- CFA Exams
- CFA Level I Exam
- Topic 1. Quantitative Methods
- Learning Module 7. Estimation and Inference
- Subject 1. Sampling Methods
CFA Practice Question
Which of the following is NOT a reason for sampling a population?
A. Inferences from a sample are the same as those which would be obtained by studying the entire population.
B. Some tests are destructive in nature and hence should be carried out only on small samples.
C. It is almost impossible to work with an entire population, in most cases.
Explanation: Samples are far easier to work with than entire populations. However, an entire population contains more information than a sample. With the population, you would get exact parameter values characterizing the population, whereas with a sample, you only get estimates.
User Contributed Comments 5
| User | Comment |
|---|---|
| chamad | Can some one explain "some tests are destructive in nature" with an example? Thanks |
| Criticull | I'll take a shot at explaining this chamad: Suppose one wants to study the effect of smoking on rhesus monkeys. You don't want to give smokes to every rhesus monkey cuz it's destructive. On another level, if every test subject is in the treatment group, then you won't have a control group against which to compare results (I'm thinking about scientific trials here). |
| Joel1980 | Comic genius. |
| Benn09 | Easier example I think, If you want to test average levels of mercury in Wild Tuna, you can't just go and catch every tuna in the world and run tests on it. And if you did the tuna population would most likely be harmed in the process |
| ascruggs92 | Criticull, I will never forget this concept now because of your explanation. Benn09, while not as entertaining, that is a fantastic example as well |