Week 6.2-7: Sampling and Quantitative Data Collection Flashcards
Population
P in Pico
The entire gorup of interest based on eligibility criteria
Sampling
selection of a portion of the population (a SAMPLE) to represent the entire population
Eligibility Criteria
The characteristics that define the population:
Inclusion and Exclusion Criteria
What is the most real proxy you can get of something in your study so its highly important to be careful when making it?
The Sample is the most real proxy
Sampling Bias
Overrepresenting or Underrepresenting population segment in terms of key characteristics
It is almost always the fault of the researcher and who they have access to (Ex: Picking a convenience sample of college sutdents)
Strata
Subpopulations of a population
ex: Male/Female, income, education, comorbidities
Target Population
the entire population of interestwe want to generalize
Accessible Population
the portion of the target population that is accessible to the researcher, from which a sample is drawn
The Goal of Studies from Sampling from the Accessible Population
generalize findings from accessible population to target population
How does sampling narrow within the population
Population (All of Life) –> Target Population –> Accessible Population –> Sample
What is the biggest and msot consistent threat to internal validity
selection threat
Sampling Frame
Through waht resource can you access the target population/accessible population?
Ex: No registrar list of all populations unlike BU - thats why they often use convenience sampling
Representative Sample
a sample whose key characteristics closely approximate those of the population
a sampling goal in quantitative research
How is a representative sample more easily achieved
probability sampling
homogeneous populations
larger samples achievement through power analysis
The goal of a sample is…
to be as reperesentative as possible to the target population
The onuse of resposibility for best sampling is on…
the researcher
Why is it very important to get a good sample before doing stats?
because the state programs assumed you did best methods and cannot tell you if a sample is not representative
Nonprobability Sampling
Does not involve selection of elements at random; it is rarely representative of the population
not great!
convenience sampling, but weakest form of sampling since it under and over represents form a function and behavior of the researcher
Probability Sampling
Involves random selection of elements: each element has an equal, independent chance of being selected
Allows researchers to estimate the magnitude of sampling error (difference between population values and sample values)
What are the elements mentioned in probability sampling
elements = whatever you are studying (in this case people)
What are some types of nonprobability sampling
convenience sampling
quota sampling
consecutive sampling
purposive sampling
Convenience Sampling
Nonprobability sampling
selecting the most conveniently available people as participants
cheap and easy but weakest form of sampling
Quota Sampling
nonprobability sampling
Identifying population strata and figuring out how many people are needed from each stratum
ex: enroll people until an amount is met - like rolling admission
Consequtive Sampling
nonprobability sampling
recruiting ALL people from an accessible population over a specific time interval
Purposive Sampling
nonprobability sampling
handpicking sample members
MOSTLY USED IN QUALITATIVE RESEARCH
What type of research is purposive sampling used in
qualitative research
How may a convenience sample and quota sample differ when recruiting a total population of 20% males and 80% females
the convenience sample will pick whatever they can get so even 5% male and 95% female
quota would keep going until 20% of men chosen and 80% are female
Types of Probability Sampling
Simple Random Sampling
Stratified Random Sampling
Systematic Smapling
Simple random Sampling
probability sampling
researchers established a SAMPLING FRAME - a list of population elements
then choses elements (people) randomly from the saampling frame
Stratified Random Sampling
Probability sampling
the population is first divided into two or more strate from the sampling frame - from which elements are randomly selected
so the elements break into groups of strata and then they select random from those groups
Systematic Sampling
Probability sampling
involves the selection of every ‘k’th case from a list, such as every 10th person on a patient list
So when they have the sampling frame / list they chosoe every 10th randomly
In order to do probability sampling what is needed first
SAMPLING FRAME
ex: Registrat, tax rolls, voting - but they cannot be biased like the voting or tax rolls too!
Sample Size
the number of study participants in the final sample
___ ___ adequacy is a key determinant of sample quality in quantitative research
sample size
Sample size needs can and should be estimated through __ __
power analysis
As the sample size gets smaller…
the risk of “getting it wrong” (Statistical conclusion validity)
too small samples are the number one reason to not get a statistically significant result - type 2 error - results not picked up but are there
The ideal sample size is…
large enought to detect differences between two groups
Sampling Strategy
- Define study pop
- Select inclusion criteria
- Develop exclusion criteria
- Design plan to recruit subjects
- Determine how many subjects are needed
- Apply the sampling plan/methodology
- Use strategies to retain subjects
Inclusion Criteria
elements must possess characteristics in order to be included in the sample
comes before the exclusion criteria
Exclusion Criteria
elements possess characteristics not desired in the study so these elements are not included in the sample
used after the inclusion criteria - so exclude from the crowd
Representativeness
How much the sample shares study characteristics with the general population of interest
How close they resemble each other
Generalizability / EXTERNAL VALIDITY
conclusions based on study findings for the “accessible” group are GENERALIZED to the population
Critically Appraising Sampling Plans - things to consider
type of sampling approach used (ex: convenience, consequitive, random)
the poopulation and eligibility criteria for sample selection
the sample size, with a rationale
a description of the sample’s main characteristics (ex: age, gender, clinical status, and so on)
*as readers we want to ID these elements
What should we assume if we cannot find the sampling method used by authors and also do not see a sampling frame?
90% of the time it means its a convenience sample they hope you did not notice
they would brag about good moethods like RCT and probability sampling methods
The author should list sample size analyzed NOT…
size began with before people dropped out
