Professional Documents
Culture Documents
07 Evaluation
07 Evaluation
Institute of Information Systems and Marketing (IISM), Karlsruhe Service Research Institute (KSRI)
utilization
technology
law
society
2 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Outline of the lecture
Basics, Filtering
Transformation Regression, Cluster Analysis
3 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
What is the value of an information?
5 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basics
Criteria to evaluate information
Decisions in uncertainty
Methods and modelling
Search for information
6 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Criteria of information evaluation
Recomm
ender
Meaning, Importance, Relevance system
Use y
Testing b
experts
Reliability
t a ti s tical
S
ds
metho
Integrity, Correctness
Timeliness …
7 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Evaluation models in practice
•Utility analysis
• AHP
2% Andere Methoden •…
Other methods
(Hedonistisches
(Hedonistic Model,
32% Modell, Real Options)
Real Options)
0% Nutzwert-
Utility analysis •End value
5% 20% Analysen
11%
•Capital value
Dynamische Methoden
Dynamical Methods •Present value
der
of investment
71% calculation
Investitionsrechnung •Internet interest
68%
Statische
Statistical Methoden
Methods rate
der
of investment
calculation
Investitionsrechnung •…
Middle class
Mittelstand Big companies
Großunternehmen •Comparison of costs
(Multiple answers or [Bernroider und Koch 2000] •Comparison of revenue
abstentions possible) •Rentability
•Amortisation period
•…
8 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Music is
in the air
9 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
10 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Utility analysis
Aim Zj
Investment
A1
Alternative Investment
Ai A2
Investment
A3
11 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Utility analysis
Aim Zj
Investment
A1
Alternative Investment
Ai A2
Investment
A3
12 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Utility analysis
Aim Zj
30 Mio. +20
Investment satisfying good satisfying
A1 fulfilled 2 * 6 = 12 4 * 3 = 12 2*1=2 26
38Mio. -50
Alternative Investment fulfilled Very good sufficient sufficient
A2 6 * 6 = 36 0*3=0 0*1=0 36
Ai
35 Mio. +40
Investment fulfilled good Very good good
A3 4 * 6 = 24 6 * 3 = 18 4*1=4 46
[Dilbert.com]
14 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process [Saaty 1980]
Decision
15 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
Calculation of the relative weighting factor by comparing and
evaluating with the following scale
Value aij
1 Factors i and j have the same importance
3 Factor i is slightly more important than factor j
5 Factor i is more important than factor j
7 Factor i is significantly more important thanfactor j
9 Factor i is definitely more important than factor j
Aim Zj
Capital value 1 2 6
Change in
Aim Zj personell 1/2 1 3
Appearance in
public 1/6 1/3 1
16 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
The weighting factors are between 0 and 1 which means that a
normation is necessary – therefore every comparison value is divided
by the sum of the columns:
Aim Zj
Capital value 1 2 6
Change in
Aim Zj
personell
1/2 1 3
Appearance in
public
1/6 1/3 1
17 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
The weighting factors are between 0 and 1 which means that a
normation is necessary – therefore every comparison value is divided
by the sum of the columns:
Aim Zj
Change in
Aim Zj
personell
0,3 0,3 0,3
Appearance in
public
0,1 0,1 0,1
18 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
The weighting factors are the mean of every row:
Aim Zj
Change in
Aim Zj
personell
0,3 0,3 0,3 0,3
Appearance in
public
0,1 0,1 0,1 0,1
Result: The capital value has a relative importance of 60% for the
investment decision, while the change in personell has a relevance of
30% and the appearance in public only 10%.
19 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
Consistency testing of the comparison
Max. value
20 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
Consistency testing of comparisons
Aim Zj
Capital value 1 2 x
Change in
Aim Zj
personell 1/2 1 3
Appearance in
public 1/x 1/3 1
21 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
Practical example: Evaluation of Cloud-Computing for
operational applications (1/3)
Flow chart of application portfolio assessment for cloud
[http://www.ibm.com/developerworks/cloud/library/cl-assessport/]
22 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
Practical example: Evaluation of Cloud-Computing for
operational applications (2/3)
Illustrative criteria hierarchy
[http://www.ibm.com/developerworks/cloud/library/cl-assessport/]
23 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
AHP – Analytical Hierarchy Process
Practical example: Evaluation of Cloud-Computing for
operational applications (3/3)
Schematic representation of AHP for evaluating cloud technical fitment
[http://www.ibm.com/developerworks/cloud/library/cl-assessport/]
24 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basics
Criteria to evaluate information
Decisions in uncertainty
Methods and modelling
Search for information
25 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model of the economical decision
making theory
Object system
Information Events
Target system
Gesetzliche Kapitalwert Veränderung Erscheinungs- Nutzgröße je Gesetzliche Kapitalwert Veränderung Erscheinungs- Nutzgröß e je
Vorschriften w1 = 6 der Belegschaft bild in der Alternative Vorschriften w1 = 6 der Belegschaft bild in der Alternative
ergonomics
w2 = 3 Öffentlichkeit w2 = 3 Öffentlichkeit
w3 = 1 w3 = 1
30 Mio. +20 30 Mio. +20
Investition befriedigend gut befriedigend Investition befriedigend gut befriedigend
erfüllt 26 erfüllt 26
A1 2 * 6 = 12 4 * 3 = 12 2* 1=2 A1 2 * 6 = 12 4 * 3 = 12 2*1=2
Alternativen
Ai
Investition
A2
erfüllt
38Mio.
sehr gut
6 * 6 = 36
-50
ausreichend
0*3=0
ausreichend
0* 1=0
36 power Alternativen
Ai
Investition
A2
erfüllt
38Mio.
sehr gut
6 * 6 = 36
-50
ausreichend
0*3=0
ausreichend
0*1=0
36
Investition
35 Mio.
gut
+40
sehr gut gut
weight Subject system Investition
35 Mio.
gut
+40
sehr gut gut
erfüllt erfüllt
A3 46 A3 46
4 * 6 = 24 6 * 3 = 18 4* 1=4 4 * 6 = 24 6 * 3 = 18 4*1=4
26 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Decision making process
27 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Decision making in uncertainty
Risk
Current or future states are not known
Probabilities are known
Methodical basis: Demand-use-maximisation
Uncertainty
Probabilities are not known
Methods: Minimax, Maximin etc.
28 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Example: Expectancy value vs. Minimax-rule
50% 50% ? ?
E1 E2 E1 E2 min
A1 42 32 37 A1 42 32 32 max
A2 63 21 42 max
A2 63 21 21
29 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basics
Methods and Modelling
Evaluation of information systems
Evaluation of IT systems
Search for information
30 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
Assumption
Information is not for free
31 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
w S j = 1
m
j=1
32 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
Example:
Return Matrix of the decision maker before information acquisition
Environment state
Expectation of profits
S1 S2 ... Sm
a priori
m
.. .. .. .. ..
. . . . .
m
33 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
A1 120 90 105
(Risk neutral) decision maker chooses the alternative with the highest
expectation of profits m
E(G)oI = max w(S j ) Gij
i
j=1
34 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
35 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
is assumed to be known
36 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
w(Sj | Ii) is the probability that the environment state Sj occurs under the
condition that information result Ii occurs
Determination of w(Sj | Ii) with the BayesTheorem
w(Ii | S j ) w(S j )
w(S j |Ii ) = m
(j=1,...,m) (i=1,...,n)
w(I | S ) w(S )
j=1
i j j
n m m
E(G)mI = w(Ii ) max w(S j |Ii ) Gij i i j
w(I ) = w(I | S ) w(S )
j
i=1
i
j=1 j=1
37 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
Assumption:
Bayes
w(I1 | S1)=0,7 w(S1 | I1)=0,7778
Theorem
w(I2 | S1)=0,3 w(S1 | I2)=0,2778
w(I1 | S2)=0,2 w(S
w(S1) = 0,52 | I1)=0,2222
w(I2 | S2)=0,8 w(S2) = 0,52 | I2)=0,7222
w(S
38 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
A1 120 90 113,3333
A1 120 90 98,18
39 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
40 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Basic model for the evaluation of information
Information value IW
IW = 106 – 105 = 1
Implication:
E(G<1> | Ii) – The return expectation value of the best (<1>) alternative with the
information result Ii
42 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Extension of the basic model
The drop out probability p reduces the expected value of the alternative
with the highest return expectation value: (1-p) * E(G<1>| Ii)
In case that the first alternative drops out ( probability p) only the second
best alternative with expected value p*(1-p)* E(G<2>| Ii) can be realized
E(G | Ii) = (1-p)*E(G<1>| Ii) + p*(1-p)* E(G<2>| Ii) + p²* (1-p)* E(G<3>| Ii)
+ p³ *(1-p)* E(G<4>| Ii) +... =
E (G | I i ) k 1 (1 p ) p k 1 E (G k | I i )
n
43 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Extension of the basic model
44 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Extension of the basic model
Summary
Opportunity costs occur because of delays in the life cycle due to information
aggregation
The opportunity costs result in missed returns due to the drop out of an
alternative
45 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)
Literature
Bamberg, Coenenberg, Krapp 2012: Betriebswirtschaftliche Entscheidungslehre.
Vahlen Verlag.
Saaty, T.L. (1980). The Analytic Hierarchy Process. McGraw-Hill, New York.
Saaty, T.L. (2000). Fundamentals of Decision Making and Priority Theory with the Analytic
Hiearchy Process. 2. Auflage, Pittsburgh.
Sassone, P.G. (1987). ACM Transactions on Office Information Systems. Vol. 5,
Cost-Benefit Methodology for Office Systems. 3, S. 273-289.
Benroider, E. und Koch, S.: Entscheidungsfindung bei der Auswahl betriebswirtschaftlicher
Standardsoftware – Ergebnisse einer empirischen Untersuchung in österreichischen
Unternehmen. Wirtschaftsinformatik. 42(4), S. 329-338.
Buxmann, P. (2001). Informationsmanagement in vernetzten Unternehmen. Gabler Verlag.
Laux, H. (1998). Entscheidungstheorie. Springer Verlag.
46 Management of Information Systems – Prof. Christof Weinhardt, Ewa Lux Institute of Information Systems and Marketing (IISM)