpected-reward condition each child received the same certificate afterward with no prior promise; in the no-reward condition no certificate was mentioned or given.

   The vertical axis is the percentage of free-play time spent drawing, running from 0 to 25%, and the horizontal axis has three pairs of bars, one pair per condition. The lighter bar in each pair is the baseline measured before the conditions began and the darker bar is the measurement taken two weeks afterward, with no certificates available. Expected reward: baseline 17%, follow-up 9%. Unexpected reward: baseline 17%, follow-up 17%. No reward: baseline 16%, follow-up 18%.

   Which conclusion is best supported by the graph?

   Only the group promised a reward in advance declined, falling from 17% to 9% while the other two groups held steady: the pattern predicted by the overjustification effect, in which an expected extrinsic reward undermines existing intrinsic motivation. Rewards reduce intrinsic motivation, since both of the rewarded groups declined. The no-reward group's rise from 16% to 18% shows that withholding rewards is an effective way to increase interest. The expected-reward group declined because a certificate was not a valuable enough reward to maintain the behavior.

   The comparison that matters is expectation, not reward: both rewarded groups received the identical certificate, and only the group promised it beforehand dropped. That isolates the mechanism: a promised reward supplies an external reason for an activity the child was already choosing freely.

   The unexpected-reward group received a certificate and stayed at 17%. Since one rewarded group declined and the other did not, "rewards reduce motivation" is contradicted by the graph's own data.

   A two-point move from 16% to 18% in a group where nothing was done is the size of change that ordinary variation produces. Reading it as an effect of withholding rewards treats noise as a finding.

   If the certificate were simply too small a reward, the unexpected-reward group, which received exactly the same certificate, should have declined as well. It did not, so the value of the reward cannot be what separates the groups.

10. Data: four studies of the same tutoring program, each measuring the same 100-point reading test. Within-group standard deviations were about 10 points in all four studies.

    Study	1	2	3	4
    Participants (n)	40	1,200	60	2,400
    Mean difference (points)	8.2	1.8	4.5	0.5
    Cohen's d	0.82	0.18	0.45	0.05
    p	.01	.002	.09	.22

    Which statement is best supported by the table?

    Study 2 found the strongest effect, because it has the smallest p-value. Study 3 found no effect, because its p-value is greater than .05. Study 1 is the most trustworthy of the four results, because it has the largest effect size. Study 2's very small p-value comes from its large sample rather than a large effect: a 1.8-point difference is under a fifth of a standard deviation, smaller than the difference Study 3 found without reaching significance.

    A p-value answers a different question from an effect size: how surprising this result would be if nothing were going on, given this sample size. With 1,200 participants even a d of 0.18 clears any conventional threshold, while the effect itself is the second smallest in the table.

    A p-value above .05 means the result did not reach the conventional threshold, not that the effect is zero. Study 3's d of 0.45 is a moderate difference measured with only 60 participants, so the honest reading is "underpowered," not "no effect."

    Study 1 pairs the largest d with the smallest sample, which is exactly the combination that should slow you down. With 40 participants the estimate is imprecise, and publication bias means the literature over-represents large effects that came from small samples.

    Dividing the mean difference by the standard deviation is what makes the four studies comparable: 1.8 points on a scale with a standard deviation of 10 is d = 0.18, a gap most students would never notice. This is the practical-versus-statistical significance distinction the exam tests, and the fix is to read d before you read p.

Lesson 4.1 · Unit 4 · CED topic 4.1

Attribution theory and person perception

Watch someone cut a line and you do not think "long day, probably late." You think "rude." Social psychology's founding observation is that people explain behavior by reaching for the actor's character and skipping the pressures acting on them, and that the skip is systematic, not random.

Lesson 4.2 · Unit 4 · CED topic 4.2

Attitude formation and attitude change

The intuitive model is that attitudes drive behavior: you believe, then you act. Half of social psychology is the discovery that the arrow also runs backwards. Get someone to act first (publicly, freely, for no good reason), and the attitude follows the act.

Lesson 4.3 · Unit 4 · CED topic 4.3

The psychology of social situations

Almost everyone believes they would speak up, refuse, or help. These experiments exist because that belief is wrong often enough to be worth measuring. What moves behavior here is rarely the kind of person you are; it is how many people are present, what they appear to think, and whether anyone can tell it was you.

Lesson 4.4 · Unit 4 · CED topic 4.4

Introduction to personality and its assessment

Personality is the part of you that travels: the pattern of thinking, feeling, and behaving that shows up across situations and over years. Four families of theory explain that pattern in incompatible ways, and each built its own way of measuring it. Learn a theory with its test: the measure is the theory made operational.

Lesson 4.5 · Unit 4 · CED topic 4.5

Psychoanalytic and humanistic theories of personality

These two families disagree about what a person fundamentally is. Freud's picture is a mind at war with itself, managing impulses it cannot admit to having. The humanists answer that people are growing toward something, and that what blocks growth is usually the conditions other people attach to their approval. Both are influential; neither is easy to test.

Lesson 4.6 · Unit 4 · CED topic 4.6

Social-cognitive and trait theories of personality

The two modern families are the testable ones. Social-cognitive theory explains personality as the loop between what you think, what you do, and the situations your behavior keeps putting you in. Trait theory skips explanation and measures: find the dimensions people reliably differ on, and see what those scores predict.

Lesson 4.7 · Unit 4 · CED topic 4.7

Motivation

Motivation theories fall into two groups, and the exam wants you to tell them apart. Push theories say an internal deficit drives you until it is corrected. Pull theories say the reward out in the world draws you toward it. Neither alone explains why someone with a full stomach eats dessert, or why a paid hobby stops being fun.

Lesson 4.8 · Unit 4 · CED topic 4.8

Emotion

Every theory here answers one question in a different order: when your heart pounds and you feel afraid, which came first? Learn them as sequences rather than names: that is how the exam tests them, with a scenario and a question about what has to happen first.

Lesson 4.9 · Unit 4 · Research methods

Sampling and generalizability

A finding is about the people it was measured on. Extending it to everyone else is a separate claim, and the only thing that supports it is how the sample was drawn. Social psychology has the sharpest version of this problem, because the behavior it studies (conformity, attribution, what counts as an insult) is exactly the kind that varies between cultures.

Lesson 4.10 · Unit 4 · Research methods

Inferential statistics: p-values and what "significant" means

Descriptive statistics summarize the sample you have. Inferential statistics answer a harder question: could a difference this large have arisen by chance, in a world where the two groups are really the same?

Unit 4 review · 10 multiple-choice

Unit 4 review: Social Psychology and Personality

Ten questions across all ten lessons in the exam's three styles (concept application, data analysis, and scientific investigation), so answer each one before you click, then read all four rationales, not just the one for the option you chose.

Multiple choice

1. Ravi's teammate Nia misses two group meetings, and Ravi concludes she is unreliable and does not care about the project. When Ravi himself misses the next meeting, he explains that his shift ran late and the bus never came. Which of the following best describes Ravi's pattern of explanation?

   He explains Nia's absences dispositionally and his own situationally, the actor-observer bias He explains Nia's absences situationally and his own dispositionally, the just-world hypothesis He overestimates how many classmates would also have skipped, the false consensus effect He lets one bad quality colour every judgment he makes of Nia, the halo effect

   Correct. Ravi reaches for Nia's character (unreliable, uncaring), and skips the circumstances, which is the fundamental attribution error; when the target is himself he reaches for the bus and the shift instead. That reversal by target is exactly what the actor-observer bias names.

   The directions are backwards: Ravi credits Nia's disposition, not her situation. The just-world hypothesis is also a different idea: the assumption that people get the outcomes they deserve, which would show up as "she must have done something to deserve it," not as an explanation of lateness.

   False consensus is overestimating how widely your own view or behavior is shared, as in assuming most people would also skip the meeting. Nothing in the scenario tells you what Ravi believes about other classmates, so this is a real bias that does not answer the question.

   A halo effect runs from one positive impression outward, letting a single good quality inflate unrelated judgments. Ravi is not generalizing from one trait across many; he is explaining one behavior, and the two explanations differ depending on whose behavior it is.

2. A club needs volunteers to hand out flyers for a cause its members feel lukewarm about. Half the members are offered a $20 gift card to do it; the other half are asked as a favour and receive nothing. Everyone hands out flyers. A week later, which group is likelier to report warmer attitudes toward the cause, and why?

   The paid group, because the gift card reinforced the behavior and reinforcement strengthens attitudes The unpaid group, because with no external reason for what they did they reduce dissonance by deciding they support the cause The paid group, because the gift card acted as a peripheral cue and increased liking through mere exposure Neither group, because attitudes change only when a message is processed through the central route

   Reinforcement raises the rate of a behavior while it is delivered; it does not reliably move the private attitude behind it. In dissonance research the payment works the other way, because a large reward supplies a ready explanation for the act and leaves no discomfort to resolve.

   Correct. Dissonance is largest under insufficient justification: the unpaid members advocated for something they barely believed, with no money or requirement to point at, so the cheapest repair is to shift the attitude toward the behavior. This is the Festinger and Carlsmith pattern, in which the $1 group reported more enjoyment than the $20 group.

   Mere exposure is increased liking from repeated contact with a stimulus itself, not from a reward attached to it, and a gift card given for handing out flyers is an incentive rather than a peripheral persuasion cue. The distractor mixes two real concepts and applies neither correctly.

   Attitude change through the central route is one path, but it is not the only one: the elaboration likelihood model also allows peripheral change, and dissonance changes attitudes with no persuasive message at all. Here the members' own behavior, not an argument, is what does the work.

3. Described data

   A research team ran a line-judgment task. On each trial the participant said aloud which of three comparison lines matched a standard line; the correct answer was obvious. Confederates seated with the participant answered first. The table reports the percentage of critical trials on which participants gave the majority's wrong answer, by condition.

   Condition	Percent of critical trials conformed
   Judging alone, no group present	1%
   Unanimous majority of 3	32%
   Unanimous majority of 7	36%
   Majority of 7, one confederate answers correctly	7%

   Which conclusion is best supported by the data in the table?

   Conformity rises steadily with each additional person added to the majority Participants conformed because the task was genuinely too hard to judge accurately Unanimity matters more than size: four extra confederates moved conformity about 4 points, while one dissenter cut it about 29 About a third of the participants conformed on every critical trial

   The table contradicts this. Going from three confederates to seven, more than doubling the majority, moved conformity only from 32% to 36%, which is the finding that majority size stops mattering quickly once a few people agree.

   The alone condition rules this out: with no group present, participants were wrong on about 1% of trials, so they could see perfectly well which line matched. That baseline is what turns the other rows into evidence about social pressure rather than about perception.

   Correct. Compare the rows the data actually let you compare: 36% minus 32% is about 4 percentage points for four extra people, while 36% minus 7% is about 29 points for one confederate who breaks the unanimity. The comparison also separates the mechanisms: a single ally mostly relieves normative pressure, the concern about being the odd one out.

   Read the column heading. These are percentages of trials, not of participants, and the two are very different claims: a third of trials can come from most people conforming occasionally rather than from a third of people conforming always.

4. Described study

   A researcher posts a 40-item online personality inventory and recruits 60 students to complete it. A week later every student receives a paragraph described as "your individual profile," and rates on a 1-to-5 scale how accurately it describes them. In fact all 60 students receive the identical paragraph, made of statements general enough to fit almost anyone. The mean accuracy rating is 4.2. The researcher concludes that the inventory is a valid measure of personality.

   Which change to the design would most directly test the researcher's conclusion?

   Recruit 600 students instead of 60 so the mean accuracy rating is estimated more precisely Have students rate the same paragraph again two weeks later to check test-retest reliability Have two trained raters score each profile independently to establish inter-rater reliability Randomly assign half the students to receive a profile generated from their own answers and compare accuracy ratings across the two groups

   A larger sample narrows the confidence interval around 4.2, but it cannot tell you what the 4.2 means, because there is still nothing to compare it against. Precision is not the problem here; the missing comparison condition is.

   Stable ratings across two weeks would show only that students agree with the same vague paragraph twice. Reliability is consistency of measurement, and a measure can be perfectly consistent while measuring nothing the researcher claims: reliability is necessary for validity but never sufficient.

   Inter-rater reliability matters when human judges score open-ended material, as with a projective test. Here the students score themselves on a fixed scale, so there are no raters to agree, and agreement would still say nothing about whether the inventory captures personality.

   Correct. With every student receiving the same generic text, a mean of 4.2 is fully explained by the Barnum effect: near-universal statements feel personal. A control group given genuinely individualized profiles supplies the comparison: only if the individualized profiles are rated clearly more accurate is the inventory adding information.

5. Marcus is passed over for a solo. He tells his friends the choir director is jealous of his voice, and that evening he snaps at his younger sister over something trivial. A psychoanalytic theorist would name these two responses, in order, as

   displacement, then projection projection, then displacement reaction formation, then rationalization rationalization, then regression

   The order is reversed. Telling everyone the director is the jealous one comes first and moves a feeling onto another person, which is projection; snapping at a sister comes second and moves a target, which is displacement.

   Correct. Projection attributes your own unacceptable feeling (here, envy) to someone else. Displacement keeps the feeling and swaps the target for a safer one, which is why the anger aimed at the director lands on a younger sibling instead. Remember the pair as: projection moves the feeling, displacement moves the target.

   Reaction formation would have Marcus acting conspicuously delighted for whoever got the solo, expressing the opposite of what he feels. Rationalization would have him producing an acceptable reason for the outcome, such as deciding he never wanted the part: neither matches what he actually does.

   Rationalization supplies a comfortable justification rather than accusing someone else of the feeling. Regression would mean returning to behavior from an earlier developmental stage, such as a tantrum, and snapping once at a sibling is ordinary displaced irritation, not a return to childhood functioning.

6. Talia is confident she can work through a hard calculus problem set if she keeps at it, and she persists for an hour when a problem resists her. She also believes she could never give a decent speech, and she avoids the one presentation in the course. Which construct does this pattern best illustrate?

   An internal locus of control High self-esteem Task-specific self-efficacy Reciprocal determinism

   Locus of control is a broad belief about what produces outcomes in general, your own effort versus luck and powerful others, so it would not flip between two courses in the same semester. The scenario gives you a belief about capability in a specific domain, not about where outcomes come from.

   Self-esteem is a global evaluation of your own worth, which is why it changes slowly and does not vary by subject. Talia's confidence is high in one task and low in another, and that variation is the clue that the item is testing a task-specific construct.

   Correct. Self-efficacy is Bandura's term for the belief that you can carry out a particular task, and it is domain-specific by definition: high for calculus, low for public speaking, in the same person. It also predicts exactly what the scenario reports: persistence in the high-efficacy task and avoidance in the low-efficacy one.

   Reciprocal determinism is the three-way loop among personal factors, behavior, and environment. It would apply if the scenario showed the avoidance changing her environment, which then fed back onto the belief; as written, you are given the belief and its immediate behavioral consequence.

7. Described data

   Students worked on an interesting spatial puzzle in three sessions. In each session the experimenter left the room for an 8-minute (480-second) break and said the students could do whatever they liked; magazines lay on the table. In session 2 only, students in one group earned $1 per puzzle solved. The table gives each group's mean number of seconds spent on the puzzle during the unmonitored break.

   Group	Session 1	Session 2	Session 3
   Paid in session 2	205 s	290 s	155 s
   Never paid	208 s	215 s	220 s

   Which statement is best supported by the table?

   The paid group spent more free-choice time on the puzzle while payment was available and less than its own baseline once it stopped, the overjustification pattern Payment permanently raised intrinsic motivation, since the paid group's session 2 mean is the highest in the table The never-paid group lost interest across the three sessions Money acted as a negative reinforcer for puzzle solving in the paid group

   Correct. Track the paid row against its own session 1 baseline: 205 to 290 while the dollar was on offer, then down to 155, fifty seconds below where it started, after payment stopped. The never-paid row barely moves, which rules out simple boredom with the puzzle and leaves the reward as the difference between the groups.

   The session 2 mean is indeed the table's highest, but the claim is about what happens afterwards, and session 3 answers it: 155 seconds is the table's lowest value. A reward can raise behavior while it is delivered and still leave the behavior lower than baseline once it is withdrawn.

   Their means rise slightly across the three sessions, from 208 to 215 to 220 seconds. That flat-to-upward trend is what makes this group a useful comparison: whatever the paid group's session 3 drop reflects, it is not the puzzle becoming dull with repetition.

   Negative reinforcement means a behavior increases because it removes something aversive, and nothing aversive is removed here. Payment is a positive reinforcer during session 2; the interesting result is the overjustification effect that follows it, which is a change in the reason for the behavior rather than a schedule of reinforcement.

8. Ana finishes a hard training run and meets a stranger at the finish line while her heart is still pounding and her breathing has not settled. She later reports finding the stranger unusually attractive. Which explanation follows Schachter and Singer's two-factor theory?

   The bodily reaction is itself the emotion, so no interpretation is involved The arousal and the felt attraction were triggered at the same moment and neither caused the other Her facial expression during the run fed back and produced the positive feeling Leftover arousal was present without an obvious cause, and she labelled it using the most salient feature of the situation

   This is James-Lange, in which the emotion just is your reading of a specific bodily pattern and no cognitive label from the situation is needed. Two-factor theory was built precisely because the same racing heart can become fear, anger, or attraction depending on what is around you.

   This is Cannon-Bard: arousal and emotional experience occur simultaneously, with the thalamus routing the signal both ways and neither one producing the other. It is a real theory but it cannot explain why the stranger, rather than the run, gets the credit.

   The facial feedback hypothesis says expressions feed back into what is felt, and it is worth treating cautiously since a large multi-laboratory replication failed to reproduce the classic result. It also does not fit: the scenario gives you exertion and a pounding heart, not a held smile.

   Correct. Two-factor theory requires both ingredients, physical arousal plus a cognitive label drawn from the situation, and the label is wrong here, which is what misattribution of arousal means. Notice the condition the explanation depends on: the arousal has to be unexplained, so if Ana consciously credited the run, the effect should disappear.

9. Described study

   Researchers post a survey link in an online forum devoted to sleep advice. Over two weeks 1,200 forum members complete it, answering questions about their electronics use and their sleep. Among respondents who report keeping a screen curfew in the hour before bed, 68% rate their sleep quality as good or very good, compared with 41% of respondents with no curfew. The researchers conclude that most adults would sleep better if they adopted a screen curfew.

   Which of the following is the most serious threat to the researchers' conclusion?

   The sample is too small for a difference of this size to be detected Respondents recruited from a sleep-advice forum are a self-selected group unlike adults in general, so the sample cannot support a claim about most adults Self-report measures are invalid and cannot be used as evidence in psychological research Without random assignment the researchers cannot generalize from their sample to the population

   Twelve hundred respondents is a large sample, and a 27-percentage-point gap is not a subtle effect, so power is not the issue. Sample size and sample representativeness are different problems: a huge sample drawn from the wrong population is still the wrong population.

   Correct. People who join a forum about sleep advice are already unusual in their sleep habits and their interest in fixing them, and those who answered are volunteers within that group, so this is volunteer bias layered on a convenience sample. Part E of an AAQ is scored on this kind of participant evidence: name who was recruited and where, then say what it does to the claim.

   Self-report has real weaknesses, social desirability and response sets among them, but it is standard, usable evidence when the construct is a private experience such as perceived sleep quality. An answer that rejects a whole class of measurement is too strong to be the best objection here.

   This confuses the two randoms. Random assignment licenses causal claims, and its absence does mean the curfew may not be what produced better sleep; random selection is what licenses generalizing to a population. The conclusion quoted is a generalization, so the sampling term is the one that applies.

10. Described data

    Students were randomly assigned to use a vocabulary app or to a control activity for four weeks, then took a 100-point vocabulary test. Reported with the table: the mean difference is 1.4 points, p = .02, and Cohen's d = 0.12.

    Group	n	Mean test score	SD
    Used the app	600	71.8	12.0
    Control activity	600	70.4	12.0

    Which statement is best supported by these results?

    The result is statistically significant, so the app produces a large improvement in vocabulary A p-value of .02 means there is a 2% chance that the app has no real effect The difference is statistically significant but the effect is small; 1,200 students is enough to make a 1.4-point gap clear the cutoff Because d = 0.12 is small, the researchers have made a Type II error

    Significance and size are separate questions, and this is the confusion data analysis items are built to catch. A p-value says a difference this large is unlikely if the app truly did nothing; only the effect size says whether the difference is worth anything, and d = 0.12 is small by any convention.

    A p-value is the probability of a result at least this extreme if the null hypothesis were true, not the probability that the null hypothesis is true, and not the probability the finding will replicate. Reversing the conditional is the single most common error with p-values.

    Correct. Both readings are in the numbers: 71.8 minus 70.4 is a 1.4-point gap, p = .02 clears the conventional .05 cutoff, and 1.4 divided by the standard deviation of 12.0 gives d ≈ 0.12. Power is why both are true at once, with 600 students per group, even a trivial difference becomes detectable, which is exactly why effect size is reported alongside.

    A Type II error is missing a real effect: failing to reject a null hypothesis that is false. These researchers detected an effect, so if they erred it would be a Type I error, a false alarm; and a small effect size is not itself evidence of any error.

Lesson 5.1 · Unit 5 · CED topic 5.1

Introduction to health psychology: stress and coping

Stress is not the exam. Stress is what your body does about the exam: an appraisal that a demand outruns your resources, then a response built for predators and aimed at a calendar. A system tuned for minutes must run for months, and the bill arrives in the immune system.

Lesson 5.2 · Unit 5 · CED topic 5.2

Positive psychology: strengths, well-being, and what lasts

For most of its history psychology studied what goes wrong. Positive psychology asks the other question, what makes a life go well, and asks it with the same tools: random assignment, control conditions, follow-up measurement. The surprise is how little of what people chase actually moves the number.

Lesson 5.3 · Unit 5 · CED topic 5.3

Explaining and classifying psychological disorders

Deciding that a pattern of thought, feeling, or behavior is a disorder is a judgment, and psychology tries to make it a disciplined one. No single sign settles it: grief is intense distress that is not a disorder, and a person can be seriously unwell without seeming odd to anyone. A diagnosis names a pattern, not a person: a working summary that guides treatment and can be revised.

Lesson 5.4 · Unit 5 · CED topic 5.4

Anxiety, obsessive-compulsive, and trauma- and stressor-related disorders

Fear is useful equipment. These diagnoses describe it running when nothing is there, running long after the danger has passed, or running so hard that avoiding the fear costs more than the feared thing would. That last part, avoidance, is what turns a bad feeling into a disorder, and it is the hinge the best treatments pull on.

Lesson 5.5 · Unit 5 · CED topic 5.4

Depressive and bipolar disorders

Sadness is a response to something. The depressive disorders are different: a persistent change in mood, energy, sleep, appetite, and self-evaluation that outlasts its trigger or arrives without one. The bipolar disorders add the other pole: episodes of elevated mood and energy just as far from an ordinary good day.

These are treatable conditions that carry real risk. Warning signs worth acting on include withdrawing from people, giving away belongings, talking about being a burden, and a sudden calm after a long low stretch. If you notice them in yourself or a friend, tell an adult you trust: help is available, and in the US the 988 Suicide and Crisis Lifeline takes calls and texts at any hour.

Lesson 5.6 · Unit 5 · CED topic 5.4

Schizophrenia spectrum disorders

Schizophrenia is a disorder of how experience is put together: perception, belief, thought, and motivation all shifting at once. It is also the diagnosis most distorted in popular use. It is not dissociative identity disorder, it does not mean a person is dangerous, and it usually arrives gradually, in late adolescence or early adulthood, rather than all at once.

Lesson 5.7 · Unit 5 · CED topic 5.4

Neurodevelopmental, dissociative, somatic symptom, eating, and personality disorders

This lesson collects the remaining DSM-5-TR chapters the exam samples from. They have little in common clinically, so learn each by the feature that distinguishes it from its nearest neighbor: that boundary is exactly what a multiple-choice item tests.

Lesson 5.8 · Unit 5 · CED topic 5.5

Treatment I: psychodynamic, humanistic, and behavioral therapies

Every therapy is a theory of disorder turned into a procedure. If the trouble is buried conflict, you excavate. If it is a person blocked from growing, you supply the conditions for growth. If it is a learned response, you arrange new learning. Read any technique backwards and you can name the theory that produced it.

Lesson 5.9 · Unit 5 · CED topic 5.5

Treatment II: cognitive, cognitive-behavioral, and group approaches

The cognitive therapies start from a claim you can test on yourself: the event is not what upsets you, your reading of it is. Two students get the same B−. One files it and moves on; the other concludes she is not cut out for this. Same event, different belief, and the belief is the part that can be examined and changed.

Lesson 5.10 · Unit 5 · CED topic 5.5

Biomedical treatment and the ethics of care

Biomedical treatments change the nervous system directly. That makes them powerful and makes the consent conversation unavoidable: side effects are real, the person taking the medication lives with them, and a competent adult may refuse. The history of this field is why those protections are written down.

Lesson 5.11 · Unit 5 · Research methods

Reading a research article: putting the AAQ skills together

The Article Analysis Question hands you one research summary and twenty-five minutes. Everything it asks for is already in the text, in a predictable place. So the skill is not reading faster; it is knowing which paragraph holds which answer.

Unit 5 review · 10 multiple-choice

Unit 5 review: Mental and Physical Health

Ten questions spread across all eleven lessons in the exam's three styles (concept application, data analysis, and scientific investigation) with DSM-5-TR criteria, treatment evidence, and the design questions that decide what a clinical result is worth.

Multiple choice

1. For three weeks Priya works long hours on a demanding project, feels wired but functional, and sleeps badly. The morning after she submits it she comes down with a heavy cold. Which stage of the general adaptation syndrome does the cold best illustrate?

   Alarm, because a new demand has triggered the fight-flight-freeze response Exhaustion, because prolonged resistance depletes resources and immune function drops, raising the risk of illness Resistance, because cortisol is holding the body at a costly plateau while the demand continues Eustress, because the deadline sharpened her performance while it lasted

   Alarm is the opening burst (sympathetic activation, adrenaline, a body mobilized in seconds), and it belongs at the start of the three weeks, not at the end. By the time the cold arrives the stressor is over, which is the giveaway that a later stage is being tested.

   Correct. Selye's third stage is where the bill comes due: cortisol held high through the resistance phase suppresses immune function, so illness tends to surface once the demand lifts. Note the coping split the lesson uses: a fixed revision schedule is problem-focused because it changes the stressor, while a run is emotion-focused because it changes her reaction.

   Resistance is the plateau during the three weeks: costly but functional, exactly the "wired but functional" stretch the scenario describes. The question asks about the cold, which comes after that plateau ends, so resistance names the wrong part of the sequence.

   Eustress is arousal that sharpens performance rather than swamping it, and it may well describe her first week. It is a label for a quality of stress, not a stage of the general adaptation syndrome, so it cannot answer a question about which stage the illness shows.

2. Described study

   Researchers recruited 240 adults through a well-being website and randomly assigned each one to one of two one-week writing exercises: listing three good things each evening along with what caused them, or writing each evening about early childhood memories. Both groups were asked to write for the same number of minutes and were told the exercise was expected to help. Happiness and depressive-symptom questionnaires were completed before the week, immediately after, and again at one, three, and six months.

   What is the main purpose of the early-memories writing condition?

   To test whether writing about childhood improves well-being over six months To increase the total sample size so the study has enough statistical power To provide a second experimental condition testing a competing theory of well-being To serve as a placebo control holding writing time and expectation constant, so any extra gain can be credited to the content of the gratitude exercise

   That question could be asked of the data, but it is not why the condition is there. A control condition is defined by its role in the comparison, it supplies the baseline the treatment is measured against, not by an interest in the control activity itself.

   Power comes from how many participants are in the study, and splitting them across conditions is how an experiment is built, not a way of manufacturing power. If sample size were the aim, the researchers would simply put all 240 in one group and have no comparison at all.

   An active comparison condition drawn from a rival theory is a legitimate design, but nothing here presents early-memory writing as a treatment anyone predicted would work. It was chosen because it looks and feels like the exercise while lacking the ingredient under test.

   Correct. A control that does nothing leaves expectancy and time-on-task free to explain any improvement; this one matches both and withholds only the active ingredient, crediting causes and noticing good things. That is what makes the comparison clean, and it is the design point a scientific investigation item usually targets.

3. A ninth grader checks his locker four or five times before every class. He knows it is locked, the checking makes him late, and he has started leaving home early to make room for it. He describes feeling ashamed and unable to stop. A counselor is weighing whether this pattern crosses the threshold for a diagnosis. Which criteria are doing most of the work?

   Distress and dysfunction, the checking causes real suffering and interferes with getting to class Deviation alone, the behavior departs from what most students do, which settles the question Danger, the pattern places the student or others at physical risk Comorbidity, more than one diagnosis is present at the same time

   Correct. Of the four criteria clinicians weigh (distress, dysfunction, deviation, danger) two are clearly present: he suffers, and the behavior costs him time and punctuality. That functional cost, not the strangeness of the behavior, is what moves a habit toward a diagnosis such as obsessive-compulsive disorder, and it is why a person can be unusual without being unwell.

   Deviation from a norm is the weakest of the four criteria on its own: plenty of statistically unusual behavior is harmless, and what counts as deviant varies by culture and setting. Using it alone is how ordinary difference gets pathologized, which is the error the criteria are designed to prevent.

   Danger means risk of physical harm to the person or to others, and nothing in the scenario describes any. It is also the rarest of the four criteria in practice: most people who meet criteria for a psychological disorder pose no danger to anyone, and assuming otherwise is the stereotype that drives stigma.

   Comorbidity describes two or more diagnoses co-occurring, such as a depressive disorder alongside an anxiety disorder. It is a real and common feature of clinical practice, but it is not a criterion for deciding whether a single pattern is a disorder in the first place.

4. Tomás is afraid of dogs. Each time he sees one he crosses the street, and his anxiety drops within seconds. He has now rerouted his commute and stopped visiting a friend who has a dog. Which principle best explains why the avoidance persists and why the fear does not fade?

   Positive reinforcement, crossing the street is rewarded by the relief that follows Punishment, the sight of a dog punishes walking on that side of the street Negative reinforcement (avoidance removes an aversive state, which strengthens it, and he never stays long enough for the fear to be disconfirmed Classical conditioning) the dog is the unconditioned stimulus and the anxiety the unconditioned response

   Relief is the removal of something unpleasant, not the addition of something pleasant, so this is negative rather than positive reinforcement. The two are the most confused pair in the unit: both raise behavior, and the sign refers to whether a stimulus is added or taken away.

   Punishment decreases the behavior it follows, and Tomás's avoidance is increasing, so the term has the wrong direction built in. Notice that the scenario reports a behavior becoming more frequent and more elaborate, which rules out punishment before any other reasoning starts.

   Correct. The avoidance is maintained because it works in the short run, and that short-run success is the trap: each crossing removes anxiety immediately, while guaranteeing he never learns that a leashed dog on the sidewalk passes without incident. Breaking that loop is exactly what exposure therapy is for, which is why it is the best-supported treatment for specific phobia.

   Classical conditioning plausibly explains how the fear was acquired, but the roles are assigned wrongly: a frightening event would be the unconditioned stimulus and the dog the conditioned stimulus. And acquisition is not the question asked: the item asks what keeps the fear alive now, which is the operant half of the two-factor account.

5. After a poor audition, Wren says: "I choked because I'm just not a performer. I never have been and I never will be. And it's the same with everything I try." Which part of her statement expresses the global dimension of a pessimistic explanatory style?

   "I choked", naming what happened in the moment "I'm just not a performer", placing the cause inside herself "I never have been and I never will be", treating the cause as permanent "It's the same with everything I try", spreading the cause across unrelated areas of her life

   This is the event being explained, not the explanation of it. Attributional style is about the causes a person assigns, so the phrase that describes the outcome cannot be the one that carries a dimension.

   This is the internal dimension: the cause is located in her rather than in the situation, the piece, or the panel. Internal, stable, and global are three separate dimensions, and items like this one ask you to keep them apart.

   This is the stable dimension: the cause is treated as permanent rather than temporary. It is the one most often mistaken for global, but stability is about time, while global is about breadth across situations.

   Correct. Global means the cause is taken to apply everywhere, so one bad audition becomes evidence about her whole life. A cognitive therapist would often dispute the global claim first, because it is the easiest to test against evidence Wren already has: the other things she does competently.

6. Described graph

   A bar chart shows the lifetime risk of a diagnosis of schizophrenia by genetic relationship to a person diagnosed with it, pooled across dozens of European twin, family, and adoption studies (Gottesman, 1991). Risk in percent runs up the vertical axis. Four bars are shown: general population, just under 1%; sibling, about 9%; fraternal (dizygotic) twin, about 17%; identical (monozygotic) twin, about 48%.

   Which conclusion is best supported by the display?

   Because the identical-twin bar is tallest, schizophrenia is caused by a single gene that a person either carries or does not Risk rises sharply with genetic relatedness, the identical-twin bar is nearly three times the fraternal-twin bar, yet stays far below 100%, so genes alone are not sufficient The sibling and fraternal-twin bars are close enough to show that shared environment has no effect on risk A risk of about 48% means an identical twin has a 48% chance of being diagnosed because the two share a household

   A single-gene disorder would produce concordance near 100% in identical twins, who share essentially all their genes. The chart shows about 48%, which is the strongest single piece of evidence against a one-gene account and for many genes of small effect plus environmental contributions.

   Correct. Read the display in both directions, which is what a data analysis item rewards: 48% against 17% for pairs who share a womb and a household is the genetic evidence, because identical twins differ from fraternal twins mainly in genes; 48% against 100% is the environmental evidence. Together they are the diathesis-stress model in one chart.

   The bars are not close, 17% is nearly double 9%, and even if they were, a chart of relatedness categories cannot isolate shared environment. That comparison needs designs the display does not contain, such as identical twins reared apart or adoption studies.

   The number is read correctly and then explained backwards. Identical and fraternal twin pairs are both typically raised in the same household, so a shared household cannot explain why one group's risk is nearly three times the other's; what differs between them is genetic similarity.

7. Described data

   Adolescent girls in a rural Fijian province completed a standard questionnaire on eating attitudes in 1995, shortly after television became available in the region, and a separate sample of about the same size completed it in 1998 (Becker and colleagues, 2002). The table gives the percentage of each sample scoring in the range that indicates disordered eating attitudes.

   Survey year	Percent scoring in the at-risk range
   1995, television newly available	about 13%
   1998	about 29%

   Which statement is best supported by these data?

   The share scoring in the at-risk range more than doubled in three years, but because exposure was measured rather than assigned the data cannot show that television caused the rise Television caused a 16% increase in eating disorders among adolescent girls in the province By 1998 about 29% of the girls surveyed had been diagnosed with anorexia nervosa Because the same girls were surveyed twice, the change is a within-person effect

   Correct. Both halves are supported: 29% against 13% is more than double, and no one was randomly assigned to watch television, so the design is correlational. The honest reading names a plausible alternative: the late 1990s also brought tourism, cash work, and schooling changes to the province, any of which could carry the effect.

   Two errors in one sentence. The causal verb is not licensed by a study with no random assignment, and the arithmetic is misreported: 13% to 29% is a rise of 16 percentage points, which is a relative increase of more than 120%, not "a 16% increase."

   A screening questionnaire measures attitudes, not diagnoses. Scoring in an at-risk range means a clinician might want to look further; a diagnosis of anorexia nervosa requires restriction leading to significantly low body weight, intense fear of weight gain, and disturbed body image, assessed by a clinician.

   The stimulus says a separate sample of about the same size completed the 1998 survey, so these are two cross-sectional samples three years apart, not the same girls followed over time. That distinction matters: a repeated cross-section describes a population shift and cannot tell you that any individual's attitudes changed.

8. Described data

   Adults who met criteria for a specific phobia of dogs were randomly assigned to eight sessions of systematic desensitization, eight sessions of cognitive restructuring alone, or a waitlist. Avoidance was measured with a behavioral approach test scored from 0 (will not enter the room) to 40 (will hold the leash), administered before treatment and again after eight weeks by an assessor who did not know the assignment. Higher scores mean less avoidance.

   Group	n	Mean before	Mean after 8 weeks
   Systematic desensitization	32	11.4	30.6
   Cognitive restructuring only	31	11.8	22.1
   Waitlist	30	11.6	13.2

   Which statement is best supported by the table?

   Cognitive restructuring produced no benefit, since its final mean is well below the desensitization group's The waitlist group's gain of 1.6 points shows that specific phobia usually resolves on its own within eight weeks Both treatments beat the waitlist and desensitization gained most: 19.2 points against 10.3 and 1.6, so graded exposure added something that changing thoughts alone did not Because the three groups began at nearly the same mean, the study has shown that the behavioral approach test is a valid measure of phobia

   Compare each group with the control, not with the best group. Cognitive restructuring gained 10.3 points against the waitlist's 1.6, which is a substantial benefit; it is simply a smaller one than desensitization produced.

   A 1.6-point move on a 40-point scale over eight weeks is close to flat, and it is the reason the waitlist is there, to show what happens without treatment. Reading it as spontaneous recovery inverts the evidence, and it would also have to explain why the treated groups moved six to twelve times as far.

   Correct. Subtract before from after in each row: 30.6 − 11.4 = 19.2, 22.1 − 11.8 = 10.3, and 13.2 − 11.6 = 1.6. The comparison is fair because random assignment left the groups nearly equal at baseline and the assessor was blind to condition, which is why the gap between the two active treatments can be read as the contribution of exposure itself.

   Similar baselines show that random assignment worked, not that the measure is valid. Validity would require the approach-test score to line up with something independent (clinician ratings, or real avoidance outside the clinic), and a table of group means cannot supply that.

9. Described study

   A company gave a new herbal supplement to 300 adults who reported persistent low mood. Everyone received the supplement, everyone knew what they were taking, and each person completed a standardized depression rating scale at enrollment and again after eight weeks. The group's mean score fell from 18.6 to 13.9. The company's press release states that the supplement "relieves depression in eight weeks."

   Which change to the design would most directly address the strongest objection to that conclusion?

   Enroll 3,000 adults rather than 300 so the change in mean score is estimated more precisely Randomly assign half the participants to an identical-looking inactive pill and keep participants and raters unaware of who received which Administer the depression rating scale every week rather than twice, so the trajectory of change is visible Replace the self-report rating scale with a clinician-administered interview to improve the validity of the outcome

   More participants would pin down the 4.7-point drop more tightly, but the problem is not how precisely the drop is measured: it is that a one-group design offers no way to tell what produced it. Thirty times the sample with no comparison condition buys a precise estimate of an uninterpretable number.

   Correct. With a single group, at least three explanations compete with the supplement: expectancy, ordinary remission over eight weeks, and regression toward the mean, since people enroll when they feel worst. A placebo arm supplies the baseline and double-blinding keeps expectancy from moving either the participants' answers or the raters' scores, and such a trial must also disclose the chance of receiving placebo in the consent form and move anyone who worsens into active treatment.

   Weekly measurement produces a nicer curve and might reveal when change happens, but every point on that curve still belongs to one group with nothing to compare it against. More frequent measurement of the wrong design does not fix the design.

   A clinician interview is generally a stronger outcome measure than self-report, and in an unblinded study it can be worse, since a clinician who knows everyone took the supplement brings expectations too. Either way, improving the dependent variable leaves the missing control group untouched.

10. A research summary contains this sentence: "Symptom severity was recorded as each participant's total score on a 17-item clinician-administered interview, completed before treatment and again after twelve weeks." Which part of an Article Analysis Question does this sentence most directly answer?

    Part B, the operational definition of the dependent variable, because it names the instrument and the score used to measure the construct Part A, the research method, because it describes what the researchers did to participants Part C, interpreting the statistic, because it reports a measurement Part E, generalizability, because it describes the participants

    Correct. An operational definition says how a construct was actually measured, here a total score on a named 17-item interview at two fixed time points, which is what part B is scored on. Writing "symptom severity means how bad the symptoms were" is a dictionary definition and earns nothing; the instrument and the score are the point.

    Part A wants the design named: experiment, quasi-experiment, correlational study, case study, naturalistic observation, survey, longitudinal design, or meta-analysis. A measures sentence describes an instrument, not a design, and you would look instead for the sentence about assignment to conditions.

    Part C asks you to say what a reported number indicates about the findings: what a mean difference, a percentage, a correlation, or an effect size tells us here. The sentence quoted reports no number at all; it describes how numbers were generated.

    Part E is answered from the participants paragraph: how many people, their ages, sex, culture, how they were recruited, and the setting. The sentence quoted names who administered the interview, which is procedure, not a participant characteristic.

AAQ 1 · 25 minutes · 7 points

Using the source, respond to parts A through F: do the reported results support or refute the claim that memory is reconstructive?

AAQ 2 · 25 minutes · 7 points

Using the source, respond to parts A through F: do the reported results support or refute observational learning as an explanation of aggression?

AAQ 3 · 25 minutes · 7 points

Using the source, respond to parts A through F: do the reported results support or refute normative social influence as the explanation of conformity?

AAQ 4 · 25 minutes · 7 points

Using the source, respond to parts A through F. Explain whether the findings support or refute neuroplasticity: the claim that experience changes the physical structure of the brain.

AAQ 5 · 25 minutes · 7 points

Using the source, respond to parts A through F. Explain whether the findings support or refute the cognitive explanation of depression: the claim that changing distorted thinking changes mood.

EBQ 1 · 45 minutes · 7 points

Using the sources provided, develop and justify an argument about whether high schools should start the school day later in the morning.

EBQ 2 · 45 minutes · 7 points

Using the sources provided, develop and justify an argument about the extent to which the situation, rather than personality, explains harmful obedience.

EBQ 3 · 45 minutes · 7 points

Using the sources provided, develop and justify an argument about whether beliefs about ability and control can be changed in ways that improve outcomes.

EBQ 4 · 45 minutes · 7 points

Using the three sources, develop and justify an argument about how much confidence a court should place in a confident eyewitness's memory.

EBQ 5 · 45 minutes · 7 points

Using the three sources, develop and justify an argument about whether self-control is best understood as a limited resource that can be used up.

Unit recap

Unit recap

Animated recap with on-screen narration. Turn on Voice to have it read aloud (uses your device's built-in voice). Pressing play counts as your one free video.