The language of experiments
Karina wants to determine if kale consumption has an effect on blood pressure. She recruits households and randomly assigns each household to either a kale-free diet plan or a kale-based diet. At the end of two months, she plans to record the original and final blood pressures for members of each household.
What is the explanatory variable?
What is the response variable?
What are the treatments?
Who or what are the experimental units?
Want to join the conversation?
- Why wouldn't the answer for problem 4 be the members of each household?(29 votes)
- The members of each household receive the same diet, but each household could have a different diet. Since the diet is consistent in the household, the household is the experimental unit.(59 votes)
- Words are slippery. The last question expects as correct answer the households, not their members. Makes me scratch my head and and ironically conclude: "Aha! So it is the households who eat (or not) the kale, not the household members".(5 votes)
- That is a funny point, but the household is the whole of the household members. We were comparing different households as a whole and not a households individual members.
Hope this helps,
- Convenient Colleague(6 votes)
- why didnt seattle just run the ball..(7 votes)
- There running back isnt good thats why lol(1 vote)
- If the household members are not the experimental units, then what are they?(4 votes)
- What can be the possible source of bias here ?(2 votes)
- what does confounding mean(1 vote)
- I THINK what differs is that if you compare households to households, you compare the average result of a household to another. It's different if the experimental unit is the members of each household, it seems like we don't take the average of the result, but we take it as individual results, which we don't want because we have to see whether or not there is a definite pattern (by averaging the final result of members in a household).
If we don't take the average of it, it would be MORE prone to bias since each individual might have different habit/some condition that might affect the result (even after we take prior action to minimize the things that we don't want to affect the result, it's still more prone to bias if we were to compare it to the method of analyzing it by taking the average of each household). I might be wrong, though. I think Khan Academy should have make it clear in the first place.(1 vote)
- Anybody got answers?(1 vote)
- number 4 should it be all of the people in the house and not just say it was a household?(1 vote)
- Isn't number 4 each individual member?(0 votes)