Level of Measurement
Quantitative clinical research requires that levels of measurements be accurate, reliable, and precise. This is the only way a research can achieve its goals and objectives in the investigation of critical issues affecting clinical medicine. Researchers assign unique numbers and values to objects of investigation including people, perceptions, and concepts (Mora, 2010). This process is described as measurement. In my view, the variables to be measured are often divided into levels of nominal, ordinal, interval or ratio. The nominal scale is asserted as a level of coding that group people, objects, perception, and attributes into common characteristics (Hamid et al., 2014). The coding is effectively achieved by assigning attributes such as numbers, letters, colors or symbols that makes it easy to different objects of investigation. Based on my view, this level of measurement may be appropriate for some research studies requiring generalization.
On the other hand, an ordinal scale level of measurement differs from nominal because it emphasizes on the ranking of research data from the highest to the lowest. This provides fundamental information in respect to data points and how each data relates to one another. Fundamentally, ordinal measurement may use non-numerical categories to define relationships between datasets. These categories include elements such as low, medium, and high. This is the level of measurement that I prefer to use in a quantitative research. My study is focused in investigating the level of smoking in society. Smoking is harmful to human health, and therefore, it is important to create an awareness of the effects of smoking to the younger generations. Based on level of measurement, my study will engage the ordinal scale to collect data from teenagers who smoke between 10 and 30 cigarettes a day. Nevertheless, ordinal level of measurement is associated with several disadvantages. First, the gaps between data values are not equal making it impossible to assess the central tendency through mean. Second, data values and responses arising from using this measurement are narrow in respect to survey questions thereby creating biases not taken into account in the survey.
Hamid R., Marjan M., Peyman A., & Dario F. (2014). Manipulating measurement scales in
medical statistical analysis and data mining: A review of methodologies. Journal of Research in Medical Sciences. Retrieved from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3963323/. Retrieved March 5, 2018.
Mora, M. (2010). Qualitative & Quantitative research: Which to use? Retrieved from:
https://www.surveygizmo.com/resources/blog/quantitative-qualitative-research/. Surveygizmo. Retrieved March 5, 2018.