Final Flashcards
measurement
process of assigning numbers to objects, events or situations representing the amount of an attribute using a specified set of rules
nominal measurement
classify or categorize data
labels
lowest of 4 levels
ordinal measurement
order/rank values w/out equal intervals
continuum of numeric values w/ small numbers representing lower levels and large number higher lvls but adding them has no meaning
3rd lowest
ex: ordering
interval measurement
give values meaning w/ equal intervals may or may not contain a sero manmade scales of measurements 2nd highest level of evidence ex. scored response survey
ratio measurement
give values meaning with equal intervals w/ absolute zero point
highest form of measurment
numbers in order
ex: weight or height
measurement error
difference between the true measure and what is actually measured
systematic error
error that occurs in the same way with each measurement
random error
errors that occur by chance w/out pattern during measurement
correlation coefficient
measures reliability (consistency)
1.00 is perfect reliability wheras 0.00 is none
needs to be at least .7
test for 3 attributes
stability, equivalence, internal consistency
methods to test for reliablity
test-retest reliability
parallel or alternate form
validity of an instrument
the extent that an instrument measures a concept accurately
types of validity
content validity (face validity/content validity index) criterion-related validity (concurrent - degree of correlation btwn two variables or predictive - measure btwn a measure and a future measure) construct validity (how much an instrument actually measures the theoretical concept)
qualitative data methods
questionnaires interviews focus groups case studies observations
appraising qualitative data collection
assess sample/setting
data collection and report
report of biases
phenomenology
indepth interviews/diaries/artwork
track an experience/phenomenon
grounded theory
observations/open-ended questions w/ individuals or small groups
develops a theory
ethnography
participant/direct observations/interviews/diagrams/documents/photographs
historical
open-ended interviews/documents/photographs/artifacts
sampling error
difference btween a sample statistic and a population parameter
sampling bias
occurs when the sample is not representative of whole pop
4 steps to ensure representative sample
define the target population and their essential characteristics
ID a pop the researcher has access to for the study and see if it resembles the target pop
develop a method to approach them (sampling frame)
select subjects from the accessible population to include in sample
simple random sampling
probability sampling
each subject has the same chance to be selected
randomized
stratified random sampling
probability sampling
strata must be mutually exclusive so a subject can be assigned to only one stratum
random sampling to select subject from each stratum
cluster sampling
probability sampling
simple random sampling selects groups/cluster then select subjects w/in each cluster
systematic sampling
probability sampling
count each kth subject by first IDing the start location
convenience sampling
nonprobability
inclusion criteria ID then subjects invited to participate
quota sampling
nonprobability
strata mutually exclusive then convience sampling used to select subject from each stratum
purposive sampling
nonprobability
research selects sample of experts
commonly used in qualitative research
research chooses based on defined criteria
theoretical sampling
nonprobability
data collection and analysis occurs simultaneously
usually for grounded theory research
power analysis
statistical analysis to determine acceptable sample size (standard power of .8)