research lecture 3 Flashcards
what level of evidence is a RCT
1B
where is the chart of the research design and outcome assessment ussuially found in the manuscript
methods : procedures
using ____ you can calculate how many participants you need to recruit
G*power
if the authors powe is .8 what does that mean
80% chance that there is a chance to finding the difference
if the p value is closer to alpha then what does that mean ? (alpha is .05)
that there is a less significant that there will be a difference
you should never see power below what number
.8
what are ways to increase power
-increase the number of participants
-decrease the variability of data by limiting people in the study
-make alpha level larger
-make effect size bigger by making difference of means bigger or make SD smaller
where do you find the consort diagram in the article
procedures
why is t-test not the best analysis
bc it requires you to divide your alpha
what kind of research design is it when participants are randomly assigned to at least two comparison groups
experimental (random)
what type of research design is it when there is no random assignment and typically involve the use of cohort groups
quasi - experimental (not random)
what is considered key for true experimental design
randomization
Some researchers consider “____ ” designs as better than having a control group since the participants serve as their own controls.
repeated measures
what measure compares 1 group over 2+ time periods
one way repeated measures
what research design handout is when the control group can be different intervention or standard of care … or simple compare without any intervention but the study sample is randomized
posttest only randomized group
what research design handout is when the control group can be different intervention or standard of care … or simple compare without any intervention but the study sample is in cohort groups
posttest only non randomized
what research design handout is - with or without randomization into groups , there are different participants in each cell
factorial fully independent (2+ IVs)
what does a 2x2 ANOVA tell you
there is 2 IV and they both have 2 pieces that go into them
what research design handout is - the subjects participate in every cell
factorial fully repeated
what is with or without randomization , groups of subjects move thru the repeated conditions
factorial mixed
what does a 2x3 mixed ANOVA mean
there are 2 IVs one with 2 levels and one with 3 that is repeated
what is the most common ANOVA
mixed
what is crossover with or without randomization
when at one point there is a switch in intervention
what provide the foundations for evaluating test , their uses and interpretations
measurement theory AKA psychometrics
what are the 2 most fundamental measurement theory concepts
reliability and validity
what study is any research study that investigates the reliability and/or validity of clinical and research measure
methodological study
what is the operational definition included in
the manuscript
variables such as aggression , cognition, pain , self esteem , strength , coordination etc must be specifically define in research .. what is this called
operational definition
what is the amount of change in an instrument that is beyond error
MDC- minimally detectable change
what is the amount for change in an instrument that results in a clinically relevant change in the pateint
MCID- minimal clinically important different
what is consider the consistency or the degree of association or covariation between 2 variables
reliability
when will reliability by high
if there is relative agreement between two varibale
what is the accuracy or the degree of correspondence between the concept being measrue and the varibale used to presents the concept
validity
when is an instrument considered valid
when the test actually measures what it is intended to measure
can the instrument be reliable without being valid >???
can it be valid without being reliable
yes and no bc if it isn’t reliable it wouldn’t be measuring the concept well
what is the difference between inter tester and intra tester
inter tester is seeing if different testers can get the same score and intra tester is seeing if the same tester can get the same score
what are the 2 different test retest reliability
seeing if successive measurements can be done by a instrument consistently
seeing if successive measurements by a patient can be consistent
what is also called equivalent forms reliability
parallel forms reliability
what type of reliability does this explain … Questions (from a test or a survey) are divided into two equivalent sets (forms). The two sets of questions are given to the same sample of people (at the same time or nearly the same time) and the reliability between the sets is calculated.
parallel forms reliability
what reliablity does this describe …The questions (from a test or a survey) are split into two sets…and both sets
are given to one group of individuals
split half reliability
what is face validity asking
does an instrument measure what it is supposed to measure based on simple observation?
in expert panels for face validity , there is typically no ___ ____
statistical analysis
what validity is The systematic examination of an instrument to determine whether it
covers the entire domain to be measured
content validity
if we ask does an IQ questionnaire have
items covering all areas of intelligence discussed in the scientific literature?… what type of validity is it
content
what is the key difference between face and content validity
face is what it appears to measure and vontent is what experts agree it measures
what validity is defined as Can an instrument be used to predict some future performance/outcome
predictive validity
this is an example of what kind of validity … Ex/ Berg Balance Scale: Score of < 45 indicates individuals may be at greater
risk of falling
predictive
what validity is the degree to which a theoretical construct is measured by an instrument.
construct validity
what validity is this an exmaple of x Ex/ Factor Analysis – statistical analysis of a survey that gives you the various
factors that are represented in an instrument.
construct
what validly is The extent to which significant changes in the participants are reflected in substantive changes in observed values
responsiveness to change
if you have a tool with these options : dependent , requires assistance and Ind is it considered responsive to change
no
if you have a tool with these options : rate your ability from 0-10 if it going to be responsive to change
yes
what is criterion validity
comparing the new tool to the gold standard
what is considered a true positive rate
sensitivity
if a person has a disease or condition what will sensitivity tell you
how often the test will be positive
what type of test LOVES to give positive test results
high sensitive test
does sensitivity rule in or out
rule out
if you have a high sensitive test and u have a negative test result what is the probability that you have the disease
you don’t have the disease it has been ruled out
what is a true negative rate
specificity
a high specific test loves to give a ___ result
negative
so if you have a high specific test and you get. a positive test result what does that men
it is highly likely you are positive , the test is ruled in
does specificity rule in or out
in
what is a true positive and false positive
true positive is if you have the conditions and the test is positive
false positive is if the test says you have the condition but you actually done (type 1 error)
what is the different between true negative and false negative
true negative is when you don’t have the condition and you test negative
false negative i when you test negative but you are positive for condition type 2 erroe
how do you find the sensitivity
true positive / condition positive (true positive + false negative )
how do you find the specificity
true negotiate / condition negative (false postive + true negative)
what does positive predictive value mean
tells us the probability that a person with a positive clinical test truly has the condition
what does a negative predictive value tell us
the portability that a person with a negative clinical test truly doesn’t have the disease
how do you find the positive predictive value
true postive / test outcome postive (true positive + false positive)
how do you find th negative predictive value
true negative / test outcome negative (false negative + true negative)
what happens to the postive predictive value as the prevalence of the disease in a population increases
it increases
negative predictive value will do what as the prevalence of the disease in the population decease
increases
what is one of the hallmarks of clinical decision making
likelihood rations
for positive likelihood ratio is it better lower or higher
higher
how do you find the positive LR
sensitivity / (100- specificity)
which likelihood ration is the the lower the better
negative
how do you find negative LR
(100- sensitivity) / specificity
if what is the positive LR quality numbe4rs
10 is the best
5 is ok
2 is poor
1 is useless
what is the negative LR quality numbers
1 is useless
.5 is poor
.2 is ok
.1 is best
if the LR+ is a 2 how much does it increase percent wise
15%
if the LR+ is a 5 how much does it increase percent wise
30
if the LR+ is a 10 how much does it increase percent wise
45
if the LR- is a .5 how much does it increase percent wise
15
if the LR- is a .2 how much does it increase percent wise
30
if the LR- is a .1 how much does it increase percent wise
45
what is a way to visually combine sensitivity and specificity
ROC curves
in the ROC curve what is on the vertical axis and horixational axis
plots sensitivity on vertical and plots (1- specificity) on horizontal axis
if you are looking at the ROC curve and u see an increased are under the line what does that represent
high sensitivity anf specificity
what number do we what the are under the red line (AUC) to be
higher to 1 the better
for the best test what would you want one theron curve
low specificity and number sensitivity
