Week 4 Flashcards
What is standard deviation?
spread form the mean
What is positively skewed data?
mean is higher than median
What is negatively skewed data?
median is higher than mean
What is the null hypothesis?
use of the word ‘no’
What is the positive hypothesis?
what you expect the outcome to be
What is the theoretical population?
the larger group the researcher wants to generalise findings to
What is the study population?
the population the researcher has access to
What is a sampling frame?
list of all the participants that usually relates to the study population
What is probability sampling?
random selection of participants from a population to ensure that all members of a target population have a chance of being selected
What are the pros of probability sampling?
more representative sample with reduced sampling errors and bias
What is non-probability sampling?
participants are chosen in a process that does not give all the participants in a population the equal chance of being selected
What are the probability sampling techniques?
simple random sampling, systematic random sampling, stratified random sampling, cluster sampling, multistage sampling
What are the non-probability sampling techniques?
cluster sampling, snowball sampling, purposive sampling, quota sampling
Simple random sampling
*every participant has an equal chance of being selected
+easiest and most common, high generalisability
-not as efficient as stratified random sampling
Systematic random sampling
*systematically selected from a list (ie every nth participant)
+easy to use and implement
-systematic biases unless the ordering of participants on the list is random
Stratified random sampling
*population divided into groups (strata), then simple random or systematic random sampling is used
+adequate sampling size
-stratas not clearly defined, complicated, time consuming
Cluster random sampling
*population randomly divided into a cluster, then a chosen cluster is sampled
+cost effective
-less efficient for large sample sizes
Multistage random sampling
*carried out in various stages with a primary population then sub populations
+used when simple random/systematic random/stratified sampling would be too expensive or complex
Convenience sampling
*participants are chosen because they are convenient (ie close proximity)
+easy, cost effective, rich qualitative data
-does not produce representative samples, hard to replicate results
Snowball sampling
*begin identifying someone who meets criteria and ask them to find other people
+used for hard to reach participants, cost effective
-not used for generalisations, relies on participants to increase sample size, can be saturated
Purposive sampling
*look for cases that provide in-depth information about the issue being researched
+can provide the researcher with justifications to make generalisations
-researcher bias
Quota sampling
*participants chosen according to pre-specified quotas regarding demographics, attitudes, behaviours etc
+ensures an adequate number of subjects with appropriate characteristics
-not used for generalisations
What are the two types of sampling error?
random and systematic - both introduce bias
What is random error?
commonly occur in a sample of over- or under- represented groups
can be reduced by increasing sampling size
What is systematic error?
cannot be reduced by increasing sampling size, usually occur as a result of inconsistencies or errors in the sampling frame
Why is sample size calculated in quantitative research?
to prevent the waste of valuable resources or unethical studies
What is the probably of an event that is possible vs impossible?
Impossible = 0 Certain = 1
What is a type I (a) error?
probability of detecting a statistically significant effect or difference between study groups in a sample when it does not exist in the target population, false positive
If the p value is set at 0.05, what can we say about it?
we can be 95% confident that a real different exists in the population and there is a 5% probability that the findings were due to chance alone
What is a type II (b) error?
when 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, false negative
How can a sample size be calculated?
by using a power analysis
What is saturation?
when little or no new data is generated (adequate sample size)
Why do sample size for qualitative research tend to be smaller than for quantitative research?
lengthy data collection processes and analysis required for qualitative methods
