1/70
Looks like no tags are added yet.
Name | Mastery | Learn | Test | Matching | Spaced | Call with Kai |
|---|
No analytics yet
Send a link to your students to track their progress
What is the concept of reliability defined as?
How can you reproduce something with consistency
What is the concept of validity defined as?
How accurate or correct something is
What example is not valid and not reliable?
marks on a target spread out and not close to the bullseye
What example is not valid but reliable?
Marks on a target close together but not close to the bullseye
What example is both valid and reliable?
When the marks are close together and on the bullseye
What is the definition of reliability?
The extent to which a measurement is consistent and free from error
Can a reliable measure be expected to repeat the same score on 2 or more different occasions?
yes it can be, as long as the characteristic of interest doesn't change
How is reliability tied to the concept of "measurement error"
in the way that it is the inverse of measurement error
- when 1 goes up, the other goes down
Examples of measurement error
bathroom scale: +/- 1lbs
range of motion: +/- 5 degrees
temp outside: +/- 1 degree C
body temp: +/- 0.2 degree C
oswestry disability sale: +/- 4 points
measurement error equation
observed score = true score +/- error
In the example of "Dr clark gives an exam," what would the observed score be?
How well you did
In the example of "Dr clark gives an exam," what would the true score be?
the actually score
In the example of "Dr clark gives an exam," what would the error be?
sleep, coffee, etc
What are the 2 types of measurement errors?
systematic error and random error
What is a systematic error?
an error that always either overestimates or underestimates
What is random error?
a variable due to chance or is unpredictable
What does the concept of regression to the mean state?
That when you repeat the same test, and nothing changes, your values should get closer to the mean value
What are the three sources of measurement error?
1. Rater or human error
2. Instrument error (faulty function)
3. variability of the characteristic being observed
What is an example of the variability of a characteristic being observed
blood glucose
- it changes over time
What is relative reliability?
The ratio of total variability between scores to variability within scores is measured as a unitless coefficient
What is the equation for relative reliability?
Rxx = true score variability/ true score variability + error variability
At the top of the equation, "true score variability" is what?
the difference BETWEEN individuals
At the bottom of the equation, "true score variability + error variability" is what?
the difference WITHIN individuals
If Rxx = 0? What does this mean?
that there is no reliability
If Rxx = 1? What does this mean?
1 is perfect reliability
If true variance is larger, what will that tell us about Rxx?
that Rxx will be larger
What is absolute reliability?
How much of a measured value is likely due to error, and the standard error of the measurement (SEM) used
How can you improve reliability?
1. standardize measurement methods (same steps every time)
2. Take multiple measurements (cancels the effects of random error)
3. train and test observers (reduces human error)
4. calibrate and improve instruments
3 types of reliability most relevant for clinicians?
test-retest
inter-rater
intra-rater
3 types of reliability of questionnaires and surveys, or for comparing different forms of testing?
internal consistency
split-hald
alternate forms
What is the test-retest reliability
Any instrument or tool that is capable of measuring a variable consistently
- usually 2 or more separate occasions, and keep conditions as constant as possible
Where do we see test-retest reliability?
used most for self-report questionnaires, and measures with mechanical/digital readouts
What is INTRA-rater reliability?
the within-rater
the degree that the examiner agrees with themself
What is rater bias?
when 1 rater takes 2 measurements and is influenced by the memory of the first score
How to avoid rater bias?
Use blinding and make assessments as objective as possible
What is INTER-rater reliability?
think "between" two different raters
- 2 or more raters who measure the same group of people
What is internal consistency?
it estimates how well the items that reflect the same construct yield similar results
What is internal consistency used for?
used to construct and evaluate scores and/or questionnaires
How do we measure internal consistency?
Cronback's alphs
- types of average inter-item correlation
What is split-half reliability (a branch of internal consistency)?
When you take all the questions, divide them, and then compare the halves
What is the purpose of alternate forms of reliability?
to see if 2 different versions of the same instrument are equivalent
ex: ROM
What are the issues affecting the validity of change?
levels of measurement
reliability (is it change or just measurement error)
stability (are there meaningless natural fluctuations)
baseline score
What is minimal detectable change (MDC)?
When the change is real but not clinically meaningful
MDC def
value at which the observed change is NOT measurement error
What is the minimal clinically important difference (MCID)
when the change is real and clinically significant
MCID def
The ability of an instrument to detect minimally important change
Measurement validity def
The extent to which an instrument measures what it is intended to measure
What are the 3 things you have to take into consideration during measurement validity?
can the text...
1. discriminate among individuals
2. evaluate change in magnitude or quality from one time to another
3. Predict an outcome
What comes first? reliability or validity?
Reliability is a prerequisite for validity
difference between reliability and validity?
Reliability = consistency of a measurement
validity = alignment of the measurement with a targeted construct
Are validity and reliability an all-or-none concept?
no they are evaluated on a continuum
What is face validity?
When an instrument appears to test what it is intended to test
- It is judged by the users of a test after the test is developed
Face validity is judged by those who _____ the instrument
USE
What are the 3 C's of validity?
contect
construct
criterion-related
What does content validity look at?
If measurements adequately represent the concept
What is content validity used for?
They are typically used in questionnaire development
What are the 3 requirements for content validity?
1. Items must represent the full scope of the construct being studied
2. The number of items is proportionate to the importance of that component
3. No irrelevant items
Content validity is judged by those who ____ the instrument
DEVELOP
What is the similarity between face validity and content validity?
There are no statistics to measure either type of validity
Criterion-related validity def
Can the outcomes of the instrument or index test be substituted for an established gold standard
What is criterion-related validity measured by?
correlation coefficients
sensitivity/specificity analyses
What are the 2 types of criterion-related validity?
concurrent and predictive validity
concurrent validity def
Scores from a new measure correlate with scores from a well-established measure administered AT THE SAME TIME
predictive validity def
establishes that the outcome of the target test can be used to predict a future criterion score or outcome
Construct validity def
reflects the ability of an instrument to measure the theoretical dimensions of a construct
What are the 3 types of construct validity?
known-group method
convergence
divergence
What does the known-group method test?
tests if the results differ between 2 groups that are known to be different
What does convergence show
If there is a correlation between 2 constructs
if R = 1, what does this mean?
1 is a perfect correlation
very strong correlation
What does divergence show?
show that there isn't a correlation between 2 constructs
If R = 0, what does this mean?
If R=0, there is no correlation