Title: Interaction Value Organizational Assessment correlation to Financial Performance Data
1Interaction Value Organizational Assessment
correlation toFinancial Performance Data
- Walid Nasrallah
- American University of Beirut
- WEHIA 2005
- Essex UK
2Overview
- Introduction
- Concepts of Interaction Value Analysis
- Survey Results
- Future
3The Big Picture
- What are we trying to do?
- Make Contingent Org. Design practical
- By considering multiple factors at once
- Why?
- Because things are changing
- But nothing changes
4What will I do today?
- Contrast two systems
- Multi-contingency by rules MYCIN
- Network-model-based (IVA)
- Partial validation
- by correlation to financial measures
- Speculation why these results ?
5Interaction Value Analysis
- Rational optimizing agents
- Described by six parameters
- Two ways to optimize ? org. forms
- Global optimum ? structured/autocratic
- Nash equilibrium ? self-org. / ad-hoc
- Model mimics org. theory results
6Interacting Agents
7Allocate Attention
8Generating Value
- Organizations Effectiveness
- Sum over all possible pairs of people (i and j)
of - Frequency of communication (pij)
- x Value of successful communication (hij)
- x Probability of successful completion (sij)
9Six Model Parameters
- Number of distinct org. roles Diversity
- Highest-to-lowest skill ratio Differentiation
- Role cooperation need Interdependence
- M/M/1 priority queue w/ reneging gives
- Arrival to departure ratio Load
- Reneging rate to service rate Urgency
- Criterion for choosing whom to serve Climate
10Diversity independent roles
x
KEY3x3 matrix 3 roles ? Low Diversity6x6
matrix 6 roles ? Medium Diversity
11Differentiation cost of settling
- One parameter fully determines H matrix.
- How
- Comparative idealization
- h(my best) h(your best) h(his best)
- Relational idealization
- h(1st)/h(2nd)h(2nd)/h(3rd) etc.
- No special treatment for value of self-interaction
12Failure mode 1 Interdependence
- Sij s1ij x s2ij Probability of failure from
over-use x Prob. of failure from waiting too
long - Over-use (s1)
- Depends only on pij
- Models neglect of other obligations
- too long (s2)
- Queuing (next slide)
13Failure Mode 2 impatient queuing
- M/M/1 queue gives failure probability at
steady state - IF response time f(load) gtimpatience time
f(urgency)THEN FAILURE - Load depends on aggregate demand
- i.e. column sum of matrix ?pij for a given j.
14Failure Mode 2. priority
- Priority reflects org. Climate
- Who gets served first?
- First come Disciplined climate
- Least Needy Capitalist climate
- Most Needy Fraternal climate
15Optimizing Effectiveness (F)
17.8
H matrix (given)
S matrix f(P)
P matrix (free variable)
18.75
16Global optimum (structured org.) vs. Nash
equilibrium (ad-hoc org.)
50
45
40
35
30
25
Percent of Time Allocated pij
20
15
10
5
0
1
2
3
4
5
6
Members Ranking of Interaction Partners
Legend
17Fraternal Climate
Climate Fraternal
Urgency Medium
Interdep. High
Diversity Low
18Capitalist Climate
Climate Capitalist
Urgency Medium
Interdep. Medium
Diversity Low
19vs. Multi-Contingency
20(No Transcript)
21Survey Description
- 23 companies surveyed
- All Publicly listed
- on a small Middle-Eastern exchange
- Directed Interview w/ CEO
- Published data on Return on Investment (ROA)
22Results
23Discussion
- We have a correlation
- T-test is OK (90 confidence level )
- Not as good as OrgCon model
- Excluding uncharted environments
- Tracked ROA growth vs. average ROA
- Why?
24Possible conclusions
- Can never control for all effects
- Fishing for right indicator?
- Mature vs. developing economy?
- IVA-subset measures opposing effect?
25Future Research Directions
- Must do more surveys
- Cross-reference old surveys
- Represent change/uncertainty in IVA
26THE END