Statistics for Environmental Engineers

Скачать в pdf «Statistics for Environmental Engineers»

+


+


У2


3



+




+



+


У3


4


+


+



+





У4


5




+


+




+


У5


6


+



+



+




У6


7



+


+




+



У7


8


+


+


+


+


+


+


+


У8

and minuses indicate the two levels of each factor to be investigated. Notice that each factor is studied four times at a high (+) level and four times at a low (-) level. There is a desirable and unusual balance across all k = 7 factors. These designs exist for N = к + 1, as long as N is a multiple of four. The analysis of these factorial designs is explained in Chapters 27 to 30.

Comments


An accurate measurement has no bias and high precision. Bias is systematic error that can only be removed by improving the measurement method. It cannot be averaged away by statistical manipulations. It can be assessed only when the true value of the measured quantity is known.


Precision refers to the magnitude of unavoidable random errors. Careful measurement work will minimize, but not eliminate, random error. Small random errors from different sources combine to make larger random errors in the final result. The standard deviation (s) is an index of precision (or imprecision). Large s indicates imprecise measurements. The effect of random errors can be reduced by averaging replicated measurements. Replicate measures provide the means to quantify the measurement errors and evaluate their importance.

Скачать в pdf «Statistics for Environmental Engineers»