1 / 24

Multilevel Research

Multilevel Research. 黃品全講授 本講義為 Mark Gavin 所撰. Bringing Attention to Multilevel Research. Calls for increased development and testing of multilevel theory

yuval
Download Presentation

Multilevel Research

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Multilevel Research 黃品全講授 本講義為 Mark Gavin 所撰

  2. Bringing Attention to Multilevel Research • Calls for increased development and testing of multilevel theory (e.g., Hackman, 2003; House, Rousseau, & Thomas-Hunt, 1995; Klein, Dansereau, & Hall, 1994; Klein & Kozlowski, 2000a, 2000b; Rousseau, 1985) • Progress has been stunted by lack of tight theorizing and methodological and analytical problems

  3. Overview of Workshop • Theoretical Issues • Methodological Issues • Analytical Issues

  4. Theoretical Issues • Typology of multiple level models • Typology of higher-level constructs • The need for multilevel research (preliminary statistical issues)

  5. Hierarchical Data Structures Hierarchical nature of organizations • Individuals nested in work groups • Work groups nested in departments • Departments nested in organizations • Organizations nested in industries/environments

  6. Hierarchical Data Structures Constructs at different levels • Individuals • Work groups • Departments • Organizations • Industries/Environments

  7. Caveat Workshop terminology usage: Discussion based on context of individuals nested within groups Terms used generically Individual = lower-level unit Group = higher-level unit

  8. Two types of Multiple Level Models Borrowing from Rousseau’s (1985) typology • Composition model • Cross-level models

  9. Single-level Model X Y Concerns the relationship between two constructs residing at the same level

  10. Examples of Single-level Models • Individual attitudes influencing job performance • Group cohesion influencing group performance • Organizational structure characteristics influencing organizational performance

  11. Composition Model X2 X1 Concerns the relationship between two (presumably) similar constructs residing at different levels where X2 is derived from X1 *Subscripts denote levels where 1 = individual level and 2= group level

  12. Composition Model • Provides mechanism linking lower-level and higher-level constructs Examples of linking mechanisms*: • Leadership theories • Social information processing • Attraction-selection-attrition • Considers the form of the relationship • Isomorphism • Partial functional identity or fuzzy composition *These play a role beyond composition models by helping us understand why we observe similarity within groups

  13. Isomorphism vs. Fuzzy Composition • Isomorphism (e.g., Rousseau, 1985) Specifies that the lower-level variable and its aggregate are conceptually and functionally identical • Partial functional identity (e.g., Rousseau, 1985) or fuzzy composition models (Bliese, 2000) • Specifies that the aggregate captures something conceptually and functionally different than its lower-level counterpart, even if we call them by the same name, largely because they pick up and carry contextual influences

  14. Examples of Composition Models • Individual mood and group affective tone (George, 1990) • Psychological/Organizational climate (James, 1982; James, James & Ashe, 1990) • Individual/Organizational learning We’ll come back to composition models shortly

  15. G G or X Y X Y Cross-level Models Cross-level main effect Cross-level moderation Concerns the influence of a higher-level variable on 1) a lower-level outcome or 2) the relationship between two lower-level variables *X and Y are individual-level variables and G is a group-level variable

  16. Examples of Cross-level Models • Cross-level Main Effect • Climate research where organizational characteristics influence individual (shared) perceptions (and subsequently behavior) • Cross-level Moderation • Leadership climate moderates the relationship between task significance and hostility incidents (Gavin and Hofmann, 2002) • Cohesion moderates the relationship between satisfaction and OCBs (Kidwell, Mossholder & Bennett, 1997)

  17. Types of Group-level Constructs • Global • Selected score • Additive • Direct consensus • Referent-shift consensus • Dispersion

  18. Global constructs • These belong to the group and have no lower-level counterparts • They are NOT derived by aggregating individual variables/responses • Examples: group size, group structure

  19. Selected Score Constructs • These originate with the individual and emerge to characterize the group • Relevant where, typically, members at the extreme place constraints on the group • The relevant individual’s score then represents the group • Examples: individual ability in certain task environments, individual personality in certain work settings, CEO tenure

  20. Additive Constructs • These originate with the individual group members and emerge to characterize the group • They are derived by aggregating individual scores • The aggregate carries no particular meaning beyond a simple statistical summation of individual characteristics • Examples: unit sales as derived by summing individual sales, group expertise as a summing of individual member tenure

  21. Direct Consensus Constructs • These originate with individual group members and emerge to characterize the group • They are derived by aggregating individual scores • They take on meaning when members share their individual characteristics or perceptions • Examples: group affective tone (George, 1990), psychological climate (James, 1982)

  22. Referent-shift Consensus Constructs • These originate with the group (though they typically have meaningful individual-level counterparts) but emerge within the group setting • They are derived by aggregating individual scores • They take on meaning when members share in their perceptions of the group characteristic • Examples: collective efficacy, team empowerment

  23. Dispersion Constructs • These originate with the individual group members and emerge to characterize the group • At the group level, these index variability or distributional properties of individual characteristics or scores within the group • They do NOT require similarity among group members – rather they often stress dissimilarity • Examples: demographic diversity, climate strength

  24. Issues with Construct Types • Only theory can determine which type a given construct fits • Depending on the theory, the same variable could fit into different types (e.g., group ability could be defined by a select score model, an additive model, a dispersion model, or …)

More Related