chapter 5/week 5 Flashcards
selecting research participants
sample
a subset of a population
sampling
the process by which a researcher selects a sample of participants for a study from the population of interest
- Many different ways
Representative sample
Representative sample
one which we can draw accurate, unbiased estimates of the characteristics of the large population
Sampling error
the extent to which characteristics of individuals selected for the sample differ from those of the population
Because of sampling error, results obtained from the sample differ from what would have been obtained using the whole population
Sampling error is important only if the sample is a probability sample
Margin of Error (error of estimation)
indicates the degree to which the data obtained from the sample are expected to deviate from the population
Factors that effect margin of error (error of estimation)
Sample size
Population size
Variance of the data
A sample will be more similar to the population (i.e., smaller measurement error) when:
The sample size is large
The population size is smaller
The variance in the data is smaller
Probability sample
a sample for which the researcher knows the probability that any individual in the population is included in the sample
Probability sampling
a sample that is selected in such a way that the likelihood that any particular individual in the population will be selected for the sample can be specified
(probability sample) obtained in four basic methods:
simple random sampling
Systematic sampling
Stratified random sampling
Cluster sampling
All cases in the population have an equal probability of being chosen for the sample
sampling decision tree
do you have population level data? –> yes (probability sampling) or no (no probability sampling)
yes –> simple random sampling, systematic sampling, stratified random sampling, cluster sampling
no –> convenience sampling, quota sampling, purposive
If you know the population parameters the you conduct probability sampling
simple random sampling
every possible member of the population has the same chance of being selected from the population
- When a sample is chosen in such a way that every possible sample of the desired size has the same chance of being selected from the population
Requires a sampling frame
Table of random numbers
sampling frame
a list of the population from which the sample is to be drawn; difficult for big samples
Table of random numbers
contains long rows of numbers that have been generated in a random order
- Typically use computers today
systematic random sampling
every kth person is selected
Not all individuals in the population have equal chance
Stratified Random Sampling
the population is divided into strata, then participants randomly selected from each stratum
Stratum
The strata should be exclusive (can’t have a person qualified for more than one group)
This method ensures that researchers have an adequate number of participants from each stratum
proportionate sampling method
stratum
a subset of the population that shares a particular characteristic
Proportionate sampling method
cases are sampled from each stratum in proportion to their prevalence in the population
cluster sampling
sample groupings or clusters of participants
Clusters are based on naturally occurring groups that are usually in close proximity
Clusters of population, then randomly select clusters (not getting data from all clusters)
Multistage Cluster Sampling
divide population into large clusters and randomly sample clusters; then randomly sample smaller clusters within those large clusters
- Then if needed, sample again from those clusters and continue until the appropriate number of participants is chosen
Difficulties in Probability Sampling
nonresponse problem
misgeneralization
nonresponse problem
failure to obtain responses from individuals that researchers select for their sample
Factors contributing:
- Personality characteristics
- Lack of time
- Literacy or language proficiency
- Sensitive topics
- Suspicion about the researcher/researcher/topic
misgeneralization
generalizing results from a study to a population that differs in important ways from the one from which the sample was drawn
Ex: a researcher studying parental attitudes may study a random sample of parents who have children in the public school system; then uses his data to make generalizations about all parents – misgeneralization because parents whose children attend private or homeschooled are not included
Nonprobability Sampling
researchers do not know the probability that a particular case will be chosen for the sample
- The error of estimation cannot be calculated
- Most research involves this type of sampling
- It is a valid method because the goal is to test hypotheses regarding how particular variables relate to behavior - not to describe how a particular population behaves
types of non probability sampling
convenience
quota
purposive
Convenience Sampling
use whatever participants are readily available
Researchers often use convenience samples of college students. Because a college sample may differ from the population at large, some results may not generalize to all people
Stopping people on the street, study patients at a local hospital, test students at a nearby school, our college
Easy
Quota Sampling
convenience sample in which the researcher takes steps to ensure that certain kinds of participants are obtained in particular proportions
Ex - you may wish to obtain an equal number of children with ADHD and children without ADHD in your study
ex - Wanting an equal proportion of male and female participants; sampling 60 women and 60 men
Purposive Sampling (Deliberate Sampling)
researchers use their judgements to decide which participants to include in the sample, trying to choose respondents who are typical of the population
Ex - pick a school district that represent the nation and work on that district alone to make predictions about the population
- Similarly, pick a county that vote along the lines of national voting behavior in past elections and sample from there
Based on previous elections, researchers have identified particular areas of the country that usually vote like the country as a whole; votes from these areas are interviewed and used to predict the outcome of an upcoming election
economic sample
a sample that provides a reasonable degree of accuracy at a reasonable cost in terms of money, time, and effort
Power
the ability of your design to find any effects of the variables being studied
High power, better chance to detect the effect. Low power failure to detect an effect that exist in the data
power’s associations
- Association between power and sample size
- The larger the sample, the larger the power - Association between power and effect size
- The larger the effect size, the larger the power
To detect smaller effect sizes you will need more power (i.e., larger sample size)
