Week 8: Research lecture notes
- final test will be on things that wasn’t covered in the midterm, mainly on new
materials (qualitative design)
- sampling: process where you want to get representative units. In research, population and
sample refers to data, inanimate objects included.
- population: entire set of individuals defined by the sampling criteria established for the
- de-limitation a.k.a exclusion criteria (your picking a sampling that’s close the population).
Only include people with certain age
- you wanna make generalization about all diabetes PT for example.
- ex of exclusion: health status and diagnosis
- Know the terms
- Elements are sampling units.
- Beware of sampling bias.
- 2 types of sampling: probability and nonprobablity (nonrandom)
- representativeness: you wanna say that the results of study which included a small group
of people that leads to conclusion for the larger group. You wanna generate things
- probability: In research terms, you wanna ask yourself whats the likelihood of someone
being selected. This is the difference between randomization and random selection.
Probability sampling is random sampling
- 4 types of probability sampling: simple random, statrified random, multistage (cluster)
Simple random: everyone has a random chance of being selected. For ex. CNO
giving you a list of RNs available for research. There will be table of random #
so like every 5 RN on the list will be on the research
Stratified: dfining subgroups and clusters. This is more specific, for ex.
Research for post bacc RNs. You have a % your sampling from
Cluster: similar to the stratified sample. You start large and you go smaller.
Systematic: you select every X number from ordered list. For ex. If you need a
sample size of 10, your class as 40, you would take every 4 student
- nonprobablity: not everyone can particpate in the study.
- there are 5 types: convenice samplng, quota sampling, purposive samplng,
network/snowball sampling Convienece sampling: can be used in quantative and quanlatative. You just happen to
be at the right time at the right place. There can be high chances of bias in this study.
Quota sampling: can be used in quantative and quanlatative. Similar to strafied
sampling but not random.
Purposive: you hand pick people (so people with rare disease or experts in stuff)
ONLY used in quanlatative
Snowballing: you start small and you ask participants for others who meet the
criteria. If its hard to find people for the study, you get other people to use word of
mouth to get the word around. For ex. Giving free parking
Matched sample: you compare two groups that have the same characteristics. For ex.
Chineese mothers and breast feeding. You pick an extraneous variable. For ex,
intervention study where you want to help students with speed reading stategies so
you match people based on
- purposive sampling: eligibility criteria. Can be done for validation of scale. If you wanna
know about certain traits, purposive sampling helps. It can also be done with a theorical
- phenomenological study is another example of purposive sampling. So it has a purpose
for a particular phemnemoia
- don’t worry about risk of bias.
- you can have a convienet sample and randomize into your groups (difference between
randomization and random study?)
- researcher decides the sample size for statisitical conclusion validity. VALIDTY DOESN’T
APPLY TO QUALATATIVE, instead it says trustworthyness
- don’t worry about effect size or measurement issues
- sample size is guided by quality o