Study: Case The Quality Assurance Operation (QAO)
Unit of international high-technology manufacturing firm f t i fi
Quality / inspection / repair function
Start-up & shut-down within about a year
Change in formal structure & mandate Changes in reporting of events
Methods
Participant-observation (1 year) Structured interviews Collection of internal company documents Qualitative & quantitative analysis
Installation Vice-President Installation Director
Vice-President of Mfg. g Director of Mfg. Quality
Old VicePresident Old Director
QAO Manager Installation Manager ICs
Operations Engineers
Process Engineers
Installers
Helpers
Formal lines of hierarchical reporting within the Quality Assurance Operation
Positive vs. Negative reporting
to Boss to Boss's Boss
Main reporting categories p g g
Root Cause Quality problems / cost savings
% "Good" vs. "Not so good" reporting
To Boss by Boss to next level: Avg. Avg 81 % positive Avg. 92 % positive
Perceptions of "how good things really were," based on interaction data
Internal interactions External interactions Avg. Avg 62 % positive Avg. 54 % positive
Reporting about "Root Cause"
Reporting Categories by Hierarchical Level g VP Mfg. Level Good Root Cause Change in QAO's mandate Root Cause success Not so Good Few problems solved at root cause Director Mfg. Quality Level Good Root Cause Change in QAO's mandate Root Cause problems solved Root Cause analysis QAO working self out of a job Not so Good Few problems solved at root cause Root Cause cost savings QAO Manager Level Good Root Cause Category Valences Given Implied Consistent?
+ + + -
0 + + -
n/a y y y
+ + + + + -
0 + + 0 + 0
n/a y y n/a y y n/a
+
0
n/a
Reporting "Problems"
Solved Positive (good to report) Neg ve Negative (bad to report) 1 307 272 46 Not Solved
Reporting "Tours"
Features of "Tour" (Type of person receiving a tour) VP Mfg. Level Good to Report (n 2) (n=2) External/Customer Very High Level Director Mfg. Quality Level Good to Report (n=5) External/Customer Very High Level Significant/High Level Not so Good to Report (n=3) Low Level Internal I l QAO Manager Level Good to Report (n=11) Significant/High Level Internal Low Level Someone (unspecified) External/Customer n=16 instances # Instances per feature Avg % Instances per feature 100 % 2 2 40 % 4 2 1 75 % 3 2 21.5 % 4 3 3 3 2
Summary of QAO Case
Positive bias in generating information about organizational events
Language use & information influenced by organizational structure (role, mandate, etc.)
People tend to report what they are expected to report
Structural constraints on formal hierarchical communication at several levels of analysis
Change in mandate from "Quality Problems / Cost Savings" to "Root Cause" % Positive vs. Negative reporting Internal structure of "Root Cause" category
Sub-categories Given & implied category valences Meaning of a concept changes at higher reporting levels Information purified as it moves up the ranks
Everything "pulls in the same direction"
Conclusions
Contributions
Influence of formal structure on organizational information processing Empirical methods & supporting evidence
Limitations
Si l case Single Range of interpretations