Measurement

Qualitative Measurement

The process of classifying within qualitative measurement entails the amalgamation of perceived organizational phenomena, also called concepts and constructs, into symbolically described categories. Applications of this type of measurement are found in, for example, ethnographic and case study research.

Theoretical constructions specify how classification categories are associated. Key concepts underlying qualitative measurement include taxonomies, typologies, and levels of analysis and aggregation.

Classification categories are embedded within a taxonomy determining their association. Those concepts that represent taxonomies reflect logic trees, with subordinate elements necessitating superordinate facets. A taxonomic characteristic of a classification category can have theoretical and practical consequences for the study of organizations. For example, if only directors are subject to classification as senior management, then senior management as a concept or construct represents directors, and other management, such as presidents, would not be considered senior management.

Typologies make a distinction between items at a specific level of taxonomy with respect to one or more of their properties. The properties that provide a basis for differentiating concepts are typically referred to as features or attributes. They describe differences of items at the taxonomic level studied. For example, in some organizational research, managers are distinguished in terms of their negotiation style (i.e., whether they display a collaborative, competitive, avoiding, or accommodating style); in other studies, they are not only characterized in terms of their negotiation style but also their gender.

Finally, organization research can focus on the measurement of concepts using both individual and aggregate units of analysis; examples of the latter are teams and organizations. Aggregate units can be examined as distinctive phenomena, where measurements specify and order properties of the aggregate unit rather than associate one aggregate unit to another. Observations and reports from individual informants with respect to the aggregate unit serve as multiple indicators of a single state, and the number of observations and type of respondents required depends on the necessity to define a phenomenon accurately. For instance, business unit managers who are knowledgeable about their business units can provide accurate and valid descriptions of these units, such as the concept tapping into organizational identification, whereas their supervising managers may not have detailed insights with respect to the business [Page 889]units studied and, hence, are likely to provide less accurate descriptions of them.

Quantitative Measurement

Measurements that are quantitative distinguish items at taxonomic levels, similarly to typological classifications, but provide greater statistical power by offering an ordering of the underlying classification categories. Stanley Stevens in 1951 put forward four stages of quantification to relate numbers to classification categories. Nominal numeration entails the arbitrary assignment of numbers to give categories unique names, such as “staff number 980962.” The categories associated with an ordinal scale are ordered monotonically according to greater-than and less-than, and the numbers assigned correspond to the rank of each category. Numerical ranking of an individual's preferences for different work conditions is an example of ordinal measurement. Differences between categories provide the basis for comparison in interval scales, and the numbers that are applied to categories reveal degrees of differences. Calendar dates can be an example of interval measurements. For example, from graduation dates we can conclude which employees have been closer to each other in time from graduation than others; this information can, in turn, be used for drawing conclusions about their university education. In a ratio scale, categories can be described in terms of magnitudes that are whole or fractional multiples of one another, and the numbers given to categories embody these magnitudes. The relative number of alliances in which an organization is involved can be an example of ratio measurements—understanding the numbers of alliances in which two organizations are involved provides the basis for stating that the focal organization has x times more alliances than the other organization.

...