Hs 570 Sampling And Additional Info

Front 1

Population

Back 1

the entire group -- you define

Front 2

Sample

Back 2

a subgroup of the population that is representative of the population

Front 3

Probability Samples

Back 3

those in which the probability of selection of each subject is known

Front 4

Nonprobability samples

Back 4

the probability of being selected is unknown

Front 5

Sampling error

Back 5

random fluctuation in scores
slight differences between your sample & the pop.

Front 6

Sampling Bias

Back 6

distortion caused by the way you select the sample. Ex. Phone book..because it is in alphabetical order

Front 7

Random Sampling

Back 7

each subject in the population has an equal chance of being selected for the sample
bias-free -- leaves selection entirely to chance

Front 8

Simple random selection

Back 8

table of random numbers
fishbowl technique with or without replacement

Front 9

Stratified Random Sampling

Back 9

divide population into nonoverlapping subgroups called strata, based on some characteristic that is important to control in the study
select a set number of subjects from each strata using simple random sampling

Front 10

Systematic Sampling

Back 10

have to have a roster or list of subjects in the population
can only be used when units in the sampling frame are random
randomly select the first subject
select every kth one (researcher selects)
quick, efficient, saves time and energy
not entirely bias-free

Front 11

Cluster Sampling

Back 11

use when:
population to be studied is infinite
a list of the members of the population does not exist
there is a widely scattered geographic distribution
use previously formed groups, but select groups at random
after the group is selected, every member becomes a subject
not entirely bias-free

Front 12

Multi-stage Cluster Sampling

Back 12

randomly select group, then randomly select part of the group
used frequently in research
if you want to generalize, this is better to use than cluster sampling

Front 13

Non probability Sampling

Back 13

Probability that a subject will be chosen is not known
Claim for representativeness of the population cannot be made
sampling error is unknown
generalizability of findings is limited
can lead to faulty conclusions
less expensive, less complicated than prob.
Use for small studies or pilot studies

Front 14

Convenience Sampling

Back 14

selecting the closest and most convenient persons
Ex. -- classrooms of Personal Health Promotion Students

Front 15

Quota Sampling

Back 15

stratified nonprobability sampling
Researcher determines strata that are relevant to the investigation
Establishes a quota for each of the strata
Obtains subjects for each strata, no randomization

Front 16

Snowball Sampling

Back 16

multi stage technique
First stage -- a person with needed characteristics is identified and interviewed
Second stage -- the person identifies others who might fit into the sample
Third stage -- those persons are interviewed
etc.

Front 17

Sample size

Back 17

Important, but not as important as representativeness of the sample to the population
Look at previous studies, pilot studies
indication of how large samples were to produce significant results

Front 18

Sample Size
Co relational Study

Back 18

atleast 30 subjects

Front 19

Sample Size
Experimental research

Back 19

15 or more (have more control over subjects)

Front 20

Sample Size
Survey Research

Back 20

need at least 100 in each major subgroup, 20 - 25 in each minor subgroup
usually won’t get more than 50% back from mail survey