2.5
A researcher records students' sleep hours and GPAs without changing anything. Is this an observational study or an experiment?
Observational Study
What is the purpose of a control group in an experiment?
To provide a baseline for comparison with the treatment group.
What type of experimental design randomly assigns all experimental units directly to treatment groups without forming blocks?
Completely randomized design.
What is the largest ocean on earth?
Pacific Ocean
Which allows researchers to generalize results to the population: random sampling or random assignment?
Random Sampling
Researchers investigate whether drinking coffee affects test scores. Identify the explanatory and response variables.
Explanatory: coffee consumption. Response: test scores.
A scientist tests a new fertilizer on only one plant. Which principle of experimental design is missing?
Replication
Researchers divide participants into age groups, then randomly assign treatments within each age group. What design is this?
Randomized block design.
What is the only U.S. state that borders exactly one other U.S. state?
Maine
Researchers run an experiment to determine whether a new medication improves recovery rates. Assuming the medication has no effect, only 4 out of 200 simulations produce results as extreme as the actual experiment.
What percentage of simulations produced these results, and what can researchers conclude?
2% of simulations produced results as extreme as the actual experiment. Since 2% is less than 5%, the results are statistically significant, providing convincing evidence that the medication improves recovery rates.
Students who play sports have higher GPAs than students who don't. Can researchers conclude that playing sports causes higher GPAs? Why or why not?
No. An observational study cannot establish causation because confounding variables may explain the association.
In a drug experiment, neither the patients nor the researchers know who received the real medication and who received the placebo. What type of blinding is used?
Double-Blind experiment
Researchers pair participants with similar GPAs and randomly assign one person in each pair to each treatment. What design is this?
Matched Pairs Design
Which U.S. president was the youngest person ever elected president?
John F. Kennedy, 43 years old
A researcher randomly selects 200 students from a school and records whether they exercise. Can the researcher generalize results to that school's students? Can they establish causation?
Generalize: yes, to the school's students. Causation: no, because there was no random assignment.
Researchers randomly assign 100 students to take an SAT prep class or not. Why is this better for determining causation than allowing students to choose their own group?
Random assignment helps create roughly equivalent groups, reducing the influence of confounding variables and allowing stronger cause-and-effect conclusions.
A coach allows athletes to choose between two training programs and compares their improvement. Identify the problem with this design.
There is no random assignment. Athletes who choose one program may differ systematically from those who choose the other.
A researcher wants to compare two diets using 80 participants: 40 athletes and 40 non-athletes. Explain how to use a randomized block design.
Create two blocks: athletes and non-athletes. Within each block, randomly assign 20 participants to each diet. Compare outcomes across treatments, accounting for the blocks.
What is the only mammal capable of true, sustained flight?
a bat
A researcher randomly assigns 50 volunteers to two different diets. One diet produces significantly greater weight loss. The researcher concludes that the diet causes greater weight loss for everyone in the United States. What is wrong with the researcher's conclusion?
The researcher can conclude that the diet caused greater weight loss in the experiment because of random assignment and statistical significance. However, the researcher cannot generalize the results to everyone in the United States because the participants were volunteers, not a random sample.
A study finds that students who eat breakfast score higher on tests. Identify a possible confounding variable and explain why.
Amount of sleep, for example, could be related to eating breakfast and could also affect test scores.
Other examples: family income, parental involvement, time management
A researcher tests a new headache medicine. The treatment group takes the medicine in the morning, while the control group takes a placebo at night. Explain the design flaw and how to correct it.
Time of administration is a potential confounding variable. Both groups should receive their treatments under comparable conditions, including the same time of day.
A researcher wants to compare two types of energy drinks using 30 students. Design a matched pairs experiment in which every student tries both drinks.
Each student receives both drinks in a randomly determined order, ideally with sufficient time between trials. Measure performance after each drink and compare the paired results.
Which country has the most time zones in the world when its overseas territories are included?
France, with 12
Researchers randomly select 100 students from a high school and randomly assign them to two different study methods. Method A produces higher average scores, but a simulation shows that a difference this large or larger would occur by chance 23% of the time if the methods were equally effective.
Can researchers conclude that Method A causes higher scores? Can they generalize their findings to the entire school? Explain both.
No, researchers cannot conclude that Method A causes higher scores because the results are not statistically significant (23% > 5%). Although random assignment allows researchers to investigate causation, there is insufficient evidence that Method A is more effective. However, researchers can generalize their findings to all students at the high school because they used a random sample, but they cannot conclude that Method A is more effective for the entire school.