| |
DSS Applications Should Shed New Light

"The diversity of functions within DSS
products makes it difficult to group all
the products into one category. By un-
derstanding, the needs of the. users and
translating those needs into software

This is part two
of a three-part
series on execu-
tive information

systems and de- | functions, there emerge five distinet
cision support types of decision support products:
systems—tech- spreadsheet products, EIS, business-mod-
nologies that en-  eling products, general DSS products
able executives and statistical-analysis products.

B ki to gain direct Of the five, spreadsheets are the most
access to infor- widely used. A spreadsheet enables us-

J AMES mation. ers to perform the first level of analysis

ba Decision sup- and to develop simple models.

MARTIN port. systems Typically, spreadsheet data is not
(DSS) can be accessed through a data dictionary; the
considered

user has the freedom to perform any
among the industry's many terms or

concepts that are either hard to grasp
or are constantly being redefined. Sim-
ply stated, a DSS is any system that can
be used to increase the efficiency and
effectiveness of the decision-making pro-
cess.

Yet, the concept of a DSS remains
confusing to many people. To some, it
means end-user computing, personal
"computing or executive information sys-

tems (ES). However, these terms are
not synonyms for DSS.

"The general characteristics of a DSS
distinguish it from other systems. Data
in a DSS is more aggregate and less
record-oriented than information sys-
tems (IS) applications. DSS data is

Using decision support tools,
managers could analyze sales

in terms of product lines,
ajii F

PG WEEKNAPPLIGATION DEVELOPMEN

"Decision Support Systems
Let Managers Examine Data in Multiple Dimensions

' Forecastii
B Techniguss.

MA 1, 1989

on a Problem

programming skills are often reguired in
order to make effective use of the |
language. |
The focus of business-modeling prod-
ucts is on a robust, modeling facility
with analysis capabilities and on tools
to transfer data into the modeling envi-
ronment from a variety of sources. Rep-
resentative business-modeling products
are IFPS/Plus, from Execucom Systems |
Corp., and System W, from Comshare.
The fourth category is the general
DSS product. These products are similar
to but more powerful than business:
modeling tools, Among the distinguish-
ing characteristics of general DSS prod-
ucts are the facilities provided to man-
age multidimensional data and the ease
in which dala can be added to the sys-
tem. These products are intended to
manage larger databases and serve a
broader set of user needs.
These types of DSS products also pro-
vide facilities for creating highly format-
ted reports and a variety of graphic dis-
plays. A very robust procedural lan-
guage is provided that reguires
proficient skill for effective use. The lan-
guage can be used to write user-oriented
procedures such as screen data entry, '
processing and reporting. Representative
products in this category are Express,
from Information Resources Inc., and

general DSS or business-modeling tool.
As described last week, executive in-
formation systems are specifically de-
signed to help executives gain insights
and track critical success factors. The
focus of an EIS is to aid a decision mak-
er in assimilating information guickly
and identifying problems or opportuni-
ties, not as an aid in problem analysis
or resolution. In many corporations,
well-developed EIS have replaced the
traditional periodic executive summary
reports. Representative products in this
category include Command Center, from
Pilot Executive Software, and Com-
mander EIS, from Comshare Inc.
Business-modeling products are used

...0r revlew sales data in "

multiple other dimensions. ||
Revlew Sales ; ij s ;
Review
Sales by
Review gije Customer.
Sales by i šišnimi
Region E

stored in the form of multidimensional
arrays. The interaction between the DSS
and the user is very flexible, immediate
and accessible to a non-IS professional.
The DSS is generally used to identify
and understand problems and opportu-
nities |

Decision support systems evolved
gradually from more structured comput-
er applications. As corporations succeed-
ed in automating wellstructured opera-
tional systems, they began to tackle the
additional reporting reguirements of
those applications. Users (analysts and
executives alike) began to demand tools
that provided flexibility and funetionali-
ty, not just reporting, to support their
problem-identification and decision-mak-
ing needs. These evolving needs were

Stratagem, from Computer Associates
International Inc.
The last category, statistical-analysis |
products, consists of products that are
used for the statistical analysis and
modeling of data. Their libraries of sta- HA
tistical routines are more extensive than
those found in the general DSS products.
Where a general DSS product may
have one or two types of regression, a
statistical analysis product has several,
including linear, multiple, polynomial
and stepwise. |
Generally, users are expected to un-
derstand the ALA of the
technigue and how to interpret the re- |
sults. Usually these are used by trained |
analysts. SAS, frora SAS Institute Ine., |

Review Marketing

Assumptions

met by DSS producis.

One important characteristic of DSS
products is their ability to support a
powerlul dala model and retrieve data
from a variety of production database-
management systems (DBMSs) as well
as [rom external sources. The decision
support structure must aid in the pro-
cesses of ad hoc dueries and analysis,
and must be flexible enough for the user
to add or delete data as the problem is
better understood.

operation, logically valid or invalid, on
any data in the spreadsheet, The lack of
control mechanisms has caused many
managers to become wary of informa-
tion reported from spreadsheets.
Spreadsheet facilities include func-

The interface must be responsive, al-
lowing a user, after reviewing pis
swer to a guestion, to immediately for-

One important characteristic of DSS products is their abili
to support a powerful data model and retrieve data
production DBMSs as well as from external

—— —iiia,: o -

and SPSS-X, frora SPSS Inc.
products in this category. mev
Micro products are now beginning to
provide capabilities previously found
only on the mainframe. Mainframe DSS
products are being ported to the micro.
As the power of the Micro continues to
be expanded and enhanced, the applica-
tions that. Can run stand-alone on the
micro will be larger and more complex.
ne K s Una that the definition of
? further l
ee Ninen r modified
revolution con

Jrom
sources.

hi a analysts to represent rela-

tionships between data elements in the

prodieni set. Models may be as straight-

lorward ni a prolit-and-loss income
ora

, ok at the Data
managers
product For net ora jo im a DSS