Beruflich Dokumente
Kultur Dokumente
Data Analysis
Periodic analysis to allow for trending on a periodic
basis to detect problems as early as possible and
implement corrective action
Statistical Process Control - preventative actions
Selection of Critical Quality Attributes
Meaningful specifications/ limits
σ,
› 68.27% of observations
0 20 40
x
60 80 100
are within µ σ
› 95.45% of observations
y
exp x / 2 2
2
are within µ 2 σ
› 99.97% of observations
2 are within µ 3 σ
2 i 1
s i 1
n n 1
Sri Noegrohati, UGM 19
“The closeness of agreement between a test result
and the accepted reference value.”
“Accuracy of a measuring instrument is the ability
of a measuring instrument to give responses close
to a true value.”
Sri Noegrohati,
Fac of Pharmacy
23
GMU
"Student's t-test deals with the problems associated with
inference based on "small" samples: the calculated mean
(Xavg) and standard deviation (σ) may by chance deviate
from the "real" mean and standard deviation (i.e., what you'd
measure if you had many more data items: a "large" sample).
For example, it it likely that the true mean size of maple
leaves is "close" to the mean calculated from a sample of N
randomly collected leaves.
At 95% of the time, the actual mean would be in the range:
› If N=5, Xavg± 2.776 σ/N1/2;
› if N=10: Xavg± 2.262 σ/N1/2 ;
› if N=20: Xavg± 2.093 σ/N1/2 ;
› if N=40; Xavg± 2.023 σ/N1/2 ;
› and for "large" N: Xavg± 1.960 σ/N1/2 .
(These "small-sample" corrections are included in the
descriptive statics report of the 95% confidence interval.)
Student" (real name: W. S. Gossett [1876-1937])
CONFIDENCE INTERVALS
The confidence interval is the expression stating that
the true mean, µ, is likely to lie within a certain
distance from the measured mean, x.
– Student’s t test
Sri Noegrohati,
26 UGM
d.f. 90% 95% 99% 99.9%
known value x
t calc n
s
tcalc > ttable at 95% confidence level
results are considered to be different
the difference is significant!
x1 x2 n1n2
t calc
spooled n1 n2
Where Spooled = pooled std dev. from both sets of data
Sri Noegrohati,
35
UGM
The equipment on which the work is being done is
broadly suited to the application. It is clean, well
maintained and within calibration.
The staff carrying out the validation are competent in
the type of work involved.
There are no unusual fluctuations in laboratory
conditions and there is no work being carried out in the
immediate vicinity that is likely to cause interferences.
The samples being used in the validation study are
known to be sufficiently stable.
Fitness for purpose is the “degree to which data
produced by a measurement process enables a
user to make technically and administratively
correct decisions for a stated purpose.”
Based on Lambert - Beer law linear correlation between
absorbance and concentration calibration curve
n xi yi xi yi
c
x y x y x
2
m
n x x
i i i i i
n xi2 xi 2 i
2
i
2
Sri Noegrohati,
41
UGM
The vertical deviation can be calculated as follows:
di = yi – (mxi + c)
Some deviations are positive (point lies above the
curve) and some are negative (point lies below the
curve) square the values so that the sign does not
play a role
di2 = (yi – mxi - c)2
Estimate the standard deviation for all y values
sy i
d 2
n2
Sri Noegrohati, UGM 42
Standard deviation for the slope
(m): 2
sy n
x
2
sm
n xi 2
i
2
s y2 xi2
x
2
sc
n xi 2
i
2
Detectability
Is the Analytical Signal distinguishable from the Blank?
need to know the uncertainty of the measurements.sc
Y detection limit = c+3sc LOD = (c+3sc)-c/m
Sri Noegrohati,
43
UGM
Sensitivity:
calibration sensitivity = slope (m) of calibration curve.= ability
to discriminate between small differences in analyte
concentration
analytical sensitivity (g) = slope (m)/standard deviation (Ss)
Performed by
manufacturer
Performed by Lab