Every physical quantity we describe in the world of science and engineering is a measurement. Whether it is the length of a microscopic bacterium or the distance between celestial bodies, the process relies on comparing a quantity to a standard unit. However, the assumption that a measurement is a perfect snapshot of reality is fundamentally incorrect. In practice, the value we record is always an approximation, and the deviation between this approximation and the true value is the source of error. Understanding what are errors in measurement is not just an academic exercise; it is the cornerstone of reliable experimentation, quality assurance, and valid data interpretation.
Defining Measurement Uncertainty
To discuss errors, we must first distinguish between mistake and uncertainty. An error is a specific, non-random deviation from the true value, while uncertainty is a quantification of the doubt regarding the result. When we characterize what are errors in measurement, we categorize them based on their origin and behavior. Systematic errors, for example, cause data to consistently lean in one direction, either always positive or always negative. Conversely, random errors cause scatter and variability in repeated measurements, making the data fluctuate around a central value without a consistent bias.
The Nature of Systematic Errors
Calibration and Environmental Factors
Systematic errors are perhaps the most insidious because they undermine accuracy rather than precision. These errors occur due to consistent flaws in the measurement system itself. A classic example is a scale that is not calibrated to zero; every measurement it produces will be offset by the same amount. Environmental factors also play a significant role here. If a metal ruler expands under high temperature, it will consistently measure shorter lengths than the true value. Similarly, a misaligned sensor or a worn-out component will introduce a fixed bias that persists across every single reading, making the data misleadingly precise but fundamentally incorrect.
Addressing Human and Technical Bias
Human factors contribute significantly to systematic deviation. If an observer consistently misreads the meniscus of a liquid in a glass tube due to parallax, or if a technician applies tension to a wire when measuring its elasticity, a pattern emerges. These are not accidents; they are consistent shifts in the data. Identifying these requires a rigorous comparison with a known standard or a different method. The goal is to isolate the instrument’s inherent defect or the observer’s consistent bias so it can be mathematically corrected or the methodology can be adjusted to eliminate the flaw.
The Impact of Random Errors
Sources of Variability
While systematic errors challenge accuracy, random errors challenge precision. These are the result of unpredictable fluctuations in the measurement process. What are errors in measurement that are random in nature? They include the natural vibration of atoms, minute variations in how a person grips a tool, or slight changes in the voltage of an electronic circuit. Even the finest instruments exhibit a "noise floor" that prevents them from reading a constant value. When you weigh the same object multiple times and get slightly different numbers each time, you are observing the effect of random error. These variations are inevitable in any dynamic system.
Statistical Mitigation
Because random errors are unpredictable, they cannot be corrected individually. However, they can be managed and reduced through statistical methods. By taking a large number of readings and calculating the mean, the positive and negative deviations tend to cancel each other out. This is why repeated trials are essential in scientific protocols. The standard deviation is then used to express the dispersion of the data, providing a quantitative measure of the precision. The central limit theorem assures us that the average of a large sample size will converge closer to the true value, effectively minimizing the impact of these random fluctuations.
