Research Subjects

Question 1

Population and Samples

Answer 1

• population: the larger group to which the results are generalized
• accessible population: potential participants within target population that is accessible to the investigators
• sample: subgroup of accessible population; serves as a reference group for drawing conclusions about the population
• target population: overall group to which the research to intends to generalize the study

Question 2

Subjects

Answer 2

• subset of accessible population
• group of subjects is called a sample
• primary data collected from subjects in real-time
• secondary data is collected during routine business or a prior research activity
• not restricted to humans

Question 3

Sample

Answer 3

• assume that the response of the sample members represents how the population would respond in similar circumstances
• heterogenous: presence of variations (physical, behavioral, psychological)
• a good sample reflects the variations of the population in same proportions they exist within the population

Question 4

Sampling Bias

Answer 4

• when individuals selected for a sample over-represent or under-represent certain population attributes related to the phenomenon under study
• conscious or unconscious
• methods of sampling are needed to make unbiased sample selections

Question 5

Samples and Validity of the Study

Answer 5

• a valid study is one that finds the truth
• 2 components
• external validity: the degree to which results of the study can generalize to individuals or settings outside the sample
• internal validity: refers to the adequacy of the study design, degree of control used when gathering the data

Question 6

How is a study’s design evaluated?

Answer 6

• identified and recruited
• selected
• managed

Question 7

Inclusion Criteria

Answer 7

-characteristics that individuals from target population must possess to be eligible for study

Question 8

Exclusion Criteria

Answer 8

-characteristics that will make individuals ineligible for the study

Question 9

Subject Selection

Answer 9

• recruitment will depend on definition of accessible population, cost, and requirements
• once potential subjects are identified a selection process must be identified
• common selection methods are categorized as probabilistic or nonprobabilistic

Question 10

Probabilistic Sampling Methods

Answer 10

• method for randomly selecting subjects for participation in a study
• every individual in population has equal chance of being chosen
• every individual has equal chance of having some of the characteristics that are present throughout the population
• minimize sampling error as well as bias
• sample is considered representative of population
• types: simple random sample, systematic sampling, stratified random sampling, cluster sampling

Question 11

Simple Random Sampling

Answer 11

• unbiased selection of sample
• draw names out of a hat
• table of random numbers

Question 12

Systematic Sampling

Answer 12

• used when accessible population is not or cannot be numbered in a list
• divide total number of individuals in accessible population by number to be selected
• 1000 in accessible population and 100 to be selected so divide 1000 by 100 get 10 and select every 10th individual in accessible population

Question 13

Stratified Sampling

Answer 13

• stratification is process of grouping individuals in population into groups based on characteristic
• sub-groups
• want to sample 100 so divide by male and female and sample 50 of each

Question 14

Cluster Sampling

Answer 14

• method to get random sampling when population is large
• divide population into large subgroups or clusters
• then randomly select sample from each of the subgroups

Question 15

Nonprobabilistic Sampling Methods

Answer 15

• used when it is difficult to obtain a true random sample
• subjects selected without randomization
• easier to implement, lower cost, greater sampling error and potential bias
• types: convenience sample, snowball sampling, purposive sampling

Question 16

Convenience Sampling

Answer 16

• subjects are chosen on basis of availability
• consecutive sampling: recruiting all patients who meet inclusion and exclusion criteria as they become available
• volunteers

Question 17

Purposive Sampling

Answer 17

-researcher hand picks subjects on basis of specific criteria to determine if they fit study

Question 18

Snowball Sampling

Answer 18

• subjects are successively recruited by referrals from other subjects
• identify a few subjects who meet inclusion/exclusion criteria
• these subjects asked to identify others who have required characteristics

Question 19

Subject Management Within the Study

Answer 19

• assignment processes allow subjects to be divided into groups of equal size and similar characteristics
• balanced groups are important to isolate the effect of the intervention rather than differences in group characteristics
• assignment may be random or non random

Question 20

Random Assignment Methods: Assignment by an Individual

Answer 20

• pull number from hat
• may lead to unequal number in groups
• groups without similar characteristics

Question 21

Random Assignment Methods: Block Assignment

Answer 21

• randomly assign subjects to one group until predetermined number is attained
• repeat for next group

Question 22

Random Assignment Methods: Systematic Assignement

Answer 22

• use list of subjects

- repeatedly count off group numbers

Question 23

Random Assignment Methods: Matched Assignment

Answer 23

• divide subjects into subgroups based on characteristics

- assign each member of subgroup to study group

Question 24

Nonrandom Assignment

Answer 24

• subjects are often members of preexisting gropus of interest OR
• investigators determine which subject goes into which group
• commonly seen in retrospective studies
• group assignment determined by presence or absence of characteristic of interest
• statistical analysis tests are used to adjust for variation due to nonrandom assignment

Question 25

Other SUbject Managment Issues

Answer 25

• standard protocols for processing subjects within a study: scripts for instructions provided to subjects, specific sequence in testing methods
• control of subjects normal daily routines: avoid changing daily routines
• masking or blinding subjects to group assignment: avoid changes in behavior due to expectation
• masking or blinding investigators to subject’s group assignment: avoid changes in investigator behavior toward group members

Question 26

Sample Size

Answer 26

• number of subjects needed to insure statistical differences may be detected
• power is the probability that a statistical test will identify a relationship or difference
• insufficient sample size may lead to false negative findings or a type II error
• accurate power calculations provide the minimum number of subjects needed