Definitions of terms in practical questions
Although students will not be ask to recall or quote these definitions in any examination, question papers will expect students to recognise these terms and answer questions involving their use.
Accuracy: An accurate measurement is one which is close to the true value. The accuracy of a measurement depends on factors such as the quality of the measuring device and the skill of the person taking the measurement. For example, if a measuring device such as a weighing balance has a zero error (in other words, it does not read zero when no mass is placed on it) then all readings will be inaccurate, unless allowance is made when the measurements are taken. Some quantities – such as g, the acceleration due to gravity, have an accepted value. An ‘accepted value’ comes from the work of many scientists who have measured the value, agreed with it and published the value. These values can be checked via text books, data tables or through the internet (remembering that some internet sources – especially those with open editing or owned by interest groups - may not always prove to be reliable).
Anomalous data: Anomalous readings are those which fall outside the normal, or expected, range of measurements. If we take a large number of readings, we can be more certain about saying which readings are anomalous (i.e. do not fit the pattern established by the others) and which are not. Anomalous readings are easy to see on a graph as a point, or points, which do not lie on or near the best-fit line. Anomalous readings should be removed from any data which is being used to calculate a mean (average) value.
Average: The arithmetical mean of a set of data is usually referred to as the average. This mean value gives you an estimate of the ‘true’ value, assuming that no reading is anomalous. For example, if you measure the length of a piece of wire four times and obtain values of 6.2 cm, 6.1 cm, 6.3 cm and 6.2 cm, the average (mean) value for the length is 6.2 cm. It is only an estimate because your measuring instruments may be giving you false readings in some way, or other variables may have affected what you were trying to measure.
Concordant readings: If readings have been taken several times and the readings are identical, or close to each other, then they are described as concordant. In the example above, the four readings for the length of the wire are concordant. However, if the readings were 6.2 cm, 7.1 cm, 6.3 cm and 6.1 cm then the readings would not be concordant. The reading of 7.1 cm is likely to be anomalous and should be checked again. Any average taken from the readings should only include the concordant readings and not anomalous ones. Usually, readings which are concordant are likely to be reliable. Concordant readings are frequently encountered in titrations in chemistry, where titre values are said to be concordant if they are within 0.20 cm3 of each other.
Control variable: A control variable is one that will affect the outcome of the investigation. Control variables must be kept constant otherwise the investigation will not be valid (a fair test) e.g. if you were investigating the effect of light on the rate of photosynthesis of a plant, you must keep the temperature around the plant constant as any change in temperature would also affect the results. If you did not keep the temperature constant, the experiment would not be valid (a fair test).
Correlation: Correlation is the relationship between the two variables in a given experiment. This is often obtained from a graph. If the gradient (slope) of a graph is positive (i.e. the graph slopes upwards) we can say there is a positive correlation. If the gradient is negative, we can say there is a negative correlation between the variables. If a straight line goes through the origin of a graph and the gradient is positive, we can say that the variables are directly proportional to each other. Even if two factors correlate very well together, remember that it is not certain that the change in one variable causes the change in the other.
Data: This is a term normally used for the set of numerical values recorded in an experiment. We usually record data in tables to make comparisons easy.
Dependent variable: The dependent variable is the quantity that changes as a result of changes made to another variable (the independent one) e.g. if we chose to vary the height of a ramp and measure the acceleration of a trolley as it runs down the ramp, the height of the ramp is the independent variable and the acceleration of the trolley is the dependent variable.
Fair test: A fair test is a series of experiments or measurements in which only the values of one variable are changed. A fair test can usually be achieved by keeping all other variables constant, or controlled. Experiments that meet these criteria are said to be valid.
Independent variable: The independent variable is the one which we vary an experiment in order to see the effect on the dependent variable e.g. we might vary the height of a ramp (independent variable) and then measure the acceleration of a trolley which rolls down it (dependent variable).
Precision: Precision is usually determined by the apparatus being used, although it can be influenced by technique. Most scientific instruments have scales – if the sub-divisions on these scales are smaller, then it is usually possible to take more precise readings. For example, it will be more precise, when measuring a small temperature rise, to use a thermometer measuring to the nearest 0.1°C than one measuring only to the nearest 1.0 °C. Note, however, that this is not always the case. Most digital stopwatches will measure to the nearest 0.01 s. This degree of “precision” is unwarranted – our own reaction times prevent readings taken to this level of precision from being valid. Precision can be improved in an experiment by using more sensitive, or better graduated, measuring devices; and by eliminating experimental error from factors such as parallax.
Reliability: The results of an investigation may be considered reliable if readings are repeated, and concordant data is obtained. The more concordant your results are, the more reliable they are likely to be. If the data collected is very unreliable, it is likely that there is something wrong with the experiment! However, a simple way to improve the reliability of data is to repeat the experiment and collect data to average. However, remember that anomalous data will need to be removed in order to improve the reliability of the data.
Validity: Data collected may be considered valid if you can say “yes” to the question, “Am I really measuring what I am trying to measure?” Validity refers to the technique and apparatus used for collecting the data. In a valid experiment all variables are kept constant apart from those being investigated. Normally only one variable is investigated at a time. Validity can be improved by reducing any uncertainties (or errors). Note that validity is not really about “human errors” caused when taking readings, it is about failing to control variables that may affect the outcome of an experiment.