Levels of Measurement

Measurement Scales

Reliable and accurate measures are required for empirical research in social sciences. In quantitative and statistics research methodology, the relationship among assigned values to a variable’s attribute is referred to as the level of measurement. According to the Research Methods Knowledge Base (2006), the level of measurement determines the type of conducted statistical analysis and research-drawn conclusions. There are four selected types of measurement scales used to conduct research questionnaires such as nominal, ordinal, interval, and ratio.

Nominal Level of Measurement

The nominal scale is the lowest and crudest form of measurement, which classifies brands, products, companies, individuals, and other entities into categories where there is no implied order (FAO, 2013). According to Garger (2010), perceptions, events, and people are placed into common trait-based categories through the nominal scale. This scale naturally suits particular data such as African American vs. Asian, redheads vs. blondes vs. brunettes, and males vs. females. It does not capture information about the central entity other than whether the entity belongs or does not belong to a group. For example, an individual can either attend or not attend college, smoke or not smoke, or have some, average, or extensive experience with computers. Categorized symbols such as labels, letters, and numbers are used to code nominal scale data (Garger, 2010). According to Tyson Gingery (2009), in questionnaires, nominal variables are categorized data groups or items that cannot be ordered or ranked. Rather than number, nominal variables are usually words. Usually, multiple choice survey questions are nominal. Examples of nominal variables include non-ordered demographic survey questions, religious affiliation, geographical region, ethnicity, gender, and yes/no questions (Gingery, 2009).

Ordinal Level of Measurement

Over the nominal scale, the ordinal scale has at least...