Dispersion-Composition Models in Multilevel Research

Abstract

Multilevel researchers have predominantly applied either direct consensus or referent-shift consensus composition models when aggregating individual-level data to a higher level of analysis. This prevailing focus neglects both theory and empirical evidence, suggesting that the variance of group members' responses may complement the absolute mean level of group members' judgments. The goals of this article are to demonstrate the application of dispersion-composition models for capturing variability among group members' collective judgments and highlight the statistical challenges (and inherent constraints) of using group means and variances as predictors of study criteria. To this end, the authors present and illustrate a six-step sequential framework for applying dispersioncomposition models using data from two independent field samples. The authors contend that the application of dispersion-composition models not only will strengthen a study's conclusions by eliminating potential rival data interpretations but may also shed new light on past findings, potentially opening new doors to a more complete understanding of multilevel phenomena.

Consequences of Collective-Focused Leadership and Differentiated Individual-Focused Leadership

“…Accordingly, critics could argue that previous results are (partially) attributable to these unmeasured reverse operationalizations or to the failure to control them (e.g., Cole et al, 2011). To gain more insight, we will explicitly integrate these reversed leadership operationalizations in our study.…”

“…Moreover, although previous theorizing highlighted the importance of positive (collective) behaviors for collective effectiveness (e.g., Kark & Shamir, 2002), no behaviors were actually modeled or tested, and despite Wu et al's (2010) call to investigate moderators, no contextual factors have been assessed in this research area so far. Finally, existing research (Wang & Howell, 2010;Wu et al, 2010) has theoretically and empirically neglected important alternative model specifications that might challenge previous results, such as the reversed operationalization of the leadership variables (e.g., Cole, Bedeian, Hirschfeld, & Vogel, 2011). Kunze et al / To Differentiate or Not 3 We address these questions by drawing on the commitment literature (e.g., Meyer, Stanley, Herscovitch, & Topolnytsky, 2002) and the theories often used within this field, namely, social-exchange theory (SET; Blau, 1964) and perceived organizational support theory (POS; Eisenberger, Huntington, Hutchison, & Sowa, 1986).…”

Individual and group‐level job resources and their relationships with individual work engagement

“…Step 5 in Table 4 indicates that group-level job resources interact with their variability within groups (γ=1.69, p <.05), indicating that the cross-level relationship of group-level job resources and work engagement is dependent on the variability within groups. In case of such a significant interaction effect, Cole et al 19) recommend to test for curvilinear effects of group-level job resources and their variability. Results however indicate no curvilinear relationships with work engagement, whereupon a significant interaction effect can be interpreted.…”

“…Again, covariates with no explanatory value were omitted from the model before the next step. Following the recommended procedure of Cole et al 19) to ensure unbiased estimates of relationships with group-level constructs, two more steps in model building were performed. This is particularly indicated in cases where ICC(2) and r WG (J) estimates indicate substantive variability within groups 19) .…”

“…Following the recommended procedure of Cole et al 19) to ensure unbiased estimates of relationships with group-level constructs, two more steps in model building were performed. This is particularly indicated in cases where ICC(2) and r WG (J) estimates indicate substantive variability within groups 19) . Thus, fourth, we included the corresponding group-level job resources variability variable to control for the varying dispersion of job resources at group level (Step 4).…”