Theory of Errors in Observations

An interactive guide to understanding errors in civil engineering measurements.

Introduction

Error is the difference between an observed value and its true value. In measurement, no true value can ever be identified due to imperfections in instruments, human limitations, and environmental factors.

Surveyors base their operations on the assumption that errors exist in every measurement. Therefore, it's crucial to understand how to minimize these errors and quantify their magnitude to determine the reliability of the measurements.

Types of Errors

1. Systematic / Cumulative Errors

Errors resulting from the observer, the instrument, and the environment. These errors follow physical laws and their magnitude and sign can often be determined. They tend to accumulate if a measuring process is repeated.

Key characteristic: They always have the same sign and magnitude under identical conditions.

2. Accidental / Random Errors

These are errors beyond the control of the surveyor. They are present in all surveying observations. They are the errors that remain after systematic mistakes have been eliminated.

Key characteristic: They occur randomly, following the laws of probability. They tend to cancel each other out over many measurements.

Precision and Accuracy

Precision refers to the degree of consistency of a group of observations. It indicates how close repeated measurements are to each other.
Accuracy refers to the closeness of a measurement to its true value.

Low Precision, Low Accuracy

Measurements are scattered (not precise) and far from the true center (not accurate).

High Precision, Low Accuracy

Measurements are clustered tightly (precise) but far from the true center (not accurate).

High Accuracy, High Precision

Measurements are clustered tightly (precise) exactly at the true center (accurate).

Key Mathematical Definitions

Most Probable Value (MPV)

The value that has more chance of being correct than any other value. For a series of observations, it is the arithmetic mean.

MPV = (Σx) / n

Where Σx is the sum of observations and n is the number of observations.

Error / Residual (v)

The difference between an observation 'x' and the most probable value 'MPV'.

v = x - MPV

Standard Deviation (Sx)

A measure of the dispersion of a set of data from its mean.

Sx = ± √(Σv²) / (n - 1)

Probable Error (PE)

A value that establishes a range within which there is a 50% chance that the true value of an observation lies.

PEs = ± 0.6745 · Sx (For any single observation)
PEm = ± 0.6745 · (Sx / √n) (For the mean)