Communication Networks

Introduction

1. Definitions (cf. Forsyth, Leavitt, Shaw, 1978)
   1. Centralization: degree to which free communication within a group is limited. Centralization is proportional to number of connections in a group/total possible number of connections.
   2. Independence is the degree to which an individual in a given position can perform his/her task autonomously (i.e. not need to depend on other members of the group). Independence is proportional to the number of sources that provide resources for a position
   3. Saturation is the total load - task, role, and message transmission - on a position in a group.

Main findings

1. Simple Tasks (e.g. transmission of information)
   1. Centralized networks (e.g. wheel) more efficient; task performed faster with fewer errors. Central members perform task once information routed.
2. Complex tasks (e.g. arithmetic problems)
   1. Decentralized networks (e.g. circle, star) more efficient. Any (e.g. best) member can perform task.

Theory (cf. Shaw, 1978)

1. Satisfaction derives from high independence and moderate levels of saturation.
2. Task performance may be aided by independence and is hindered by saturation.
   1. Both independence and saturation are affected by the task and the communication pattern.
   2. Independence is higher in decentralized networks.
   3. Saturation is lower in centralized networks for simple tasks (few messages need to be sent); higher in centralized networks for complex tasks (because the task requires more messages, they flow through a central position, and peripheral members ask for more information).

Reinterpretation of main findings

1. Efficiency.
   1. Simple tasks. Centralized networks should be more efficient than decentralized networks on simple tasks because saturation is low and independence is unnecessary.
   2. Complex tasks. Centralized networks should be less efficient than decentralized networks on complex tasks because saturation is high and independence useful.
2. Satisfaction.
   1. Simple tasks. Networks may not differ in satisfaction on simple tasks (but central member more satisfied than peripheral in centralized network because more independent).
   2. Complex tasks. Decentralized networks (all of whose members are relatively independent but not saturated) should have higher satisfaction than centralized networks (whose peripheral members have low independence and all of whose members are likely to be highly saturated).

Direction of Communication Exercise

Introduction (application of communication network theory to direction of communication)

1. Complex task used
2. One-way communication
   1. More Centralized
   2. Center saturated task performance hindered, less satisfaction.
   3. Peripheral members very dependent more frustrated.
   4. Takes less time
3. Two-way communication
   1. More Decentralized
   2. Less saturation better task performance, greater satisfaction.
   3. Peripheral members less dependent less frustrated.
   4. Takes longer

Method

1. Subjects
2. Procedure
   1. Random assignment to roles (leader, workers, observer)
   2. 1-way, 2-way instructions
   3. Observer duties
3. Measures
   1. Time
   2. Estimated scores
   3. Actual scores
   4. Worker satisfaction (x point scales, x high)
      1. Self report
      2. Observer rating
   5. Leader satisfaction (x point scales, x high)
      1. Self report
      2. Observer rating

Results

1. Table of measures for 1-way vs. 2-way
2. Report results for measures for each experimental hypotheses (see above)
3. One-way communication
   1. Center saturated task performance hindered, less satisfied.
   2. Peripheral members very dependent more frustrated.
   3. Takes less time
4. Two-way communication
   1. Less saturation better task performance, more satisfied.
   2. Peripheral members less dependent less frustrated.
   3. Takes longer

Discussion

1. Implication of results for hypotheses (confirmed/disconfirmed).
2. Implication of results for communication network theories
3. Speculation about why hypotheses confirmed/disconfirmed.