Final (Part 1) Flashcards
(40 cards)
features of sampling
group of people selected to represent target pop, must define the studied population,
fidelity and validity
f- reproducibility of results
v- reliablity of results
interval scale
order, 0 means someghing (temp)
what is sampling bias
selection of participants who differ from population, does not reflect target population
what is magnitude of effect
large (small sample size), small (large sample size)
how to determine size of quantitative sample
predetermined, study needs power of 80 percent (p value x4, subtract from one to find power)
what is data reduction
classify and categorize data, find double meanings, code data to explain
simple vs systematic random sampling
simple- each item has equal chance of being included
system.- items selected at reg intervals
ordinal scale
no relative order (stage of cancer)
what is selection criteria
inclusion and exclusion criteria, to narrow down who is in study (disease, age, etc)
nominal scale
one answer, no number (gender, yes/no, etc)
what is probabilistic sampling
every member of population has equal or non-zero chance of being selected
correlation coeeficiant
measure of association, -1 to 1, 0.7 to 0.9 is ideal
ratio scale
0 means nothing (mg of drug)
what is qualitative analysis
non-statistcal analysis, observe and intrepret patterns
measurments have…
concepts (dimensions) and indicators- heart health and bp
simple dicotonimus scale, simple scale, likert scale, differential semantic scale, visual analog classification
answer in two categories, 3 or more predefined answers, level of agreement, 7 point scale (-3 to +3), long line contrinuos
what is sampling error
diff between sample and what would have been obtained from general population, random variations that affect representativness/ generalization
a population is initially…
heterogeneous, different elements make it up
a sample is a…
supgroup of the accesbile population, must represent target population
Type 1 and type 2 error
1- reject null when null is true
2- do not reject null when null is false
types of non-prob sampling
convience, snowball (social networks), quota (divided based on age, gender), reasoned choice (selected based on knowledge), theoretical (based on contribution to theory)
4 steps of qualitative data analysis
data cleaning, data reduction, data analysis, data presentation
cluster sampling
used for larger studies, choose groups (neighborhood, school, etc), individuals studied
