Brancati, measurement2018

Social Scientific Research Dawn Brancati - ®SAGE Los Angeles | London | New Delii Singapore | Washington DC | Mebboume 2019

Measurement Quality The process by which concepts are transformed into measures is known as operationaliza- tion. The commonly heard phrase, 'How is that operationalized?', means how is a concept measured. In practice, it is not always possible to measure well important concepts. There are two standard metrics used to evaluate the quality of measures, namely, validity and reliability. To these two, I would add a third, which I refer to as discriminatory power. These three criteria are summarized in Table 15.2. validity than others. Figure 15.1 illustrates the concept of validity. The figure on the top depicts a measure with a low level of validity. The overlap between the concept and the measure is small. The figure in the middle depicts a measure with a moderate level of valid- ity, while the figure at the bottom represents one with a high leve] of validity, as demonstrated by the large overlap berween the concept and the measure. To better understand validity, consider the example of civic engagement. Civic engage- ment refers to the ways in which citizens participate in the Jife of their community either measure of civic engagement is voting (Putnam 2000). Voting in local elections can be a way in which people participate in the life of their community and voting in local elections suggests that self-interest is the primary motivation for voting (Downs 1957; Riker and Ordeshook 1968). These benefits include the ability to influence policy to one's own advantage, but are limited to them. They also include psychological benefits, such as a sense of empowerment, wellbeing, and self-satisfaction. Research also finds that people vote for normative reasons and habit (Gerber et al. 2008 ) Criterion Definition flidity extent to which a measure captures the concept it is intended to represent Seliability extent to which a measure produces consistent and dependable resulis Zsca'iminatory power extent to which a measure is able to distinguish between two concepts Table 15.2 Criteria for measurement quality

Low validity Measure - Moderate validity Measure High validity Concept Measure | | L Figure 15.1 Validity of measures Volunteering in one's local community and donating to local groups and causes ar< arguably more valid measures of civic mindedness than voting. Survey research shows thas community-mindedness is a significant reason why people volunteer and donate to causcs (Unger 1991; Clary et al. 1996). However, neither volunteering nor donating are perfectls valid measures of civic mindedness cither since people engage in both activities for reasoms other than a desire to improve the lives of others (Anderson and Moore 1978; Clary et a 1996). These motivations include a desire to look like a good person, self-actualization. loneliness, skill acquisition, and so forth. Reliability Reliability refers to the degree to which a measure produces consistent and dependable reszis A reliable measure yields the same data each time that it is used. Validity and reliabilicy 2 independent of each other. Measures may have high validity and low reliability, or vice verss