Predict21

1
Predict-21 Performance Predictor Tool
Randy Brou, Ph.D. Navy Personnel Research,
Studies, and Technology (BUPERS-1) Bureau of
Naval Personnel Millington, TN Presented at the
Eighth Annual Navy Workforce Research Analysis
Conference Arlington, VA 5 May 2008
The opinions expressed are those of the authors.
They are not official and do not represent the
views of the U.S. Navy Department
2
Outline

• Function
• Issues
• Objectives
• Approach
• Deliverables Payoff

3
Key Capabilities of Enhanced, Integrated HR
Decision-making
Integrated policy, process and resource
allocation analysis
Richer description of total force and work
requirements
Career planning
Job negotiation
Optimal person-job-team matching
Unit-level decision making
Unit planning
Right PersonRight Place/WorkRight TeamRight
Cost
4
Determine FIT

• Search across the Total Force
• Identify qualified candidates for Skills Fit
• Assess for Individual / Team FIT

Readiness
Resource Planner
HR Requirements
Workforce Data Bank
Candidates
Measure FIT in the context of an existing team
5
What Can We Predict Today?

• Success in training

Today we predict training outcomes
6
What Do We Need to Predict?

• Success in training
• Success in career
• Success in teams
• Success during emergencies

We need to predict performance in the Fleet
6
7
Technical Objective

• Develop quantifiable metrics of team processes /
  performance
• Investigate links between cognitive
  non-cognitive characteristics and team processes
  / performance
• Simulate teams performing tasks to measure team
  member FIT

Need measures to quantify predict FIT
7
8
Virtual Assessment

• 3 ways to view virtual assessment tools
• Assess team effectiveness
• Needs Task simulation with opportunities to
  measure team behaviors

Virtual assessments can be used at multiple levels
9
Virtual Assessment

• 3 ways to view virtual assessment tools
• Assess team effectiveness
• Needs Task simulation with opportunities to
  measure team behaviors
• Assess individual teamwork skills
• Additional Needs Agents capable of receiving
  inputs and responding realistically

Virtual assessments can be used at multiple levels
10
Virtual Assessment

• 3 ways to view virtual assessment tools
• Assess team effectiveness
• Needs Task simulation with opportunities to
  measure team behaviors
• Assess individual teamwork skills
• Additional Needs Agents capable of receiving
  inputs and responding realistically
• Predict team performance
• Additional Needs Accurate models of real people
  driving agent behavior

Virtual assessments can be used at multiple levels
11
Enhancing Decision-Making
New measures will enhance assignment decision
making
11
12
Salas Big 5 of Teamwork
Team Orientation
Team Leadership
Mutual Performance Monitoring
Adaptability
Back-up Behavior
Shared Mental Models
Closed-Loop Communication
Mutual Trust
Existing research will guide our efforts
13
Technical Approach

• Experiments
• Individuals complete cognitive / noncognitive
  assessment tests
• Individuals assigned to teams
• Control composition of traits of interest
• Teams complete tasks designed to emphasize
  particular team processes
• Performance process metrics analyzed as a
  function of team composition

Experiments will result in model of optimal team
composition
14
Technical Approach

• Virtual Experiments
• Individuals complete cognitive / noncognitive
  assessment tests
• Individuals complete task with computer agents
• Complete tasks designed to emphasize particular
  team processes
• Performance process metrics analyzed as a
  function of individual traits
• Compared to real team setting results
• Use results to make team simulator

Further experiments will allow team performance
to be simulated
15
Project Deliverables

• Predict-21 will provide two major deliverables
• Tool for predicting optimal team composition
• Virtual environment for testing team performance
• FY-08 deliverables will include development of
  testable composition model
• FY-09 deliverables will include a prototype
  composition tool prototype virtual environment

Development of decision support tools
15
16
Measures of Success and Payoff

• Predict-21 will provide guidance for forming
  optimized teams
• Predict-21 will impact NAVY success by informing
  optimal team design and classification,
  ultimately improving FIT

Predicting FIT will lead to superior unit
performance
16
17
Project Team
NPRST Team Dr. Randy Brou, Principal
Investigator Ronald Bearden Kimberly Crayton Sara
Rayne Daniel Riner Dr. Jacqueline Mottern
Academic / Contractor Team University of Central
Florida Dr. Eduardo Salas Mississippi State
University Dr. Stephanie Doane University of
Memphis Dr. Ron Landis University of Southern
California Dr. Stephen Read Dr. Lynn Miller
18
Contacts

• Dr. Randy Brou
• Institute for Selection Classification (NPRST)
• randy.brou_at_navy.mil 901 874-4955 DSN 882
• Dr. Jacqueline Mottern
• Director, Institute for Selection
  Classification (NPRST)
• jacqueline.mottern_at_navy.mil 901 874-4656, DSN
  882

19
Backup
20
Cognitive / Noncognitive Assessment

• Bell (2007) performed a meta-analysis on
  predictors of team performance and highlighted
• Lab setting predictors include
• General Mental Ability
• Emotional Intelligence
• Field setting predictors include
• Agreeableness
• Conscientiousness
• Openness to Experience
• Collectivism
• Preference for Teamwork