Measures are ways to quantify and compare quantities. They are statistical in nature and can be inexact, but they reduce uncertainty. The main system of measurement is the SI units, but many countries use their own customary systems for measuring length, volume, etc.
Metrics are more sophisticated measures that offer more information than measures alone. They combine measures with various data points and often repackage them in useful-yet-easily digestible ways.
Units of Measurement
A unit of measure is a way to express a physical quantity. Units are based on standards, which are special objects that allow comparisons between different quantities. Large measurements like the length of a galaxy use larger units of measurement, while smaller measurements like the weight of a grain of salt use smaller units of measurement.
Different traditional systems of measurement exist, including the metric system and United States customary units. The metric system is most widely used in the world.
The metric system uses units of length, mass and capacity. Each unit of measurement has a symbol that represents it. Each unit also has a name that helps identify it. Students should become familiar with the names of these units and their symbols, as well as how to convert between them. It is important to understand that converting from one unit of measurement to another does not change the value of the measurement; it simply changes the measurement system.
Measurement Scales
The way you measure data affects the statistical analysis you can perform. The four measurement scales – nominal, ordinal, interval and ratio – each have different properties.
For example, a nominal scale names categories that data can fall into but does not give any numerical meaning. This data is often used for categorizing information – such as ranking students in reading classes.
Interval and reporting scales have the properties of a nominal scale but provide a quantitative meaning for differences between data points. These scales allow you to order data like heaviest to lightest or ranks of athletes, for instance. They also let you use arithmetic operations, so you can calculate mean, median and mode.
Ratio scales have the properties of interval and reporting scales, but they also include a true zero point. A zero on a ratio scale is meaningful and means the absence of something – such as elapsed time if there is no elapsed time.
Measurement Instruments
A measurement instrument is a device that measures some quantity or property. It is used in a laboratory or industrial setting. Examples include a radar speed gun, a calibrated scale, and a check weigher. Precision is the closeness of a measurement to its ideal value; accuracy is how far a measured quantity deviates from that ideal, or right value.
The reliability of a measure is the degree to which it consistently yields the same results under the same conditions. This includes internal consistency reliability and test-retest reliability. It is important to note that the level of reliability does not imply that an instrument has validity. In fact, it is possible for a measure to have large reliability and still have zero validity.
A “measures” section in a research article typically identifies the variable that is being measured, describes the measurement method, and provides information about the psychometric properties of the measurement instrument such as its reliability and validity. This information enables researchers to compare instruments that claim to measure the same construct.
Measurement Methods
Measurement methods identify how data will be collected and used to track project indicators. They are a key component of a project logframe.
Quantitative measurement methods collect data that can be counted and analyzed statistically to calculate percentages, averages, ranges and variances. Examples include tracking logs, questionnaires, structured observation and knowledge and achievement tests.
Compensation method compares the value of the measured quantity with that of a standard reference. If the measuring instrument has a certain deviation from its normal state, the value of a compensation variable can be determined by replacing it with individual quantities of different sizes until one comes close to the measured quantity. This method is less influenced by external factors than substitution methods. It is also suitable for periodic measurements.