DATA MEASUREMENT 3
Thescale of measurement or level of analysis is the nature of factswithin the numbers assigned to variables. A variable has any of thefour levels of frequencies, which are ordinal, nominal, interval, andratio as developed by psychologist Stanley Smith Stevens. It isimportant to know the different levels of measurements in statisticsbecause it will determine which method of statistical analysis isappropriate for the particular research being carried out. How theinvestigation will be conducted depends on the variable involved inthe study (Gareth, Tibshirani, Witten & Hastie, 2013). In thefirst level, which is nominal, the numbers in the variables are usedto classify data. Words, letters and alpha-numeric symbols areutilized. The ordinal level depicts ordered relationships among thevariable observation. The third level, interval level, ranks, ordersmeasurement, and specifies the distance between each level on thescale from a small range to high. In the last level, which is theratio level, observations possessing equal intervals can have a valueof zero. By understanding the levels of measurements, it is easy todecide which one will yield better outcomes. The interval level ofmeasurements and ratio level are precise while, on the other hand,nominal classifies data.
Theknowledge of different levels of measurements in statistics isparamount in avoiding statistical errors. Data is delicate, andthrough it, conclusions are to be drawn based on findings. The use ofthe right level of measurement in analysis gives correct results(Gareth et al., 2013). Thus, it is safe to say that data cannot bemeasured until one understands the different levels of measuredbecause of the operationalization problem that could arise. This isthe process regarding how a concept will be measured. Data cannotjust be measured anyhow rather, it has to fall under any of the fourlevels of measurement. It is easy to classify the collected data byunderstanding the given levels.
James,G., Witten, D., Tisbshirani, R., & Hastie, T. (2013). Anintroduction tostatistical
learning:With application in R. NewYork, NY: Springer