| Terms |
Definitions |
|
SYSTEMATIC SAMPLING
|
Divides total number of elements by the number of elements to be selectedThis is the sampling intervalStarting point is determined randomly
|
|
Accessible or Experimental population:
|
The portion of the population that has a chance to be selectedThe sample is chosen from the accessible population
|
|
CONVIENENCE SAMPLING
|
Chosen on basis of availabilityOften used because it’s so easyCan cause significant biasParticipants often self-select
|
|
How does Disproprtional sampling affect Groups?
|
Group scores will be appropriately weightedGroups means won’t be affected but the relative contribution of these scores to the overall data interpretation will be controlled
|
|
PURPOSIVE SAMPLING
|
Participants hand picked based on specific criteriaGo out and hunt for them and beg for them to participate
|
|
Simple random sampling
|
Unbiased because each selection is independentThe sample is selected from the accessible populationUse random numbers table or computer for selection process
|
|
QUOTA SAMPLING
|
Stratification is usedControls for potential confounding effect of known characteristicsGuided process to assure adequate # of participants obtained from each stratum
|
|
Sampling bias:
|
Over or under representation of certain population attributes
|
|
Which type of Probability sampling Can provide a morre representative sample?
|
Proportional Stratified Sampling
|
|
PROPORTIONAL STRATIFIED SAMPLING
|
Selection based on the percentage of representation in accessible populationStratification increases precision of estimates only when the stratification variable is closely related to the variables of experimental interest
|
|
SNOWBALL SAMPLING
|
Stage 1 – a few participants selected and testedStage 2 – these participants are asked to identify othersContinues until adequate sample is obtained
|
|
CLUSTER SAMPLING
|
Successive random sampling of a series of units in the population (3 stages)Convenient & efficient with a large populationIncreases potential sampling error because of the 3 stages
|
|
NONPROBABILITY SAMPLING- name each of the types
|
Convenience samplingConsecutive samplingQuota samplingPurposive samplingSnowball sampling
|
|
Population:
|
The larger group to which results are generalizedDefined aggregate of persons, objects, or eventsCharacterized and meet a specified set of cri
|
|
Target or Reference population:
|
The overall group to which generalizations will be made
|
|
Disproportional sampling
|
Random samples of adequate size from each category are selectedProportional weights are calculatedProbability of selection is determinedInverse of this probability is takenEach score is multiplied by this number
|
|
DEFINING THE POPULATION-Inclusion criteria:
|
Primary traits of the target & accessible populationThese traits qualify someone as a participantThe more restricted the sample the harder it will be to generalize
|
|
What does probability sampling have to do with validity?
|
Gives the greatest possible confidence in the sample’s validityThis is because it has the greatest likelihood of accurately reflecting the population’s characteristics
|
|
Nonprobability sampling:
|
Sample chosen some way other than randomlyAll elements do not have equal chance of selectionCannot assume sample represents the target population
|
|
DEFINING THE POPULATION-Exclusion criteria:
|
actors that preclude someone from being a participantFactors that are generally confounding factorsFactors may interfere with interpretation of results
|
|
Which type of Probability Sampliing has an area probablilty sampling (geographical), and random-digital dialing (multistage samplin of area codes and telephone exchanges)
|
CLUSTER SAMPLING
|
|
RECRUITMENT
|
Probably the most difficult and time consuming part of doing researchMust be well thought-out in relation to inclusion/exclusion criteriaRecruitment methods will depend on research questionMay have to get IRB approval from various facilities in orde
|
|
What does Recruitment have to do with Power?
|
The size of the sample is important for POWER of the studyPower is the ability to find significant differences when they existThe larger the sample the greater the power to detect differences
|
|
PROBABILITY SAMPLING, name the different types
|
Simple random samplingSystematic samplingStratified random samplingProportional stratified samplingDisproportional samplingCluster sampling
|
|
Sample:
|
A subgroup of the populationServes as a reference groupUsed for estimating characteristics or drawing conclusions about the population
|
|
CONSECUTIVE SAMPLING
|
Process of recruiting participants who meet criteria as they become availableAlso very easy but can create significant bias
|
|
STRATIFIED RANDOM SAMPLING
|
The researcher first identifies the population characteristicsThen partitions members of population into homogeneous nonoverlapping subsets of strata based on these characteristics
|
|
Probability sampling:
|
Samples created through random samplingSample averages = statisticsPopulation averages = parametersDifference between the 2 groups is sampling error
|
