/
4-
>^—
"^^
^i
\
HD28
.M414
U).
133br
6Z
Center for Information Systems Research
Massachusetts
Institute of
Sloan School of
n
Technology
Management
Massachusetts Avenue
Cambridge, Massachusetts, 02139
JAN 111983
THE EFFECTS OF STRUCTURED DEVELOPMENT METHODS
ON THE JOB SATISFACTION OF PROGRAMMER/ANALYSTS;
A THEORETICAL MODEL
David K. Goldstein
May 1982
Sloan WP
//
CISR WP
D. K.
1330 - 82
//
90
Goldstein
1982
Center for Information Systems Research
Sloan School of Management
Massachusetts Institute of Technology
.
ABSTRACT
This paper presents
the
use
structured
of
programmer/analysts.
methods
structured
principle
the
task
-
-
is
development
the model,
that explains
methods
on
the independent variable
variable
-
the
role conflict,
role ambiguity,
job
test the model
job
use
of
satisfaction
of
intervening
skill
variety,
These are variables that will be significantly
and
satisfaction of programmer/analysts.
the theoretical model
the
-
task achievement,
affected by the introduction of structured methods
the
of
job satisfaction of
This can be best explained through several
identity, and autonomy.
affect
the
effects
the
hypothesized to have an overall positive effect on
dependent
programmer/analysts.
variables
In
theoretical model
a
will
This paper describes
and outlines a quas -exper iment that can
i
significantly
be
used
to
Over the past ten
been advocated as
a
various
of
of
types
structured
of
to structured systems analysis,
way of improving productivity
Proponents
development.
use
the
programming
structured
from
methods,
years,
structured
and
methods
quality
claim that they simplify
description and communication of systems requirements and
allows
phases
designs.
systems
to develop systems more quickly;
programmer/analysts
it
reduces systems errors, and leads to the development of systems that
meet user needs
(Goldstein,
on the job satisfaction of programmer/analysts
reduces the skill
-
received
has
a
level
decrease
programmer/analysts
of
in
(Kraft,
its effect
-
very
little
reduces
some
of
1977).
which
could
However, we could argue that the
job satisfaction.
use of structured methods reduces conflict between
users,
better
Some researchers have argued that the use of structured methods
attention.
to
also
1982a).
One potential consequence of the use of structured methods
lead
of
lifecycle and by providing a language that facilitates the
development
This
systems
in
systems development by providing guidelines for carrying out the
the
has
programmer/analysts
the ambiguity in systems development, and
and
leads to
increased job satisfaction.
Level
managers.
In
in
(Locke,
for
MIS
I976).
Turnover
is
of
special
MIS, due to the shortage of programmer/analysts and the high
cost of training new programmer/analysts.
decreases
job
benefits.
Alternatively,
satisfaction,
structured methods,
i
consequences
many studies job satisfaction has been shown to be negatively
related to absenteeism and turnover
importance
important
of job satisfaction could have
if
then
its
If
the use of structured
methods
personnel costs could outweigh its
job satisfaction
increases
with
the
use
of
this would provide further evidence for those advocating
ts use.
Page
2
This paper presents a model that explains the effects
structured
methods
variable
-
explained
the
-
use
structured
of
the job satisfaction of programmer/analysts.
through
In
the
-
is
methods
several
intervening
variables
This
can
a
set of environmental variables
best
be
variables that will be
-
methods
will
and
significantly affect the job satisfaction of programmer/analysts.
includes
of
have an overall positive effect on the principle dependent
to
significantly affected by the introduction of structured
also
use
on the job satisfaction of programmer/analysts.
model, the independent variable
hypothesized
the
of
The model
factors that could affect
-
the validity of the model.
This paper is divided into four sections.
research
relevant
the
second
discusses
theoretical
the
quasi -experimental
structured
on
design
that
The
methods
model.
first
section
reviews
and job satisfaction.
The
The
outlines
third
can be used to test the model.
a
The fourth
discusses the significance of the research.
REVIEW OF RELEVANT LITERATURE
There
development
is,
no research that directly examines the effects
is
however,
structured
methods
a
on
of
the
literature
are
satisfaction
also
in
a
few
that
discusses
the
There
impacts
of
systems development process and an even larger
body of literature that discusses
There
structured
on the job satisfaction of programmer/analysts.
large body
methods
of
the
articles
that
programmer/analysts.
Page 3
determinants
examine
the
of
job
satisfaction.
determinants
of job
Impacts of Structured Methods on Systems Development
Many authors have described how structured methods are used
Various
development.
1976),
types
structured
of
methods, such as H PO
design
(Jackson,
programmer/analysts
^^^^)
have been used to aid
197'*)
(I98O),
Canning
Jones
(1979a),
(197^),
describe how different structured methods are used
(1979)
data
and
,
systems analysis, design, and programming.
in
Goldstein (I98I), Mendes
Winters
Warnier,
1975;
(Jones,
I
structured design (Stevens, Myers, and Constantine,
structure
systems
in
and
systems
in
analysis.
Programmer/analysts use these methods to model both the functions
performed
and data used by
solution
will
do.
business, as well as what the proposed systems
articles,
the
In
communicate
their
understanding
systems analysis errors
problem
business
solution.
caused
users
the
or
They
better
are
These two factors lead to
a
misunderstanding
analysts
by
that
able
a
to
the business problem and the proposed
of
systems solution to systems users.
claim
systems analysis develop
in
problem.
better understanding of the business
authors
the
structured methods
using
programmer/analysts
a
misunderstanding
reduction in
users'
the
analysts' proposed
the
This should lead to the development of systems that
meet
better
the requirements of users.
Canning
Block
(1979b),
(1979)
discuss
In
system
will
and
(1978),
use of structured methods
in
Hamilton
and
the systems design
to
describe
how
a
They can describe the modules the system will use,
function.
the interfaces between
simplify
the
Menard
systems design, structured methods are used
phase.
developed.
Bernstein (1972),
modules,
and
the
data
structures
that
must
be
The authors claim that the design aids are used to structure and
the
easier to turn
programming
into
process.
programs
when
The
output
these
aids
Page k
of
are
the design process
used.
This
is
allows
organizations
productivity
errors
is
less
are
programmers
experienced
experienced
more
their
of
programming
management
use
to
programmers.
maintenance
reduced,
or
easier when structured methods are
is
increase the
to
The
number
of
simplified, and project
used
to
support
systems
des gn.
i
Canning
Baker
(197'+a,b),
(1972),
I
nmon
(1976),
and
Rader
(1978)
describe the impact of structured methods on the programming phase.
design and programming methods are used together.
cases
The
benefits of these methods in programming are the same as they are
the
methods.
programming methods are used to structure the programming process.
in
some related research, Kraft
use
structured
of
programming
personnel.
the programming task.
carried
be
to
and Greenbaum
out
in
design.
claim
(1979)
that
less
by
This makes it
experienced
easier
and
the
effect
the
of
for
and less expensive
They draw analogies between the effect of structured methods
programmer/analysts
The
such as structured programming and Chief
methods,
Programmer Teams, de-skill
(1977)
many
The design methods
programming
structured
produce modules that are programmed using
In
on
assembly line and scientific
management techniques on factory workers.
Determinants of Job Satisfaction
There
is
determinants.
great
a
Locke
deal
(1976)
of
research
provides
determinants of job satisfaction.
a
on
job
summary
satisfaction
of
and
its
the research on the
He divides the determinants
events
into
and conditions, and agents.
Among
events
and
conditions,
he
describes
the
effects
on
job
satisfaction of the job itself, the pay, promotion, recognition, and working
conditions.
Hackman
and Oldham
(I98O)
Page 5
identify several characteristics of
the job
itself that positively affect job satisfaction.
They are:
a
job requires a variety of
to which
var iety - The degree
carrying out the work, involving a number of
in
different activities
different skills and talents of the person.
Ski
1
1
requires completion of a
is,
doing a job from
that
Task identity - The degree to which a job
identifiable piece of work,
whole and
beginning to end with a visible outcome.
substantial
has
a
the job
Task significance - The degree to which
lives of other people, whether those people are in the
the
impact on
immediate organization or in the world at large.
freedom,
Autonomy - The degree to which the job provides substantial
and discretion to the individual in scheduling the work
independence,
and in determining the procedures to be used in carrying it out.
activity
the work
The degree to which carrying out
individual with direct and clear
the
job provides
(Hackman
information about the effectiveness of his or her performance
and Oldham, I98O, pages 78-79)
Job feedback
required by
-
the
•
Hdckman
Oldham found significant positive correlations between each of
and
these job characteristics and job satisfaction.
One aspect of work not considered by Hackman and Oldham is
of task achievement found
workers
in the job.
amount
the
Locke claims that the degree to which
overcome the challenges of their jobs positively affects their
can
He also stresses that
job satisfaction.
individual differences can moderate
the effects of these factors on job satisfaction.
Among
agents,
co-workers,
and
Locke
describes
the
effects
the organization on job satisfaction.
of structured methods changes the job performed by
can
change
their
sel
f
-percept on.
i
The
-percept on,
i
Inasmuch as the use
literature
on
the
it
impacts
of
could
also
the relationship of programmer/analysts to users and managers.
Role
constructs
that
ambiguity and role conflict
to
conflicting
et al.,
(Kahn,
measure the degree of ambiguity
subject
f
programmer/analysts,
structured methods suggests that the use of structured
affect
sel
of
in
a
demands
196^)
are
methods
two
job and the degree to which
from
Page 6
co-workers
and
a
worker
managers.
is
Both
constructs have been shown to be negatively
This
especially true for boundary spanning jobs
is
deal
of
related
intra- or
i
job
to
satisfaction.
jobs requiring
-
a
great
nterorgani zat onal contact.
i
Job Satisfaction of Programmer/Analysts
few researchers have attempted to examine
A
satisfaction
a
programmer/analysts.
in
reinforcement
needs
satisfaction of
highest
when
model
to
Awad
(1977)
examine
programmer/analysts.
Willoughby
model
programmers
to both
in
significant
job
(1972)
used
affecting
job
satisfaction
was
factors
found
high levels of ability utilization, achievement, advancement,
supervisor fairness were present.
differences
of
and Willoughby
the
creativity, recognition, responsibility, company
and
determinants
the
in
one
company.
needs between programmers and analysts.
correlation
social
status,
Awad applied the needs reinforcement
analysts
and
fairness,
found
He
no
He also found that a
between needs-reinforcement correspondence and job
sat! sf act ion.
Bostrom (I98O)
examined the effects of role conflict and role ambiguity
on job satisfaction of system designers.
He considered the effects of these
variables on 75 user-designer pairs involved
found
that
role
systems
in
maintenance.
He
conflict and role ambiguity were significantly negatively
correlated with job satisfaction.
Couger and Zawacki
the
effects
(I98I)
of
job
programmer/analysts.
They
programmers
and
analysts
used Hackman and Oldham's
characteristics
conducted
large
on
job
sample
model
to
examine
satisfaction
survey
of
with
1000
at many companies and government agencies.
They
a
found that characteristics of the job, as measured by the
potential, correlated positively with job satisfaction.
Page
7
job's
motivating
The job satisfaction literature identifies
determinants
satisfaction
of
factors
several
that
programmer/analysts
in
including
characteristics of their job and characteristics of their relationship
others
the
in
organization.
The
are
with
structured methods literature provides
insight into how the methods are used and how they might affect the jobs and
the interactions of programmer/analysts.
areas
will
drawn
be
on
next
In the
these
section,
two
develop a model of how the use of structured
to
methods affects the job satisfaction of programmer/analysts.
THEORETICAL MODEL
Figures
1
and
2
methods
structured
variable
-
overall
positive
present the theoretical model
satisfaction.
job
to
the use of structured
effect
on
methods
principal
the
satisfaction of programmer/analysts.
examining
variables
task
the
-
effects
the
of
-
and
autonomy
satisfaction in other studies.
they
are
programmer/analysts.
-
performance
variables
quality of their work.
methods
will
will
-
It
positively
2
is
hypothesized
dependent
1,
of
the independent
have
to
variable
task achievement,
use
-
an
the job
explained
by
intervening
variety,
skill
which have been shown to be related to job
This research will
significantly
Figure
Figure
independent variable on a set of
are significantly affected by the introduction
that
In
This effect can be best
role conflict, role ambiguity,
identity,
that relates the
related
relates
the
show that these variables
of
to
methods
and
satisfaction
in
structured
job
independent
variable
to
two
the
productivity
of programmer/analysts and the
is
hypothesized
that
the
use
of
structured
affect these performance variables, but that this
have no significant effect on job satisfaction.
Page 8
This section describes
ROLE
CONFLICT
ROLE
AMBIGUITY
TASK
ACHIEVEMENT
USE OF
STRUCTURED
METHODS
SKILL
VARIETY
TASK
DENTITY
AUTONOMY
FIGURE 1: HYPOTHESIZED EFFECTS OF STRUCTURED
METHODS ON JOB CHARACTERISTICS AND JOB SATISFACTION
the dependent variable,
research
hypotheses,
the
and
independent
the
and
intervening
environmental variables.
variables,
the
The environmental
variables measure factors that could affect the validity of the model.
Dependent Variables
Job satisfaction of programmer/analysts is the main dependent
in
the model.
variable
Locke defines job satisfaction as "a pleasurable or positive
Page 9
.
ns
PRODUCTIVITY
JOB
USE OF
SATISFACTION
STRUCTURED
METHODS
ns
QUALITY
FIGURE 2: HYPOTHESIZED EFFECTS OF STRUCTURED
METHODS ON PERFORMANCE AND JOB SATISFACTION
(Locke,
1976,
page I3OO)."
related strongly
(positively)
In
the job satisfaction literature,
to turnover and absenteeism and
(negatively)
also
be
valuable
methods
on
structured
methods
claims
programmer/analysts.
-
examine
to
structured
examined
it
less
has been
strongly
to performance.
Although job satisfaction will be the focus of
will
experience
state resulting from the appraisal of one's job or job
emotional
other
effects
the
that
their
of
variables.
dependent
use
investigation,
our
the
The
improves
the
it
introduction of
literature
on
performance of
There are two main aspects of performance that can
be
the effects of structured methods on the quality of work done by
programmer/analysts
effects
their
and
on
productivity
the
of
programmer/analysts
These dependent variables will not be emphasized in our study
reasons.
First,
there
is
little
dispute
improves the quality of systems and that it
programmer/analysts.
Second,
there
is
a
two
that using structured methods
improves
the
productivity
of
great deal of dispute about how to
measure program quality and programmer productivity.
Page 10
for
This
will
negatively
affect the credibility of any results obtained
in
this area.
Independent Variable
The use of structured methods
We will define structured methods as
to
application
business
develop
representing how
a
the independent variable in the model.
is
They
systems.
system functions and
include
to describe how the business functions before the
how
the
how the system itself functions.
It
introduction of
generally
is
using
for
The language
business will function after the system
made
is
up
how
language for
a
set of guidelines
a
language for systems analysis, design and programming.
system,
describe
set of procedures that
a
a
the
used
is
computer
developed, and
of
set
a
of
diagrams that describe functional and data hierarchies, and data flows.
The guidelines describe how to get information about the business
the
system
user,
from
how to describe to the user what the system will do, and
how to describe the modules and data structures that make up the system.
some cases,
the methods are supported by automated tools
generation
of
diagrams,
the
check
the
that
aid
In
the
in
consistency of the diagrams, and
generate code from the lowest level of the diagrams.
There are
advocated
by
several
various
examples
of
the
data needs.
methods
that
been
have
One example is the data structure approach
groups.
developed by Michael Jackson (1975)'
describe
structured
Data structure diagrams
are
used
to
data being used by the business and to describe the business'
The system
structure
built
is
around
the
data
structure.
Jackson's method as well as three other development methods are described in
a
recent
article
by
Bergland
(I98I).
Exxon
has
enhanced the Jackson
approach and added more specific guidelines and automated aids.
the
analysis
method
They
SSA and the design and programming method PST.
Page
11
call
Other
•
examples include the PRIDE/ASDM method described by
consists
aids,
of
Softec's SADT
(Ross,
methods
programming
the
group
(e.g.,
methods.
programming),
development
These
methods
methods
analysis
and
above
the
in
programming
formatters),
screen
information analysis), would not be included
structured
structured
of
(1975)
aids
structured
(e.g.,
and
1977).
developed and marketed by Yourdon
Automated
which
(I98I),
development methodology supported by automated
phase
nine
a
Canning
(e.g.,
definition
of
aids do not meet the requirements
described above.
structured
There are several reasons for choosing the use of
defined
as
above,
as
the
independent variable.
are currently used by many organizations.
a
a
cause changes
it
effect
large
easier
detect
to
could
many
they
the jobs performed by programmer/analysts.
They
This should
make
design, and coding phases.
changes
in
job
that
the
use
of
Third,
satisfaction.
Based on the research of Kraft
argue
people.
Second,
the analysis,
in
controversial.
we
on
structured methods
First,
This facilitates the selection of
site and also makes the results relevant to
have
methods,
and Greenbaum
(1977)
(1979).
de-skills
methods
structured
programmer/analysts and reduces their job satisfaction.
Based on the
discussed
structured
above,
we
could
argue
that
the
use
of
facilitates systems development, reduces role conflict
and
are
they
cases
methods
ambiguity,
and
leads to an increase in job satisfaction.
intervening Variables
The literature on job satisfaction identifies
determinants
of
job
satisfaction.
affected by the introduction of
number of variables
as
Some of these variables should also be
structured
Page
a
12
methods.
We
will
call
them
.
variables.
intervening
allow us to better understand how the use of
They
structured methods leads to changes in job satisfaction.
variables
The six
are described below.
Rol e conf
1
i
ct
Role
-
"the
is
degree
incongruity
of
in
person
I98O,
page 92)."
in
jobs that require worl<ers to deal with
many departments within an organization or that require dealing with
outside
organization,
the
development requires
a
conflict
role
is
great deal of contact
These
team.
technical
support staff, and operations staff.
(I96'*)
include
users
at
people
levels,
all
people
The task of systems
common.
with
project
Kahn et al.
or
the expectations or requirements communicated to a focal
incompatibility
(Bostrom,
conflict
outside
d.
p.
of
the
management,
have identified several components of role conflict.
They are:
Person-role Conflict incongruent with the
person.
the extent
to
which role expectations are
orientations, standards, or values of the focal
Intrasender Conflict - the extent to which role requirements are
incompatible with the resources or capabilities of the focal person.
Intersender Conflict expectations from one
part es
the
party
extent to which role requirements or
oppose those from one or more other
i
Role Overload - the extent
to which
the various
role expectations
communicated to the focal person exceed the amount of time available
for their accomplishment (Bostrom, I98O, page 93)-
Bostrom examined the effects of different components of role conflict on the
job satisfaction of programmer/analysts.
relationship between person-role,
and
job
satisfaction.
communication
between
organization,
they
He found
a
significant
negative
intrasender, and intersender role conflict
Since
structured
programmer/analysts
and
methods
should
other
members
facilitate
of
the
should reduce the level of role conflict experienced by
the programmer/analysts.
Page 13
.
1
-
Role ambiguity
expectations
the requirements
found
Bostrom
93)."
degree
"the
is
which
to
vague, ambiguous, or unclear, thereby making
are
for the person to fulfill
page
ambiguity
Role
the
satisfaction.
I98O,
perceived
ambiguity
programmer/analysts was significantly negatively related to their
job
difficult
(Bostrom,
role]
role
of
level
his
[of
it
desired
by
level
of
Since using structured methods should clarify the tasks
to be performed by programmer/analysts,
should reduce the level
it
role
of
ambiguity that they experience.
Task achievement
Researchers have found
-
degree
the
that
to
which
workers can overcome the challenge of their job positively affects their job
(Locke,
development,
programmer/analysts
1976).
methods
structured
Since
satisfaction
perceive
should
simplify systems
increase
an
task
in
ach evement
i
Three of the variables identified by Hackman and Oldham
var iety
task
,
identity
structured methods.
By
,
autonomy
and
reduce
task
-
sk
i
1
should be affected by the use of
-
limiting the tasks performed by programmer/analysts,
reduce
the use of structured methods should
should
(I98O)
identity
and
skill
variety.
Further,
it
autonomy by encouraging the division of
development projects into relatively independent modules.
Hypotheses
The
hypotheses
satisfaction
use
through
structured
of
relate
the
methods
1)
the
use
structured
of
methods
intervening variables described above and
to
productivity
and
quality
in
to
2)
job
the
systems
development.
HI
the
job
:
The use of structured methods will cause an
satisfaction of programmer/analysts
Page
U
.
It
will
overall
increase
in
simplify and structure
the tasks performed by programmer/analysts and
programmer/analysts
between
that their job
better
is
and
users.
improve communication
will
it
Thus, system developers will
more manageable and react more favorably to
feel
We
it.
can
explain the positive effects of the use of structured methods on job
satisfaction by examining their effects on some of the intervening variables
discussed above.
H2
(A)
The use of
:
perceived
by
structured
programmer/analysts
methods
programmer/analysts and others involved
in
conflict
role
improve communication between
will
It
.
decrease
will
systems development.
H2
(B)
:
Role conflict will be negatively related to job satisfaction
H3
(A)
:
The use of structured
perceived by programmer/analysts
methods
It
.
will
will
decrease
role
.
ambiguity
lead to a better specification of
the tasks to be performed by programmer/analysts.
(B)
:
Role ambiguity will be negatively related to job satisfaction
Hi* (A)
:
The task achievement of programmer/analysts will
H3
the use of structured methods
increase
.
with
The structured methods will make the task of
.
systems development easier and hence programmer/analysts will
feel
that they
can better do their jobs.
Hi* (B)
Task
:
sati sf action
achievement
will
be
positively
related
to
.
Some of the job characteristics defined by Hackman and Oldham
negatively
job
affected
by
the
introduction of structured methods.
moderate the positive effects of structured
methods
on
job
will
be
This will
satisfaction.
Specifically, three job characteristics will be negatively affected.
H5
(A)
perceived
The use of structured methods will
:
by
programmer/analysts
the programmer/analyst
in
.
It will
systems development.
Page 15
reduce
the
skill
variety
limit the options available to
•
H5
(B)
:
Skill variety will be positively related to
job
H6
(a)
:
The use of structured methods will
the
perceived
programmer/analysts
by
.
reduce
satisfaction
.
identity
task
The development task will be split into
smaller tasks when structured methods are used.
H6
(B)
:
Task identity will be positively related to job satisfaction
H7
(A)
:
The use of structured methods will
perceived
project
programmer/analysts
by
managers
specify
better
to
Using
.
the
limit the level
methods
tasks
be
to
autonomy
of
structured
.
allow
will
performed
by
programmer/analysts.
H7
(B)
:
H8
(a)
:
Autonomy will be positively related to job satisfaction
The
introduction
of
structured
productivity of programmer/analysts and
work
This
.
it
methods
.
improve
will
the
improve the quality of their
will
based on the previous research on the impact of structured
is
methods on productivity and quality.
H8
(B)
Changes
:
in
productivity and quality will not have
effect on job satisfaction
.
a
significant
The research with other types of jobs has shown
little relationship between job satisfaction and performance
(Locke,
There
between current
is
evidence
some
performance
satisfaction
future
and
and
that
there
satisfaction,
current
performance
individual differences, measurement
play a major role
satisfaction.
in
is
a
but
relationship
not
(Wanous,
problems,
between
197'«)
and
•
current
1976)
job
Locke claims that
contextual
differences
moderating the relationship between performance and job
Controlling
for
these
study.
Page 16
factors
is
out of the scope of this
.
Environmental Variables
The hypotheses discussed above can be tested with
That
variables
intervening
between
programmer/analysts
not
measure
to
many
as
of
consider
validity
three
of
these
of
types
of
we
This will allow us to
.
in
a
characteristics
can
will
call
determine
the
characteristics
-
of
We
this section, we will
In
variables
environmental
programmer/analysts,
individual
group
a
site.
which
factors,
quasi -exper iment
the
-
of
However,
particular research
a
environmental variables, as possible.
degree
group
a
methods.
the
in
for differences between the treatment and
group or for peculiarities of
attempt
-
and a control group
-
structured
using
quasi -exper iment we cannot control
control
group
treatment
a
programmer/analysts using structured methods
of
i
examine differences in the dependent variable and
can
we
is,
quas -exper iment
a
project
of
and
teams,
characteristics of the research site.
We will assume the quas -exper iment
i
group
That
design.
is,
we
is
pretest-posttest with
with
concerned
be
will
a
satisfaction between the pretest and posttest and with
control
changes
in
job
differences
in
the
amount of the changes between the treatment and control groups.
The environmental variables can threaten both the internal and external
validity of the quasi -exper iment
programmer/analysts
us to attribute
structured
a
methods
the treatment and control
in
change
-
example,
For
.
a
in
job
satisfaction
differences
between
the
groups could falsely lead
to
threat to internal validity.
the
In
introduction
addition,
of
there
if
were significant differences between the programmer/analysts at the research
site and programmer/analysts at other organizations, the general zab
i
the quasi -exper iment would be affected
this
section
we
will
-
a
1
i
threat to external validity.
rely on Cook and Campbell's
Page 17
i
(1979)
ty of
In
classification of
.
.
threats to val
idi ty
Characteristics
characteristics
programmer/analysts
of
programmer/analysts
of
in
Differences
-
the treatment and control
could affect both internal
and external validity.
variables,
longevity,
such
job
as
experience,
tenure
structured
methods,
and
level
motivation
of
systems development
of
amount
group,
groups
background
This includes
amount
age,
project
the
in
several
in
with
experience
of
satisfaction with job
and
contexts
For example, differences
validity
in
job longevity could threaten
the quas -exper iment
of
i
.
Suppose the treatment group contained
after
that the job satisfaction of the new hires would decline
the
novelty
new
their
of
wears
job
months
six
change
This
off.
a
We would expect
much larger percentage of new hires than the control group.
when
internal
the
job
in
satisfaction would be independent of the introduction of structured methods.
We could,
structured
however,
falsely attribute this
methods.
Cook
change
the
to
introduction
of
and Campbell would classify this as a threat to
internal validity due to the interaction of selection and maturation.
Background differences between the treatment and control
threaten
also
external
validity.
For example,
if
older or more experienced in systems development or
structured
methods
than
the
i
(1977)
containing
less
the treatment group was
more
experienced
experienced
be
generalized
to
programmer/analysts.
other
treatment
Further,
Katz
has shown that age and job longevity differences moderate the effects
of job characteristics on job satisfaction.
If
most of the members
treatment group were working at the company for at least ten years,
be
with
control group, we could not be sure that the
results of the quas -exper iment could
groups
could
groups
difficult
to
generalize
the
of
it
the
would
results to organizations containing less
Page I8
experienced programmer/analysts.
Cook and Campbell would classify this as a
threat due to the interaction of selection and treatment.
Other
individual
the
Hackman
level
the control
the
and
Oldham
effects
of
(I98O)
job
motivation
i
ty of
as
job
on
motivated
We would then be uncertain that any differences in
group.
methods
1
and
these
identify
characteristics
the level of job satisfaction between the groups were
structured
of
Suppose the treatment group was significantly more
satisfaction.
than
moderate
that
as
i
quasi -exper iment.
constructs
such
contexts, could also affect the general zabi
job
with
satisfaction
differences,
or
to
the
interaction
of
due
to
structured
use
the
of
methods
with
teams
could
motivation.
Project team characteristics
affect
external validity.
-
Differences
in
project
One aspect of team differences
style of the project leader and of the programmer/analyst's
and Seashore
(I966)
have identified four
the leadership
is
peers.
Bowers
leadership characteristics:
Support - behavior that enhances someone
worth and importance.
else's
feeling
of
encourages members
Interaction Facilitation - behavior
that
group to develop close, mutually satisfying relationships.
personal
of
the
Emphasis - behavior that stimulates an enthusiasm for meeting
groups goal or achieving excellent performance.
the
Goal
Work Faci
tation - behavior that helps achieve goal attainment by such
providing
planning,
and
by
activities as scheduling,
coordinating,
resources such as tools, materials, and technical knowledge (Bowers and
Seashore, page 2*47)
1
i
.
These
characteristics
can be applied to both the project
members of the project team.
leaders and other
Bowers and Seashore found significant positive
correlations between measures of these characteristics and job
and performance.
to
Yunger and Hunt
(1976)
satisfaction
found these characteristics similar
the characteristics identified in the Ohio State LBDQ leadership scales.
Page 19
In
addition, differences in
differences
longevity
found
-
amount
the
in
background
the
the
time
of
of
project
the
team has worlced together
could affect could affect external validity.
longevity
group
that
had
significant
a
leaders
Katz
and
-
team
(1979)
has
effect on performance in
research and development groups.
Project team differences between the treatment and control groups would
have the same effect on external validity
individual
as
differences.
We
would be uncertain whether to attribute changes in the dependent variable to
independent variable or, to the interaction of the independent variable
the
and the project team differences.
Research site characteristics
affect
could
site
both
-
internal
The
characteristics
of
and external validity.
the
research
For example,
the
target organization could select only the most satisfied programmer/analysts
to be used
in
the treatment group.
satisfaction
the
and the posttest
We
this were the case, we
would
expect
this group to regress to the mean between the pretest
of
-
If
independent of the introduction
structured
of
methods.
could then falsely attribute their change in job satisfaction to the use
of structured methods.
Cook and Campbell
call
this
threat
a
to
internal
validity due to statistical regression.
Differences
could
how structured methods were used at
in
threaten
also
programmer/analysts
were
a
using
valuable
some
of
in
the
validity
the control
tool,
they
structured
the
the
of
the
research
quas -exper iment.
i
site
If
group determined that structured
the
methods
could start imitating the treatment group by
techniques.
The
differences
job
in
satisfaction between the treatment and control group could then be moderated
by
the
imitation
of
the treatment.
A
treatment group did not become proficient
Page 20
similar problem would occur
in
the use of
structured
if
the
methods
between the pretest and the posttest.
Organizational characteristics of the research site could
validity.
external
of changes
For example,
top management
in
compensation scheme, or
if
the d. p. organization has had
the last year,
in
if
affect
number
a
they have introduced
number
they have had a large
if
also
new
of
a
new
hires
or
this could affect programmer/analyst's reaction to the introduction
layoffs
of structured
methods.
This
would
limit
general zabi
the
i
1
i
of
ty
the
quas -exper iment to other organizations.
i
QUASI-EXPERIMENTAL DESIGN
The model described above can be tested using
control group design.
Cook and Campbell
(1979)
pretest-posttest
a
with
diagram this design as:
0X0
The
design
introduce
can
implemented
be
structured
programmer/analysts.
programmer/analysts
introduced
to
methods
We
in
measure
later
(the
finding
to
some,
then
company
the
variables (the pretest).
months
can
by
the
but
administer
right
independent,
a
before
not
structured
The questionnaire can then be
posttest).
about to
of
its
methods
are
environmental
and
readministered
six
The group of programmer/analysts who do not
control
posttest.
the
section
all,
is
questionnaire to all the
intervening,
use structured methods can serve as a
This
one company that
outlines
the
pretest
and
quas -exper imental
design.
It
group
at
i
discusses the requirements of the research site, the experimental procedure,
the measures, and the method of analysis.
detailed discussion of the design.
.
Page 21
Goldstein
(1982b)
provides
a
more
Research Site
is
about
-
The research site should be a large
introduce
to
Ideally,
the organization should have 100 or more
should
develop
transaction
large
programmer/analysts.
processing
and
information systems that span boundaries within the organization.
the types of systems most suited for development
organization
The
should
have
not
with
suffered
These are
structured
recent
any
methods.
organizational
overshadow the effects of the introduction of structured methods.
able
find
the
organization
through
It
management
disruptions, such as the ones discussed in the previous section, that
be
that
methods on some of its project groups.
structured
mostly
organization
could
We should
company involved in marketing
a
structured methods.
Experimental procedure
the
all
The questionnaire should
-
programmer/analysts
the organization and their project
in
right before the introduction of structured methods
introduction.
their
administered
be
and
six
to
leaders
months
after
structured methods should be used by at least 30
The
prog rammer /anal ysts.
Measures
-
Measures
already
have
variables
in
al.
and modified by Bostrom
(1970)
conflict
and
the
model.
ambiguity.
example,
For
(I98O)
JDS
also
contains
can
Hackman and Oldham's
items
motivation, and satisfaction
(1966)
to
with
used
be
(I98O)
to
many
the
of
job
measure
and
satisfaction,
Bowers
contexts.
et
role
Job Diagnostic Survey
identity,
task
measure
job
for
the scales developed by Rizzo,
contains scales that measure skill variety,
The
developed
been
and
autonomy.
level
of
Seashore's
measure can be used to measure the leadership characteristics of each
programmer/analysts'
peers and project leader.
the questionnaire to measure task achievement,
programmer/analysts
identified
above,
Page 22
and
Other
items can be added to
the other characteristics
the
of
programmer/analysts'
proficiency with structured methods.
Project leaders should be given
used
obtain
to
information
programmer/analysts,
separate questionnaire.
productivity
on
obtain
to
a
quality
and
should be
It
their
for
information on their team and on their own
background, and to verify the intervening variable
measures
obtained
from
their programmer/analysts.
Method of Analysis
hypotheses.
This
differences
in
assumption
is
is
Simple gain score analysis can be used to test the
-
an analysis of variance
changes in
a
technique
examines
the
variable from the pretest to the posttest.
The
that the treatment will
lead to more
that
less)
(or
change
treatment group than in the control group (Cook and Campbell,
in
the
1979).
The characteristics of the treatment and control groups can be examined
to determine the validity
between
our
of
results.
treatment
and
differences,
and
team
differences
validity
our
results.
the
of
control
Any
measures
could
affect
characteristics
the
differences
on the background,
groups
Further,
significant
population will give us some idea of the general zabi
i
1
i
the
individual
internal
the entire
of
ty of our results.
DISCUSSION
This research will provide some insights into the effects of the use of
structured methods on the job satisfaction of programmer/analysts.
pretest-posttest
of
structured
explanation
a
with control group design allows us to isolate the effects
methods.
as
Using
to
The
how
the
intervening
use
of
sati sfaction.
Page 23
variables
structured
provide
methods
further
affects
job
There are, however, some potential problems with the study.
six month time frame might not be
enough
long
become familiar with structured methods.
for
First,
the
programmer/analysts
to
However, a longer time frame would
create problems with experimental mortality.
Turnover would cause people to
drop out of the experiment between the pretest and the posttest.
Second, differences between treatment and control
the
experimental
control
for
in
a
1
i
between
and
and
the
population of programmer/analysts could
ty of
the
results.
groups
affect the general izabi
groups
Although
design,
quas -exper imental
i
this
is
measures of individual and team differences should point out
to
of multiple
presence
the
hard
potential
any
problems.
A
larger sample survey could
developed
instruments
in
eliminate
some
of
methods
in
structured methods.
If
The
this quas -exper iment could be used to compare the
i
programmer/analysts that use structured methods to those
structured
problems.
the
several
companies
that
do
not
use
that have partially implemented
the results of the pretest-posttest quasi -exper iment
were confirmed with this posttest only design, this would
provide
stronger
evidence that the hypotheses were true.
Finally,
First,
this research should lead to further research
this research will
satisfaction
in
programmer/analysts.
Research
two
areas.
provide some insights into the determinants of job
research could explore the
Further
effects of the factors identified here, as well as
satisfaction.
in
could
also
explore
the
other
factors,
effects
on
job
job enrichment
programs or other work redesign efforts on programmer/analysts.
This
is
an
important research, because of the relationship between job satisfaction and
job outcomes such as productivity and turnover.
in d.
p.
organizations.
Page 2U
These are critical problems
Second,
changes
on
the research provides some insights
job
satisfaction.
computers
workers.
could
research
the
effects
of
job
The introduction of new office technologies
and the more widespread use of
Further
into
technologies on job satisfaction.
will
examine
change
the
This would give us
the
effects
a
better
jobs
of
of
these
idea
of
possible costs and benefits of office automation or other technologies.
Page 25
many
new
the
REFERENCES
Awad,
Elias M., "Prediction of Satisfaction of Programmers, Analysts,"
Management January, 1977. 12-18.
Data
,
Production Programming,"
Baker, F. T., "Programmer Team Management of
Systems Journal
Volume 11, Number 1, 1972, pages 56-75'
IBM
,
Bergland, G. D., "A Guided Tour
October, I98I, pages 13-37-
of
Program
Methods,"
Design
Computer
.
Implementation of
the Jackson
Bernstein, C. M., "Experience with Exxon's
Program Design Method," National Computer Conference Proceedings
1978,
pages 636-637,
Variables
in
Critical
Bostrom, Robert, P., "Role Conflict and Ambiguity:
User-Designer Relationship," Proceedings of the Seventeenth Annual
the
I98I, pages 88-112.
Computer Personnel Research Conference
,
Organizational
"Predicting
David G. and Stanley E. Seashore,
Effectiveness with a Four-Factor Theory of Leadership," Admi ni strat ive
Science Quarterly
1966, pages 238-263.
Bowers,
,
Canning, R. G., Editor, "The Search for Software Reliability," EDP Analyzer
Volume 12, Number 5. May, 197'«a-
Structured
Editor,
"The Advent of
Canning,
R. G.,
Volume 12, Number 6, June, 197'»b.
Ana yzer
1
Programming,"
,
EDP
,
Canning, R. G., Editor, "The Analysis of User Needs," EDP
17. Number 1, January, 1979a-
Analyzer
Canning, R. G., Editor, "The Production of Better Software,"
Volume 17. Number 2, February, 1979bCanning, R.G., Editor,
Volume 19, Number
"Application System
October, I98I
10,
Design
Aids,"
Volume
,
EDP
Analyzer
,
EDP
Analyzer
,
.
Quas -Exper imentat ion Design and
Cook, Thomas D. and Donald T. Campbell,
Houghton Mifflin Company, Boston,
Analysis
Issues
for Field Settings
1979.
i
,
Couger,
Daniel
Personnel
,
Zawacki,
Motivating and
and Robert
John Wiley and Sons, New York, I98I.
Goldstein, David K., "Exxon's Use of Structured Systems
I98I.
Study," Unpubl shed paper
i
,
Page 26
Managing
Analysis:
Computer
A
Case
Goldstein, David K.,
"The
Impact of
Development Aids on the Systems
Development Process," Proceedings of the Conference on Human Factors in
Computer Systems
1982a, pages 130-13't.
,
Goldstein, David K., "The Effects of Structured Development Methods on the
Job Satisfaction of Programmer/Analysts:
A Quasi -Exper imental Design,"
Unpubl shed paper
1982b.
i
,
Greenbaum, Joan M., In the Name
Philadelphia, 1979.
Of
Efficiency
,
Temple
Redesign
,
Addison
Hackman, J. R. and G. R. Oldham, Work
Mass.. 1980.
University
Wesley,
Hamilton, Kenneth and Arthur Block, "Programmer Productivity in
Environment," nfosystems April, 1979,
pages 4i»-50 and
pages 62-68.
I
,
Bill, "An Example of Structured
pages 82-86.
Inmon,
Design,"
Jackson, Michael A., Principles of Program Design
Datamation
Katz,
R. L.
Wolfe D.,
Quinn R.,
Organizational
Stress:
Studies
John Wiley, New York, I96i*.
Reading,
Structured
a
May,
1979,
March,
1976,
Academic Press,
,
Jones, Martha N., "HlPO for Developing Specifications,"
1976, pages 112-125.
Kahn,
,
Press,
Datamation
1975.
.
March,
Snoek J. D.,
and
Rosenthal
R.,
in Role Conflict and Role Ambiguity
,
,
Ralph, "Satisfaction and Behavioral Reactions of
Employees to Task
Characteristics as a Function of Job Longevity. Age. and High-Order
Needs," Sloan Working Paper 932-77. MIT, Cambridge, Mass, 1977.
,
Katz.
Ralph, "The Influence of Group Longevity on Project Communication and
Project Performance," Sloan Working Paper
1179-80. MIT. Cambridge,
Mass, 1980.
.
Kraft, Philip, Programmers and Managers
Locke,
.
Spr nger-Verlag, New York.
i
1977.
A., "The Nature and Causes
of
Job Satisfaction,"
in
Dunnette,
D.,
Ed., Handbook of Industrial and Organizational Psychology
Rand
McNally, I976.
E.
M.
,
Menard, Jayne B.,
"Exxon's Experience with the Michael
Jackson Design
Method," Data Base
Volume
II, Number 3. Winter-Spring. I98O. pages
88-92.
,
Mendes, Kathleen S., "Structured Systems Analysis:
A
Technique to Define
Busines Requ rments, " Sloan Management Review Volume Ik, Number *,
Summer, 198O, pages 51-63i
,
Page 27
1
Rader
'EB
^^^"Ct
,
J.
A.,
National
9 n in
"Experience with an Application of Structured Design,"
Computer Conference Proceedings
1978, pages 633"636.
,
Rizzo, J. R., R. J. House, and S. J. Lirtzman, "Role Conflict and Ambiguity
in Complex Organizations," Administrative Science Quarterly
Volume 15.
1970, pages 15O-I63.
,
Ross,
Douglas T., "Structured Analysis (SA)
A Language
for
Communicating
Ideas,"
IEEE Transactions on Software Engineering
Volume SE-3. Number
1, January,
1977. pages 16-21.
:
,
Stevens, W. P., G. J. Myers, and L. L. Constantine, "Structured Design,"
Systems Journal
Volume I3. Number 2, 197'*. pages 115"139'
I
BH
,
Wanous, J. P., "A Causal-Correlational Analysis of the Job Satisfaction and
Performance Relationship," Journal
of Applied Psychology
Volume 59.
197'*, pages 139-Ki».
,
Warnier, J. D., Logical Construction of Programs
New York, 1974.
.
Willoughby, T. C, "Staffing the MIS Function,"
Volume U, Number I4, 1972, pages 2l»l-259.
Van
Surveys
Computing
ACM
Winters, Edward W., "Experience with Problem Statement Language:
Structured
Documentation,"
Appl cat ion
Deve opment
Proceedi ngs
1979. pages ]}'J-]k].
1
i
Reinhold,
Nostrand,
A
,
Tool for
Sympos urn
i
,
Yourdon, E., and
York, 1975.
L.
L.
Constantine, Structured Design
,
Yourdon
New
Press,
Yunger, G. W. and J. G. Hunt,
"An
Empirical
Comparison of the Michigan
Four-Factor and the Ohio State LBDQ leadership Scales," Organ zat ona
Behavior and Human Performance Volume 17, 1976, pages 'S-^Si
,
Page 28
i
BASfiM^fe
^G0 Z'SlB
SEP
I
'68
Lib-26-67
HD28.M414 no.l330- 82
Goldstein,
745062
Oav/The effects of
D*BKS
structur
0Q136
III'
3
TDflD
002 DM? SM3