Laboratory Statistics Flashcards
AVERAGE or arithmetic mean
mean (x̄)
MIDPOINT of data
median
most FREQUENTLY occurring value in a data set
mode
distribution of data points AROUND THE MEAN
standard deviation
square of std deviation
variance
best indicator of PRECISION
↓CV ↑precision
coefficient of variation
difference between the HIGHEST and LOWEST values
range
Bell-shaped curve
Gaussian/Normal Distribution
empirical rule of Gaussian
68% - 95% - 99.7%
± 1s
68%
± 2s
confidence interval
95%
± 3s
99.7%
CV rule of thumb
<10%
control values before qc data analysis
minimum of 20
pair of medical decision points that span the LIMITS OF RESULT expected for a defined HEALTHY POPULATION
reference interval
don to CONFIRM VALIDITY of an EXISTING OR PUBLISHED RI for an analyte
verifying RI
VERIFYING RI
- requires at least __ study individual
- RI is adopted if __% of the subjects fall outside the range
20
<10%
done when there is NO EXISTING RI for an analyte or when transference studies fail
establishing RI
ESTABLISHING RI
- requires __ study individuals
- RI is set based on the __%
120
95%
x̄ ±2S
confidence interval
done by running TWO CONTROL materials TWICE A DAY over a 10-DAY PERIOD
precision study
involves spiking a sample with a known amount of an analyte and determining how much of it can be detected by the method in the presence of other compounds in the matrix
recovery study
compares the MEAN of two groups of data or the ACCURACY of two methods
T test
compares the SDs of two groups of data or the PRECISION of two procedures
F test
used to compare two methods using the best fit line through the data points
linear regression
- independent variable
- standard/reference method
X
- dependent variable
- new method
Y
AXIS
Abscissa
X-axis
Independent variable
Standard/reference method
VOYD
Vertical / X-axis
Ordinate
Y-axis
Dependent variable
ability of a method to detect the SMALLEST CONCENTRATION of an analyte
analytical sensitivity
ability of a method to detect ONLY the ANALYTE OF INTEREST
analytical specificity
- ability of a test to detect a given DISEASE or CONDITION
- proportion of individuals with the DISEASE who have a positive test result
diagnostic sensitivity
- ability of a test to detect a given ABSENCE OF A GIVEN DISEASE or CONDITION
- proportion of individuals with NO DISEASE who have a negative test result
diagnostic specificity
- PROBABILITY that a positive test result indicates DISEASE
- proportion of individuals with POSITIVE results who truly have the disease
Positive Predictive Value
- PROBABILITY that a negative test result indicates ABSENCE OF DISEASE
- proportion of individuals with NEGATIVE results who truly do not have the disease
Negative Predictive Value
TP/TP+FNx100
%Sensitivity
TN/TN+FPx100
%Specificity
TP/TP+FPx100
%PPV
TN/TN+FNx100
%NPV