Research Flashcards
Nominal
Name
Mutually exclusive ie type of arthritis, blood type
Ordinal
Ranking
Ie MMT, assistance level, pain, joint laxity
Interval
Intervals are equal
NO true zero
Degrees, functional status test
Ratio
Intervals equal
TRUE zero
ROM, distance, height, etc
Alternate forms reliability
Consistency of measurements obtained from a different form of test.
Internal consistency
Items or elements reflect one basic phenomenon or dimension
Intrarater
Same person over time
Interrater
agreement of measures made by More than one person
Test-rest reliability
Consistency of repeat measures
Face validity
Does it test what it’s supposed to
Content validity
Does it reflect the content domain of interest and not extraneous elements?
Construct validity
Degree to which theoretical construct is measured
Criterion-related validity
Validity when compared to “Gold standard”
Includes: concurrent, predictive and prescriptive validities
Concurrent validity
Type of criterion-related
Comparing measurement to a gold standard at the same time
Predictive validity
Type of criterion related
Measurement valid because it is predictive of future behavior
Prescriptive validity
Type of criterion related
Suggests form of tx a person should receive
Sampling error
Chance between statistic calculated from sample and true parameter in population
Probability sampling
Random selection. Each members has same chance of getting picked
Includes: simple random, systemic, stratified random, cluster
Simple random sampling
Type of probability sampling
Subjects have equal chance of being selected
Systematic sampling
Type of probability sampling
Taking every nth subject
Stratified random sampling
Type of probability sampling
Population decided into strata then simple random sample from each
Cluster sampling
Type of probability sampling
Decided into clusters (geographic) and random sample from each
Non-probability sampling
No random selection
Convenience sampling
Type of non-probability
Subjects who are readily available
Purposive sampling
Type of non-probability
Deliberately selected based on predefined criteria
Quota sampling
Type of non probability
Similar to stratified random, but subjects from subgroups are not random
Snowball sampling
Type of non probability
Subjects ID by asking existing subjects to ID names of other potential participants
Single blind
Subjects blinded
Double blind
Subjects and researches
Triple blind
Subjects, researchers, fatal analyzers
Active control
Known effective form of tx as control
Placebo control
Inactive substance that looks same as tx
Unethical to use if effective treatment is available as active control
P value
Probability that a particular statistic result could have happened by chance
P is smaller than alpha
Reject the null hypothesis
P is larger than alpha
The null hypothesis is not rejected
Alpha level
Significance level
Probability of rejecting the null hypothesis when it is true, or chance of Type 1 error
Type 1 error
Alpha error
Wrongly rejecting null hypothesis
Concluding there is a difference when there is not (false positive)
If alpha set at 0.01
There is a 1% chance of a type 1 error occurring
Type II error
Beta error
Wrongly deciding not to reject the null
Concluding there is no difference or relationship when there is (false negative)
Normal distribution
Bell shaped
Mean, median, mode are the same
Normal distribution- % of all values within 1 SD Above and below mean
68%
Normal distribution- % of all values within 2 SDs above and below mean
95%
Normal distribution- % of all values within 3 SDs above and below mean
99%
Mean and medial to left of mode
Negatively skewed
Left tail is elongated
Mean and median to right of mode
Positively skewed
Right tail is elongated
Coefficient of variation
Ratio of SD of a distribution to mean
CV = (s/ mean)*100
Standard deviation
Square root of variance
Bar graph
Magnitude or frequency of categories
Length of bar is proportional to magnitude or frequency
Box and whisker plot
Illustrates the distribution of values within a group through 5 numbers with outliers shown as circles
Numbers in box and whisker plot
Minimum score Lower quartile (q1) Median (q2) Upper quartile (q3) And maximum score
Histogram
Displays distribution by plotting frequency (y axis) for each interval represented (x axis).
Line graphs
Relationship between variables with dependent on y axis and independent on x axis
Idea for showing trends in data over time
Scatter plot
Relationship between two quantitative variables
Line of best fit generated to show relationship
Sensitivity
Percentage of people who test positive for disease amount people who have the disease
-airport metal detectors
Specificity
Percentage of people who test negative for a specific disease among group of people who do not have it
-airport pat down
Incidence
Number of new cases in the population at risk during specified Tim interval
Prevalence
Number of existing cases at a point in time including new and pre-existing
Relative risk
Measure of risk of certain event happening I one group compared to risk of same happening in another
RR 1.0 means that is is equally probable in both groups.
Greater than one implies increase, less than one implies decrease risk
Odds ratio
Measure of odds of event happening in one group compared to odds of same event happening on other group (usually case control for exposure)
1.0 means exposure probably does not increase risk
>1 means may increase risk
<1 means may reduce risk
