sampling+large data set

Question 1

5 types of sampling

Answer 1

simple random, systematic, stratified, opportunity, quota

Question 2

describe simple random

Answer 2

assign number to every unit. use random number generator to pick units, ignore repeats

Question 3

pros and cons of simple random

Answer 3

+bias free, easy and cheap
- not suitable when population too big, sampling frame needed

Question 4

describe systematic

Answer 4

find interval k (pop/sample), take every kth unit

Question 5

pros and cons of systematic

Answer 5

+simple, quick, suitable for big pop
-sampling frame needed, can introduce bias if frame not random

Question 6

describe stratified

Answer 6

divide population into strata, simple random in each strata

Question 7

pros and cons of stratified

Answer 7

+reflects pop structure, proportional representation of groups
-must be clearly stratified, (same an simple random)

Question 8

describe quota

Answer 8

pop divided into characteristics, quota of items in each group, recruit samples until quota is hit

Question 9

pros and cons of quota

Answer 9

+representative samples, no sampling frame, easy comparison between groups
-introduces bias, non responses not recorded, needs to be divided

Question 10

descirbe opportunity

Answer 10

sample taken from people available at time of study who fit criteria

Question 11

pros and cons of opportunity

Answer 11

+easy, cheap
-unlikely to be representative, dependent on individual researcher

Question 12

pros and cons of census

Answer 12

+accurate, representative
-time consuming, expensive

Question 13

pros and cons of sampling

Answer 13

+quicker
-may not be representative

Question 14

cities in large data set

Answer 14

heathrow, leeming, leuchars, camborne, hurn, jacksonville, beijing, perth

Question 15

lds dates

Answer 15

mar-oct 1987 and 2015

Question 16

cleaning data

Answer 16

replacing trace with 0 or 0.025. if n/as, remove the entry