Control Charting in

Chemical Analysis
Labs
PREPARED BY: EMAD EDDADU
LAB QUALITY EXPERT
 ‫عماد عبد اللطيف الددَ‬
‫ميميائي‪ ،‬خبير جُدي جحاليل‬
‫مُلُد في مديىة الزرقاء في األردن عام ‪1960‬‬ ‫‪‬‬

‫بنالُريُس ميمياء مه جامعة اليرمُك (إربد ‪ /‬األردن) عام ‪1983‬‬ ‫‪‬‬

‫بروامج الزمالً لمىظمة حظر األسلحً النيميائيً (الٌاي ‪ٌُ /‬لىدا) عام ‪2008‬‬ ‫‪‬‬

‫محقاعد حاليا ً َاعمل لىفسي‬ ‫‪‬‬

‫عملث في المرمز العربي للدراسات الٍىدسيً (فرعي الرياض َجدي في السعُديً) ‪2017-2016‬‬ ‫‪‬‬

‫عملث في الجمعيً العلميً الملنيً في عمان ‪2014-1987‬‬ ‫‪‬‬

‫عملث في الشرمة األردويً لإلسحثمارات الصىاعيً في مديىة الملل عبد هللا الثاوي الصىاعيً ‪1987-1985‬‬ ‫‪‬‬

‫عملث في القيادي العامً للقُات المسلحً األردويً ‪1985-1983‬‬ ‫‪‬‬

ISO/IEC 17025: 2017, 7.7
Ensuring validity of results

The laboratory shall have a procedure for monitoring the validity

of results. The resulting data shall be recorded in such a way
that trends are detectable and, where practicable, statistical
techniques shall be applied to review the results. This
monitoring shall be planned and reviewed and shall include,
where appropriate, but not limited to:
…ISO/IEC 17025: 2017, 7.7
Ensuring validity of results

A. Use of reference materials or quality control materials;

B. Use of alternative instrumentation that has been
calibrated to provide traceable results;
C. Functional check(s) of measuring and testing
equipment;
D. Use of check or working standards with control charts,
where applicable;
E. Intermediate checks on measuring equipment;
…ISO/IEC 17025: 2017, 7.7
Ensuring validity of results
F. Replicate tests or calibrations using the same or
different methods;
G. Retesting or recalibration of retained items;
H. Correlation of results for different characteristics of
an item;
I. Review of reported results;
J. Interlaboratory comparisons;
K. Testing of blind sample(s).
References:
 NORDTEST REPORT TR 569, INTERNAL
QUALITY CONTROL, Nordic Innovation
Center. http://www.nordtest.org
 Dux, J. P., “Handbook of Quality
Assurance for the Analytical Chemistry
Laboratory”, Second Edition, 1991.
 “Standard Methods for the Examination
of Water and Wastewater” online,
APHA, AWWA and WEF,
www.standardmethods.org
 Youden, W. J., Steiner, E. H., “Statistical
Manual of the AOAC”, AOAC, 1975.
 It has been shown that
laboratories with effective
Validation
IQC systems performed
better in proficiency
testing schemes

Proficiency
Accreditation
Testing
Quality

Internal Certified
Quality Reference
Control Materials
Types of errors in chemical analysis:

Systematic

Random

Blunders
Systematic Errors

 Alwaysin the same direction (either positive or

negative)
 Either
constant (blank value in colorimetric
analysis), or relative (extraction recovery)
 Canbe confirmed by validation tools, such as
analysis of CRMs or analysis of real samples
using standard reference methods
Sources of Systematic Errors:

 Instrumental errors

 Method errors

 Personal errors
Precision

Is the closeness of agreement between the

repeated observations.
It is usually expressed as a standard deviation
Repeatability (sr)

Measurement precision under a set of

repeatability conditions of measurement.
Repeatability condition of measurement refers
to measurements being made on the same
material by a single analyst, using the same
procedure, under the same operating
conditions over a short time period. The whole
procedure should be repeated from taking a
new test portion of a sample to the final reading
or calculation of result.
Within-laboratory reproducibility
(Intermediate precision) sb or sRw
The degree of agreement between individual
results determined in a laboratory on a sample
with the same measurement procedure over a
long-time period, i.e. at least a year. The time
period could be shorter if enough data are
collected but in many cases a year is suitable to
encompass all variations in reagents, personnel,
instrument service, etc. Also called intermediate
precision.
Reproducibility (sR)

Measurement precision under a set of

reproducibility conditions of measurement.
Reproducibility conditions of measurement refers
to measurements being made on the same
material using the same procedure but by
different analysts working in different locations..
 Different types of precision estimates
1 Repeated injection of Injector precision
the same solution to
Reproducibility
the analytical
instrument
2 Repeat determination Repeatability
Intermediate
of a complete method (short term precision) precision
by one analyst on the
same day
3 Repeat determination Long term precision Repeatability
of a complete method (intermediate precision)
by several analysts
over several days or
weeks in the same
laboratory Injector
precision
4 Repeat determination Reproducibility
of a complete method
in different
laboratories
Trueness

Is the closeness of agreement between

the average value obtained from a
large set of test results and an
accepted reference value.
What Is Bias?

Bias()  x  
A visual illustration of analytical bias
Accuracy

Is the closeness of agreement between a test

result and the accepted reference value .
i.e. it’s a qualitative concept involves a
combination of random component (precision)
and a common systematic error or bias
component.
Precise or biased?
Systematic Bias
error Trueness

Total error Accuracy Uncertainty

Any spread
Random Precision
measure
error
IQC:

Internal quality control: set of procedures

undertaken by laboratory staff for the continues
monitoring of operation and results of
measurements in order to decide whether
results are reliable enough to be released
Control material:

 Control material: material used for the

purposes of internal quality control and
subjected to the same or part of the same
measurement procedure as that used for the
test materials.
(ISO Guide 80:2014, Guidance for the in-house
preparation of quality control materials (QCMs))
Reference Material (RM)

material, sufficiently homogeneous and stable

with respect to one or more specified properties,
which has been established to be fit for its
intended use in a measurement process
(ISO/Guide 30:2015)
Certified Reference Material (CRM)

reference material (RM) characterized by a

metrologically valid procedure for one or more
specified properties, accompanied by an RM
certificate that provides the value of the
specified property, its associated uncertainty,
and a statement of metrological traceability
(ISO/Guide 30:2015)
Average

𝑥𝑖
𝑋=
𝑛
‫االنحراف المعياري‬
(Standard Deviation)

(X i X )
2

S  i 1

n 1
n  1 : degrees of freedom
Variance (V)

V s
2

(X i X )
2
n
V 
i 1 n 1
Relative Standard Deviation
(RSD)

S
RSD 
X
Coefficient of Variation (CV)

S
CV  100  
X
Normal distribution
properties
 What is a control chart ?

A statistical tool for the analysis of data

obtained during a continuing process.
Process Control

Has two main features:

 The detection of a change in the performance
of the process
 Identification
of the causes of the change and
appropriate corrective action
Control charting in Chemical Analysis

In Analytical Chemistry Control Charts

are used to monitor both:
Trueness
Precision
of analytical methods
Analytical Control Program

 Type of QC sample
 Type of QC chart
 Control limits (warning and control)
 Control frequency
Types of control samples

 Matrix certified reference material

 Standard solutions
 In-house reference material
 Routine sample
 Blank
CRMs as control materials

Use of them is subject to a number of constraints:

 No closely matching CRMs available
 Cost
 Not applicable if either the analyte or the matrix is
unstable
 Not available in sufficient amount
 Not all of CRMs are of equal quality
When identifying a suitable source of material the
following issues should be considered:

 Matrix of interest
 Properties/analytes of interest
 Appropriate concentration levels or physical values
 Intended end use
 Degree of homogeneity required
 Degree of stability required (shelf life)
 Quantity of material required
 Suitability of materials already available
 Ease of sourcing and sampling
 Available storage facilities
 Health and safety implications
Kinds of control charts used
in chemical analysis:

 Charts of single value for a characterized

control material
1. Mean (X) charts
2. x charts
 Chart of duplicate analysis of real samples
(range-R- chart)
 Chart of spiking recovery
 Chart of blank determination
Other types

 Moving average charts

 Cumulative sum (CUSUM) charts
 J charts
Practical points in setting up the QC

 Concentration range
 Range chart with test samples
 Frequency of control analysis
 Position of control samples in an
analytical run
 A good balance between QC and test
samples
Design of x chart
Calculation of the five lines:
UCL  X  3S
UWL  X  2 S
CL  X
LWL  X  2 S
LCL  X  3S
X control chart
Uncertainty associated with estimating
standard deviation
D𝐞𝐬𝐢𝐠𝐧 𝐨𝐟 𝑿 𝑪𝒐𝒏𝒕𝒓𝒐𝒍 𝑪𝒉𝒂𝒓𝒕

Date 𝑋𝑖1 Xi2 𝑋 𝐼1:𝑋 𝐼2

𝑋𝑖 =
2
 Calculation of the five lines for chart of duplicates X

x   x i

n
3s 2s
UCL  x  , UWL  x 
2 2
3s 2s
LCL  x  , LWL  x 
2 2
 R chart for duplicates

Date Xi1 Xi2 RX

i i1
 X i2
 ‫المدى‬R %
<R 57.5 Youden, W. J., Steiner, E. H.,
“Statistical Manual of the
1.5R-R 19.4 AOAC”, AOAC, 1975.

2.0R-1.5R 12.1
2.5R-2.0R 6.4
> 2.5R 4.6
Total 100
Nordtest R-chart or r%-chart

 For duplicates:
Central line is the mean range
𝑚𝑒𝑎𝑛 𝑟𝑎𝑛𝑔𝑒
s (standard deviation) =
1.128
warning limit = 2.83 s = 2.51 mean range
action limit = 3.69 s = 3.27 mean range
Nordtest example 1: determination
of Ni in steel alloy with XRF
Sample type Control chart Control limits Central line
Steel routine sample X-chart Target Mean value

Mean value = 4.58 % (abs, (mean of a control sample over

a year)and standard deviation = 0.026% (abs)
s target= 0.0458 % (abs)(half of standard uncertainty (2%
relative))
CL = 4.58 % (abs)
WL = 4.58 ± 2 ● 0.0458 = 4.67 and 4.49 % (abs)
AL = 4.58 ± 3 ● 0.0458 = 4.72 and 4.44 % (abs)
Nordtest example 2: determination
of Co in low-alloy steel with XRF
Sample type Control chart Control limits Central line
Steel routine sample X-chart Target Mean value

Mean value = 0.0768 % (abs) (mean of a control

sample over a year)and standard deviation =
0.00063% (abs)
s target= 0.001 % (abs)(LOQ required is 0.01% abs)
CL = 0.0768 % (abs)
WL = 0.0768 ± 2 ● 0.001 = 0.0788 and 0.0748 %
(abs)
Nordtest example 3: determination of
N-NH4 in water with indophenol blue
method
Sample type
Standard solution
Control chart
X-chart
Control limits
Statistical
Central line
Mean value
R-chart Statistical Mean range value

Mean = 19.99 µg/l, s =0.521µg/l

CL = 19.99 µg/l
WL = 19.99 ± 2 ● 0.521 = 21.03 and 18.95 µg/l
AL = 19.99 ± 3 ● 0.521 = 21.55 and 18.43 µg/l
Mean range = 0.559 µg/l, s =0.559/1.128 = 0.496 µg/l
CL = 0.559 µg/l
WL = 2.83 ● 0.496 = 1.40 µg/l
AL = 3.69 ● 0.496 = 1.83 µg/l
Nordtest example 4: determination
of Pb in water with ICP-MS
Sample type Control chart Control Central line
limits
In-house lake water X-chart Statistical Mean value

Mean = 0.294 µg/l, s =0.008 µg/l

CL = 0.294 µg/l
WL = 0.294 ± 2 ● 0.008 = 0.310 and 0.278 µg/l
AL = 0.294 ± 3 ● 0.008 = 0.318 and 0.270 µg/l
Nordtest example 5: determination of
As in biological material with ICP-MS
Sample type Control chart Control Central line
limits
CRM X-chart Target Certified value

Certified value = 18.0 µg/l, s target = 0.05 ● 18.0 = 0.9 µg/l

CL = 18.0 µg/l
WL = 18.0 ± 2 ● 0.9 = 19.9 and 16.2 µg/l
AL = 18.0 ± 3 ● 0.9 = 20.7 and 15.3µg/l
 Nordtest example 6: determination
of phosphorus in water using
spectrophotometric method
Sample type Control chart Control limits Central line
Routine r%-chart Statistical Mean relative range
samples

Mean range = 1.88 %, s = 1.88 / 1.128 = 1.67 %

CL = 1.88 %
WL = 2.83 ● 1.67 % = 4.73 %
AL = 3.69 ● 1.67 % = 6.16 %
 Nordtest example 7: determination
of b-HCH in biological material with
Gas Chromatography
Sample type Control chart Control limits Central line
CRM X-chart Target Reference value

Certified value = 16.0 µg/l, s target = 0.15 ● 16.0 = 2.4 µg/l

CL = 16.0 µg/l
WL = 16.0 ± 2 ● 2.4 = 20.8 and 11.2 µg/l
AL = 16.0 ± 3 ● 2.4 = 23.2 and 8.8 µg/l
 Nordtest example 9: determination
of Zn in hydrogen peroxide with ICP
method-blank sample
Sample type Control chart Control limits Central line
Blank sample X-chart Statistical Mean

Mean= 0.039 mg/l, s = 0.045 mg/l

CL = 0.039 mg/l
WL = 0.039 ± 2 ● 0.045 = 0.129 and – 0.051 mg/l
AL = 0.039 ± 3 ● 0.045 = 0.174 and – 0.096 mg/l
Control Chart of
Spiking Recovery

 A %=100 . {S.(V+W)-U.V.} / C.W

 C: concentration of the spiking solution
 V: volume of the sample
 W: volume of the spike
 S:measured concentration in the spiked
sample
 U:measured concentration in the unspiked
sample
Instructions for daily
Evaluation
of
Control charts
Is the method within statistical control?

Can we report results?

Control charting rules

Control charting rules chosen should

provide a high rate of error detection
with a low rate of false rejection
Western electrical rules for evaluating control charts

1 point beyond 3 SD
9 consecutive points above or below central line
6 consecutive increasing or decreasing points
 14 consecutive alternating point
2 of 3 consecutive points beyond 2 SD
4 of 5 points beyond 1 SD
 15 consecutive points within 1 SD
9 consecutive points beyond 1 SD
Stop!
1 point beyond 3 SD
Stop!
2 of 3 consecutive points
beyond 2 SD
Stop!
4 of 5 points beyond 1 SD
Stop!
6 consecutive increasing points
Stop!
6 consecutive decreasing points
Seven successive samples are on one
side of the mean:
How does this control charts looks?
What about this?
So:

 Method is in control
 Method is in control but can be regarded
as “out of statistical control”
 Method is out of control
Long term evaluation of control
charts
 Review of current quality
 How often should control limits be
evaluated?
 What makes a change in control limits
necessary?
 Review of spread and bias
Significance
Testing
F-test
s 2
F  a
s 2
b
Va
F 
Vb
t-test
X1  X 2 N N
t  1 2

s p N N
1 2

N 1
: number of readings of group 1

N 2
: number of readings of group 2
X 1 : mean of group 1
X 2 : mean of group 2
s p
: pooled standard deviation
 Pooled Standard Deviation (sp)

N1 N2 Nk

 ( X i1 X 1)   (X i 2  X 2)  .....   ( X ik  X k )
2 2 2

sp  i 1 i 1 i 1

N-k
N  N 1  N 2  .....  Nk
N  k : Degrees of freedom
degrees of freedom(v)  n1  n2  2
s (n1  1)  s (n2  1)
2 2
sp  1 2
(n1  n2  2)
 Exercise
Update the following control chart based on the following data:
Group Standard UCL UWL Mean LWL LCL
deviation
First 20 0.28 50.5
points
Second 20 0.26 50.4
points
Notes: There is no "out of control points" according to control rules specified
Unit is mg/L
Significant difference in spread of data?

Significant difference in means of the two

sets?
 0.282
 F= = 1.16 , F critical two tailed, 95 % con𝑓𝑖𝑑𝑒𝑛𝑐𝑒 ≅ 2.5
0.262

(0.262 ∗19):(0.282 ∗19)

 sp = = 0.27
20:20;2

50.5 ;50.4 20 ∗20

 t=± ● = 1.17, t critical (two tailed, 95 %) =2.024
0.27 20:20
Combine two sets?

Group Standard UCL UWL Mean LWL LCL

deviation
1 0.28 51.34 51.06 50.5 49.94 49.66
2 0.26 51.18 50.92 50.4 49.88 49.62
1+2 0.27 51.26 50.99 50.45 49.91 49.64
Nordtest example 8: determination
of Cu in water using ICP-OES
Sample type Control chart Control Central line
limits
In-house synthetic X-chart Statistical Mean value
standard R-chart Statistical Mean range
value

 . Mean = 1.054 mg/l, s =0.0667 mg/l, CL = 1.05 mg/l

WL = 1.05 ± 2 ● 0.0667 = 0.92 and 1.19 mg/l
AL = 1.05 ± 3 ● 0.0667 = 0.85 and 1.255 mg/l
 Mean range = 0.11 mg/l, s =0.11/1.128 = 0.0975 mg/l
CL = 0.11 mg/l
WL = 2.83 ● 0.0975= 0.28 mg/l
AL = 3.69 ● 0.0975 = 0.36 mg/l
Nordtest example 8: determination
of Cu in water using ICP-OES
Sample type Control chart Control Central line
limits
In-house synthetic X-chart Statistical Mean value
standard R-chart Statistical Mean range
value
0.08342
 F= = 1.563 , F critical two tailed, 95 % con𝑓𝑖𝑑𝑒𝑛𝑐𝑒 = 1.67
0.06672
(0.06672 ∗59):(0.0834 2 ∗58)
 sp = = 0.07545
60:59;2
1.055 ;1.041 60∗59
 t=± ● = 1.012, t critical (two tailed, 95 %) =1.98
0.07545 60:59
0.097522
 For R-chart, F= =1.037, F cri two tailed, 95 % con𝑓
0.09572
= 1.671.67
Other uses of quality control data and control charts

 Measurement uncertainty
 Method validation
 Method comparison
 Estimation of limit of detection (LOD)
 Person comparison or qualification
 Evaluation of proficiency tests
 Environmental parameters and similar checks
Other kinds of
control charts
Moving Average chart

 This chart equilibrate changes happen to the x chart

to show significant changes more clearly
 Each point is determined by averaging a preselected
number of last measurements
(here the average of last four measurements was
selected).
 The more numbers is selected the more smoothing
effect, but also with little response for significant
changes.
 Warning and control lines are divided by the square
root of the number selected for averaging
X chart with a step change
Moving average chart for the previous data
X chart with continuous upwarding trend
Moving average chart for the previous data
 J chart

If z ≥ 3 then J = 8.
If 2 ≤ z < 3 then J = 4.
If 1 ≤ z < 2 then J = 2. If z ≥ 3 then J = 8. 
If –1 < z < 1 then J = 0.
If 2 ≤ z < 3 then J = 4. 
If –2 < z ≤ –1 then J = - 2.
If –3 < z ≤ –2 then J = - 4. If 1 ≤ z < 2 then J = 2. 
If z ≤ –3 then J = - 8.
If –1 < z < 1 then J = 0. 

If –2 < z ≤ –1 then J = - 2. 

If –3 < z ≤ –2 then J = - 4. 

If z ≤ –3 then J = - 8. 

J 4 6 4 4 4 4 6 8 10 10 6 6 6 8 8 4 4 2