The Levels of Measurement of Data Research

347 Words1 Page
Level of Measurement The majority of the data collected in this survey is solely nominal or categorical in its level, without any numerical values assigned (though numbers could be assigned to the categories if desired, so long as it is understood that the numerical identifiers have no significance beyond their identifying purpose) (Rubin, 2009). Several of these elements are binary nominal variables (i.e. "yes/no" questions), and there are also four elements that progress beyond the nominal level to the ordinal level of measurement, however none of the elements progress beyond ordinal to interval or ratio measurements. Each of the individual items on the survey instrument is assessed for its level of measurement below, with explanations: 1. Nominal, binary (yes/no). No numerical value to the question or its responses. 2. Nominal and ordinal categories are based on rank, which is numerically based. Though numbers are based on a ratio scale (distance), the fact that ranges are given for each response means no interval (and thus no ratio) measurement can be taken. 3. Nominal and ordinal. Again, an interval scale would be possible if ranges were not given for each response, which limits the responses to ordinal. The nominal level provides some additional information, especially with the last possible response. 4. Nominal and ordinal categories are based on rank, which is numerically based. Again, a range exists for each response, and would could be a ratio scale (currency)
Open Document