A systematic error refers to a nonrandom statistical error which causes the mean of several separate measurements to differ significantly from the actual value (Muema , 2008). Systematic error is used to define the bias between the mean of two measured values. Systematic error can also be regarded as the difference between a computed or measured value and the true value stated theoretically . for instance, in an experiment to determine the acceleration due to gravity , the measured value can be less or greater than the theoretical value which is 9.8 N/KG (Muema , 2008).
Systematic errors can either be constant or related to the true value. Constant errors are not simple to handle because their effects are only noticed when done away with. They cannot be removed though through repeating the measurements or by obtaining averages. They can only be removed by calibrating the measuring instrument in order to eliminate the zero error (Muema , 2008). Systematic errors which change during an experiment are easier to detect. Firstly ,a Systematic error in a measurement can be caused by poor/imperfect calibration of the measuring instrument . This is what is popularly referred to as the zero error (Fisher ,1966). This results to a constant zero error. In this case , the first mark in the measuring instrument could be ahead of its usual position. For instance, if a stop watch used to measure time in an experiment starts with one second on the clock instead of zero, then all readings obtained are off by one second ( Fisher , 1966). If the experiment is repeated several times and the timer starts at the same position each time , then a big percentage error is recorded .The final result ends up being larger than the actual value. Secondly , systematic error can be attributed to changes in the environment which interferes with the measurement process . For instance , when a ruler is use d to measure length , expansion of the ruler occasioned by high temperatures can lead to a wrong value. If on the other hand the ruler contracts due to low temperatures , then a value smaller than the true value is recorded (Fisher , 1966).
Systematic errors can also be caused by an estimate based on the physical law. For instance during an experiment involving pendulum oscillations , if the movement of the support is not taken into account when estimating the pendulum’s frequency , then an error occurs (Fisher , 1966). A systematic error can also occur if readings are taken at any other angle other than 90 degrees to the instrument’s scale. This is what is called a parallax error (Muema , 2008).
A systematic error attributable to zero error can be mitigated by resetting the instrument immediately before measurements are taken. If this error is not corrected at this stage , then one can mitigate it by subtracting its value from the readings (Muema , 2008). For instance in the case of an experiment involving measurement of time whereby the timer starts with one second on the clock , then to get the true values , one has to subtract one second from all values if the error cannot be avoided by resetting the timer (Muema , 2008).
To recognize these systematic errors , one can compare the measurements with those obtained using a different apparatus perceived to be more accurate. For instance in the above experiment involving measurement of time , if the stop watch used gives readings which are randomly distributed about the mean, a systematic error can be realized by checking the stopwatch against a speaking clock of the telephone system which is considered to be more accurate(Muema , 2008) . If the stopwatch is found to be running slower or faster than the stopwatch , then there is a systematic error. Systematic errors can also be detected if quantities which are already known are measured using the instrument. In the case of parallax error , taking measurements from an angle of 90 degrees to the scale can mitigate the error (Muema , 2008).