← Back to course
Statistics 9-12 / Week 06 / Tuesday
2/6
Week 06 · Samples, Surveys and Studies

Tuesday

Three kinds of study
// Asking a few runners about the course
⏱ about 20 min

Tuesday: Three Kinds of Study

"Three questions on the board," Wren says. "Do runners want a warm-up? Does a warm-up make laps faster? Did the early birds run faster?"

"Same word, warm-up," Comet says. "But those are three different studies. What do you notice?"

"The first asks a sample and reports," Wren says. "A sample survey. Chance picks who is asked."

"The second assigns a warm-up," Comet says. "Short or long, by a fair draw. Then we time laps. An experiment."

"The third just watches," Wren says. "The runners chose for themselves. An observational study. Nothing assigned."

Nova projects three cards with a different glow on each. "Would you like a hint? Ask where the chance is."

"Survey: chance picks the sample. Experiment: chance picks the treatment. Observational: no chance at all," Comet says.

Way 1: ask what the crew did

A sample survey asks or measures a sample to describe a population. The crew does nothing to the runners except ask.

An experiment assigns a treatment, like a short or a long warm-up, and then measures a result.

An observational study measures groups that already exist. The crew assigns nothing; it only watches.

Way 2: ask where the chance is

In a sample survey, random selection decides who is in the sample. That keeps the sample fair for the population.

In an experiment, random assignment decides who gets which treatment. That spreads lurking variables evenly between groups.

In an observational study, there is no random assignment. The groups chose themselves, so lurking variables can hide.

Both ways land on the same three names. Use whichever question is easier to answer first.

StudyWhat the crew doesWhere the chance isCan it show a cause?
sample surveyasks 10 random runners about a warm-upwho is pickedno, it describes
experimentassigns long or short warm-up by a fair draw, times lapswho gets which treatmentyes
observational studycompares runners who chose long or short warm-upsnowhereno, lurking variables
NAME THE STUDY
  • Read the question.
  • Tap your answer.
The crew draws 10 sign-up cards at random and asks each runner about the start time. What kind of study is this?
The crew assigns runners by a fair draw to a long or a short warm-up, then times their laps. What kind of study is this?
The crew compares lap times of runners who chose a long warm-up with runners who chose a short one. What kind of study is this?
The crew measures the stride length of 10 randomly chosen runners to describe all 40. What kind of study is this?
WHERE IS THE CHANCE?
  • Read the question.
  • Tap your answer.
How does randomization work in a sample survey?
How does randomization work in an experiment?
How does randomization work in an observational study?
Which kind of study can show that a warm-up causes faster laps? Type one word.
WHY THIS EXERCISERandom assignment is what lets a study say "because of" instead of "together with".
StatementTrue or false?
In a sample survey, chance decides who is in the sample.?
In an experiment, runners choose their own treatment.?
An observational study assigns no treatment.?
An observational study can prove a cause as well as an experiment can.?
WHY THIS EXERCISEKnowing the kind of study tells you what its result can and cannot claim.
Try it
Write three questions about your home, one for each kind of study. Say where the chance is in each.
For the experiment, describe the fair draw you would use to assign the treatment.
Draw three boxes labeled survey, experiment and observational. In each, sketch where the fair draw happens, if anywhere.

Three studies, sorted. Tomorrow is Data Lab: twenty samples from the bag, and how much a statistic wobbles.

← Monday