What this lesson covers
- Find the list the random draw actually came out of. That list is the population.
- Why a conclusion cannot travel WIDER than the list
- Why it cannot narrow to a subgroup either, which almost nobody teaches
- Why a mean is never a claim about each person
- The three traps and the one question to ask every time
- A random sample represents the group it was drawn from, and nothing wider
- A subgroup needs its own random sample, not a slice of yours
- A mean describes the group, not any individual in it
- Find the list first, then keep the conclusion on exactly that list
Chapters
- 0:00Intro
- 0:04A number, and a question about who it covers
- 0:14The study, and four conclusions
- 0:34The list is the population, and B docks
- 0:55Watch the other three miss
- 1:20Three traps
- 1:49Recap
Lesson transcript
The narration of the video, word for word, under its chapter headings.
0:14The study, and four conclusions
Two hundred students chosen at random out of the fourteen hundred enrolled at one high school. Average sleep on a school night, six point eight hours. Four conclusions are offered. Only one of them stays inside what this sample can support, and the other three each stretch it in a different direction.
0:34The list is the population, and B docks
The random draw came out of exactly one list. The fourteen hundred students enrolled at this school. That list is the population, and it is exactly how far the conclusion is allowed to travel. So the mean for all fourteen hundred students at this school is about six point eight hours. That one docks.
0:55Watch the other three miss
Now watch the other three miss. All high school students in the state is wider than the list, and nobody outside this school ever had a chance of being picked. Students who play a sport is narrower than the list, and the sample was never drawn from that group on its own. And every student sleeps about six point eight hours turns an average into a claim about each person, which no average can carry.
1:20Three traps
Three traps. One, going wider than the list. A random sample only represents the group it was drawn from. Two, going narrower. A subgroup needs its own random sample, not a slice of yours. Three, reading the mean as a rule about individuals. Six point eight is the average of the group, and plenty of students in it sleep four hours, or nine. The fast check. Find the list the researcher drew from, then keep the conclusion on exactly that list.
1:49Recap
The conclusion travels exactly as far as the list you sampled. No wider, and no narrower.
Read the written version: the Evaluating statistical claims strategy guide, then try 25 hard Evaluating statistical claims questions with full explanations. Both are free, no account needed.