Transcript Simplification Issues

Global Surveying via the Web: Better,
Faster, Cheaper ... and More Chaotic
Presented by
L. Allen Slade and Nicholas Mills
at the
Society for Industrial and Organizational Psychology,
14th Annual Conference
April 30 - May 2, 1999
Overview
• Introduction - The Ford Pulse Survey System
• Globalization
• Chaos in Surveys
• Electronic Surveying
Survey Quality Standards
• Trustworthiness
– Provide valid and reliable data.
• Usefulness
– Support effective decisions.
• Efficiency
– Use least resources possible.
The Pulse Vision
To pursue feedback about people
as aggressively as
we pursue feedback about profits and products.
-----------------------------------------------------• Provide a people metric for management.
• Support work-group level feedback and action
planning.
Pulse Core Dimensions
Dimension Name (# Questions, 
•
•
•
•
•
•
•
Training & Development (4, .84)
Stress (4, .87)
Reward & Recognition (4, .84)
Workload (4, .82)
Job & Company (4, .80)
Quality (5, .80)
Work Group/Teamwork (4, .73)
•
•
•
•
•
Supervision (6, .91)
Empowerment (5, .83)
Diversity (6, .84)
Business Issues (10, .88)
Survey Process (4, .92)
Pulse Process:
Supplemental Questions
•Country Questions - 5 (optional)
•Function Questions (optional):
– 3 levels
– 5 Questions per level (optional)
– Up to 2 functions per employee
Some Key Decisions
• Web Browser vs. E-mail Based Surveys
• Anonymity vs. Integrity
• Timing is Key
• Universal Suffrage vs. Efficiency
Survey Support Tools
• Web Survey System
– Central Item Bank
• Promote common items
• Allows data sharing
– Survey Results Data Base
• Store survey results in Ford data base
• Ford software generates reports from results data base
• Allows quicker and cheaper reaction time to requests for survey
information
• Maintain confidentiality and anonymity - no individual data on
Company system - rather an electronic storage of paper reports we
currently get from vendor
Does Technology Pay Off?
96 Census 97 Sample 98 Census
(Web)
(Web)
(Paper)
Response Rate
(n of surveys)
65.9%
(70,000)
Time for Initial Reports 3 months
(200)
(n of Reports)
Time for All Reports
(n of Reports)
Cost Savings
5 months
(7,500)
50%
(4,000)
63.4%
(66,000)
1 month
(100)
18 days
(600)
2 months
(3,000)
33 days
(22,000)
28.5%
vs. 1996
Linguistic and Cultural Diversity
on the Pulse Survey
• Chinese (2)
• Czech
• Dutch
• English (2)
• Finnish
• French (2)
• German
. . . in 46 Countries
• Hungarian
• Italian
• Japanese
• Polish
• Portuguese (2)
• Spanish (4)
• Thai
Impact of Language, Culture and Distance on
Employee Surveys
• Survey Content
– Translation of words
– Translation of constructs
• Cultural Differences in Survey Use
– Trust of Management
– Familiarity with Surveys
– Use and Appropriateness of Anonymous Feedback
• Delays Caused by Distance
Effective Management of
Linguistic and Cultural Diversity
• Goal = Survey robust in design, interpretability and
acceptance across Ford
• Communication among Survey Coordinators
– Monthly Conference Call
– Annual Global Conference
– Incessant e-mails
Translation:
Standards
• Use of External Translators and Internal Reviewers
• Standards for Translators
– Certification
– Survey Translation Experience
– Ford Experience - 3 years or more
• Standards for Reviewers
– good understanding of country’s language & culture
– reading and writing proficiency in both English and target
language
– orally communicate effectively in target language
– knowledge of survey process
Translation:
Process
• Web Interface
• Translation Steps
– Question approved in English.
– Translator does initial translation.
– Lead translator approves translation
– Reviewer checks phrasing and Ford lingo.
– “Back-translation” - second/external review, not blind
back translation.
Translation:
The Need for Statistical Analysis
• Maintain translator and reviewer
accountability
• Depersonalize criticism
• Create Order out of (n questions*n
languages) questions to review
Simplification
• Principles
– Try to please everybody and you will please nobody.
– Common as possible, different only when necessary.
• Simplification Planned for the 1999 Pulse
– Limiting supplemental (non-core)
– Reducing demographic questions
– Simplifying report definitions
– Reducing number of preliminary reports
• Growth is natural - Pruning is necessary.
Causes of Chaos
• Speed of Change
–In the external environment
–Of management expectations
• “Chaos Among our Customers”
• Increased Complexity
–Globalization
–Centralization
Reactions to Chaos
Ineffective Reactions
• Panic
• Over-planning
Effective Reactions
• Simplification
• Effective Management of Chaos
• Speed (Electronic Surveying)
Effective Management of Chaos
Model
Planning
Staff Skills
Leadership
Focus
From:
Scientific Model:
 Understand, Predict &
Control
Plan, Then Manage to Plan
 Contingency Planning
 Minimize Time in Plan
 Missing a Deadline is
Catastrophe
 Technical Expertise
 Accept Direction
 Manage People
 Plan Projects
 Image of Control
To:
Political Model:
A nalyze Possibilities &
Influence Outcomes
Plan, But Expect Change
+ Issue Management
Plan Slack Time
U se Internal Deadlines as a
Tool For Influence
+ Durability and Innovation
A ccept Empowerment
+ Lead People
+ Manage Projects
Image of Calm in the Storm
Ford Pulse: Web Process Flow
ID Survey
Participants
Survey Design
•Timing
•Participation
•Questions
Survey
Administration
Vendor
Processes
Data
Organization
Mapping
Report
Definition
Translation
ID Managers
Receiving
Reports
Ford Intranet Server
www.pulse.ford.com
Database/Software
•Text - Questions, instructions, menus, etc.
•Organization Mapping Information
•Report Definitions
•Processed Results (Group Data)
Vendor Server
Inside Ford firewall
•Oracle Database
•Raw Survey Data
•Processing and Data
Cleaning Rules
Report
Delivery
•Browser
•Print File
Process Flow =
2-Way Data =
Exchange
1-Way Data =
Exchange