Scribd
Upload
39 views

Lecture 5

Lecture

Uploaded by

hossamragab1443
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
39 views

Lecture 5

Lecture

Uploaded by

hossamragab1443
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 16

Course name: Chemical and statistical measurements

Course code: Chem 1103

1
Lecture 5:
Significance test
Rejection of a result
Detection of Gross Errors

• There are times when a set of data contains an


outlying result that appears to be outside the range
that the random errors in the procedure would give.

• It is generally considered inappropriate and, in some


cases, unethical to discard data without a reason.
However, the questionable result, called an outlier,
could be the result of an undetected gross error.
Detection of Gross Errors

• Hence, it is important to develop a criterion to decide


whether to retain or reject the outlying data point.

• The Q test is generally acknowledged to be an appropriate


method for making the decision.
• An outlier is a result that is quite different from the others
in the data set and might be due to a gross error.
Significance test

• Significance test can be used to know if the difference


between the two results is significant, or whether it
can be accounted for merely by random variations.

• Significance tests are widely used in the evaluation of


experimental results.
Types of significance test

1. t-test

2. F-test

3. Q-test
Q-test

• Dixon's Q test, or simply the Q test, is used for the


identification and rejection of outliers.
• It is usually conducted for small data sets; between 3-7 items.
• It’s commonly used if data sets include one suspect
observation that’s much lower or much higher than the other
values; an outlier.
• Keeping an outlier in data affects calculations like the mean
and standard deviation, so true outliers should be removed.
Q-test

Q- test:
𝑋𝑎 −𝑋𝑏
Qt =
𝑅

xa is the suspect value.


xb is the closest value to xa
R is the range of the dataset.

Notes
• The data need to be sorted from smallest to highest.
• The calculated Q value must be compared with the
tabulated Qcritical value.
Q-test

Notes
• The data need to be sorted from smallest to highest.
• The calculated Q value must be compared with the tabulated
Qcritical value.
• The data need to be sorted from smallest to highest.
• The calculated Q value must be compared with the tabulated
Qcritical value.
• If Qcalculated > Qcritical value , the suspected value is not an
outlier.
• If Qcalculated > Qcritical value , the suspected value is an outlier.
Q-test

A table containing the critical Q values for CL 90%, 95% and 99% and N=3-10
.[D.B. Rorabacher, Anal. Chem. 63 (1991) 139]

N: number of values
Example 1
The following table represents the data for children age against weigh

13.9 ,13.1,13.5,12.9,15.1 kg

One value appears suspect. Determine if it can be ascribed to accidental error,


at the 95 % confidence level.

1- Arrange the data


12.9-13.1-13.5-13.9-15.4

2- Calculate the range (w)


W= 15.4-12.9 = 2.5

3- Calculate the difference between abnormal point and the nearest point
(a)
Xa-xb= 15.4-13.9 = 1.5

11
Example 1

𝟏.𝟓
4- calculate Q = = 0.6
𝟐.𝟓

5- Using the confidence level, then compere the calculated result(Q) with
the theoretical result(from Q table)
Q crit = 0.712

Q cal less than Q crit the abnormal data is is not an outlier at 5% level

12
Example 2
The analysis of a city drinking water for arsenic yielded
values of

5.60. 5.64,5.70, 5.69, and 5.81 ppm.

The last value appears anomalous; should it be rejected at


the 95% confidence level?

The difference between 5.81 and 5.70 is 0.11 ppm


The spread (5.81-2 5.60) is 0.21 ppm. Thus,
Q = 0.11/ 0.21 = 0.52

For five measurements, Q crit at the 95% confidence level is


0.71. Because 0.52 less than 0.71, we must retain the
outlier at the 95% confidence level.
Example 3
The following values were obtained for the nitrite concentration
(mg l-1) in a sample of river water:

0.403, 0.410, 0.401, 0.380

The last measurement is noticeably lower than the others and is


thus suspect: should it be rejected?

The difference between 0.380 and 0.401 is 0.021 ppm


The spread (0.410 - 0.380) is 0.03 ppm. Thus,
Q = 0.021/ 0.03 = 0.7

For five measurements, Q crit at the 95% confidence level is 0.829.


Because 0.7 less than 0.829, we must retain the outlier at the 95%
confidence level.
Example 4
Three further measurements were added to those given in
the previous example

0.403, 0.410, 0.401, 0.380, 0.400, 0.413, 0.408

Should the value 0.380 still be retained?

The difference between 0.380 and 0.400 is 0.02 ppm


The spread (0.413 - 0.380) is 0.033 ppm. Thus,
Q = 0.02/ 0.033 = 0.6

For five measurements, Q crit at the 95% confidence level is


0.568. Because 0.6 more than 0.568, we must reject the
outlier at the 95% confidence level.
Recommendations for Treating Outliers

There are several recommendations for the treatment of a small


set of results that contains a suspect value:

1. Reexamine carefully all data relating to the outlying result to


see if a gross error could have affected its value. This
recommendation demands a properly kept laboratory notebook
containing careful notations of all observations .
2. If possible, estimate the precision that can be reasonably
expected from the procedure to be sure that the outlying result is
questionable.
3. Repeat the analysis if sufficient sample and time are available.

You might also like