Anda di halaman 1dari 125

Bahan Kuliah

• http//:Byudono.wordpress.com
• yudonob@hotmail.com
Statistics for Analytical
Chemistry

Recommended textbook:
“Statistics and Chemometrics for Analytical Chemistry”
James N Miller & Jane C Miller
(Pearson Prentice Hall)
STATISTIKA :
Ilmu dan atau seni yang berkaitan
dengan tata cara (metode) pengumpulan
data, analisis data, dan interpretasi hasil
analisis untuk mendapatkan informasi
guna penarikan kesimpulan dan
pengambilan keputusan
1. Konsep Statistika

STATISTIKA :
Kegiatan untuk : KEGUNAAN
• mengumpulkan data
• menyajikan data
• menganalisis data dengan metode tertentu
?
• menginterpretasikan hasil analisis

Melalui fase

STATISTIKA DESKRIPTIF :
Berkenaan dengan pengumpulan, pengolahan, dan penyajian sebagian
atau seluruh data (pengamatan) tanpa pengambilan kesimpulan

dan fase

STATISTIKA INFERENSI :
Setelah data dikumpulkan, maka dilakukan berbagai metode statistik untuk
menganalisis data, dan kemudian dilakukan interpretasi serta diambil kesimpulan.
Statistika inferensi akan menghasilkan generalisasi (jika sampel representatif)
2. Statistika & Metode Ilmiah

METODE ILMIAH :
Adalah salah satu cara mencari kebenaran yang bila ditinjau dari
segi penerapannya, resiko untuk keliru paling kecil.

LANGKAH-LANGKAH DALAM METODE ILMIAH :


1. Merumuskan masalah
2. Melakukan studi literatur
3. Membuat dugaan-dugaan, pertanyaan-pertanyaan atau
hipotesis

4. Mengumpulkan dan mengolah data, menguji hipotesis, atau


menjawab pertanyaan
INSTRUMEN
5. Mengambil kesimpulan
SAMPEL

SIFAT DATA
PERAN STATISTIKA
VARIABEL

METODE ANALISIS
3. Data

DATA terbagi atas DATA KUALITATIF dan DATA KUANTITATIF

DATA KUALITATIF : DATA KUANTITATIF :


Data yang dinyatakan dalam Data yang dinyatakan dalam
bentuk bukan angka. bentuk angka
Contoh : jenis pekerjaan, Contoh : lama bekerja,
status marital, tingkat jumlah gaji, usia, hasil
kepuasan kerja ulangan

DATA

KUALITATIF JENIS KUANTITATIF


DATA

NOMINAL INTERVAL
ORDINAL RASIO
4. Data
DATA NOMINAL :
Data berskala nominal adalah data yang diperoleh dengan cara kategorisasi atau klasifikasi.
CIRI : posisi data setara
tidak bisa dilakukan operasi matematika (+, -, x, :)
CONTOH : jenis kelamin, jenis pekerjaan

DATA ORDINAL :
Data berskala ordinal adalah data yang dipeoleh dengan cara kategorisasi atau klasifikasi, tetapi di
antara data tersebut terdapat hubungan
CIRI : posisi data tidak setara
tidak bisa dilakukan operasi matematika (+, -, x, :)
CONTOH : kepuasan kerja, motivasi

DATA INTERVAL :
Data berskala interval adalah data yang diperoleh dengan cara pengukuran, di mana jarak antara dua
titik skala sudah diketahui.
CIRI : Tidak ada kategorisasi
bisa dilakukan operasi matematika
CONTOH : temperatur yang diukur berdasarkan 0C dan 0F, sistem kalender

DATA RASIO :
Data berskala rasio adalah data yang diperoleh dengan cara pengukuran, di mana jarak antara dua titik
skala sudah diketahui dan mempunyai titik 0 absolut.
CIRI : tidak ada kategorisasi
bisa dilakukan operasi matematika
CONTOH : gaji, skor ujian, jumlah buku
5. Pengolahan Data

PROSEDUR PENGOLAHAN DATA :

A. PARAMETER : Berdasarkan parameter yang ada statistik dibagi menjadi

• Statistik PARAMETRIK : berhubungan dengan inferensi statistik


yang membahas parameter-parameter populasi; jenis data interval
atau rasio; distribusi data normal atau mendekati normal.

• Statistik NONPARAMETRIK : inferensi statistik tidak membahas


parameter-parameter populasi; jenis data nominal atau ordinal;
distribusi data tidak diketahui atau tidak normal

B. JUMLAH VARIABEL : berdasarkan jumlah variabel dibagi menjadi

• Analisis UNIVARIAT : hanya ada 1 pengukuran (variabel) untuk n


sampel atau beberapa variabel tetapi masing-masing variabel
dianalisis sendiri-sendiri. Contoh : korelasi motivasi dengan
pencapaian akademik.

• Analisis MULTIVARIAT : dua atau lebih pengukuran (variabel)


untuk n sampel di mana analisis antar variabel dilakukan bersamaan.
Contoh : pengaruh motivasi terhadap pencapaian akademik yang
6. Pengolahan Data

MULAI

Statistik NOMINAL INTERVAL


Jenis Statistik
Non Parametrik ORDINAL Data RASIO
Parametrik
?

Analisis SATU DUA / LEBIH Analisis


Jumlah
Univariat Variabel Multivariat
?
The General Analytical Problem

Select sample
Extract analyte(s) from matrix

Separate analytes

Detect, identify and


quantify analytes

Determine reliability and


significance of results
Errors in Chemical Analysis

Impossible to eliminate errors.


How reliable are our data?
Data of unknown quality are useless!

•Carry out replicate measurements


•Analyse accurately known standards
•Perform statistical tests on data
Mean Defined as follows:
N
 xi
i=1
x =
N

Where xi = individual values of x and N = number of replicate


measurements

Median
The middle result when data are arranged in order of size (for even
numbers the mean of middle two). Median can be preferred when
there is an “outlier” - one reading very different from rest. Median
less affected by outlier than is mean.
Illustration of “Mean” and “Median”
Results of 6 determinations of the Fe(III) content of a solution, known to
contain 20 ppm:

Note: The mean value is 19.78 ppm (i.e. 19.8ppm) - the median value is 19.7 ppm
Precision

Relates to reproducibility of results..


How similar are values obtained in exactly the same way?

Useful for measuring this:


Deviation from the mean:

di  xi  x
Accuracy

Measurement of agreement between experimental mean and


true value (which may not be known!).
Measures of accuracy:

Absolute error: E = xi - xt (where xt = true or accepted value)

Relative error: x x
E  i t 100%
r x
t
(latter is more useful in practice)
Illustrating the difference between “accuracy” and “precision”

Low accuracy, low precision Low accuracy, high precision

High accuracy, low precision High accuracy, high precision


Some analytical data illustrating “accuracy” and “precision”

HN NH3+Cl-
S
H H

Benzyl isothiourea
hydrochloride

OH

N
Analyst 4: imprecise, inaccurate
Analyst 3: precise, inaccurate Nicotinic acid
Analyst 2: imprecise, accurate
Analyst 1: precise, accurate
Bias and precision – dot-plots of the data in Table 1.1.
In summary, precision describes random error, bias
describes systematic error, and the accuracy, i.e.
closeness to the true value of a single measurement or
a mean value, incorporates both types of error.

• Repeatability describes the precision of within-run


replicates
• Reproducibility describes the precision of between-run
replicates
•The reproducibility of a method is normally expected to
be poorer (i.e. With larger random errors) than its
repeatability
Types of Error in Experimental
Data
Three types:
(1) Random (indeterminate) Error
Data scattered approx. symmetrically about a mean value.
Affects precision - dealt with statistically (see later).

(2) Systematic (determinate) Error


Several possible sources - later. Readings all too high
or too low. Affects accuracy.
(3) Gross Errors
Usually obvious - give “outlier” readings.
Detectable by carrying out sufficient replicate
measurements.
Sources of Systematic Error
1. Instrument Error
Need frequent calibration - both for apparatus such as
volumetric flasks, burettes etc., but also for electronic
devices such as spectrometers.
2. Method Error
Due to inadequacies in physical or chemical behaviour
of reagents or reactions (e.g. slow or incomplete reactions)
Example from earlier overhead - nicotinic acid does not
react completely under normal Kjeldahl conditions for
nitrogen determination.
3. Personal Error
e.g. insensitivity to colour changes; tendency to estimate
scale readings to improve precision; preconceived idea of
“true” value.
Systematic errors can be
constant (e.g. error in burette reading -
less important for larger values of reading) or
proportional (e.g. presence of given proportion of
interfering impurity in sample; equally significant
for all values of measurement)
Minimise instrument errors by careful recalibration and good
maintenance of equipment.

Minimise personal errors by care and self-discipline

Method errors - most difficult. “True” value may not be known.


Three approaches to minimise:
•analysis of certified standards
•use 2 or more independent methods
•analysis of blanks
Tackling systematic errors:
Foresight: identifying problem areas before
starting experiment
Careful experimental design, e.g. use of
calibration methods
Checking instrument performance
Use of standard reference materials and other
standards
Comparison with other methods for the same
analyte(s)
Participation in proficiency testing schemes
Statistical Treatment of
Random Errors
There are always a large number of small, random errors
in making any measurement.

These can be small changes in temperature or pressure;


random responses of electronic detectors (“noise”) etc.

Suppose there are 4 small random errors possible.


Assume all are equally likely, and that each causes an error
of U in the reading.
Possible combinations of errors are shown on the next slide:
Combination of Random Errors

Total Error No. Relative Frequency

+U+U+U+U +4U 1 1/16 = 0.0625

-U+U+U+U +2U 4 4/16 = 0.250


+U-U+U+U
+U+U-U+U
+U+U+U-U

-U-U+U+U 0 6 6/16 = 0.375


-U+U-U+U
-U+U+U-U
+U-U-U+U
+U-U+U-U
+U+U-U-U

+U-U-U-U -2U 4 4/16 = 0.250


-U+U-U-U
-U-U+U-U
-U-U-U+U

-U-U-U-U -4U 1 1/16 = 0.01625

The next overhead shows this in graphical form


Frequency Distribution for
Measurements Containing Random Errors

4 random uncertainties 10 random uncertainties

This is a
A very large number of Gaussian or
random uncertainties normal error
curve.
Symmetrical about
the mean.
Mean and Standard Deviation

Find the mean and standard deviation of A’s results.


Replicate Data on the Calibration of a 10ml Pipette

No. Vol, ml. No. Vol, ml. No. Vol, ml

1 9.988 18 9.975 35 9.976


2 9.973 19 9.980 36 9.990
3 9.986 20 9.994 37 9.988
4 9.980 21 9.992 38 9.971
5 9.975 22 9.984 39 9.986
6 9.982 23 9.981 40 9.978
7 9.986 24 9.987 41 9.986
8 9.982 25 9.978 42 9.982
9 9.981 26 9.983 43 9.977
10 9.990 27 9.982 44 9.977
11 9.980 28 9.991 45 9.986
12 9.989 29 9.981 46 9.978
13 9.978 30 9.969 47 9.983
14 9.971 31 9.985 48 9.980
15 9.982 32 9.977 49 9.983
16 9.983 33 9.976 50 9.979
17 9.988 34 9.983

Mean volume 9.982 ml Median volume 9.982 ml


Spread 0.025 ml Standard deviation 0.0056 ml
Calibration data in graphical form

A = histogram of experimental results

B = Gaussian curve with the same mean value, the same precision (see later)
and the same area under the curve as for the histogram.
Results of 50 determinations of nitrate
ion concentration, in µgml-1
The normal distribution
SAMPLE = finite number of observations
POPULATION = total (infinite) number of observations
Properties of Gaussian curve defined in terms of population.
Then see where modifications needed for small samples of data

Main properties of Gaussian curve:

Population mean (m) : defined as earlier (N  ). In absence of systematic error,


m is the true value (maximum on Gaussian curve).
Remember, sample mean ( x ) defined for small values of N.
(Sample mean  population mean when N  20)

Population Standard Deviation (s) - defined on next overhead


s : measure of precision of a population of data,
given by:
N
 i
( x  m ) 2

i 1
s
N

Where m = population mean; N is very large.

The equation for a Gaussian curve is defined in terms of m and s, as follows:

 ( x  m ) 2 / 2s 2
e
y
s 2
Two Gaussian curves with two different
standard deviations, sA and sB (=2sA)

General Gaussian curve plotted in


units of z, where
z = (x - m)/s
i.e. deviation from the mean of a
datum in units of standard
deviation. Plot can be used for
data with given value of mean,
and any standard deviation.
Normal distributions with the same mean
but different values of the standard
deviation.
Area under a Gaussian Curve

From equation above, and illustrated by the previous curves,


68.3% of the data lie within s of the mean (m), i.e. 68.3% of
the area under the curve lies between s of m.

Similarly, 95.5% of the area lies between s, and 99.7%


between s.

There are 68.3 chances in 100 that for a single datum the
random error in the measurement will not exceed s.

The chances are 95.5 in 100 that the error will not exceed s.
Sample Standard Deviation, s

The equation for s must be modified for small samples of data, i.e. small N

 i
( x  x ) 2

i 1
s
N 1
Two differences cf. to equation for s:

1. Use sample mean instead of population mean.

2. Use degrees of freedom, N - 1, instead of N.


Reason is that in working out the mean, the sum of the
differences from the mean must be zero. If N - 1 values are
known, the last value is defined. Thus only N - 1 degrees
of freedom. For large values of N, used in calculating
s, N and N - 1 are effectively equal.
Alternative Expression for s
(suitable for calculators)

N
(  xi ) 2
(  xi 2 )  i 1

i 1 N
s
N 1

Note: NEVER round off figures before the end of the calculation
Log-normal distribution
Reproducibility of a method for determining
Standard Deviation of a Sample the % of selenium in foods. 9 measurements
were made on a single batch of brown rice.
Sample Selenium content (mg/g) (xI) xi2
1 0.07 0.0049
2 0.07 0.0049
3 0.08 0.0064
4 0.07 0.0049
5 0.07 0.0049
6 0.08 0.0064
7 0.08 0.0064
8 0.09 0.0081
9 0.08 0.0064

Sxi = 0.69 Sxi2= 0.0533


Mean = Sxi/N= 0.077mg/g (Sxi)2/N = 0.4761/9 = 0.0529

0.0533  0.0529
Standard deviation: s  0.00707106  0.007
9 1
Coefficient of variance = 9.2% Concentration = 0.077 ± 0.007 mg/g
Standard Error of a Mean

The standard deviation relates to the probable error in a single measurement.


If we take a series of N measurements, the probable error of the mean is less than
the probable error of any one measurement.

The standard error of the mean, is defined as follows:

sm  s
N
Pooled Data

To achieve a value of s which is a good approximation to s, i.e. N  20,


it is sometimes necessary to pool data from a number of sets of measurements
(all taken in the same way).

Suppose that there are t small sets of data, comprising N1, N2,….Nt measurements.
The equation for the resultant sample standard deviation is:

N1 N2 N3

 i 1  i 2  i 3 ....
( x  x ) 2
 ( x  x ) 2
 ( x  x ) 2

i 1 i 1 i 1
s pooled 
N 1  N 2  N 3 ......t

(Note: one degree of freedom is lost for each set of data)


Pooled Standard Deviation Analysis of 6 bottles of wine
for residual sugar.

Bottle Sugar % (w/v) No. of obs. Deviations from mean


1 0.94 3 0.05, 0.10, 0.08
2 1.08 4 0.06, 0.05, 0.09, 0.06
3 1.20 5 0.05, 0.12, 0.07, 0.00, 0.08
4 0.67 4 0.05, 0.10, 0.06, 0.09
5 0.83 3 0.07, 0.09, 0.10
6 0.76 4 0.06, 0.12, 0.04, 0.03

(0.05) 2  (010
. ) 2  (0.08) 2 0.0189
s1    0.0972  0.097
2 2
and similarly for all sn .
Set n  ( x  x )
i
2
sn
1 0.0189 0.097
2 0.0178 0.077 01326
.
3 0.0282 0.084 spooled   0.088%
4 0.0242 0.090 23  6
5 0.0230 0.107
6 0.0205 0.083
Total 0.1326
Two alternative methods for measuring the precision of a set of results:

VARIANCE: This is the square of the standard deviation:

N
 i
( x 2
 x ) 2

i 1
s2 
N 1

COEFFICIENT OF VARIANCE (CV)


(or RELATIVE STANDARD DEVIATION):
Divide the standard deviation by the mean value and express as a percentage:

s
CV  ( )  100%
x
Use of Statistics in Data
Evaluation
How can we relate the observed mean value ( x ) to the true mean (m)?
The latter can never be known exactly.

The range of uncertainty depends how closely s corresponds to s.

We can calculate the limits (above and below) around x that m must lie,
with a given degree of probability.
Define some terms:

CONFIDENCE LIMITS
interval around the mean that probably contains m.

CONFIDENCE INTERVAL
the magnitude of the confidence limits

CONFIDENCE LEVEL
fixes the level of probability that the mean is within the confidence limits

Examples later. First assume that the known s is a good


approximation to s.
Percentages of area under Gaussian curves between certain limits of z (= x - m/s)

50% of area lies between 0.67s


80% “ 1.29s
90% “ 1.64s
95% “ 1.96s
99% “ 2.58s

What this means, for example, is that 80 times out of 100 the true mean will lie
between 1.29s of any measurement we make.

Thus, at a confidence level of 80%, the confidence limits are 1.29s.

For a single measurement: CL for m = x  zs (values of z on next overhead)

For the sample mean of N measurements ( x ), the equivalent expression is:


CL for m  x  zs
N
Values of z for determining
Confidence Limits

Confidence level, % z

50 0.67
68 1.0
80 1.29
90 1.64
95 1.96
96 2.00
99 2.58
99.7 3.00
99.9 3.29

Note: these figures assume that an excellent approximation


to the real standard deviation is known.
Confidence limits of the mean for large samples
If we assume that this distribution is normal then 95% of the sample means will lie in
the range given by:
The sampling distribution of the mean,
showing the range within which 95% of
sample means lie.
Values of t for confidence intervals

A more complete version of this table is given in Table A.2 in Appendix 2.


Values of t for various levels of probability

Degrees of freedom 80% 90% 95% 99%


(N-1)
1 3.08 6.31 12.7 63.7
2 1.89 2.92 4.30 9.92
3 1.64 2.35 3.18 5.84
4 1.53 2.13 2.78 4.60
5 1.48 2.02 2.57 4.03
6 1.44 1.94 2.45 3.71
7 1.42 1.90 2.36 3.50
8 1.40 1.86 2.31 3.36
9 1.38 1.83 2.26 3.25
19 1.33 1.73 2.10 2.88
59 1.30 1.67 2.00 2.66
 1.29 1.64 1.96 2.58

Note: (1) As (N-1)  , so t  z


(2) For all values of (N-1) < , t > z, I.e. greater uncertainty
Confidence Limits when s is known
Atomic absorption analysis for copper concentration in aircraft engine oil gave a value
of 8.53 mg Cu/ml. Pooled results of many analyses showed s  s = 0.32 mg Cu/ml.
Calculate 90% and 99% confidence limits if the above result were based on (a) 1, (b) 4,
(c) 16 measurements.
(a) (b)

(164
. )(0.32) (164
. )(0.32)
90% CL  8.53   8.53  0.52 mg / ml 90% CL  8.53   8.53  0.26mg / ml
1 4
i.e. 8.5  0.5mg / ml i.e. 8.5  0.3mg / ml
(2.58)(0.32)
99% CL  8.53 
(2.58)(0.32)
 8.53  0.83mg / ml 99% CL  8.53   8.53  0.41mg / ml
1 4
i.e. 8.5  0.8mg / ml i.e. 8.5  0.4 mg / ml

(164
. )(0.32)
90% CL  8.53  . mg / ml
 8.53  013
16
. mg / ml
i.e. 8.5  01
(c)
(2.58)(0.32)
99% CL  8.53   8.53  0.21mg / ml
16
i.e. 8.5  0.2 mg / ml
If we have no information on s, and only have a value for s -
the confidence interval is larger,
i.e. there is a greater uncertainty.
Instead of z, it is necessary to use the parameter t, defined as follows:

t = (x - m)/s

i.e. just like z, but using s instead of s.

By analogy we have: CL for m  x  ts


N
(where x = sample mean for N measurements)

The calculated values of t are given on the next overhead


Example
The sodium ion content of a urine specimen was determined by using an
ion-selective electrode. The following values were obtained: 102, 97, 99, 98,
101, 106 mM. What are the 95% and 99% confidence limits for the sodium
ion concentration?
The mean and standard deviation of these values are 100.5 mM and 3.27 mM
respectively. There are six measurements and therefore 5 degrees of freedom.
From Table A.2 the value of t5 for calculating the 95% confidence limits is 2.57
and from equation (2.9) the 95% confidence limits of the mean are given by:
Confidence Limits where s is not known
Analysis of an insecticide gave the following values for % of the chemical lindane:
7.47, 6.98, 7.27. Calculate the CL for the mean value at the 90% confidence level.

xi% xi2
7.47 55.8009
6.98 48.7204
7.27 52.8529

Sxi = 21.72 Sxi2 = 157.3742


x
 x

i 2172
.
 7.24
N 3
(  xi ) 2
x  N
2
i 157.3742 
. )2
(2172
3 90% CL  x  ts  7.24 
(2.92)(0.25)
s  N 3
N 1 2
 0.246  0.25%  7.24  0.42%

90% CL  x  zs
(164
. )(0.28)
If repeated analyses showed that s s = 0.28%:  7.24 
N 3
 7.24  0.27%
Testing a Hypothesis

Carry out measurements on an accurately known standard.

Experimental value is different from the true value.

Is the difference due to a systematic error (bias) in the method - or simply to random error?

Assume that there is no bias


(NULL HYPOTHESIS),
and calculate the probability
that the experimental error
is due to random errors.

Figure shows (A) the curve for


the true value (mA = mt) and
(B) the experimental curve (mB)
Bias = mB- mA = mB - xt.

Test for bias by comparing x  xt with the


difference caused by random error

Remember confidence limit for m (assumed to be xt, i.e. assume no bias)


is given by:
ts
CL for m  x 
N
 at desired confidence level, random
errors can lead to:
ts
x  xt  
N
ts
 if x  xt  , then at the desired
N
confidence level bias (systematic error)
is likely (and vice versa).
Detection of Systematic Error (Bias)
A standard material known to contain
38.9% Hg was analysed by x  37.8%  x  xt  11%
.
atomic absorption spectroscopy.  xi  113.4  xi2  4208.30
The results were 38.9%, 37.4%
4208.30  (113.4) 2 3
and 37.1%. At the 95% confidence level, s   0.943%
2
is there any evidence for
a systematic error in the method?

Assume null hypothesis (no bias). Only reject this if

x  xt   ts N
But t (from Table) = 4.30, s (calc. above) = 0.943% and N = 3

ts N  4.30  0.943 3  2.342%


 x  xt   ts N

Therefore the null hypothesis is maintained, and there is no


evidence for systematic error at the 95% confidence level.
Are two sets of measurements significantly different?

Suppose two samples are analysed under identical conditions.


Sample 1  x1 from N1 replicate analyses
Sample 2  x2 from N 2 replicate analyses

Are these significantly different?


Using definition of pooled standard deviation, the equation on the last
overhead can be re-arranged:

N1  N 2
x1  x2  ts pooled
N1 N 2
Only if the difference between the two samples is greater than the term on
the right-hand side can we assume a real difference between the samples.
Test for significant difference between two sets of data

Two different methods for the analysis of boron in plant samples


gave the following results (mg/g):
(spectrophotometry)
(fluorimetry)
Each based on 5 replicate measurements.
At the 99% confidence level, are the mean values significantly
different?
Calculate spooled = 0.267. There are 8 degrees of freedom,
therefore (Table) t = 3.36 (99% level).
Level for rejecting null hypothesis is

 ts N1  N2 N1 N2 - i.e.  (3.36)(0.267) 10 25
i.e. ± 0.5674, or ±0.57 mg/g.
But x1  x2  28.0  26.25  1.75mg / g
i.e. x1  x2  ts pooled N1  N 2 N1 N 2

Therefore, at this confidence level, there is a significant


difference, and there must be a systematic error in at least
one of the methods of analysis.
Example
Other uses of confidence limits
Significance tests
Comparison of an experimental mean with a known value

In order to decide whether the difference between and μ is significant, that


is to test H0: population mean = μ, the statistic t is calculated:

where x = sample mean, s = sample standard deviation and n = sample size.


Example
In a new method for determining selenourea in water, the following values
were obtained for tap water samples spiked with 50 ng ml−1 of selenourea:

50.4, 50.7, 49.1, 49.0, 51.1 ng ml−1


Is there any evidence of systematic error?
The mean of these values is 50.06 and the standard deviation is 0.956.
Adopting the null hypothesis that there is no systematic error, i.e. μ = 50, and using
equation

From Table A.2, the critical value is t4 = 2.78 (P = 0.05). Since the observed value of is less
than the critical value the null hypothesis is retained: there is no evidence of systematic
error. Note again that this does not mean that there are no systematic errors, only that they
have not been demonstrated.
Comparison of two experimental means
In order to decide whether the difference between two sample means and is
significant, that is to test the null hypothesis, H0: μ1 = μ2, the statistic t is
calculated:

where s is calculated from:

and t has n1 + n2 − 2 degrees of freedom.


This method assumes that the samples are drawn from populations with equal standard
deviations.
Example
In a comparison of two methods for the determination of chromium in rye
grass, the following results (mg kg−1 Cr) were obtained:
Method 1: mean = 1.48; standard deviation 0.28
Method 2: mean = 2.33; standard deviation 0.31
For each method five determinations were made.
Do these two methods give results having means which differ significantly?

There are 8 degrees of freedom, so (Table A.2) the critical value t8 = 2.31 (P = 0.05): since the
experimental value of is greater than this the difference between the two results is significant
at the 5% level and the null hypothesis is rejected. In fact since the critical value of t for P =
0.01 is about 3.36, the difference is significant at the 1% level. In other words, if the null
hypothesis is true the probability of such a large difference arising by chance is less than 1 in
100.
Example

Does the mean amount of tin found differ significantly for the two boiling times?
The mean and variance (square of the standard deviation) for the two times are:

There are 10 degrees of freedom so the critical value is t10 = 2.23 (P = 0.05). The
observed value of t (= 0.88) is less than the critical value so the null hypothesis
is retained: there is no evidence that the length of boiling time affects the recovery rate.
In order to test H0: μ1 = μ2 when it cannot be assumed that the two samples
come from populations with equal standard deviations, the statistic t is calculated,
where
Example

Substitution in equation (3.4) gives t = −8.48 and substitution in equation (3.5)


gives 5.3, which is truncated to 5. The critical value is t5 = 4.03 (P = 0.01) so the
null hypothesis is rejected: there is sufficient evidence to say that the mean
concentration of thiol differs between the groups.
Paired t -test

+1.48, +0.66, +0.24, +0.21, −0.10, −0.61, −0.10, +0.09, −0.07, −0.21
Paired t -test
One-sided and two-sided tests
It is suspected that an acid–base titrimetric method has a significant indicator
error and thus tends to give results with a positive systematic error (i.e. positive
bias). To test this an exactly 0.1 M solution of acid is used to titrate 25.00 ml of
an exactly 0.1 M solution of alkali, with the following results (ml):
25.06 25.18 24.87 25.51 25.34 25.41

mean = 25.228 ml, standard deviation = 0.238 ml

From Table A.2 the critical value is t5 = 2.02 (P = 0.05, one-sided test). Since the
observed value of t is greater than this, the null hypothesis is rejected and there
is evidence for positive bias.
It is interesting to note that if a two-sided test had been made in the example above
(for which the critical value for t5 = 2.57) the null hypothesis would not have been
rejected!
F-test for the comparison of
standard deviations
Example

For each method eight determinations were made.

The critical value is F7,7 = 3.787 (P = 0.05), where the subscripts indicate the
degrees of freedom of the numerator and denominator respectively. Since the
calculated value of F (4.8) exceeds this, the variance of the standard method is
significantly greater than that of the proposed method at the 5% probability
level, i.e. the proposed method is more precise.
Example
In a comparison of two methods for the determination of chromium in
rye grass, the following results (mg kg−1 Cr) were obtained:
Method 1: mean = 1.48; standard deviation 0.28
Method 2: mean = 2.33; standard deviation 0.31
For each method five determinations were made.

From Table A4, taking the number of degrees of freedom of both


numerator and denominator as 4, the critical value is F4,4 = 9.605. The
calculated value is less than this, so there is no significant difference
between the two variances at the 5% level.
Outliers
For example, if the following results were given for a titration:

Grubbs’ test for an outlier

where and s are calculated with the suspect value included

The critical values for G for P = 0.05 are given in Table A.5.
Example
Dixon’s test for an outlier

The critical values of Q for P = 0.05 for a two-sided test are given in Table A.6. If the
calculated value of Q exceeds the critical value, the suspect value is rejected.
Detection of Gross Errors

A set of results may contain an outlying result


- out of line with the others.
Should it be retained or rejected?
There is no universal criterion for deciding this.
One rule that can give guidance is the Q test.

Consider a set of results

The parameter Qexp is defined as follows:

Qexp  x q  x n /w

where xq = questionable result


xn = nearest neighbour
 w = spread of entire set
Qexp is then compared to a set of values Qcrit:

Qcrit (reject if Qexpt > Qcrit)

No. of observations 90% 95% 99% confidencelevel

3 0.941 0.970 0.994


4 0.765 0.829 0.926
5 0.642 0.710 0.821
6 0.560 0.625 0.740
7 0.507 0.568 0.680
8 0.468 0.526 0.634
9 0.437 0.493 0.598
10 0.412 0.466 0.568
Rejection of outlier recommended if Qexp > Qcrit for the desired confidence level.

Note:1. The higher the confidence level, the less likely is


rejection to be recommended.
2. Rejection of outliers can have a marked effect on mean
and standard deviation, esp. when there are only a few
data points. Always try to obtain more data.
3. If outliers are to be retained, it is often better to report
the median value rather than the mean.
The following values were obtained for
Q Test for Rejection the concentration of nitrite ions in a sample
of Outliers of river water: 0.403, 0.410, 0.401, 0.380 mg/l.
Should the last reading be rejected?
Qexp  0.380  0.401 ( 0.410  0.380)  0.7
But Qcrit = 0.829 (at 95% level) for 4 values
Therefore, Qexp < Qcrit, and we cannot reject the suspect value.
Suppose 3 further measurements taken, giving total values of:
0.403, 0.410, 0.401, 0.380, 0.400, 0.413, 0.411 mg/l. Should
0.380 still be retained?

Qexp  0.380  0.400 ( 0.413  0.380)  0.606


But Qcrit = 0.568 (at 95% level) for 7 values
Therefore, Qexp > Qcrit, and rejection of 0.380 is recommended.

But note that 5 times in 100 it will be wrong to reject this suspect value!
Also note that if 0.380 is retained, s = 0.011 mg/l, but if it is rejected,
s = 0.0056 mg/l, i.e. precision appears to be twice as good, just by
rejecting one value.
Obtaining a representative sample

Homogeneous gaseous or liquid sample


No problem – any sample representative.
Solid sample - no gross heterogeneity
Take a number of small samples at random from throughout the bulk - this will
give a suitable representative sample.

Solid sample - obvious heterogeneity


Take small samples from each homogeneous region and
mix these in the same proportions as between each
region and the whole.

If it is suspected, but not certain, that a bulk material is heterogeneous, then


it is necessary to grind the sample to a fine powder, and mix this very
thoroughly before taking random samples from the bulk.

For a very large sample - a train-load of metal ore, or soil in a field - it is always
necessary to take a large number of random samples from throughout the whole.
Analysis of variance
Comparison of several means
Dot-plot of results in Table 3.2.
Generalization of Table 3.2
Within-sample variation
Between-sample variation
Summarizing our calculations
Table A.3 the critical value of F is 4.066 (P = 0.05).
Since the calculated value of F is greater than this the
null hypothesis is rejected: the sample means do differ
significantly.
A simple way of deciding the reason for a significant result is to arrange the means
in increasing order and compare the difference between adjacent values with a
quantity called the least significant difference. This is given by:
The general formula
The arithmetic of ANOVA
calculations
One-way ANOVA tests for a significant difference between means when
there are more than two samples involved. The formulae used are:
The chi-squared test
Testing for normality of distribution
109, 89, 99, 99, 107, 111, 86, 74, 115, 107, 134, 113, 110, 88, 104
the Kolmogorov–Smirnov method
Calibration methods:
regression and correlation
This general procedure raises several
important statistical questions:
1. Is the calibration graph linear? If it is a curve, what is the form of the
curve?
2. Bearing in mind that each of the points on the calibration graph is subject
to errors, what is the best straight line (or curve) through these points?

3. Assuming that the calibration plot is actually linear, what are the errors
and confidence limits for the slope and the intercept of the line?
4. When the calibration plot is used for the analysis of a test material, what
are the errors and confidence limits for the determined concentration?
5. What is the limit of detection of the method? That is, what is the least
concentration of the analyte that can be detected with a predetermined level
of confidence?
The product–moment correlation
coefficient

r can take values in the range -1 ≤ r ≤ +1


Example 5.3.1
Calibration plot for the data in Example
y = 1.93x + 1.52

The calculated value of t is compared with


the tabulated value at the desired
significance level, using a two-sided
t-test and (n - 2) degrees of freedom.
The line of regression of y on x
Errors in the slope and intercept of the
regression line
The y-residuals of a regression line
Calculate the standard deviations and confidence limits of the
slope and intercept of the regression line
Calculation of a concentration
and its random error

n = jumlah pengukuran
m = jumlah perulangan
Example 5.6.1
General form of the confidence limits for a
concentration determined by using an
unweighted regression line
Limits of detection
y = 1.93x + 1.52

2,82 = 1,92x + 1,52

X = (2,82 – 1,52)/1,92

X = 0,67
The method of standard additions
Example 5.8.1
Use of regression lines for comparing
analytical methods

Figure 5.10 Use of a regression line to compare two analytical methods: (a) shows
perfect agreement between the two methods for all the samples; (b)–(f) illustrate the
results of various types of systematic error (see text).
Perbandingan Metode CF dan EP
Comparison of two analytical methods:
data from Example 5.9.1

Anda mungkin juga menyukai