Statistics Flashcards
(23 cards)
Define accuracy
How close a number is to what it should be, by comparing a number to known/accepted value
Define precision
Number of decimal places of a measured number (reproducibility)
Describe accuracy and precision (bull’s eyes)
Good accuracy - Closer to the bull’s eye
Good precision - Closer each point is to eachother
Calculating absolute error
Difference between measured and true value
Calculating relative error
Absolute error divided by true value
What method is used to measure precision?
Standard deviation
What is s^2?
Variance/RSD (%)
What is Gaussian rounding?
Avoiding biased results by rounding to the nearest even number
What is standard deviation?
Measure of variation
What are the 3 types of errors?
Gross, Systematic, Random
Describe gross errors
Occurs sometimes, large and cause results to be high/low (lead to outliers)
Describe systematic errors
Definite values, lead to bias, affects all data
Describe random errors
Uncontrollable variables, undetectable uncertainties
What is associated with systematic error?
Poor accuracy
What is associated with large random error?
Poor precision
What is random error a representation of?
The distribution around the mean (outer areas of mean in distribution curve)
Describe Dixon’s Q-test
Used for gross errors by comparing difference between samples and nearest measurement
What is the formula for the Q-test?
(Suspect value - Nearest value) / (Largest value - Smallest value)
Define confidence interval (CI)
The probability that interval estimate includes the population mean (u)
What is the equation for confidence interval for the mean?
u = x +/- t (s/ 2/n) u = Population mean x = Sample mean s = Standard deviation t = Student's t-value n = Number of measurement
What is SEM?
Standard error of the mean - (s/ 2/n)
What is the best confidence percentage to use?
95% unless told
How to calculate t-value?
Find degree of freedom by number of samples - 1
