Final Exam

Question 1

What is a Simple Random Sample? (SRS)

Answer 1

SRS is when every sample size n has an equal chance of being selected

Question 2

What is a stratified random sample?

Answer 2

When you define non-overlapping groups of elements, within every group take a random sample

Question 3

What is a systematic random sample?

Answer 3

Randomely selecting one element from the 1st k elements in the frame and every kth element thereafter

Question 4

What is a cluster random sample?

Answer 4

Taken by observing all elements within random sample of clusters

Question 5

Benefits of stratified random sample over SRS?

Answer 5

Can be more cost efficient

Reduces Variance

Question 6

Why would cluster sample be preferred over SRS?

Answer 6

Used when a good frame isn’t available/costly

Can save time and money

Question 7

Why would systemeatic random sample be preferred over SRS?

Answer 7

Doesn’t require sampling frame

Can provide more info per unit cost

Question 8

What are some sampling methods that aren’t probability samples?

Answer 8

.

Question 9

What are benefits to R over SAS?

Answer 9

.

Question 10

What things can affect # of samples that will be chosen for each strata in stratified random sample?

Answer 10

Total # of elements in each stratum
Variability of observations in each stratum
Cost of sampling observations in each stratum

Question 11

When considering capture/recapture methods, what is the benefit of the Chapman estimator over the Lincoln-Peterson estimator?

Answer 11

Chapman estimator is nearly unbiased as opposed to the Lincoln-Peterson estimator

Question 12

Does the expected value of a random variable have to be an integer?

Answer 12

No

Question 13

Does the overall sample size needed for an entire sample have to be an integer?

Answer 13

Yes

Question 14

In what scenario would it be better to use a regression estimator instead of a ratio estimator?

Answer 14

When the line doesn’t appear to go through the origin

Question 15

What are three advantages of using a ratio estimator?

Answer 15

You don’t ruin the product (Opening Peanuts)
Can compensate for nonresponse
Don’t have to sample entire population (Peanut example: dont have to open every peanut)

Question 16

When estimating a regression line to fit a response variable based on an explanatory variable, what criteria determines the “best” model?

Answer 16

We minimize the sum of the squared residuals

Question 17

Neyman allocation is a method of selecting the sample sizes for each stratum under a certain condition. What is this condition

Answer 17

Because the costs are equal so c of i cancels out of numerator and denominator

Question 18

Snowball sampling can help the research

Answer 18

Collect data cos effectively
Overcome the problem of not having an accessible sampling frame
Access difficult or hidden populations

Question 19

Capture Recapture: t

Answer 19

t= the number of individuals marked from the original sample

Question 20

Capture Recapture: N

Answer 20

N= the size of the population (this is what we’re trying to estimate)

Question 21

Capture Recapture: s

Answer 21

s= the number of marked individuals in the second sample

Question 22

Capture Recapture: n

Answer 22

n= the size of the second sample