There are many way to determine and correct systematic error which are;

1. Composition sample analysis: on this way, the method used should provide a known answers, if the method used is not able to provide that, then the error could be trace to the type of method or instrument used in carrying out the analysis.
2. Analyze blank sample: blank sample (sample containing none of the parameter) may use to carry out the analysis, if no trace of error found using the blank sample for the analysis, it simply mean that there is a problem with the method used in the analysis.
3. Different analytical method should be used to measure the same quantity: when different is used, yet you cannot get the same result, then there is a problem with the method used for the analysis.
4. Consult different operator in different laboratory to carry the same measurement using the same, of different method. If disagreement or high variation detected, then the error should be trace to the operator or the equipment used for the analysis.

Random ErrorError in in measuring Analytical chemistry

This type of error is also called indeterminate errors. They are error that occur as result of shortage of physical measurement, this type of error in unavoidable. But good experiment or repeated experiment can limit the size of the errors, but cannot be totally eradicated. In another way these type of error can also be called accidental errors. The error are determine via little variation in successive measurement carried out by the same analyzer under similar experimental conditions. Random error cannot be forecasted or predicted. The error could either be addition  of figure from the true value or reduction of figure from the true value, when is addition of figure to the true value, is said to positive random error, if it is a reduction of figure from the positive value, is said to be negative random error.

Example

1. The type of change associated with the same analyzer reading the same scales many times.
2. Change associated with three of four different analyzer reading the same measurement scales or reading the minimum measurement of a volumetric flask.

Unfortunately, they will report changing value showing subjective interpolation between markings.

It has been examined that these type of errors frequently go through random distribution; as such probability law of mathematics may help us to arrive at a conclusions regarding the most probable in a series of measurement result.

Point to note; the error that affect the precision and accuracy of a measure value, thereby raise a questions, on the accuracy of the value reported.

Expressing Accuracy of a Measurement

There are lot of way to express the accuracy of measurement of analytical result which include;

1. Absolute Error or Absolute Uncertainty

Variation of different seen between true value and measure value, is reported in the measurement units.

Example: if a 4.99mg of sample is analyzed as 4.94mg, the absolute error is 0.05, it become a mean error when measure value is the average measurement of the several value.

2. Relative Error or Relative Uncertainty

This is an expression that compare the absolute uncertainty to the size of its associated measurement or expressing the true value as the percentage of absolute errors.

Example

On the above analysis of relative which is

0.05/4.99 x 100%/1 = 1.01

Now the relative accuracy can be deducted as follow;

4.94/4.99 x 100%/1 = 98.99

NB:  relative accuracy and relative error must give 100% when summed together.

Relative error shown above may be express as pph (part per hundred) or ppt (part per thousand)

Example

If we were given the result of analysis, to be 34.75μm when compare to true value of 35.15μm. Calculate the relative error in part per hundred and part per thousand?

Solution

Absolute error = 34.75 – 35.15 = -0.4

Relative error in pph = -0.4/34.75  x 100/1 = -1.15pph

Relative error in ppt = -0.4/35.15 x 1000/1 = -11.37