Digital SAT Math · Problem-Solving and Data Analysis

SAT Observational Studies and Experiments Practice Questions

These questions ask whether a study's design supports a conclusion: whether it shows cause and effect, and to which population its results apply.

College Board skill name: Evaluating statistical claims: Observational studies and experiments

Questions in the bank
200
Easy / Medium / Hard
60 / 90 / 50
Subtopics
10
Typical time
~1.2 min each

Updated . Questions are original BlitzSAT items written to the College Board's published skill specifications. They are not official SAT questions.

How Observational Studies and Experiments questions appear on the SAT

Descriptions of studies, followed by 'Which conclusion is appropriate?' The answer depends on random assignment and random sampling.

Observational Studies and Experiments is part of the Problem-Solving and Data Analysis domain, which makes up about 15% of the Math section (5–7 questions per test). A calculator is allowed on every Math question, and the built-in Desmos graphing calculator is available in Bluebook.

How to approach Observational Studies and Experiments questions

  1. 1Random assignment to treatment groups is what allows a cause-and-effect conclusion.
  2. 2Random selection from a population is what allows you to generalize to that population.
  3. 3An observational study (no assignment) supports association only, not causation.
  4. 4Check both conditions separately; a study can satisfy one and not the other.

Common mistakes

  • Claiming causation from an observational study.
  • Generalizing from volunteers or a convenience sample to a whole population.
  • Confusing random sampling with random assignment.

10 SAT Observational Studies and Experiments practice questions with answers

Pick an answer to check it. Each explanation covers the correct answer and why each of the other choices is wrong. The set runs from easy to hard.

0 / 10 answered
1EasyLimit a conclusion to what the study design supports
In a hypothetical study at a design workshop, researchers observe participants creating geometric logos. Participants decide for themselves whether to use a grid template. Among the participants observed, those using a template make fewer revisions on average than those not using one. Which conclusion is supported by this study?
Show answer and explanation

Answer: B

The observation supports an association between template use and fewer revisions in the participants studied. Because participants chose whether to use a template, other differences between the groups could explain the association. The study also provides no basis for extending the finding to all designers.Why the other choices are wrongAParticipants were not randomly assigned to use templates, so the observed association does not establish that templates caused fewer revisions.CThe study does not establish that the workshop participants represent all geometric-logo designers.DAn association does not establish causation in the reverse direction either. The study does not show that revision counts caused template choices.
2EasyExplain the role of a control group
In a hypothetical recreation-center study, members are randomly assigned to two groups. One group receives a new weekly message listing available activities; the other receives no message. After a month, both groups are surveyed about how many activities they attended. What is the primary role of the group that receives no message?
Show answer and explanation

Answer: C

The no-message group is the control group. Its attendance provides a comparison with attendance in the message group, helping researchers assess whether receiving the messages affects attendance.Why the other choices are wrongAA control group provides a comparison; it does not ensure that study participants represent all center members.BHaving a control group does not verify the accuracy of members' survey responses.DControl-group members' attendance can change for reasons unrelated to the messages. The group measures attendance without the new messages rather than preventing change.
3EasyIdentify selection bias in a survey
In a hypothetical algebra course, every student can access an online practice tool. To estimate the percentage of all students who find the tool easy to use, the instructor displays a survey only after a student completes ten practice sessions. Which change would most directly address the selection bias in this plan?
Show answer and explanation

Answer: C

Students who struggle with the tool may stop before completing ten sessions and therefore never see the survey. Selecting students randomly from the complete roster and contacting them outside the tool gives those students a chance to be included.Why the other choices are wrongAMore response time does not include students who never reach ten sessions.BSurveying everyone in the restricted group still excludes students who complete fewer than ten sessions.DAdding a question changes what is measured but leaves the biased selection process unchanged.
4MediumDetermine an appropriate population for generalization
In a hypothetical recreation survey, a town has a complete list of adults who registered for at least one town-run recreation class during the past year. The town randomly selects 120 people from that list. All selected people respond to a question about whether they would prefer weekend classes to weekday classes. Which is the largest population to which the survey results can reasonably be generalized?
Show answer and explanation

Answer: B

The random sample was drawn from the complete list of adults who registered for classes during the past year, and everyone selected responded. The results can therefore reasonably be generalized to that population. Adults outside that list and future registrants were not the population sampled.Why the other choices are wrongATown residents who did not register for a class were not eligible for selection, so the sample does not represent all adult residents.CThis is unnecessarily restrictive. Random sampling supports generalization beyond the respondents to the population from which they were selected.DNext year’s registrants may differ from last year’s registrants. They are not the population from which this sample was drawn.
5MediumIdentify a plausible confounding variable
In a hypothetical geometry project, students choose to build either arched or flat cardstock bridges and also choose their cardstock thickness. The arched bridges generally use thicker cardstock and support greater maximum loads than the flat bridges. Every bridge spans the same distance between supports, and bridges are tested in a randomly selected order. A student concludes that the arched shape caused the greater load capacity. Which variable is the most plausible confounding variable?
Show answer and explanation

Answer: C

Cardstock thickness differs systematically between the shape groups and could also affect how much load a bridge supports. Therefore, the observed difference in load capacity cannot be attributed to shape alone: thickness is a plausible confounding variable.Why the other choices are wrongAThe distance between supports is the same for every bridge, so it does not explain the difference between the arched and flat groups.BMaximum supported load is the outcome being compared, not a separate variable that could confound the relationship between shape and load capacity.DTesting order was randomized, and the problem provides no systematic association between order and bridge shape. Thickness, in contrast, is explicitly associated with shape.
6MediumIdentify the benefit of random assignment
In a hypothetical transport-planning study, volunteers test one of two station maps by finding a specified platform. Some volunteers already know the station well. Researchers randomly assign volunteers to the two map groups and measure the time each volunteer takes. What is the main benefit of random assignment in this study?
Show answer and explanation

Answer: A

Random assignment tends to distribute participant characteristics, including station familiarity, across treatment groups without systematically favoring either map. This helps separate the maps' effects from preexisting differences, although exact balance is not guaranteed.Why the other choices are wrongBAssignment determines which map volunteers use; it does not make a volunteer sample representative of all station users.CRandom assignment can produce unequal numbers of familiar users by chance.DFamiliarity can still affect individual completion times. Random assignment addresses its distribution between groups, not its effect on individuals.
7MediumEvaluate voluntary-response sampling
In a hypothetical town survey, every household receives a link to an optional questionnaire asking whether outdoor odors frequently interfere with use of its yard. Most responding households answer yes. An organizer argues that collecting many more responses through the same optional questionnaire will ensure an unbiased estimate for all town households. Which evaluation of this argument is most appropriate?
Show answer and explanation

Answer: D

Participation is voluntary, so households bothered by odors may be especially motivated to respond. Collecting more responses through the same process does not necessarily remove this self-selection bias. The information does not establish the actual size or direction of the bias.Why the other choices are wrongAReceiving an invitation is not the same as being randomly selected as a respondent. Households decide for themselves whether to participate.BA larger sample does not automatically correct systematic differences caused by voluntary participation.CVoluntary response does not necessarily produce an underestimate. If affected households are more likely to respond, the result could instead be an overestimate.
8HardIdentify when a causal claim is justified
In a hypothetical investigation, managers of 18 facilities volunteer to test a cartridge intended to reduce particles in collected runoff. At each facility, researchers divide one runoff sample equally between two identical tanks. A random process assigns one tank to use a pump with the cartridge and the other to use the same pump without the cartridge. All other procedures are identical. Particle concentrations are consistently lower in the cartridge tanks, with a difference too large to reasonably attribute to chance. Which conclusion is best supported?
Show answer and explanation

Answer: B

Random assignment within each facility allows a causal comparison between using the pump with the cartridge and using the pump alone. Thus, the evidence supports an effect of adding the cartridge under the tested conditions. Because there was no untreated group and the facilities volunteered, the investigation does not establish the effect of the entire system versus no treatment or justify a conclusion about all facilities.Why the other choices are wrongABoth groups used a pump. Without an untreated comparison group, the experiment does not establish the combined effect of the pump and cartridge relative to leaving runoff untreated.CThe facilities volunteered rather than being randomly sampled from all facilities, and the experiment tested only particular conditions. A claim about all facilities is not justified.DVoluntary participation limits generalization, but it does not eliminate the causal evidence provided by random assignment of the cartridge within each facility.
9HardDistinguish random sampling from random assignment
Two hypothetical school projects investigate whether daily reminders affect assignment completion. Project 1 randomly selects students from the school's complete enrollment list; everyone selected participates. These students choose whether to receive reminders. Project 2 recruits volunteers from the same school and randomly assigns each volunteer either to receive reminders or not to receive them. Both projects compare assignment completion between their reminder and no-reminder groups. Which statement correctly describes what the designs can support?
Show answer and explanation

Answer: C

Project 1 uses random sampling, which supports generalization to the school, but students' self-selection into reminder groups prevents a causal conclusion. Project 2 uses random assignment, which supports evaluating a causal effect among participating volunteers. However, those volunteers were not randomly sampled from the school, so schoolwide generalization is not automatically justified.Why the other choices are wrongAThis reverses the roles of the two random processes. Random sampling in project 1 does not eliminate differences caused by students choosing their groups, while random assignment in project 2 supports causal inference.BProject 1 cannot support a causal conclusion because reminder use is self-selected. Its random sample addresses representativeness, not treatment-group comparability.DProject 2 can support causal inference because it randomly assigns treatment. That assignment does not make its volunteer participants representative of the school.
10HardCompare observational and experimental designs
Two hypothetical studies examine a panel coating and scratch resistance.Study 1 randomly samples panels from a factory's monthly output. Researchers compare coated and uncoated panels that were produced using methods chosen by factory managers.Study 2 uses panels supplied by one manufacturer, without randomly sampling them. Researchers randomly assign half to receive the coating followed by high-temperature drying and half to receive no coating followed by low-temperature drying. All other processing and testing conditions are identical.Which statement best describes what these designs can establish?
Show answer and explanation

Answer: B

Study 1 is observational: managers, not the researchers' random process, chose the production methods. Random sampling does not rule out other explanations for an association. Study 2 randomly assigns two treatment packages, so it can assess the causal effect of coated/high-temperature processing versus uncoated/low-temperature processing for the experimental panels. Because coating and temperature change together, their separate effects cannot be identified. The lack of random sampling limits generalization, not the randomized comparison itself.Why the other choices are wrongARandom sampling in Study 1 helps with representation of monthly output, but it does not make coated and uncoated panels comparable in other production characteristics.CIn Study 2, drying temperature changes along with coating status. Random assignment permits a causal comparison of the packages, not attribution to the coating alone.DRandom sampling is not required for a randomized experiment to assess a causal treatment effect among its experimental units. Its absence instead limits how broadly that effect can be generalized.

These are 10 of 200 questions on this skill in the BlitzSAT question bank.

Practice the rest by difficulty, timed or untimed, with your history saved. Free, no credit card.

Practice all 200 questions

What the 200 Observational Studies and Experiments questions cover

The BlitzSAT bank splits this skill into 10 subtopics. Counts are from the current bank.

SubtopicQuestionsEasyMediumHard
Compare observational and experimental designs20695
Determine an appropriate population for generalization20695
Distinguish random sampling from random assignment20695
Evaluate voluntary-response sampling20695
Explain the role of a control group20695
Identify a plausible confounding variable20695
Identify selection bias in a survey20695
Identify the benefit of random assignment20695
Identify when a causal claim is justified20695
Limit a conclusion to what the study design supports20695

Other Problem-Solving and Data Analysis skills

More Math skills

Keep going