biosensore lecture 2- statistics for bioinstrumentatiom

accuracy of the measured data from an instrument

closeness of the data to the true value

• agreement between two data sets

accuracy of a new bioinstrument if there is no known true value

the closeness of the data from a new instrument to a reference standard

what does statistics involve

collecting, classifying, summarizing, organizing, analyzing, and interpreting numerical data

key concepts in statistics

• population

• sample

• parameter

• statistic

population

entire population of interest

sample

a subset of a population selected for analysis

parameter

a numerical characteristic of a population

statistic

a numerical characteristic of a sample

mean

measure of central tendency

measures of dispersion

• variance

• standard deviation

statistical and graphical approached to determine accuracy

• data plot

• correlation coefficient (r )

high r value example

low r value example

example of finding correlation coefficient, r

• steps to get from data set to r value

is the correlation coefficient a good measure of accuracy?

• strength of relation does NOT equal accuracy

• a data set that is in poor agreement with the reference data set can have a high correlation coefficient as long as they are related

what does the correlation coefficient tell us?

• how two sets of data are related

• the strength of relation

Bland-Altman plot

plot of the difference between the measured data and the reference data against their mean

statistical values used in Bland-Altman plots

• bias: the mean difference

• limts of agreement (upper and Iower)

bias

mean of the difference of the measured and reference data

good agreement Bland-Altman plot

• small bias

• small LOA

• no systematic trend

example of creating a Bland Altman plot