9. Sampling Methods and External Validity

Question 1

Population & sample

Answer 1

the entire set of persons, objects, or
events that the researcher intends to study.
▪ Impossible or extremely costly to study complete
populations
▪ therefore, a sample is drawn from the target population
▪ Generalize from sample to entire population

Question 2

external validity is closely related to

Answer 2

inductive reasoning

Question 3

Inference:

Answer 3

the process of generalizing from sample data to the population

Question 5

Representative Samples

Answer 5

representative sample → can confidently generalize to rest of
population
* If sample biased (not representative), generalize → less validly
and may lead to incorrect conclusions

Question 6

Selection of appropriate sampling method depends on the __
and ______of the researchers

Answer 6

aims resources

Question 7

Incidental Sampling

Answer 7

• Cheapest and easiest sampling method for conducting clinical studies
• selection of most accessible and available members of the target population

Question 8

Quota Sampling

Answer 8

Sometimes it is known in advance there are important subgroups within the population
-need to know which population groups are important to a question,
and the exact proportion of different groups in the population

Question 9

Random Sampling

Answer 9

• One of the best but probably more expensive methods
• all members of population have an equal chance of
  selection
• More likely to be representative than incidental sampling

Question 10

procedure for drawing a random sample

Answer 10

1. constructing a list of all members of the population
2. using a method like dice, coins, a hat, or random number
   tables to select randomly from the list
   Computers can do efficiently, even with large populations
   * if need to select replacements for some cases (e.g. refusal to
   participate, dropping out, failure to satisfy the sample inclusion
   criteria) may introduce bias

Question 11

advantages of random sampling

Answer 11

• since sample and population sizes
  are known, we can estimate how
  representative the sample is.
• more representative than non-
  random samples
• Smaller sample size necessary

Question 12

disadvantages of random sampling

Answer 12

• need to list every member of the
  population
• usually easier and less expensive to
  use conveniently available groups

Question 13

Stratified Random Sampling

Answer 13

Like quota sampling but with random sampling for each
subgroup
* These groups collectively are called the strata

Question 14

stratified random sampling advantages

Answer 14

-important groups are proportionately
represented
-know exact representativeness of the
sample – important for statistics

Question 15

stratified random sampling disadvantages

Answer 15

-A list of all members of the population,
their characteristic(s), and the
proportions of the important groups
* Cost
* gain in sample accuracy small
relative to simple random sampling

Question 16

area sampling

Answer 16

• one sample on basis of location of cases
• effective, inexpensive method for social surveys
• area samples are not considered truly random

Question 17

systemic sampling

Answer 17

Working through a list of the population and choosing every n-th
(e.g. 10th or 20th) case for inclusion in the sample.
- The assumption is that the list is arranged randomly.
- If the list is organized in a patterned way that coincides with the
sampling interval, there is a risk of bias.
▪ easy, convenient

Question 18

Sample size

Answer 18

Sample size chosen is not random
* Optimal sample size is one adequate to make correct
generalization from the sample to the target population
* Balance number with costs of data collection (or danger/pain to
patients or lab animals)
* There are some principles to guide researchers

Question 19

sampling error

Answer 19

Discrepancy between the true population parameter
(e.g. average age) and the sample statistic (e.g.
average age of sample)
If we do not know the actual population parameters, we can
only estimate the probable sampling error.

cost/benefit at larger sample sizes

Question 20

sampling error equation

Answer 20

= 1/sq root of n
greater n, smaller error

Question 21

External validity

Answer 21

the extent to which the results of an
investigation can be accurately generalized to other samples or situations
Types: population and ecological

Question 22

population validity

Answer 22

Generalizing the findings from the sample to the population
from which it was drawn.
-The accessible population may not be the same as the target population

Question 23

Ecological Validity

Answer 23

The extent to which the findings of a research study can be generalized to real-world settings or situations
need caution to generalize to other situations