1.7
Standard Deviation of Calculated Results
Standard deviation measures the spread of data around the mean value. Many large data sets follow a Gaussian distribution, also known as a normal dist…
A plot of relative deviation from the mean and its frequency of occurrence appears as a Gaussian curve. This probability distribution curve of a population can be described by the population mean and standard deviation.
The standard deviation measures the spread of clustered data about the mean.
When the standard deviation is lower, the distribution about the mean is narrower, leading to high precision.
For a finite number of measurements – the population subset – the mean and standard deviation are denoted as the sample mean and sample standard deviation, respectively.
A set of N measurements can have N deviations from a reference point – true mean, which means N degrees of freedom. However, if the estimated mean is set as a reference point, only N-1 independent deviations are required, as the last one is predetermined.
The standard deviation expressed as a fraction of the mean is called the relative standard deviation. When given as a percentage, it is the coefficient of variation. Variance is the square of standard deviation.
View the full transcript and gain access to JoVE Core videos
Q1: What does standard deviation measure in a data set?
Standard deviation measures the spread of data around the mean value. A lower standard deviation indicates data clustered closely to the mean, reflecting high precision. A higher standard deviation shows data spread widely from the mean, indicating lower precision. This metric is fundamental for understanding data distribution and measurement reliability in analytical chemistry.
Q2: How does a Gaussian distribution relate to standard deviation?
A Gaussian distribution, or normal distribution, is a bell-shaped curve where the mean represents the most frequently observed value. Standard deviation determines the width of this curve: a narrow curve indicates a small standard deviation and high precision, while a broad curve reflects a large standard deviation and lower precision. The population mean and standard deviation fully describe this probability distribution.
Q3: Why is sample standard deviation calculated using N-1 instead of N?
A set of N measurements has N degrees of freedom when referenced to the true mean. However, when using the estimated sample mean as reference, only N-1 independent deviations exist because the last deviation is predetermined by the others. This adjustment corrects bias in estimating population standard deviation from a finite sample.
Q4: What is the difference between population and sample standard deviation?
Population standard deviation describes the spread of an entire data set using the population mean. Sample standard deviation describes a subset of data using the sample mean. For finite measurements, sample standard deviation uses N-1 degrees of freedom to provide an unbiased estimate of the true population standard deviation.
Q5: How do you express standard deviation as a relative measure?
Relative standard deviation expresses standard deviation as a fraction of the mean, providing a dimensionless measure of precision. When expressed as a percentage, it is called the coefficient of variation. This metric allows comparison of precision across measurements with different scales or units in analytical work.
Q6: What is the relationship between variance and standard deviation?
Variance is the square of standard deviation. Both measure data spread, but variance uses squared units while standard deviation uses original units. Standard deviation is often preferred in analytical chemistry because it directly reflects measurement precision in the same units as the data.
Q7: How does precision relate to the width of a distribution curve?
High precision measurements produce a narrow distribution curve with a small standard deviation, clustering data tightly around the mean. Low precision measurements create a broad distribution curve with a large standard deviation, spreading data widely from the mean. The width of the Gaussian curve directly reflects measurement reliability and reproducibility in analytical procedures.