Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
110 Cards in this Set
- Front
- Back
empirical
|
denotes experience and gathering of data
based on observation |
|
scientific method
|
define problem
create hypothesis test with experiment analyze results draw conclusion hypothesis true or false report results |
|
qualitative data
|
non-numerical
descriptive |
|
quantitative data
|
numerical data
|
|
prospective
|
looking ahead
|
|
retrospective
|
looking back
|
|
variable
|
any entity that can take on different values for different people or the same person at different times
|
|
variables not always numerical
|
e.g. gender
|
|
independent variable
|
manipulated, changed is on y axis
|
|
dependent variable
|
what's affected is on x axis
|
|
types of research
|
descriptive (documentation)
relational (comparisons) causal |
|
relationships
|
positive (one up the other up)
negative (one up the other down) curvilinear (changes over range of both variables) |
|
deductive
|
more general to more specific - top- down
|
|
inductive
|
specific to general - bottom - up
|
|
research fallicies
|
ecological - assumption about an individual based on group data
exception - generalizations based on one observation |
|
validity
|
best available approximation of truth of a given proposition, inference, or conclusion
|
|
measures, samples, designs don't have validity
|
only propositions / conclusions have validity
|
|
4 types of validity
|
conclusion validity
internal validity construct validity external validity |
|
conclusion validity
|
there is a relationship b/w cause and effect
|
|
internal validity
|
assuming there is relationship, is it causal
|
|
construct validity
|
assuming causal relationship, is it generalizable to constructs
|
|
external validity
|
is it generalizable to other persons places, times
|
|
sampling
|
population - group you want to generalize about
sampling frame - list from which you draw population |
|
standard deviation
|
spread scores around the average (1,2,3 sd's = 65%, 95%, 99.8% of population
|
|
Standard error
|
spread of scores around the average of the averages in sampling distribution
|
|
N
|
# cases from sampling frame
# cases in sample |
|
Simple random sampling
|
every unit has equal chance of selection
|
|
systematic RS
|
choose every 10th, for example
|
|
Stratified RS
|
divide population into homogeneous subgroups and then take SRS in each group
|
|
Cluster RS
|
divide pop into clusters (geographic borders) and randomly sample clusters.
ex: 10 clusters, randomly select 6. measure all units in each of 6 |
|
Multistage RS
|
combinations of previous examples
|
|
non-prob. sampling
|
doesn't involve random sampling
|
|
convenience sampling
|
selection based on ease of inclusion
|
|
Modal sampling
|
based on most frequently occurring
|
|
expert
|
group w/ expertise
|
|
quota sampling
|
based on predetermined quota
|
|
heterogeneity sampling
|
interested in including all opinions and not concerned w/ proportionality
|
|
snowball sampling
|
asking members of study to ID other potential members
|
|
construct validity
|
measure of how well your program reflects your theory
Translational Validity: face content criterion -related validity predictive concurrent convergent discriminant |
|
translation validity
|
face validity - 'face value' of translation of construct
content validity - checking the operationalization against relevant content domain for construct |
|
face validity
|
face value of translation of construct
|
|
content validity
|
checking the operationalization against relevant content domain for construct
|
|
criterion-related validity
|
check your operationalization against some criterion
predictive - ability of operationalization to predict concurrent-operation - ability to distinguish b/w groups convergent - degree to which operation is similar to others discriminant - opposite of convergent |
|
predictive -
|
ability of operationalization to predict
|
|
concurrent operation
|
ability to distinguish b/w groups
|
|
convergent
|
degree to which operation is similar to others
|
|
discriminant
|
opposite of convergent
|
|
threats to construct validity
|
anything that causes incorrect conclusion about operation
inadequate preoperational explication of constructs mono-operation bias mono-method bias interaction of diff treatments |
|
inadequate preoperational explication of constructs
|
bad job defining construct
|
|
mono-operation bias
|
reliance on a single implementation of independent variable
|
|
mono-method bias
|
only use one measure of a construct
|
|
interaction of different treatments
|
participants receive 2 confounding treatments
|
|
interaction of testing and treatment
|
is testing making groups more sensitive to treatment
restricted generalizability across constructs confounding constructs and levels of constructs |
|
restricted generalizability across constructs
|
unintended consequences
|
|
confounding constructs and levels of constructs
|
label not a good description of what you implemented
|
|
social threats to construct validity
|
reactive behavior to treatment
|
|
Measurement
|
random error
systematic error 4 levels: nominal, ordinal, interval, ratio |
|
random error
|
no consistent effects across sample
|
|
systematic error
|
consistently either positive or negative
|
|
4 levels of measurement
|
nominal or categorical - naming quality
ordinal measurement - ranking interval- continuum of equally spaced intervals ratio - continuum with zero representing the bottom |
|
reliability
|
consistency or repeatibility of measure assuming characteristic doesn't change over time
|
|
true score theory
|
measurement is sum of true level and random error
|
|
4 classes of reliability estimates
|
inter-rater: extent to which you & friend agree on measurement
Test-retest: consistency of repeated measures parallel forms: Parellel-forms reliability is gauged by comparing two different tests that were created using the same content. This is accomplished by creating a large pool of test items that measure the same quality and then randomly dividing the items into two separate tests. The two tests should then be administered to the same subjects at the same time. Alternate forms: his form of reliability is used to judge the consistency of results across items on the same test. Essentially, you are comparing test items that measure the same construct to determine the tests internal consistency. |
|
inter-rater
|
extent to which you and friend agree on measurement
|
|
test-retest
|
consistency of repeated measures
|
|
parallel form
|
Parellel-forms reliability is gauged by comparing two different tests that were created using the same content. This is accomplished by creating a large pool of test items that measure the same quality and then randomly dividing the items into two separate tests. The two tests should then be administered to the same subjects at the same time.
|
|
alternate form
|
This form of reliability is used to judge the consistency of results across items on the same test. Essentially, you are comparing test items that measure the same construct to determine the tests internal consistency. When you see a question that seems very similar to another test question, it may indicate that the two questions are being used to gauge reliability.
|
|
internal consistency
|
how well items on instrument that reflect same construct yield similar results
|
|
qualitative measures
|
any measure not numerical
|
|
qualitative methods
|
participant observation, direct observation, unstructured interviewing, case studies
|
|
unobstructive measures
|
allow researcher to gather data w/o becoming involved
|
|
indirect measures
|
occur naturally in research setting
|
|
content analysis
|
systematic analysis of text
|
|
secondary analysis
|
reanalysis of quantitative data
|
|
Ethical Research
|
voluntary participation
informed consent risk of harm privacy right of service -non treatment group has right to treatment terminate participant if wanted |
|
survey question types
|
dichotomous - y/n
nominal - applying a number to category ordinal - order of preference interval level - likert 1-10 systematic differential |
|
systematic differential
|
sets of bipolar differential (don't agree, agree, strongly agree
|
|
experimental design
|
characteristics: randomization, control, manipulation of independent variable
Observation = O Treatment = X |
|
post test only
|
R X O
R O |
|
Pre-Post
|
R O X O
R O O |
|
Factorial design
|
2 x 2
|
|
cross over
|
R O X O O
R O O X O |
|
proxy-pretest
|
N O1 X O2
N O1 O2 |
|
separate
|
pre/post
double pretest switching replication single participant |
|
treats to internal validity
|
single group threats
history maturation testing instrumentation mortality regression to mean multiple group threats selection history - factors occurring b/w pre and post selection - maturation selection-testing selection-instrumentation selection-mortality selection-regression |
|
single group threats
|
threats affecting assumption treatment did in fact cause effect
|
|
threat to validity: history
|
everything outside of study that affects particpant
|
|
threat to validity: maturation
|
time passage affecting participant
|
|
threat to validity: testing
|
testing gives participant experience
|
|
threat to validity: instrumentation
|
change in way outcome measured
|
|
threat to validity: mortality
|
participants die
|
|
threat to validity: regression to mean
|
higher or lower score in initial test results in higher or lower score in subsequent tests
|
|
threat to validity: multiple group threats
|
anything other than treatment that affects post test
|
|
threat to validity: selection history
|
maturation, testing, instrumentation, mortality, regression
|
|
social interaction threats
|
diffusion/imitation of treatment: control hears about treatment from treatment group
avis effect: control competes w/ treatment resentful demoralization: control group gives up after learning what treatment group is getting Compensatory equalization: investigators provide different treatment to control placebo effect: participation in study elicits effect separate from independent variable halo effect: researcher preconceived notions hawthorne effect: when you know you're being studies you change behavior |
|
Epidemiology
|
incidence - number of new cases of disease that occur in pop w/in a period of time
Prevalence - proportion of population affected by disease at given time: new and existing cases |
|
relative risk
|
proportion in unexposed group w/ disease compared to proportion in the exposed group
|
|
Relative risk calculation
|
cumulative incidence of exposed / divided by cumulative incidence of unexposed
|
|
odds ratio
|
relative risk calculated when disease is rare
OR = exposed cases x non-exposed cases / div. by unexposed cases x exposed cases |
|
descriptive research
|
ecological studies
case reports survey/cross -sectional surveillance |
|
descriptive research: ecological studies
|
used to examine patterns relating to risk of disease
|
|
descriptive research: case reports
|
one person examined
|
|
descriptive research: survey/cross-sectional
|
collecting wide range of data at one point in time (NHANES)
|
|
descriptive research: surveillance
|
systematic survey designed to monitor specific health outcome
|
|
analytic nutrition epidemiology
|
diet and disease
|
|
analytic epidemiologic study designs
|
cross sectional
cohorts case control randomized controlled trials |
|
cross sectional design
|
participants measured at same point in time
descriptive, not good for causal relation |
|
cohorts
|
track health info of individuals over time
good for tracking disease development assess exposures as they occur |
|
case-control
|
participants w/ disease vs. controls
quick and smaller sample sizes only investigate one disease |
|
randomized controlled trials
|
used to establish causal relationships
gold standard in research |