Sampling

Sample

the population for your study

define the target population as clearly as possible

inclusion criteria → characteristics that population must exhibit to be part of the sample

exclusion criteria → characteristics that would preclude population form being part of the sample

Sampling Bias

overrepresening or underrepresenting population segments in terms of key characteristics

Strata

subpopulations of a population

ex: male and female

Target Population

the entire population of interest

Accessible Population

the portion of the target population that is accessible to the researcher, from which a sample is drawn

Representaive Sample

a sample whose key characteristics closely approximate those of the population, sampling goal in quantitative research

more easily achieved by → probability sampling, homogenous populations, larger samples achieved through power analysis

Probability

prob of being included in the sample is known

can be more certain the sample is representative of the general pop

can calculate needed sample size

reduces sampling bias

must be an available list of all member of the pop for participants to be drawn from

Nonprobability

may or may not accurately represent the pop

convenient and economical

no way to ensure all members of a pop have a chance to be part of the sample

used when total pop is unknown or not available

Simple Random Sampling

a specified number or percentage of the pop is determined in advance

all members of a pop assigned a number

numbers are selected at random until the target number is reached

Stratified Random Sampling

pop is divided into groups based on characteristics

criteria for groups determined by the study

everyone is assigned number and equal amounts of participants are drawn from each group

ex: males, females, ethnic groups

Systematic Sampling

involves the selection of every kth case from a list, such as every 10th percent on a patient list

Convenience Sampling

every person who meets criteria is asked to participate

uses sample that is available

Quota Sampling

like convenience sampling

draws from an available population

list of criteria needed to divide samples into groups → age, gender, education

decide on number of participants in each group and then select that number of participants from each group

Consecutive Sampling

recruiting all people from an accessible pop over a specific time interval

Purposive Sampling

handpicking sample members

Sample Size

use the largest one possible for your study

decreases risk of error, increases generalizability, and external validity

power analysis → stat test to determine needed sample size for your study

sample must be available to you

homogeneity vs heterogeneity → heterogenous pop needs larger sample

if pop is small, use entire pop

Quantitative Data Sources

basic decision is use one of the following:

new data collected specifically for research purposes

existing data → records, historical data, existing data set

self reports → patient reported outcome

observations

biophysiological measures → biomarkers

Structured Self Reports

data are collected with a formal instrument

interview → questions asked orally either face to face or via telephone

survey → questions prespecified in written form to be self administered

interview questions → closed ended (yes or no) or open ended

Questionnaire Pros

less costly

reach a greater number of people

offer the possibility with anonymity which may be crucial in obtaining records about certain opinions

Interview Pros

higher response rates

appropriate for more diverse audiences

some people cannot fill out a questionnaire

opportunities to clarify questions or to determine comprehension

opportunity to collect supplementary data through observation

Scale

device that assigns a numeric score to people along a continuum

used to make fine quantitative discriminations among people with different attitudes, perceptions, traits

Likert Scales

summated rating scales

please rate how often you use nicotine products on a scale of 1-never, 2-sometimes

Summated Rating

total score is sum of item scores

Response Bias

biases reflecting the tendency of some people to respond to items in characteristic ways, independently of item content

Social Desirability Bias

misrepresent oneself by giving answers that are consistent with social views

Extreme Response Set

consistency select extreme alternates (strongly agree or strongly disagree), not representative of actual feelings

Acquiescence Response Set

agree with statements regardless of the content

opposite can occurs but happens less frequently

Observation

structures observation of pre-specified behaviors

involves use of formal instruments that dictate what to observe, how long, and how to record data

Time Sampling

sampling of time intervals for observation

ex: random sampling intervals over a given length

In Vivo Measurments

performed directly within or on living organisms

ex: blood pressure

In Vitro Measurements

performed outside the organisms body

preformed in a lab

ex: urinalysis

Testing Instruments

psychometric assessment → evaluates measurers measurement properties

reliability → extent to which scores are free from error

Validity

face → whether the instrument looks like its measuring target

content → extent to which instruments content adequately captures the construct

criterion → extent scores on a measure are a good reflection of gold standard

construct → degree to which evidence about a measures scores in relation to other variable supports the inference construct has been well represented