Unit 2 follows information through the mind, from perceiving a scene and thinking through a problem to storing and retrieving what happened. The exam treats those steps as connected, and this guide is organized around that connection.

Unit 2 is one path, not eight separate topics

The framework lists eight topics, from perception (2.1) through intelligence and achievement (2.8). Studying them as eight separate lists misses how a question about an eyewitness error can also involve attention at encoding and suggestion at retrieval. The topics are connected, and exam questions often require more than one at once.

Think of the unit as a sequence. Perception interprets incoming information, thinking works with it to solve problems or make decisions, and memory stores and retrieves it. Intelligence concerns how psychologists measure differences in those abilities and evaluate the measures.

Perception is interpretation, not recording (2.1)

Your eyes and ears deliver raw input. Perception is the brain building a usable scene out of it, and the building goes two ways at once. Bottom-up processing starts with the sensory input and works upward toward meaning, the way you sound out an unfamiliar word letter by letter. Top-down processing starts with experience, expectations, and context and works downward, the way you read a messy handwritten note because you already know what it probably says.

Exam questions usually present a brief scenario and ask you to identify which type of processing is involved. If the scenario emphasizes prior knowledge, a schema, or a perceptual set, it describes top-down processing. If it emphasizes the features of the stimulus itself, it describes bottom-up processing. A schema is a mental framework that organizes what you expect to experience. A perceptual set is a tendency to perceive one interpretation rather than another. Both schemas and perceptual sets make perception more efficient, and both can also introduce predictable biases.

Gestalt grouping, attention, and depth

Gestalt psychologists argued that we organize scenes into wholes, not lists of parts. You separate a figure (the object you focus on) from its ground (the background), and you group elements by proximity (near each other), similarity (alike), and closure (filling gaps to complete a shape). When a question shows dots arranged in clusters and asks why you see groups instead of single dots, it is testing these principles, not eyesight.

Attention limits how much incoming information you can process at one time. Selective attention allows you to follow one conversation in a noisy room, which is known as the cocktail party effect, but it also means other information may be missed. Inattentional blindness refers to failing to see a visible object because your attention was directed elsewhere. Change blindness refers to failing to notice a change in a scene. If you miss an object that was visible the entire time, that is inattentional blindness, and if you miss a change where one thing was replaced by another, that is change blindness.

Depth perception depends on visual cues. Binocular cues require both eyes and include retinal disparity and convergence. Monocular cues can be used with one eye and include relative size, interposition, texture gradient, and linear perspective. If a scenario describes judging distance with one eye closed, binocular cues cannot explain that judgment.

Thinking runs on shortcuts, and the shortcuts have names (2.2)

An algorithm is a step-by-step method that guarantees a correct answer if you finish it. A heuristic is a shortcut that is usually right and sometimes wrong in a specific, testable way. Almost every thinking question in this unit asks you to spot which shortcut a person used, or which trap the shortcut led them into.

The representativeness heuristic involves judging how likely something is by how well it matches a prototype. For example, a quiet student who likes puzzles may be judged more likely to be a chess club member than a cheerleader, even if the cheer squad is ten times larger. In cases like this, information about how common each group actually is tends to be ignored. The availability heuristic involves judging likelihood by how easily examples come to mind. After local news covers a plane crash, people often rate flying as riskier than driving because images of crashes are easy to retrieve. This is an example of a memory process affecting how a thinking task is performed.

The traps that follow

• Confirmation bias - seeking and remembering evidence that supports what you already believe.
• Mental set - approaching a new problem with strategies that worked on old problems, even when a simpler approach exists.
• Functional fixedness - seeing an object only in its usual function, so a box that holds tacks never becomes a platform for a candle.
• Framing - the way a choice is worded can change the decision. For example, a treatment described as having a 90% survival rate tends to sound better than the same treatment described as having a 10% mortality rate.
• Gambler's fallacy - treating independent random events as if past outcomes change the odds. Five heads in a row does not make tails due.
• Sunk cost fallacy - staying with a failing choice because of what you already spent on it, rather than what it will cost and give you from here.

Two additional distinctions are also covered in this topic. Convergent thinking involves narrowing options toward a single correct answer, which is what most standardized tests require. Divergent thinking involves generating many possible answers, which is what brainstorming tasks and many measures of creativity require. Executive functions is the general term for mental processes involved in planning, inhibition, and task switching that help coordinate the steps of a multi-step problem.

Memory is sorted by what you can say and what you can do (2.3-2.5)

Explicit memory holds facts and events you can state in words. Inside it, episodic memory covers events from your own life, and semantic memory covers general knowledge that is not tied to one moment, like the capital of France. Implicit memory shows up in performance without conscious recall. Procedural memory, the exam workhorse inside implicit memory, covers skills such as typing, riding a bike, or playing a scale on the piano. If a patient can learn a skill but cannot remember the practice sessions, implicit memory is intact and explicit memory is not.

These types of memory differ in how long information is held and how much can be held. Sensory memory is the brief initial holding area, called iconic memory for visual input, lasting a fraction of a second, and echoic memory for sound, lasting about three to four seconds. Short-term memory holds roughly seven items, give or take two, for about twenty to thirty seconds without rehearsal. Working memory refers to the same brief storage system when information is also being actively used. Holding a phone number briefly is an example of short-term memory, while holding it and also reversing it or adding it to a total involves working memory, which is coordinated by the central executive with a phonological loop for speech-based material and a visuospatial sketchpad for images and spatial information.

Long-term memory has no practical capacity limit for the purposes of this course, so the important question is how information moves into long-term storage. Automatic processing encodes information about space, time, frequency, and familiar word meanings without deliberate effort. Effortful processing involves deliberate strategies such as rehearsal, organization, and connecting new material to what you already know. The framework also refers to long-term potentiation, which is the strengthening of synapses after repeated firing, as the biological process that supports storage.

Encoding decides what retrieval can find (2.4, 2.6)

Levels of processing describes how encoding affects retention. Structural encoding focuses on appearance, phonemic encoding on sound, and semantic encoding on meaning, and memory lasts longer as processing moves in that order. Rereading notes keeps processing shallow and builds familiarity with the page, which is recognition rather than the recall a test requires.

Three encoding strategies are tested often and are useful for studying. Chunking groups items into meaningful units so the roughly seven-item limit applies to chunks rather than digits. Mnemonic devices attach new material to vivid cues, and the method of loci attaches it to places on a familiar route. Spacing spreads study across sessions, and distributed practice generally outperforms massed practice on delayed tests. In the serial position effect, early items benefit from extra rehearsal (primacy) and late items are still active at recall (recency), while middle items get neither advantage. Rotating study order keeps the same items from always falling in the middle.

Retrieval, forgetting, and memory going wrong (2.6, 2.7)

Recall produces an answer with few cues, as on an FRQ, while recognition identifies an answer among options, as on multiple choice. Retrieval is better when test cues match encoding conditions, which includes context-dependent and state-dependent effects and the encoding specificity principle. The testing effect is the finding that retrieval practice strengthens later retrieval more than additional study. Ebbinghaus showed forgetting is steep at first and then levels off, so early review is more useful than waiting until before the exam.

Encoding failure means information never entered long-term memory, and storage decay means the trace weakened over time. Proactive interference is earlier learning interfering with newer recall, and retroactive interference is newer learning interfering with older recall. Note which memory is blocked to tell them apart.

Memory is reconstructive rather than a replay. The misinformation effect is misleading information after an event altering recall, and constructive memory is the broader tendency to fill gaps from schemas, as in Loftus's eyewitness research. Flashbulb memories for emotional public events feel vivid but can still be inaccurate. Anterograde amnesia impairs forming new explicit memories after damage, often involving the hippocampus, while retrograde amnesia is loss of memories formed before the damage.

Intelligence questions are really measurement questions (2.8)

Spearman proposed g, a general intelligence factor that underlies performance across cognitive tasks, identified through factor analysis of correlations among tests. Later theories split ability further. The Cattell-Horn-Carroll tradition separates fluid intelligence, reasoning and solving new problems, from crystallized intelligence, accumulated knowledge and vocabulary. Fluid performance tends to decline in later adulthood. Crystallized knowledge tends to hold or grow. A question describing an older adult who is slower on novel puzzles but excellent at crosswords is testing exactly that split.

Intelligence tests are also used for different purposes. An aptitude test is designed to predict future performance. An achievement test measures what has already been learned. Many exam questions about this topic do not focus on the purpose of a test but on whether the test is sound, which is evaluated using three properties.

Three additional findings are covered in this topic. The Flynn effect refers to the rise in average IQ scores across the twentieth century. Because this increase occurred too quickly to be explained by genetic change, it is generally attributed to environmental factors such as schooling and nutrition. Stereotype threat refers to the concern that awareness of a negative stereotype about one's group may lower performance in the area being tested. Mindset research distinguishes between a fixed mindset, in which intelligence is viewed as set at birth, and a growth mindset, in which intelligence is viewed as capable of developing through effort and strategy. The framework also includes emotional intelligence, which involves the ability to perceive, understand, and manage emotions in oneself and others.

For this course, focus on the definitions, the distinctions between types of intelligence, and the reasoning behind test measurement rather than on broader debates about the nature of intelligence.

Where students lose points

Errors in Unit 2 usually do not come from missing an entire topic. They more often involve two terms that sound similar, with a question written to test whether you can distinguish between them. The seven pairs below account for many of the most common errors in this unit.

When you read a practice question, look for the detail that distinguishes the two terms in the pair. In most questions, that detail appears in a single clause in the stem.

Those seven pairs are built into the Cognition deck in Rycal. Drill the pair that keeps tripping you instead of rereading the whole table.

Practice question 1

Practice question 2

The explanations identify each incorrect choice and explain why it does not fit the scenario. If you missed one, rate that card Still Learning in the Cognition deck in Rycal and drill it spaced over the next few days.

Eight questions to answer without looking

Answer each question out loud or in writing before you check your notes. Looking at the answer first turns the exercise into rereading, which involves the shallower processing described earlier in this guide. If writing helps, do a quick Brain Dump in Rycal first and then check what you left out.

1. A student misses a gorilla costume crossing the stage during a counting task. Name the phenomenon and state why it is not a memory failure.
2. Why does reading a handwritten shopping list go faster than sounding out each word, and which processing direction is that?
3. A voter judges a candidate likely to be corrupt because news stories about corruption come to mind easily. Which heuristic is that, and how does it differ from representativeness?
4. Give one example each of mental set and functional fixedness, using the same box of thumbtacks.
5. Your old locker combination keeps displacing your new one. Which interference is that, and which memory is being blocked?
6. Explain why cramming the night before loses to the same hours spread across a week, using the spacing effect and the forgetting curve.
7. An older adult is slower on brand-new puzzles but beats everyone at trivia night. Name the two intelligences and say which is doing which job.
8. A scale always reads exactly four pounds heavy. Is it reliable, valid, both, or neither? Defend your answer in one sentence.

Anything you cannot say out loud without looking belongs in the Cognition deck. In Rycal, mark that card Still Learning and it will return sooner while the cards you know move farther apart.

Questions students ask about Unit 2

How much of the exam is Unit 2?

The framework weights cognition at 15 to 25 percent of the multiple-choice questions, the same range as the other units. Memory accounts for a substantial portion of that range, perception and thinking appear regularly, and intelligence is usually tested through questions about measurement.

What is the difference between a heuristic and an algorithm?

An algorithm guarantees a correct answer if you complete its steps. A heuristic is a faster shortcut that is usually correct but can lead to errors in specific situations. Most decision questions describe a heuristic.

Why does rereading feel like it works?

Rereading mainly builds recognition of the page. A test usually requires recall of the ideas in a new scenario. That difference helps explain why students can feel prepared and score lower than expected. Self-testing practices the recall the test requires.

Can a test be reliable without being valid?

Yes. A scale that always reads four pounds heavy is consistent and wrong. Reliability is necessary for a good test but it is not sufficient. Validity asks whether the test measures what it claims to measure.

About this guide. Written for Rycal and aligned to the College Board AP Psychology course framework, Unit 2, topics 2.1-2.8. All questions and explanations are original Rycal writing. Rycal is independent and is not affiliated with or endorsed by the College Board.

Key terms for this unit

Accommodation (schemas), Achievement tests, Algorithm, Alzheimer's disease, Anterograde amnesia, Apparent movement, Aptitude tests, Assimilation (schemas), Attention, Autobiographical memory, Automatic processing, Availability heuristic, Binocular depth cues, Bottom-up processing, Central executive, Change blindness, Chronological age, Chunking, Closure, Cocktail party effect, Concept, Construct validity, Constructive memory, Context-dependent memory, Convergence, Convergent thinking, Creativity, Distributed practice, Divergent thinking, Echoic memory, Effortful processing, Elaborative rehearsal, Encoding, Encoding failure, Episodic memory, Executive functions, Explicit memory, External factors in perception, Figure-ground, Fixed mindset, Flynn Effect, Forgetting curve, Framing, Functional fixedness, Gambler's fallacy, General intelligence (g), Gestalt psychology, Growth mindset, Heuristic, Highly superior autobiographical memory, Historical misuse of intelligence testing, Iconic memory, Identifying students for educational services, Imagination inflation, Implicit memory, Infantile amnesia, Intelligence quotient (IQ), Interposition, Levels of processing model, Linear perspective, Long-term memory, Long-term potentiation, Maintenance rehearsal, Massed practice, Medulla, Memory consolidation, Mental age, Mental set, Metacognition, Method of loci, Misinformation effect, Mnemonic devices, Monocular depth cues, Mood-congruent memory, Multi-store model, Multiple abilities view of intelligence, Perception, Perceptual constancy, Perceptual set, Phonemic encoding, Phonological loop, Predictive validity, Primacy effect, Priming, Proactive interference, Procedural memory, Prospective memory, Prototype, Proximity, Psychometric principles, Recall, Recency effect, Recognition, Relative clarity, Relative size, Reliability, Representativeness heuristic, Repression, Retinal disparity, Retrieval practice, Retroactive interference, Retrograde amnesia, Schema, Semantic encoding, Semantic memory, Sensory memory, Serial position effect, Short-term memory, Similarity, Socio-culturally responsive assessments, Sociocultural biases in intelligence testing, Source amnesia, Spacing effect, Split-half reliability, Standardization, State-dependent memory, Stereotype lift, Stereotype threat, Structural encoding, Sunk-cost fallacy, Test-retest reliability, Testing effect, Texture gradient, Tip-of-the-tongue phenomenon, Top-down processing, Selective Attention, Inattentional Blindness, Grouping, Depth Perception, Color Constancy, Perceptual Adaptation, Stroboscopic Movement, Phi Phenomenon, Autokinetic Effect, Cognition, Nudge, Insight, Fixation, Intuition, Overconfidence, Memory, Neurogenesis, Shallow Processing, Deep Processing, Storing, Retrieval, Flashbulb Memory, Relearning, Retrieval Cues, Encoding Specificity Principle, Interleaving, Reconsolidation, Deja Vu, Intelligence, Cattell-Horn-Carroll Theory (CHC), Savant Syndrome, Grit, Emotional Intelligence, Stanford-Binet, Wechsler Adult Intelligence Scale (WAIS), Psychometrics, Normal Curve, Content Validity, Cohort, Validity, Visuospatial sketchpad, Within-group vs. between-group variation in IQ scores, Working memory model.