August, 2009 | 1 Lea Ellwardt University of Groningen / ICS Joe Labianca University of Kentucky Rafael Wittek University of Groningen / ICS Who Are the.

August, 2009 | 1

Lea EllwardtUniversity of Groningen / ICS

Joe LabiancaUniversity of Kentucky

Rafael WittekUniversity of Groningen / ICS

Who Are the Objects of Positive and Negative Gossip at Work?

| 2August, 2009

Why Study Gossip?

› Informal, sozializing ties affect organizational outcomes

› Workplace gossip as mechanism to reinforce solidarity and ostracism of colleagues from network

› Gossip = “informal evaluative talking about absent colleagues”

› Need to distinguish between positive and negative gossip

› To date hardly research on objects of gossip

| 3August, 2009

Gossip Triad: Who Are the Objects?

gossip object

gossip sender gossip receiver

| 4August, 2009

Main Propositions

Types of Gossip› Positive gossip as means for interpersonal affection

and solidarity› Negative gossip as means for social control and

ostracism

Gossip as Group Process› Employees simultaneously embedded in work groups

and organizational network› Group membership and social position in the broader

network determine likelihood of being gossiped about

| 5August, 2009

Group-level Hypotheses› Employees of a group are goal interdependent› Benefits of socializing within group› Costs of negative gossip high within groups

(solidarity norm), but much lower outside groups› Use negative gossip against rivaling groups for

social comparison, group identity, discrimination

H1: Positive gossip is about employees from the gossipers’ work group.

H2: Negative gossip is about employees outside the gossipers’ work group.

| 6August, 2009

Network-level Hypotheses

› Again, invest solidarity in important others: people high in social status (bask in glory, hope for reciprocity)

› High costs of negative gossip due to supporters and retaliation

› Low costs of negative gossip about low-status people → no defenders (“picking on the weak”)

H3: Positive gossip is about people high in social status. H4: Negative gossip is about people low in social status.

| 7August, 2009

› Negative gossip as means for norm enforcement and ostracism by the entire group

› Uneven distribution among colleagues, focus on single individuals (“picking on a few”)

› Social norm to dislike someone black sheep

H5: Negative gossip in organizational networks is concentrated on a small number of objects (“scapegoats” or “black sheep”).

Network-level Hypotheses

| 8August, 2009

Model

Ostracism

Negative gossip about an employee

Solidarity

Positive gossip about an employee

Solidarity

Positive gossip about an employee

-

+

+

-

Social status in the informal network

Centrality in friendship network

Social status in the informal network

Centrality in friendship network

Shared group membership

Formal work units in the organization

Shared group membership

Formal work units in the organization

Ostracism

Negative gossip about an employee

Ostracism

Negative gossip about an employee

H1

H2

H3

H4 H5

| 9August, 2009

Research Design

› Data collection in Spring 2008› One department in a medium-sized Dutch child

care organization› Population: mostly female social workers in part-

time› Flat hierarchy: one line-manager supervises all

others

› Computer-assisted questionnaire› Response rate = 83%› N = 30

| 10August, 2009

Measures› Dependent variables

Gossip about colleagues (negative and positive)› Independent variables

Formal work groupsSocial status = in-eigenvector centrality in

friendship network (UCINET VI)

› ControlsSocial relationships (valued network)Contact frequency (valued network)Job satisfaction (attribute)

| 11August, 2009

Peer-rated gossip about colleagues

› “Alter gossiped about object”

› Reduced biases of social desirability and self-serving attribution

› Reduced impact of non-response

› cf. research on bullying

object

ego(informant)

alter

| 12August, 2009

Method of Analysis

› Exponential Random Graph Modeling (ERGM), also called p* model

› Statistical package: SIENA-p* in STOCNET

› Two models: • Negative/mixed gossip network• Positive gossip only network overlap was eliminated

| 13August, 2009

Negative Gossip – Scapegoating

Circle sizes represent social status (in-eigenvector

friendships)

| 14August, 2009

Positive Gossip – broadly distributed

Circle sizes represent social status (in-eigenvector

friendships)

| 15August, 2009

Positive Gossip

About Colleagues Negative Gossip

About Colleagues

Parameter Estimate SE Estimate SE

Controls: Job satisfaction Objects -0.09 0.07 -0.17 0.11 Gossipers 0.10 0.06 -0.42** 0.14 Similarity 0.10 0.30 0.20 0.45

Dyadic relationships Shared group membership 0.78*** 0.20 0.58* 0.27 Relationship quality 0.14+ 0.07 -0.28** 0.11 Contact frequency 0.01 0.04 0.30*** 0.08

Social status in network Objects 0.07 0.10 -0.31* 0.13 Gossipers 0.26** 0.10 0.19 0.16 Similarity 0.06 0.29 0.08 0.40

Network statistics Alternating in-k-stars -0.12 0.38 1.04*** 0.27 Alternating out-k-stars 0.63* 0.30 0.47 0.30 Reciprocity 1.01** 0.34 1.03* 0.42 Transitive triplets 0.21*** 0.02 0.22** 0.08 3-cycles -0.44*** 0.09 -0.24 0.17 Alternating k-triangles 0.13 0.18 -0.11 0.21

ERGM Results

picking on the weak

scape-goating

| 16August, 2009

Conclusion & Discussion

› Choice of gossip object driven by group solidarity and social status

› Social status affects positive and negative gossip differently group picking scapegoats (bullying)

› Shared group membership leads to both positive and negative gossip showing and reinforcing solidarity within groups (social control) inderdependence creates any type of gossip

| 17August, 2009

Conclusion & Discussion

Limitations› Context specific findings› Cross-sectional design

Future research› Elaborate on outcomes of gossip:› Ostracism and integration of individuals over

time (structural holes vs. closure)

August, 2009 | 18

Thank you for listening!

…Questions or comments?