Methods of minimizing errors in analysis
Although an analyst cannot prevent random errors, the following methods can reduce systemic errors.
Calibration of apparatus - When instruments are calibrated, errors are minimized, and the original measurements are corrected as necessary.
Control determination - An experiment using a standard substance under similar experimental conditions is designed to minimize errors.
Blank determination - As a result of not using the sample, a test is conducted in identical conditions, which minimizes errors caused by impurities in the reagent.
In chemical analysis, there are several uncertainties that yield a large number of 'errors' that can be broadly categorized, namely:
1. Determinate (systemic) errors
Identifiable, avoidable errors may be accounted for in principle by determining their overall value and assigning their reasonable causes. Errors of this type include:
Personal errors - Errors related to 'personal equation' appear exclusively in personal errors, as they are not related to the prescribed procedure or methodology.
Instrumental errors - There is a risk of measuring substance concentrations with incorrect calibration and faulty instruments, such as pH meters, single pan electric balances, UV-spectrophotometers, and potentiometers.
Reagent errors - These errors are caused by the reagents themselves, such as impurities present in reagents, platinum (Pt) vaporizing at high temperatures, and foreign substances unintentionally introduced due to reactions between reagents and porcelain or glass.
Constant errors - Unless the measured amount is very large, a constant error does not differ much from measurement to measurement, but it becomes less significant as the size of the measurement increases.
Proportional errors - Proportional errors vary in absolute value with a sample size so that the relative error remains constant no matter how many samples are analyzed. The chemical is usually incorporated into a substance that interferes directly with an analytical process.
Errors in methodology - The method of sampling (incorrect) and the completeness of the reaction are often the reasons for serious measurement errors.
Additive errors - This type of error does not depend on the number of substances available in the sample.
2. Intermediate (random) errors
It is impossible to pinpoint specific well-defined reasons for indeterminate errors. Usually, these problems arise from minute variations that occur inadvertently during several successive measurements that are carried out under identical experimental conditions by the same analyst. There is a high probability of both high and low results due to these errors, which are mostly random. In other words, they are neither correctable nor erasable and thus represent the 'ultimate limitation' of the measurements. Repeating measurements of a variable, subsequently analyzing the results statistically, has the effect of 'decreasing' the importance of the results to a substantial degree.
Accuracy
As a consequence of this, the terms 'accuracy' and 'precision' are frequently used interchangeably about scientific data, but there is an important difference between the two as follows:
According to traditional thinking, an accurate measurement is very near to the true value of the amount being measured. In most cases, the accuracy of a study is inversely proportional to the error of the study, that is, the greater the accuracy, the smaller the error. An experimental value is subtracted from a true value to arrive at its value.
Precision
In terms of correlation, it may be described as the relationship between two groups of experimental results, although it does not necessarily indicate the relationship with ‘true value’. An accurate measurement is deemed to have a high degree of reproducibility, while a high degree of precision is the same as a reproducible measurement. Accuracy is a factor that contributes to precision, though it is not a prerequisite for it.
The term "significant figures" applies to anything with more than zero digits. Zeros at the start or end are not considered significant figures. Scientific notation must have the same number of significant figures as decimal notation. If the answer contains significant figures, it should equal the least significant number. No matter whether you are adding or subtracting, it is always a good idea to use the same number of decimal places.
Rules to determine significant numbers
Significant digits include all digits other than zero. Example – 1, 2, 3, …......, n.
Those digits that fall between two non-zero ones (trapped zeros) are significant. Example – 501.128 this number has sic significant figures as 5, 0, 1, 1, 2, 8.
Zeroes at the beginning of non-zero numbers have no significance. Example – 0.00341 has three significant figures as 3, 4, 1.
A number with a decimal (zeroes following a non-zero number) is generally not significant if it has no trailing zeros (see below for more details). Example – 6000 has only one significant figure such as 6. No trailing zeros are counted.
There is also significance in the number called zero. The number 0 follows the decimal point. The number 0 follows the decimal point. Example – 0.0 has two significant numbers.
Numerical values are considered significant if they have scientific notation.
Although an analyst cannot prevent random errors, the following methods can reduce systemic errors.
Calibration of apparatus - When instruments are calibrated, errors are minimized, and the original measurements are corrected as necessary.
Control determination - An experiment using a standard substance under similar experimental conditions is designed to minimize errors.
Blank determination - As a result of not using the sample, a test is conducted in identical conditions, which minimizes errors caused by impurities in the reagent.
In chemical analysis, there are several uncertainties that yield a large number of 'errors' that can be broadly categorized, namely:
- Determinate errors - Determining the root cause of (systemic) errors;
- Random (indeterminate) errors.
1. Determinate (systemic) errors
Identifiable, avoidable errors may be accounted for in principle by determining their overall value and assigning their reasonable causes. Errors of this type include:
Personal errors - Errors related to 'personal equation' appear exclusively in personal errors, as they are not related to the prescribed procedure or methodology.
Instrumental errors - There is a risk of measuring substance concentrations with incorrect calibration and faulty instruments, such as pH meters, single pan electric balances, UV-spectrophotometers, and potentiometers.
Reagent errors - These errors are caused by the reagents themselves, such as impurities present in reagents, platinum (Pt) vaporizing at high temperatures, and foreign substances unintentionally introduced due to reactions between reagents and porcelain or glass.
Constant errors - Unless the measured amount is very large, a constant error does not differ much from measurement to measurement, but it becomes less significant as the size of the measurement increases.
Proportional errors - Proportional errors vary in absolute value with a sample size so that the relative error remains constant no matter how many samples are analyzed. The chemical is usually incorporated into a substance that interferes directly with an analytical process.
Errors in methodology - The method of sampling (incorrect) and the completeness of the reaction are often the reasons for serious measurement errors.
Additive errors - This type of error does not depend on the number of substances available in the sample.
2. Intermediate (random) errors
It is impossible to pinpoint specific well-defined reasons for indeterminate errors. Usually, these problems arise from minute variations that occur inadvertently during several successive measurements that are carried out under identical experimental conditions by the same analyst. There is a high probability of both high and low results due to these errors, which are mostly random. In other words, they are neither correctable nor erasable and thus represent the 'ultimate limitation' of the measurements. Repeating measurements of a variable, subsequently analyzing the results statistically, has the effect of 'decreasing' the importance of the results to a substantial degree.
Accuracy
As a consequence of this, the terms 'accuracy' and 'precision' are frequently used interchangeably about scientific data, but there is an important difference between the two as follows:
According to traditional thinking, an accurate measurement is very near to the true value of the amount being measured. In most cases, the accuracy of a study is inversely proportional to the error of the study, that is, the greater the accuracy, the smaller the error. An experimental value is subtracted from a true value to arrive at its value.
Precision
In terms of correlation, it may be described as the relationship between two groups of experimental results, although it does not necessarily indicate the relationship with ‘true value’. An accurate measurement is deemed to have a high degree of reproducibility, while a high degree of precision is the same as a reproducible measurement. Accuracy is a factor that contributes to precision, though it is not a prerequisite for it.
The term "significant figures" applies to anything with more than zero digits. Zeros at the start or end are not considered significant figures. Scientific notation must have the same number of significant figures as decimal notation. If the answer contains significant figures, it should equal the least significant number. No matter whether you are adding or subtracting, it is always a good idea to use the same number of decimal places.
Rules to determine significant numbers
Significant digits include all digits other than zero. Example – 1, 2, 3, …......, n.
Those digits that fall between two non-zero ones (trapped zeros) are significant. Example – 501.128 this number has sic significant figures as 5, 0, 1, 1, 2, 8.
Zeroes at the beginning of non-zero numbers have no significance. Example – 0.00341 has three significant figures as 3, 4, 1.
A number with a decimal (zeroes following a non-zero number) is generally not significant if it has no trailing zeros (see below for more details). Example – 6000 has only one significant figure such as 6. No trailing zeros are counted.
There is also significance in the number called zero. The number 0 follows the decimal point. The number 0 follows the decimal point. Example – 0.0 has two significant numbers.
Numerical values are considered significant if they have scientific notation.
Get subject wise printable pdf documentsView Here
No comments:
Post a Comment
Please don't spam. Comments having links would not be published.