Skip to content

Free-response practice

The Evidence-Based Question (EBQ)

7 points2 original prompts4 scored examples

The Evidence-Based Question (EBQ) is the second of the two free-response questions on the AP Psychology exam. Section II gives you 70 minutes for both questions and counts for 33.3 percent of your score. Each question is worth 7 points, so the EBQ alone is about one-sixth of your total. The exam suggests spending about 45 minutes on it, which leaves the other 25 for the Article Analysis Question.

On this page · 6 sections

You get a question on one topic, such as whether sleep helps memory, plus three sources. Each source is a short summary of a published study that lists the participants, the method, the results and what the authors concluded. Your job is to build a small argument in five labeled parts.

Part A (1 point) is your claim, a specific position that answers the question, and Part B(i) (1 point) is one piece of specific evidence from one source. Part B(ii) (2 points) explains how that evidence supports your claim, using psychology from the course. Part C(i) (1 point) is a second piece of evidence from a different source. Part C(ii) (2 points) explains that evidence, using a different piece of psychology than you used in B(ii).

Count where the points sit: the two explanation parts carry 4 of the 7. Citing evidence is the easy part, but explaining it through a named concept, such as memory consolidation, extinction or the sympathetic nervous system, is where the score is decided.

Readers score each row of the rubric for the specific move it names; they do not grade style, and a longer answer earns nothing extra. A plain sentence that names the concept and states the link beats a polished paragraph that only hints at it. The exam is given in the Bluebook app, so you type your answer and should label every part A, B(i), B(ii), C(i), C(ii) so the reader can find each point.

The rubric, in plain terms

7 points, earned line by line. Readers score what is on the page, nothing else.

Part A: defensible claim1 pt

States a clear position that answers the question and that the sources can actually support. It must say what happens and in which direction, such as "Sleeping soon after learning strengthens memory for both facts and skills." A claim may include a condition if the sources show one. Restating the topic ("Sleep and memory are related") does not earn the point, and neither does a hedge like "it works for some people but not others" when no source tests individual differences.

Part B(i): evidence from one source1 pt

Cites a specific, accurately reported result from one named source that bears on the claim. Good evidence names the groups or variables and gives the result, such as a mean, a percentage, a correlation or a before-and-after change. "Source 1 shows sleep helps" is a conclusion, not evidence, so it does not earn the point, and misreporting a number or a group also loses it.

Part B(ii): explanation using course psychology2 pts

First point: explains accurately how the cited result supports the claim. It must spell out the link rather than repeat the result, and it must not draw a conclusion the study design cannot support, such as cause from a correlation. Second point: makes that explanation through a psychological perspective, theory, concept or research finding from the AP Psychology course. The concept must be named, described correctly and applied to this evidence; a misdefined concept, or one dropped in without being applied, loses the second point. A sound explanation with no concept can still earn the first.

Part C(i): evidence from a different source1 pt

Meets the same standard as B(i), with one extra condition: the evidence must come from a different source than the one used in B(i), so citing a second number from the same source does not earn the point. A vague gesture at a source ("Source 3 shows some people relapsed") with no result or condition named also falls short.

Part C(ii): explanation using different course psychology2 pts

Meets the same standard as B(ii). First point: an accurate explanation of how the C(i) evidence supports the claim. Second point: the explanation runs through a psychological perspective, theory, concept or research finding that is different from the one used in B(ii), named and applied correctly. Reusing the B(ii) concept caps this part at the first point.

Practice prompts

Prompt 1

Does sleep after learning strengthen memory?

Question: Using the sources below, develop and justify an argument about whether sleeping after learning improves memory.

The three sources are fictional studies written for practice.

SOURCE 1 Introduction: Researchers tested whether a night of sleep protects newly learned material. Participants: 60 university students in the Netherlands, ages 18 to 24, recruited with campus flyers. All of them reported normal sleep and no sleep disorders. Method: At 9 p.m., each participant studied 40 pairs of unrelated words, such as cloud and hammer. They kept studying until they could recall at least 32 of the 40 pairs, and participants were then randomly assigned to one of two groups. The sleep group slept in the lab while EEG recorded their sleep stages. The wake group stayed awake in the lab all night, supervised, watching films. Both groups slept normally at home the next night. At 9 p.m. two days after learning (48 hours later), each participant saw the first word of each pair and had to recall the second. Results: The sleep group recalled an average of 30 of 40 pairs, while the wake group recalled an average of 23. The difference was statistically significant (p < .01). In the sleep group, minutes spent in slow-wave sleep (NREM stage 3) were positively correlated with recall (r = +.45). Discussion: The authors concluded that the first night of sleep after learning helps stabilize new memories. A later night of recovery sleep did not make up the loss.

SOURCE 2 Introduction: Researchers asked whether teenagers who sleep more learn new words better. Participants: 418 students, ages 13 to 17, at two public high schools in Australia. Parents gave consent and students agreed to take part. Method: Students wore wrist activity monitors for 14 school nights. Average sleep was 6.9 hours a night, ranging from 5.1 to 9.0 hours. During the second week, students studied 30 unfamiliar vocabulary words on Monday morning and were tested on them Friday. Results: Average nightly sleep was positively correlated with Friday recall (r = +.34, p < .001). Students who averaged under 7 hours recalled an average of 15 of 30 words, while students who averaged 7 hours or more recalled 19. Students who slept less also reported more caffeine and more evening screen time. Discussion: The authors noted that their design could not rule out other factors that differ between short and long sleepers.

SOURCE 3 Introduction: Researchers tested whether a daytime nap improves a physical skill. Participants: 48 paid adult volunteers in Canada, ages 20 to 35. Method: At noon, every participant practiced typing a five-key finger sequence as fast and accurately as possible. Participants were randomly assigned to take a 90-minute nap in the lab, with EEG recording, or to rest awake for 90 minutes while watching a nature video. At 4 p.m., everyone was tested on the sequence again with no further practice. Results: The nap group typed the sequence 16 percent faster than at the end of training, while the rest group was 3 percent faster. The difference was statistically significant (p < .01). Among nappers, the number of sleep spindles recorded in NREM stage 2 correlated with improvement (r = +.52). Discussion: The authors concluded that sleep strengthens memory for skills, not only memory for facts.

  1. (A)State a specific, defensible claim, grounded in psychological science, that answers the question. [1 pt]
  2. (B(i))Support your claim with one specific piece of evidence from one of the sources. [1 pt]
  3. (B(ii))Explain how the evidence in B(i) supports your claim, using a psychological perspective, theory, concept or research finding from AP Psychology. [2 pts]
  4. (C(i))Support your claim with one specific piece of evidence from a source other than the one you used in B(i). [1 pt]
  5. (C(ii))Explain how the evidence in C(i) supports your claim, using a psychological perspective, theory, concept or research finding different from the one you used in B(ii). [2 pts]
Prompt 2

How well does exposure therapy treat phobias?

Question: Using the sources below, develop and justify an argument about how effective exposure-based therapy is for treating specific phobias.

The three sources are fictional studies written for practice.

SOURCE 1 Introduction: Researchers tested a single long session of exposure therapy for children afraid of spiders. Participants: 72 children in Sweden, ages 8 to 14, each diagnosed with a specific phobia of spiders and referred by local clinics. Parents gave consent and children agreed to take part. Method: Children were randomly assigned to treatment or a waitlist, and the treatment was one 3-hour session. The therapist and child built a fear hierarchy, from looking at photos, to holding a closed jar with a spider, to letting a spider walk across the child's hand. The child moved up a step only when fear at the current step had dropped. Fear was measured with a behavioral approach test: the number of steps, out of 12, the child completed toward a live spider. Results: One week later, treated children completed an average of 10.1 of 12 steps. Waitlist children completed 3.2 (p < .001). On a 0 to 10 scale, treated children rated their fear 3.1 and waitlist children 7.8. One year later, treated children still averaged 9.4 steps. Discussion: The authors concluded that one session of gradual exposure produced large, lasting gains.

SOURCE 2 Introduction: Researchers compared exposure in virtual reality with exposure in real places for fear of heights. Participants: 60 adults in Spain, ages 19 to 58, diagnosed with a specific phobia of heights. Method: Participants were randomly assigned to six sessions of virtual reality exposure (virtual balconies, glass elevators and rooftops) or six sessions of real-world exposure on actual balconies and rooftops. Before and after treatment, each person stood on a real fifth-floor balcony while heart rate was recorded. Results: In the virtual reality group, average heart rate on the balcony fell from 104 to 82 beats per minute; in the real-world group, it fell from 106 to 80. The two groups did not differ significantly. 18 percent of the virtual reality group quit treatment early, compared with 33 percent of the real-world group. Discussion: The authors concluded that virtual exposure works about as well as real exposure and that more patients are willing to finish it.

SOURCE 3 Introduction: Researchers tested whether fear returns after successful exposure therapy when the setting changes. Participants: 45 adults in the United States with a specific phobia of snakes, recruited through newspaper ads. Method: All participants had exposure therapy with a caged snake until their fear rating fell below 2 out of 10. Participants were randomly assigned to one of three groups. Group 1 was treated in one room and tested two weeks later in that same room, while Group 2 was treated in one room and tested in a new room. Group 3 was treated across three different rooms and tested in a new room. Results: Average fear ratings at the test were 2.5 for Group 1, 5.6 for Group 2 and 3.0 for Group 3 (out of 10). The difference between Group 2 and the other two groups was statistically significant (p < .01). Discussion: The authors concluded that fear reduced by exposure can return in a new setting, and that treating patients in several settings protects against this.

  1. (A)State a specific, defensible claim, grounded in psychological science, that answers the question. [1 pt]
  2. (B(i))Support your claim with one specific piece of evidence from one of the sources. [1 pt]
  3. (B(ii))Explain how the evidence in B(i) supports your claim, using a psychological perspective, theory, concept or research finding from AP Psychology. [2 pts]
  4. (C(i))Support your claim with one specific piece of evidence from a source other than the one you used in B(i). [1 pt]
  5. (C(ii))Explain how the evidence in C(i) supports your claim, using a psychological perspective, theory, concept or research finding different from the one you used in B(ii). [2 pts]

Scored examples

Two responses per prompt — one near the top of the rubric, one in the middle — with every point accounted for. The mid response earns its score through the mistakes students actually make.

Does sleep after learning strengthen memory? · high-scoring response7/7

A. Sleeping soon after learning strengthens memory for both facts and physical skills, so cutting sleep after studying means losing part of what you learned.

B(i). In Source 1, students who slept the night after studying word pairs recalled an average of 30 of 40 pairs two days later. Students who were kept awake that night recalled only 23, and the difference was significant (p < .01), with students randomly assigned to the groups.

B(ii). This supports my claim through memory consolidation. A new memory starts out fragile, and during slow-wave sleep the hippocampus replays it and helps move it into stable long-term storage. The wake group missed that first night of consolidation, and Source 1 says one night of recovery sleep later did not make up the loss. Within the sleep group, more minutes of slow-wave sleep went with better recall (r = +.45), which is what consolidation predicts. Since sleep was randomly assigned, the experiment lets us say sleep caused the better recall.

C(i). In Source 3, adults who took a 90-minute nap after practicing a finger-tapping sequence typed it 16 percent faster four hours later, while adults who rested awake improved only 3 percent.

C(ii). This supports the skills half of my claim through procedural memory. Procedural memory is a kind of implicit memory for how to do things, like typing or riding a bike, which you show by performing rather than by consciously recalling it. The nappers got faster with no extra practice, so the skill improved while they slept. That means sleep strengthens implicit skills as well as the explicit word-pair memories in Source 1.

How it scored

1/1A — A clear, specific position (sleep after learning strengthens memory for facts and skills) that two experiments in the sources can back.
1/1B(i) — Names the source, both groups and the exact result, with significance. Mentioning random assignment sets up the causal reasoning later.
2/2B(ii) — First point: explains why the wake group remembered less and ties the slow-wave correlation to the claim. Second point: memory consolidation is named, described correctly and applied to this study's design.
1/1C(i) — Different source, specific numbers for both conditions, accurately reported.
2/2C(ii) — First point: explains that improvement without practice means the gain happened during sleep. Second point: procedural (implicit) memory is a different concept from B(ii), defined correctly and applied to the finger-tapping task.
Does sleep after learning strengthen memory? · mid-scoring response3/7

A. Sleeping after you learn something helps you remember it better later on.

B(i). Source 1 says the sleep group remembered 30 out of 40 word pairs and the wake group remembered 23 out of 40.

B(ii). This supports my claim because the people who slept remembered more than the people who stayed up. When you sleep you go into REM sleep, which is stage 3, where you have delta waves and vivid dreams, and in REM the brain goes over what you learned that day, which is why the sleep group did better.

C(i). In Source 2, teens who slept more remembered more vocabulary words, with a correlation of r = +.34.

C(ii). This proves that getting more sleep causes better memory, because if teenagers slept 7 hours or more they would remember 4 more words each week, so schools should start later so students can sleep more.

How it scored

1/1A — Thin but defensible, because it takes a position the sources support. A stronger claim would say what kind of memory, or when the sleep has to happen.
1/1B(i) — Specific and accurate: names the source, both groups and their scores.
0/2B(ii) — No points. The first sentence only repeats the result, and the rubric asks for the link, not the result again. The attempt at a link rests on a wrong account of sleep, since REM is not stage 3: delta waves belong to NREM stage 3, while vivid dreaming belongs to REM. Source 1 also tied recall to slow-wave sleep, not REM. A stronger answer would say the wake group lost the chance to stabilize the memory that first night.
1/1C(i) — A different source, with an accurately reported correlation in the right direction.
0/2C(ii) — No points. Source 2 is correlational, and its own discussion warns that caffeine and screen time differ between short and long sleepers. "Proves that more sleep causes" is a conclusion the design cannot support. No psychological concept is named, and the school start-time point drifts away from the claim.
How well does exposure therapy treat phobias? · high-scoring response6/7

A. Exposure therapy is an effective treatment for specific phobias, and its gains carry over to new settings best when exposure happens in more than one setting.

B(i). In Source 1, children with a spider phobia who had one 3-hour exposure session completed an average of 10.1 of 12 steps toward a live spider a week later. Children on the waitlist completed 3.2. A year later, the treated children still averaged 9.4 steps.

B(ii). This supports my claim through extinction in classical conditioning. For these children, the spider is a conditioned stimulus that has become linked with fear, and in exposure the child meets the spider over and over and nothing bad follows, so the spider is no longer paired with anything harmful. The conditioned fear response weakens, and the child can approach, while the waitlist children had no such experience, so their fear stayed high. Keeping 9.4 steps after a year shows the new learning lasted.

C(i). In Source 3, patients treated in one room and then tested in a new room rated their fear 5.6 out of 10, compared with 2.5 for patients tested in the room where they were treated. Patients treated in three rooms and tested in a new room rated it only 3.0.

C(ii). This supports the second half of my claim, also through extinction. Extinction does not erase the original fear; it adds new learning that the snake is safe, and that new learning was tied to the treatment room, so when the setting changed, fear came back. Treating in three rooms kept fear low in the new room, which is why exposure works best across settings.

How it scored

1/1A — Specific and defensible, because it answers "how effective" and adds a condition that Source 3 actually tests: whether fear returns when the setting changes. Source 3 compares settings, not how long gains last, so the claim rightly talks about new settings, not durability.
1/1B(i) — Exact results for both groups plus the one-year follow-up, all accurately reported.
2/2B(ii) — First point: clearly explains why the treated group could approach the spider. Second point: extinction is named, the spider is correctly called a conditioned stimulus, and the mechanism is applied to this study.
1/1C(i) — Different source, with all three groups' results reported accurately.
1/2C(ii) — First point earned: the explanation of why fear returned in a new room is accurate and tied to the claim. Second point lost because extinction was already the concept in B(ii). Stimulus generalization would have worked here, since training in three rooms let the new safety response spread to an unfamiliar room.
How well does exposure therapy treat phobias? · mid-scoring response1/7

A. Exposure therapy can work for some people with phobias but not for others, depending on the person.

B(i). Source 2 says the people who did virtual reality therapy had their heart rate on the balcony go from 104 to 82 beats per minute.

B(ii). This shows exposure therapy works, because a lower heart rate means the person is less afraid, and a high heart rate comes from the parasympathetic nervous system, which is the fight-or-flight system. After therapy, the person's sympathetic nervous system calmed them down, so they were not as scared on the balcony.

C(i). Source 3 shows that for some people the fear came back after therapy.

C(ii). This is stimulus generalization, because the patients were scared of the snake in the first room, and that fear generalized to the new room, so the therapy did not work for them. This proves therapy only works for some people.

How it scored

0/1A — Not earned. "It depends on the person" is a hedge, not a position, and none of the sources compares types of people, so the sources cannot back this claim as stated.
1/1B(i) — Specific and accurate: names the source, the group and the before-and-after heart rates.
0/2B(ii) — No points. The first point needs an explanation of how the evidence supports the claim in Part A. This one reads the drop in heart rate as less fear, which argues that the therapy worked, so it does not back the student's claim that therapy works only for some people. The second point is lost because the divisions are reversed: the sympathetic system drives fight-or-flight and raises heart rate, while the parasympathetic system calms the body.
0/1C(i) — Not earned. It names a different source but gives no result or condition, and it does not say that fear returned in a new room (5.6) and stayed low in the treatment room (2.5).
0/2C(ii) — No points. The explanation says fear returned because of the person. Source 3 shows the setting made the difference: Group 3 was treated in three rooms and stayed low (3.0) in a new room. That makes the explanation inaccurate. "This proves therapy only works for some people" is an overclaim the sources cannot support. Generalization is also misapplied, because the fear was never tied to the room; what failed to transfer was the new safety learning.

Where points die

Hedging the claim in Part A, such as "exposure therapy works for some people and not others" or "sleep and memory are connected." Take a side the sources can carry and say which way the effect runs. You may add a condition, but only one a source actually tests, such as "especially when exposure happens in several settings."

Treating a correlational source as proof of cause, such as "Source 2 proves more sleep causes better word recall." Check how each source got its data, and if nothing was randomly assigned, say the variables are linked and name a possible third variable, like caffeine or screen time. Save causal language for the sources that are experiments with random assignment.

Citing a conclusion instead of evidence, such as "Source 1 shows sleep helps memory." Report the result itself: which groups, what was measured and the numbers. "The sleep group recalled 30 of 40 pairs and the wake group 23" earns B(i), but the author's conclusion does not.

Naming a concept that sounds right but describing it wrong, such as calling REM sleep "stage 3 with delta waves," or saying the parasympathetic system causes fight-or-flight. Know the pairs cold: REM brings vivid dreams, rapid eye movements and paralyzed muscles. NREM stage 3 is the deepest sleep, with slow delta waves, and is when sleepwalking and night terrors happen. The sympathetic division speeds the heart and mobilizes energy, while the parasympathetic division slows the heart and restarts digestion.

Mislabeling the parts of classical conditioning, such as calling the feared spider the unconditioned stimulus. The unconditioned stimulus triggers a response with no learning at all, like the food in Pavlov's dogs' mouths. A feared object learned through a bad experience is a conditioned stimulus. If a child bitten by one terrier now fears every dog, that is stimulus generalization: the response spreads to similar stimuli.

Reusing the B source in C(i), or the B(ii) concept in C(ii). Plan before you write by picking two sources and two different concepts, such as extinction for Source 1 and stimulus generalization for Source 3. A repeated source loses C(i), and a repeated concept caps C(ii) at 1 point.

Score your own response

0 / 7

Be the harshest reader in the room: award a point only when your response does what the line says, on the page — not what you meant to write.