participant recruitment and sampling
steps in sampling
1. define the population of interest 2. consider the unit of analysis 3. identify a performance site 4. develop a sampling approach 5. implement the sampling procedures
4 components of power analysis
1. sample size 2. statistical power (the likelihood the finding of statistical significance actually exists-- the mini acceptable level is .8 3. effect size (the magnitude of statistical significance) 4. level of significance (typically set at .05); (when three of the components are known, the fourth one can be calculated)
Systematic sampling (probability sampling)
Considered equivalent to random sampling as long as there is no reoccurring pattern of order in the listing-- the number of subjects for the sample is known and divided into the number of the population
Purposive sampling (non-probability sampling)
Deliberate selection of individuals by the researcher based on certain predetermined criteria (usually stated as inclusion and exclusion criteria)
Convenience sampling (non-probability sampling)
Most problematic, yet widely used. The use of volunteers or easily available subjects such as a group of students in a program or clientele in a clinic
everyday philosophers,
Those who think a great deal about the setting and can give insight to what is going on, but they are not as forthcoming with their ideas
Quota sampling (non-probability sampling)
Used when different proportions of subject types are needed so that there is appropriate representation in the sample that may not be attainable with purposive or convenience sampling
systematic errors
a flaw and serious problem- sources of this error include using volunteers and using groups that are available and convenient
secondary selection
a strategy a qualitative researcher may use when he or she does not know who the best participants are
Simple random sampling (probability sampling)
all individuals in a defined target population have an equal and independent chance of being selected for a sample
Stratified random sampling (probability sampling)
all individuals in identified subgroups of the population have an equal and independent chance of being selected for a sample
sampling errors
differences bw the study sample and the larger population it represents; usually calculated by estimation
adequacy
extent to which a qualitative study obtains enough data to provide a rich description of the phenomena of interest
external validity
extent to which findings of a study can be generalized to a larger population; major threats are people, places and time
probability sampling methods
include simple random sampling, stratified random sampling, systematic sampling, cluster sampling
non-probability sampling
includes convenience sampling, purposive sampling, snowball/network sampling, quota sampling
cluster sampling (probability sampling)
individuals are not randomly selected; groups or programs are selected, and every member of the group or program is invited to participate in the study
key informants
individuals who are in positions or have information that will be more beneficial compared with others in the organization
population
larger group of people that have the characteristics for participating in research study
Snowball/ network sampling (non-probability sampling)
method in which initially identified subjects provide names of others who may meet the study criteria
sampling in quantitative research
must define population from start; main purpose is to accurately draw conclusions about the population; sample size is established after the research design is determined
sampling in qualitative research
two principles of appropriateness and adequacy; randomization and number of participants are not of interest
selection of subjects
revolves around the nature of research question, the research approach, and practical considerations such as access, budget, time, space, etc.
theory-based selection
seeking individuals exemplifying a particular theoretical construct
maximum variation
seeking individuals with extremely different experiences of the phenomena; exploring the broadest range
homogenous selection
seeking informants with the same experience; exploring depth of phenomena
sample
smaller group of people from the population who participate in the study; represent the larger population
Types of individuals useful to qualitative research
social gadflied, constant observers, everyday philosophers, marginal people
general rule of thumb for most OT research is determining sample size
striving for the largest sample possibe; problems with sma;; sample size; not worth undertaking due to possible Type 2 Error; no standards or set rules for the right number of informants; the quality of data drives the sampling process-- when no new information is obtained, the right number has been achieved
exclusion criteria
the characteristics that a person cannot have in order to participate in a given research study; typically factors that could potentially confound the results of study or prevent individual from participating
inclusion criteria
the characteristics that a person must have in order to participate in a given research study
appropriateness
the extent to which a qualitative study identifies participants who can best inform the phenomena under study
statistical power
the liklihood the finding of statistical significance actually exists; min acceptable level is .80
effect size
the magnitude of statistical significance
target population
the population to which a quantitative researcher wants to generalize the study findings; determined by inclusion and exclusion criteria
social gadflied
the well-liked and lively individuals who mix with and talk to everyone in the group
subject pool
those who are eligible to participate in study; must possess inclusion criteria and not possess the exclusion criteria and are available for selection
constant observers
those who are the longer, well-established members of the group who will freely speak of the details of past events
marginal people
those who do not feel like they really belong to the group of feel ambivalent about the group. Because they do not have strong loyalty they will often talk to outsiders and be able to give valuable insights about the group