The Fallacy of Formative Measurement

By J. Edwards. 2011

483 citations recorded by the source corpus.

Subfields: Management Information Systems, Organizational Behavior and Human Resource Management

View Paper: https://www.semanticscholar.org/paper/a82e5c11822f7fa5d2e2187f6f0ad8489f5e6a65

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Management Information Systems

In Organizational Behavior and Human Resource Management

Citations in the corpus

Measuring Procedural Justice in Police-Citizen Encounters (2015, background): “Procedural Justice as a Formative Index Our decision to treat procedural justice in police behavior as a formative index needs explanation, particularly given the ongoing debate on formative vs. reflective measures (Diamantopoulos, Riefler, & Roth, 2008; cf. Edwards, 2011).”
Researching Long-Term Orientation (2014, methodology): “In addition to theory’s primary role in explaining how constructs are interrelated, auxiliary theories—theories that connect latent constructs to observable phenomena (Blalock, 1968)—allow empirical research to test theory (Edwards, 2011).”
The MIMIC model and formative variables: problems and solutions (2013, methodology): “2011), there are significant areas of disquiet around the formative model (e.g. Borsboom et al. 2003; Cadogan and Lee 2013; Edwards 2011; Franke et al. 2008; Howell et al. 2007a; Lee and Cadogan 2013; Wilcox et al. 2008).”
The Impact of Management Control on Employee Motivation and Performance in the Public Sector (2018, methodology): “The scales for each of the types of MC are formative constructs (Edwards, 2011), because the theory dictates that they together form one category of MC elements.4 To minimize a potential bias caused by sequencing effects (cf. Speklé & Widener, 2018), we randomized the order of appearance of the surv”
Measuring entrepreneurial passion: Conceptual foundations and scale validation (2013, methodology): “…all three domains of EP, the fit of a CFAmodel treating hopefulness as 5 In order to test this measurement model, we followed the guidelines set forth in Edwards (2011) that “to achieve identification, the loadings relating to (the formative relationships) must be fixed to constant (2011: 382–3).””
Partial least squares path modeling: Time for some serious second thoughts (2016, methodology): “…can assess the unidimensionality of the indicators, that is, whether they measure the same construct, which renders the resulting composite conceptually ambiguous (Edwards, 2011; Gerbing and Anderson, 1988; Hattie, 1985) as well as makes reliability indices uninterpretable (Cho and Kim, 2015).”
Jack of all, master of all? CEO generalist experience and firm performance (2019, background): “Conceptually, the formative measure has a causal assumption because it treats indicators as causes of constructs (Edwards, 2011), which often makes it difficult to interpret the effect sizes of indicators measuring different attributes that are, at times, on incompatible measurement scales.”
Narcissism and Leadership: A Meta-Analytic Review of Linear and Nonlinear Relationships (2015, unknown): “First, an individual trait (such as narcissism) has multiple behavioral manifestations (or reflective indicators; Edwards, 2011; Edwards & Bagozzi, 2000; Joreskog, 1971), and these behavioral indicators have different thresholds for enactment.”
Discovering Unobserved Heterogeneity in Structural Equation Models to Avert Validity Threats (2013, background): “While there is an ongoing discussion on the interpretation and use of formative measures (Aguirre-Urreta and Marakas 2012; Diamantopoulos 2011; Edwards 2010; Jarvis et al. 2012; Petter et al. 2012), there is general consensus that the theoretical meaning of a construct should correspond to its…”
Entrepreneurial fear of failure: Scale development and validation (2020, methodology): “In order to ensure model identification, we followed the procedures outlined by Edwards (2011) in a multiple indicator-multiple causes model using the five items of the Conroy PFAI to load directly on the second order factor for fear of failure.”