New Jersey Institute of Technology (NJIT)
Computer and Information Science Department (CIS)
CIS677:
Information System Principles
Professor: Michael
Bieber
CIS677 - Notes for Lecture 7 - Professor Bieber
Information Value
Guest:
Murray Turoff, Distinguished Professor, NJIT CIS Department
(turoff@adm.njit.edu;
http://eies.njit.edu/~turoff)
Definition of Information
When Information Has Value
- surprise: the information tells me something I didn't
know
- change in decision: I actually decide something differently
because of this new information
- payoff: the change results in an increased payoff
Example 1. Value of developing a market research information
system:
- Suppose the information it produces changes our view of the
market. Suppose it tells us that the probability of $50m sales for
a $3m newspaper advertising campaign
- Altered Decisions:suppose this causes us to use TV instead of
newspaper advertising
- increase in profit?
Example 2: Value of determining the best sales price:
- Surprises: tells us to sell our product at a lower price,
increase demand by 25%
- Altered Decisions: we would produce more, but our factory is
at full capacity, and 25% increase in sales won't justify building
a new factory.
- Since capacity is low, the new information doesn't matter
Example 3: value of purchasing a shipping (delivery) information
system:
- Surprises: tells us which shipper to use for a given customer
address and priority
- Altered Decisions: we used to always use one shipper
- Increase profits, or just pass these savings on to the
customer?
Uncertainty and Equivocality/Ambiguity
Daft & Lengel: Ambiguity
{Daft, R. L., and Lengel, R. H., "Organizational
Information Requirements, Media Richness and Structural Design,"
Management Science 32(5), 1986, 554-571.}
Scott M. asks regarding ambiguity and
equivocality (16:10):
If you look in the dictionary, there is no
definition of "equivocality". It seems that they made this word
up.
Rational Actor
Calculating the Value of Information
Normative Value of Information
Normative: rational, theoretically correct value of
information.
Assume the information you have is perfect: 100% accurate and
unbiased; thus optimal result
Value of Info = Expected Value of Information - Cost of
Information
Expected Value (written on overhead):
25% probability * $100m outcome = $25m
75% probability * $200m outcome = $150m
Expected Value = $175m
Calculating the Realistic/Revealed Value of Information
Value of Information =
Payoff with the information - Payoff without the information - Cost
to get information
Assumes:
- you can measure the results
- you can deal with qualitative results
- you know the cost of the information
- the experiments/prototype approximates reality
Note:
This measurement doesn't require people to be rational, as it only
measures results.This, e.g., allows people to satisfice instead of
coming up with the optimal solution.
Susan C. asks, regarding needing to know the objective for your
prototype (34:35):
You say the objective is for a prototype, then you set up
another objective for building the final system, so you need to set
up two objectives?
Susan C. continues (35:25):
And that objective covers the cost of what you can
afford?
Calculating the Subjective Value of Information
Estimate or guess the value of information to the best of your
ability, perhaps based on your past experiences.
example approaches:
- multi-criteria decision analysis
- subjective value through ranking alternatives subjectively;
through what-if analysis; by asking: how much would you be willing
to pay for this information or result
- by asking the question: How much would you be willing to
pay for this report/feature
- ranking alternatives qualitatively
* Scaling Theory
* Common Biases
(discussed more during the session: Cognitive Aspects of
IS)
Information Attributes & Characteristics
{slide/handout: Information Characteristics}
{slide/handout: Attributes of Quality Information}
Media Richness
{Daft & Lengel - see reference above}
Value of Information & Designing Information Systems
last updated: 2/15/2000
This page: http://www.cis.njit.edu/~bieber/CIS677/lecture-notes/lecture7.html