Assessment of Measures of Central Tendency and Dispersion Using Likert-Type Scale

Abstract

This research aims to analyzing measures of central tendency (mean, median, and mode), and measures of dispersion (variance and standard deviation) and to graphically present the simulated data using Likert scale measurement. Mean (μ), Median (Md), Mode (Mod), Variance ( ), and Standard deviation (σ) have been calculated on simulated data using 5 and 4 points Likert scales with four different statements. Data analysis is key in any development especially when there is need to understanding people's opinions on a particular issue that has attracted the attention of a specific population. Likert scale of measurement is one of the methods that use statement problems in collecting statistical data in which the respondent agrees with the opinion or disagrees.  Considering the 5-point Likert scale, the measure of dispersion in which statement two had the highest standard deviation of 1.53 and the mean value of 2.35 being the least of the mean values implies the significant spread of the data across opinions. Statement one of the 4-points Likert scale has two modes (bimodal) where the opinions “Strongly Agree” and “Agree” have ties of 20 responses each. This implies that interpretation of such data may be difficult, because the data can be roughly split into two different groups, and can be useful for further research to find a mode to draw conclusions. The best form of Likert scale can be recommended for 5-point Likert scale so as to give room for undecided opinion. Other measures of central tendency and dispersion can be useful in interpretation but the recommended statistical tool is standard deviation as it shows the spread of the opinions about the mean.