WEEK 8 - SAMPLING Flashcards
What does participant mean?
someone chosen to take part in a study
What is the theoretical population (or target population)
The larger group that the researcher wants to generalise their findings to
What is the study population (or accessible population)
Is the population that the researcher has access to draw participants from
What is a sampling frame
is a detailed list of all participants, which can include people, locations or events in a population.
What is a sample?
is the group of participants who have been chosen to be a part of the current study
What is sampling?
the process of selecting participants, so that researchers can attempt to generalise their results back to a theoretical population.
What are the types of sampling?
- Probability sampling
- Non-probability sampling
What type of sampling in quantitative
- In quantitative research (aim is to recruit a sample that is representative of the theoretical population to permit generalisation) = probability samples
What type of sampling in qualitiative?
- In qualitivate research (aim to tare individual cases that are data rich) = non-probability samples
Probability Sampling approaches
- simple random
- systematic random
- strafed random
- cluster
- multi-stage
Non-Probability Sampling approaches
- convenience
- snowball
-quota - purposive
What is Sampling error?
- Measurement error refers to errors that take place with the data, whereas sampling error occurs when the group of participants chosen (the sample) is inadequate or not random enough
What are the two types of sampling errors?
- random errors
- systematic errors
- both can introduce bias
What are random errors?
- Random errors are common
- Occur randomly in a sample as a result of under‐ or over representation of certain groups.
-Likelihood of random errors can usually be reduced by increasing the sample size
What are systematic errors?
- Systematic errors usually occur as a result of inconsistencies or errors in the sampling frame
Researchers should take care when designing the sampling frame
Power of the Study
- Most studies will use 80% power, which means they will have an 80% probability of detecting a real difference if it exists. This means there will be a 20% probability of not detecting a real difference
- If a power calculation is not conducted, confidence in the findings is reduced because it is unknown whether the sample was too small to detect differences that exist or may detect differences that, in fact, do not exist in the population
What is Statically signitifcanr effect
The significance of research findings relates to the probability of detecting a statistically significant effect or difference between study groups in the sample when it does not exist in the target population
Probability — Type I Error
- If an event is impossible, then its probability is 0. If an event is certain, its probability is 1.
- researchers will say that the intervention works or that a real difference between groups exists when it does not
the significance (α) level
- If the α level is set at 5%, a significant result means that we can be 95% confident that a real difference exists in the population and there is a 5% probability that the finding was due to chance alone
- clinical trials of medicines or interventions, the α level can be set more stringently at 1% (a 1% probability that the finding was due to chance alone)
type II (β) error
- the probability of finding that there is no effect or difference between groups in the sample when, in the population, there is a true effect or difference
What is Statistical ‘power’
- is used to describe the probability of detecting a real difference that exists in the population
- required sample size can be calculated using a power analysis
Power Analysis
- requires some information about the expected range of scores (variance or standard deviation) that would be expected within the population being studied
Sampling in Qualitative Research
- The literature sometimes suggests a range or a minimum and maximum number of participants or cases for qualitative research to be valid
- Ranges depend considerably on the type of topic that is being studied and other factors (e.g. the nature of the research questions, the homogeneity or heterogeneity of the population)
What is Heterogeneity
looks at how diverse the results are for studies in a review
What is a confidence interval?
