Measurement Levels

Measurement

In the physical sciences, measurement has been established as the process of assigning numbers to objects or observable events according to some formalized set of rules. This becomes troubling for communication researchers as they commonly seek to measure concepts and phenomena that are more abstract than objects and are not as easily observed as events. Many prominent communication phenomena fall into this troubling category, including media effects, nonverbal communication, media literacy, and health and risk messaging, among many others.

To measure such abstract and largely unobservable phenomena using only empirical observations remains the true challenge of social scientific measurement. For instance, a communication researcher might be interested in studying how health messages about skin cancer risks might influence people’s intention to wear sunscreen. While studying this topic, the researcher will confront a host of measurement challenges, not the least of which will be deciding how to measure abstract concepts like message attention and understanding, self-efficacy, and behavioral intention using empirical observations that represent the abstract concepts.

This process of measurement in the social sciences involves empirical and theoretical considerations. Empirically, the communication researcher will observe physical responses and assign numbers to those responses, formally known as indicators. Indicators can take many physical forms including observed behaviors and responses (e.g., people choosing to wear sunscreen after being exposed to a risk message about skin cancer), or indicators can be observed responses from study participants on survey questionnaire items or responses to interview questions. The communication researcher will then confront the challenge of linking these empirical observations to the more abstract concept they represent. The process of measurement focuses on this crucial junction between the empirically observable physical world and the latent unobservable concepts communication researchers commonly study. When empirically observed indicators strongly reflect the abstract concept under investigation, valuable inferences and knowledge claims can be made; however, poor assumptions and incorrect inferences can be made when indicators are more weakly tied to abstract concepts.

In communication research, measurement is noted to be a process of assigning values to concepts. In this process, the researcher denotes what values to assign or classify to each quality or aspect of a concept. It is the duty of the researcher to identify each quality and appropriately define its parameters for the assignment of values or classifications. To assist with the assignment and classification of values to concepts, four scales of measurement, also known as measurement levels, were defined by Stanley Stevens in 1946. Each level of measurement has distinct characteristics that enable researchers to compare similar variables using a variety of statistical tests detailed later in this entry. What follows is a discussion of the four levels of measurement.

The Four Levels of Measurement