This paper examines the role of measurement and scaling in quantitative research and dissertation design. It defines scaling as a branch of measurement that associates qualitative constructs with quantitative metric units to yield a single numerical score representing a construct of interest. The paper distinguishes between unidimensional and multidimensional scales, using examples such as height measurement and SAT scoring to illustrate each type. It also outlines the primary purposes of scaling, including hypothesis testing, exploratory research, and attitude measurement, and traces the historical origins of scaling techniques from psychology and education in the early to mid-twentieth century.
Measurements that incorporate scale data are frequently used in quantitative research projects and dissertations. For that reason, it is critical to understand what scaling is and how it can best be utilized in these types of projects. Scaling is based on specific rules and assigns numbers to objects, making it an effective way to measure and compare abstract constructs. As Trochim (2008) explains, "Scaling is the branch of measurement that involves the construction of a measure based on associating qualitative judgments about a construct with quantitative metric units. Like an index, a scale is typically designed to yield a single numerical score that represents the construct of interest" (p. 129). Scales are organized under two broad categories: unidimensional and multidimensional.
These two categories indicate that scales can have any number of dimensions, though most scales that are developed in practice tend to limit the number of dimensions used. A straightforward way to understand what a dimension is involves thinking of each dimension as a number line. Measuring constructs then requires the researcher to determine in advance whether a construct is best captured by one number line or whether it requires more than one.
A clear example of a unidimensional construct is height, which is measured along a single axis. Hunger, by contrast, may be multidimensional because a person can be more or less hungry at any given time of day and for varying reasons. A strong example of a two-dimensional construct is human intelligence, which is commonly assessed using more than one criterion. The SAT, for instance, measures both math and verbal skills. To compare students with one another, both a math score and a verbal score are required — plotted as x and y coordinates — to locate a student's performance relative to others. Scales can have many dimensions; the examples of single and two-dimensional scales discussed here are simply illustrative. There is no fixed limit to the number of dimensions a scale may incorporate.
Scaling serves several important purposes in research. Rather than relying on a list of text statements or a simple set of questions, scaling can be used to test specific hypotheses where those other approaches fall short. Scaling is also considered a major component of exploratory research and can address questions such as what dimensions underlie a given set of ratings.
However, one of the most valuable uses of scaling is for scoring purposes. Scaling provides an effective medium for collecting responses to a set of items in a way that allows a single number to be assigned, with that number representing an overall attitude or belief. This capacity to reduce complex responses to a meaningful numeric score is what makes scaling particularly useful in survey-based and attitudinal quantitative research.
Scaling can be understood as a branch of measurement concerned with the construction of instruments that equate qualitative constructs with quantitative metric units. This approach to measurement descended from the fields of psychology and education, both of which required methods for measuring things that were, prior to the development of scaling, technically not measurable. It is therefore not surprising that many individuals find the concept of scaling unfamiliar or difficult to grasp.
"Traces scaling's roots in psychology and education"
As Trochim (2008) notes, "it is important to realize that although a scale is an instrument that can be used alone, it is often integrated into a larger and more complex instrument, such as a survey" (p. 131). Scaling is a tool that transforms abstract ideas into measurable, quantifiable constructs. Understanding the role of scales in research and research design is essential for those engaged in quantitative projects, particularly at the dissertation level. Scaling as a science of measurement operates according to specific rules, assigning numerical values to objects so that abstract constructs can be compared in a systematic and meaningful way.
Trochim (2008). Chapters 5 and 7.
You’re 88% through this paper. Sign up to read the remaining 1 section.
Sign Up Now — Instant Access Already a member? Log inAlways verify citation format against your institution’s current style guide requirements.