Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
V
o
l.
6, N
o
. 1
,
Febr
u
a
r
y
201
6,
pp
. 24
2
~
24
8
I
S
SN
: 208
8-8
7
0
8
,
D
O
I
:
10.115
91
/ij
ece.v6
i
1.8
247
2
42
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Key S
o
ftware Metrics and its Impa
ct on each other for Software
Devel
o
p
m
ent P
r
oj
ect
s
Mri
dul
B
h
ardw
aj, Ajay R
a
n
a
Amity
School of
Engineerin
g and
Technol
ogy
, Amity
University
, Noida Up, I
ndia
Article Info
A
B
STRAC
T
Article histo
r
y:
Received J
u
n
1, 2015
Rev
i
sed
O
c
t 16
, 20
15
Accepte
d Nov 5, 2015
Ever
y
softwar
e
development pr
oject is
uniqu
e and differ
e
nt fro
m repeatable
manufacturing
process. Each software
proj
ect
s
h
are differ
e
nt
chall
e
nge
s
rela
ted to te
chn
o
log
y
, peopl
e a
nd tim
e
lines. If ever
y
pro
j
ect is unique, how
projec
t m
a
nager
can es
tim
a
t
e pro
j
ec
t in
a consistent way
b
y
apply
i
ng his past
experi
enc
e
. One
of the m
a
jor chall
e
nges
faced by
the pro
j
ect manager is to
identif
y
th
e ke
y softwa
re m
e
t
r
ics to
contro
l
and m
onitor
the pro
j
ec
t
execu
tion. Each s
o
ftware
deve
lo
pm
ent
project may
b
e
unique but share some
com
m
on m
e
tric
that
can
be us
ed
to cont
ro
l
and m
onitor th
e
project execution.
These metr
ics are software size, e
ffor
t
, pro
j
ect duration
and p
r
oductiv
ity
.
These metrics tells project manager about
what
to deliver (size)
,
how it was
deliv
ered in pas
t
(productiv
it
y
)
and how long will it t
a
ke to
deliv
er with
current
te
am
ca
pabili
t
y
(t
im
e a
nd effo
rt)
.
In this paper, we
explain the
relationship among these key
metrics and how th
ey
st
atistically
impact each
other. Th
ese relationships have
been
deriv
e
d based on the data
published in
book “Practical Software
Estimati
on” b
y
International Software
Benchmarking
Group. This paper also
exp
l
ain
s
how these metrics can be
used in predictin
g the total number of de
fects. Stud
y
suggest
s that out of th
e
four key
software metrics
softw
a
re
size
significa
ntly
impact th
e other thr
ee
metrics (project effort, duratio
n
and productivity
)
. Productiv
ity
does not
significantly
depend on th
e s
o
ftware s
i
ze but
it repres
ents
t
h
e nonlinea
r
relationship with software size a
nd maximum team size, hence,
it is
recommended not to have a ver
y
big t
eam
s
i
ze as
it m
i
ght im
pact the overa
l
l
productivity
.
To
tal project dur
ation only
dep
e
nds
on the software size and it
does not d
e
pend
on the maximum team size
. I
t
im
plies th
at
we
can
not redu
ce
projec
t duration
b
y
incr
eas
ing t
h
e team
s
i
ze
. T
h
is
fact is
contr
a
r
y
to th
e
percep
tion
that
we can
redu
ce
t
h
e proj
ect
dura
t
i
on b
y
in
creas
ing
the
proje
c
t
team
s
i
ze. W
e
c
a
n conclud
e
that
s
o
ftware s
i
ze is
the im
portant m
e
tri
c
s
and a
significant
effort must be
put dur
ing project
in
iti
a
tion ph
ases to
es
tim
ate
the
projec
t siz
e
.
As
software siz
e
wi
ll h
e
lp
in
estim
at
ing th
e proj
ec
t d
u
ration
and
projec
t efforts so error in estim
ating th
e s
o
ftwa
re s
i
ze wil
l
hav
e
significant
impact on the accuracy
of project durat
ion and
effort. Al
l these
ke
y
m
e
tr
ics
m
u
st be re-
cal
ibr
a
ted
during
the
p
r
ojec
t dev
e
lopm
ent
life
c
y
c
l
e
.
Keyword:
Effort,
Produ
ctiv
ity
Pr
oj
ect du
r
a
tion
Soft
ware
de
vel
opm
ent
m
e
t
r
i
c
s
Soft
ware
size
Copyright ©
201
6 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
Mridul B
h
ardwaj,
A
m
ity
School
of Engi
neeri
ng and Technol
og
y
A
m
ity
University
,NOIDA UP, INDIA
Em
a
il: m
r
id
u
l
2
7
07@g
m
ail.co
m
1.
INTRODUCTION
Every
s
o
ft
ware
devel
o
pm
ent
pr
o
j
ect
i
s
u
n
i
q
ue a
nd
di
f
f
ere
n
t
from
repeat
a
b
l
e
m
a
nufact
u
r
i
ng
pr
ocess
.
Each softwa
re project share differe
nt challenges relate
d t
o
t
echn
o
l
o
gy
, pe
opl
e an
d t
i
m
e
l
i
nes. I
f
eve
r
y
pro
j
ec
t
is un
iqu
e
,
ho
w pr
oj
ect m
a
n
a
ger
can
estim
ate
pr
oj
ect in a con
s
isten
t
w
a
y by ap
p
l
ying
h
i
s
p
a
st ex
p
e
r
i
en
ce. On
e
of t
h
e m
a
jor c
h
al
l
e
nge
s face
d by
t
h
e
pr
oje
c
t
m
a
nager i
s
to
id
en
tify th
e k
e
y software
metrics to
co
n
t
ro
l and
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
6, No
. 1, Feb
r
uar
y
20
1
6
:
24
2 – 24
8
24
3
m
oni
t
o
r t
h
e p
r
oject
e
x
ec
ut
i
o
n. Eac
h
s
o
ft
wa
re de
vel
o
pm
ent
pr
oject
m
a
y
be u
n
i
q
ue
but
share
som
e
com
m
on
metric th
at can
b
e
u
s
ed
to
co
ntro
l and
m
o
n
i
t
o
r th
e
project e
x
ecution.
As per Put
n
am
, La
wre
n
ce H., and W
a
re
[1], these m
e
trics are soft
ware si
ze, effort, project dura
tion and produ
ctiv
ity.
Th
ese m
e
trics tells
p
r
oj
ect
man
a
g
e
r abou
t wh
at to
d
e
liver (size),
ho
w
it wasd
eliv
er
ed
in p
a
st
(p
ro
du
ctiv
ity) and
ho
w lon
g
will it tak
e
to
d
e
liv
er
with
cu
rren
t team
cap
ab
ility (ti
m
e an
d effort).
1.1. Softw
are Siz
e
–Measu
r
ement of
Func
tionality
So
ft
ware size is th
e
m
easu
r
emen
t o
f
so
ft
ware fun
c
tio
n
a
lity th
at is
b
e
in
g
d
e
liv
ered
or exp
ected
to
b
e
delivere
d
by s
o
ft
ware. Soft
ware size is a num
erical
measure
of soft
ware re
quirement that are define
d
q
u
a
litativ
ely (in
m
o
st case in
word do
cu
m
e
n
t
) b
y
u
s
er. Software
size is mo
st natu
ral m
e
t
r
ic of so
ft
ware as i
t
i
s
i
nde
pe
nde
nt
fr
om
al
l
ot
he
r m
e
t
r
i
c
s. So
ft
ware
si
ze
dep
e
nde
d
o
n
l
y
o
n
w
h
at
t
o
del
i
v
er
rat
h
er
on
h
o
w
t
o
deliver. As s
o
ft
ware
size is a num
e
rical
m
eas
ure
of s
o
ft
ware fun
c
tion
a
l requ
irem
en
ts
so
pr
oj
ect m
a
n
a
g
e
r m
u
st
rem
e
m
b
er th
at
n
o
two
so
ft
ware proj
ect will b
e
sam
e
in
fu
nctio
n
a
lity b
u
t
th
ey m
a
y h
a
v
e
sam
e
so
ftware
size.
In size
-bas
ed
project estim
a
tion,
details of
soft
ware
req
u
i
r
em
ent
are not
im
port
a
nt
.
W
h
at
i
m
port
a
nt
i
s
t
h
e
relative size of project in c
o
m
p
arison
to the already c
o
mpleted
projects. Fo
r e
x
am
pl
e we m
a
y
not
be
abl
e
t
o
esti
m
a
te
th
e so
ftware size of
p
r
op
o
s
ed
pro
j
ect as d
e
ta
il so
ftware requ
iremen
ts are
not
available, but we ca
n
co
m
p
are th
e com
p
lex
i
t
y
o
f
th
e p
r
o
p
o
s
ed
project with
th
e alread
y d
e
liv
ered
p
r
oj
ects. Th
is
will h
e
lp
th
e pro
j
ect
m
a
nager t
o
p
r
e
d
i
c
t
t
h
e p
r
o
j
ect
per
f
o
r
m
a
nce b
a
sed
on
past
p
e
rf
orm
a
nce of
t
h
e si
m
i
l
a
r pro
j
ect
. T
h
ere a
r
e
m
a
ny
soft
ware si
zi
n
g
m
e
t
hod vi
z.
fu
nct
i
o
n p
o
i
n
t
s
, use case
poi
nt
s, st
o
r
y
p
o
i
n
t
s
(fo
r agi
l
e
p
r
oject
s
)
,
ob
ject
base
d
cou
n
t
et
c.
IF
P
U
G
f
unct
i
on
p
o
i
n
t
i
s
t
h
e m
o
st
wi
del
y
use
d
s
o
ft
ware
si
zi
ng
m
e
t
hod.
F
o
r
f
u
nct
i
o
n
p
o
i
n
t
s
,
y
ou ca
n
fi
n
d
i
n
d
u
st
ri
al
dat
a
on s
o
ft
wa
re pr
o
j
ect
m
e
t
r
i
c
publ
i
s
hed b
y
Int
e
rnat
i
o
nal
Soft
wa
re B
e
n
c
hm
arki
ng St
a
nda
r
d
Gro
u
p
(ISB
S
G) [2
], Qu
an
titativ
e Software
Man
a
g
e
m
e
n
t
(QSM)
[3
] bu
t
still th
ere is a
ch
allen
g
e to
measu
r
e
the softwa
re s
i
ze in the earl
y
stage
of
project life cycle as detail requ
irem
ents are not a
v
ailable.Mridul
B
h
ar
dwa
j
a
nd
Ajay
R
a
na [
4
]
i
n
hi
s p
a
pe
r “
E
st
im
at
e Soft
ware F
u
nct
i
o
n
a
l
Si
ze bef
o
re
R
e
qui
rem
e
nt
p
h
ase
of
Devel
opm
ent
Li
fe C
y
cl
e”
sug
g
est
e
d t
h
e
m
e
t
hod t
o
est
i
m
a
t
e
soft
ware f
u
nct
i
o
n
a
l
si
ze when
det
a
i
l
s
requirem
ents are
not a
v
ailabl
e.
1.2. E
f
fort and Time
Project effort is the cu
m
u
lative tim
e
spends
by the entire project team
on t
h
e pr
oj
ect
. Effo
rt
gene
ral
l
y
m
e
asure
i
n
per
s
o
n
h
o
u
r
s,
pe
rso
n
m
ont
hs o
r
pe
rso
n
day
s
but
pers
o
n
ho
ur
s i
s
m
o
st
sui
t
a
bl
e an
d
unam
b
i
g
u
o
u
s
uni
t
as
ot
he
r
uni
t
s
re
q
u
i
r
e c
o
n
v
e
r
si
o
n
f
r
o
m
hou
r t
o
da
y
or m
ont
h.
T
h
i
s
co
n
v
ersi
on
i
s
not
st
anda
rd
bec
a
u
s
e i
n
s
o
m
e
cou
n
t
r
y
(
s
peci
al
l
y
devel
ope
d
coun
tries) th
ere is
8
h
ours
work
in
g in
a day while in
so
m
e
co
un
tr
y (d
ev
el
o
p
i
n
g
co
un
tr
y)
it is m
o
r
e
th
an 8 hou
r
s
.
Tim
e
represe
n
t
s
the calendar
duratio
n
of th
e p
r
oj
ect. It tells th
e exp
ected
project start and end
dates
.
Though it appears that Tim
e
and effort
s are in
terch
a
n
g
e
ab
le
m
e
tric b
u
t
actu
a
lly it
is n
o
t. Th
ere is n
o
lin
ear
rel
a
t
i
ons
hi
p
be
t
w
een
t
i
m
e
an
d e
f
f
o
rt
.
P
r
o
j
ec
t
m
a
nager m
u
st unde
rstand t
h
at we ca
n re
duce the
overall
project
d
u
ration
b
y
add
i
ng
m
o
re team
m
e
m
b
ers b
u
t
b
e
yond
cert
a
in
po
in
t in
crease in
team
s
i
ze w
ill n
o
t
resu
lt in
redu
cing
th
e
proj
ect du
ration b
u
t
it will in
crease proj
ect
du
ration
.
To
und
erstan
d
it in
m
o
re d
e
tail, if to
tal
est
i
m
a
t
e
d pro
j
ect
effo
rt
i
s
1
2
pe
rs
on m
ont
hs an
d t
h
e
r
e i
s
pr
o
j
ect
t
e
am
of
4 pe
o
p
l
e
. I
n
t
h
i
s
case est
i
m
a
t
e
d
p
r
oj
ect d
u
ration
will b
e
3
m
o
n
t
h
s
pro
v
i
d
e
d
t
h
ere is no
p
l
ann
e
d
id
le ti
m
e
.
Can
we d
e
liv
er th
is p
r
o
j
ect in
on
e
m
o
n
t
h
with 12 m
e
m
b
er tea
m
?
Obv
i
o
u
s log
i
cal an
swer
is “No
”
b
ecau
s
e ad
d
i
n
g
m
o
re me
m
b
er
will add
n
e
w
com
m
uni
cat
i
on c
h
an
nel
s
a
n
d al
so
i
n
c
r
eas
e t
h
e i
n
t
e
grat
i
o
n
ef
fo
rt
. P
r
oj
ect
m
a
nager
m
u
st
unde
rst
a
nd
t
h
i
s
rel
a
t
i
ons
hi
p a
n
d sh
o
u
l
d
c
h
o
o
s
e
opt
i
m
al
t
e
am
si
ze and s
h
ou
l
d
pr
ef
er
con
s
tan
t
tea
m
size th
r
oug
hou
t th
e pr
oj
ect
life cycle. Figu
re
1
ex
p
l
ai
n
th
e relation
s
h
i
p b
e
tween
team size an
d project d
u
ratio
n till p
o
i
n
t
o
f
reflectio
n
in
crease in
team size will h
e
l
p
in
redu
cing
th
e ov
era
ll p
r
o
j
ect d
u
ratio
n
but in
crease in
tea
m
size b
e
yo
nd
th
e
p
o
i
n
t
of reflectio
n
will in
crease th
e
p
r
o
j
ect
du
ration
.
Pr
oj
ect
m
a
n
a
g
e
r m
u
st u
n
d
e
rstand
this p
o
i
n
t
of refl
ectio
n
as it will h
e
lp h
i
m
to
co
mmit ti
melin
es with
proj
ect st
akeh
o
l
d
e
rs. It is n
o
t
easy t
o
id
en
tify th
e
poin
t
o
f
refl
ect
i
o
n b
u
t
wo
rk
b
r
ea
k
do
wn
st
r
u
ct
ure
o
r
Del
phi
t
e
c
h
n
i
ques c
a
n
hel
p
pr
o
j
ect
m
a
nager i
n
i
d
e
n
t
i
f
y
i
ng t
h
e
optim
al tea
m
s
i
ze. Since cost of pe
ople
is th
e
m
a
j
o
r co
st in an
y so
ftware d
e
v
e
l
o
p
m
en
t t
h
at is wh
y in
m
o
st
o
f
the Agile s
o
ft
ware
de
velopment projects
the team
si
ze rem
a
i
n
const
a
nt
. P
r
oject
m
a
nage
r m
u
st
u
n
d
erst
a
n
d
t
h
ese rel
a
t
i
o
ns
hi
ps a
n
d base
d
on t
h
e p
r
oject
ob
ject
i
v
e
(as
d
e
fi
ne
d i
n
pr
oj
e
c
t
chart
e
r
#
)
he
sho
u
l
d
m
a
ke cor
r
ect
b
a
lan
ce
of th
ese m
e
trics (co
s
t
,
effort and
du
ratio
n
)
.
#Project cha
r
ter de
fines the
project
objective approve
d
by
project s
p
ons
o
r. In som
e
ca
ses (product
devel
opm
ent
p
r
o
j
ect
s w
h
e
r
e t
i
m
e
t
o
m
a
rk
et
i
s
t
h
e key
)
,
pr
oj
ect
end
dat
e
i
s
sacrosanct
whi
l
e project sc
ope and
cost
ca
n
be c
h
a
nge
d.
I
n
som
e
cases (t
a
x
base
d s
o
ft
ware
a
p
p
l
i
cat
i
on),
p
r
o
j
e
c
t
sco
p
e i
s
sacr
osa
n
ct
w
h
i
l
e
pr
ojec
t
end
dat
e
c
a
n
b
e
cha
nge
d.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Key Software
Metrics and its Impact
on ea
c
h
other
for Software Development Projects
(
M
ri
dul
Bh
ar
d
w
aj
)
24
4
Fi
gu
re 1.
How
in
crease i
n
Team
size will i
m
p
act pro
j
ect
duratio
n
?
1.
3. Pro
ducti
v
i
ty or R
a
te o
f
Del
i
v
ery
Produ
ctiv
ity can
b
e
d
e
fin
e
d as effort requ
ired
to
d
e
liver un
it so
ftware size. It h
e
lp
s th
e proj
ect
man
a
g
e
r to
pred
ict th
e
ov
erall effort req
u
i
re t
o
d
e
liv
er
th
e
pro
j
ect. Produ
ctiv
ity d
e
p
e
nd
s
on
th
e team
ex
pertise
o
r
exp
e
rien
ce t
o
work in
sim
i
lar tech
no
l
o
g
y
, Team
will b
e
m
o
re p
r
o
d
u
c
tiv
e if t
h
ey h
a
ve wo
rk
ed on
si
milar
technology earlier. Productivi
t
y will
also de
pend
on the
business proc
ess
unde
rstanding, if team
understand
th
e b
u
s
i
n
ess pro
cesses, it will
h
e
lp
th
em
to
t
r
an
slat
e the business re
quire
ments to technical requirem
e
nts. If
sam
e
tea
m
h
a
d earlier
work
ed
tog
e
th
er t
h
en
it will h
e
l
p
i
n
im
p
r
ov
ing
the produ
ctiv
ity as it will reduce th
e
ti
m
e
req
u
i
red
t
o
reso
l
v
e th
e co
llab
o
ration
issu
es. Proj
ect
man
a
g
e
r m
u
st u
n
d
e
rstand
th
e critical facto
r
th
at can
in
flu
e
n
ce th
e t
e
a
m
p
r
odu
ctiv
ity. Fo
llo
wi
n
g
are th
e critical
poi
nt
s t
h
at
p
r
oject
m
a
nager
m
u
st
consi
d
er
whi
l
e
esti
m
a
t
i
n
g
th
e
tea
m
p
r
odu
ctivity.
Team
experience to
work
on s
i
m
i
lar technology
Un
de
rst
a
n
d
i
n
g
of
b
u
si
ne
ss
pr
o
cesses
of s
o
ft
w
a
re a
ppl
i
cat
i
o
n
u
nde
r
de
vel
o
p
m
ent
.
Ex
peri
enc
e
of t
e
am
wo
rki
n
g
t
oget
h
er
Und
e
rstand
ing
o
f
clien
t
env
i
ron
m
en
t as it wi
ll h
e
lp
to fact
o
r
in
th
e clien
t
dep
e
nd
en
cy, if an
y
It
i
s
chal
l
e
n
g
i
n
g t
o
de
fi
ne t
e
a
m
prod
uct
i
v
i
t
y
an
d
bi
g
g
er c
h
al
l
e
nge t
o
gi
ve
pr
o
duct
i
vi
t
y
a n
u
m
b
er b
u
t
it an
essen
tial metrics th
at n
o
t o
n
l
y
n
eed to
b
e
p
r
ed
icted
bu
t also con
s
ist
e
n
tly
m
o
n
ito
red
d
u
ring
t
h
e
proj
ect
life cycle. In
Ag
ile
p
r
oj
ect
p
r
od
u
c
tiv
ity is term
ed
as
“Velocity” and
define
d as t
h
e
num
b
er of story
points
del
i
v
ere
d
pe
r i
t
erat
i
on.
Vel
o
c
i
t
y
depe
nds
on
t
eam
si
ze
an
d nu
m
b
er
of
stor
y po
in
ts d
e
liver
ed, as m
o
st o
f
th
e
agile projects
have
constant t
e
a
m
si
ze throughout the
proj
e
c
t life cycle and length
of
the
each iteration is also
co
nstan
t
so
h
i
g
h
e
r
th
e v
e
l
o
city, h
i
g
h
e
r
th
e tea
m
p
r
od
u
c
tiv
i
t
y. I
n
A
g
ile pro
j
ect v
e
l
o
city
o
r
team p
r
oductiv
ity
i
m
p
r
ov
es as pro
j
ect
prog
ress.
Velo
city will be h
i
gh
i
n
later i
t
eratio
n
o
f
th
e
p
r
oj
ects.
Proj
ect m
a
n
a
ger shou
ld
u
s
e t
h
e
h
i
sto
r
ical data o
f
sim
ilar
p
r
oj
ects to
esti
matio
n
th
e prod
u
c
tiv
ity of the tea
m
.
As m
o
st
of t
h
e pr
o
duct
i
v
i
t
y
num
bers a
r
e p
ubl
i
s
hed i
n
ra
n
g
e s
o
p
r
o
j
ect
m
a
nager s
h
o
u
l
d
care
f
ul
l
y
ch
ose t
h
e
p
r
od
u
c
tiv
ity nu
m
b
er with
i
n
t
h
at rang
e. Figu
re 2 exp
l
ain
h
o
w
produ
ctivity can
b
e
u
s
ed
t
o
d
r
i
v
e th
e
o
v
e
rall
project efforts.
Fi
gu
re
2.
D
r
i
v
i
n
g
Ef
f
o
rt
s
Usi
n
g
Si
ze a
n
d
P
r
od
uct
i
v
i
t
y
Ag
ile or Iterativ
e d
e
v
e
lop
m
e
n
t m
o
d
e
l are m
o
re p
r
ef
erab
l
e
to
waterfall m
o
d
e
l if th
ere is h
i
gh
er
u
n
c
ertain
ty to
d
e
fi
n
e
th
e team
p
r
o
d
u
c
tiv
ity. In
waterf
all
m
o
d
e
l actu
a
l produ
ctiv
ity o
f
th
e team
will o
n
l
y
b
e
k
now
n
af
ter
the co
n
s
tru
c
tion
p
h
a
se so
pr
oj
ect
m
a
n
a
g
e
d
o
e
s f
i
n
d
an
y oppo
r
t
u
n
ity to
r
e
def
i
n
e
th
e pr
odu
ctiv
ity
num
ber a
n
d
he
nce t
h
e
ot
her
key
m
e
t
r
i
c
s (eff
ort
a
n
d t
i
m
e)
, h
o
w
eve
r
,
i
t
e
rat
i
on a
n
d
agi
l
e
de
vel
o
pm
ent
m
odel
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
6, No
. 1, Feb
r
uar
y
20
1
6
:
24
2 – 24
8
24
5
p
r
ov
id
e p
r
oj
ect
m
a
n
a
g
e
r
th
e o
ppo
rt
u
n
ity
to
re-d
efi
n
e
th
ese
key m
e
trics.Mridul B
h
ardwaj and
Ajay
Rana [4]
explains the
st
eps t
o
calculat
e
these
key m
e
trics.
Fi
gu
re 3.
2.
R
E
SEARC
H M
ETHOD
As
we di
sc
uss
e
d i
n
sect
i
on
1
si
ze, eff
o
rt
, d
u
r
at
i
on a
n
d p
r
o
duct
i
v
i
t
y
are t
h
e key
so
ft
wa
re
m
e
t
r
i
c
s t
h
at
can
h
e
l
p
pro
j
ect
m
a
n
a
g
e
r in
p
r
oj
ect m
o
n
ito
ring
and
ex
ecutio
n
bu
t it will h
e
lpfu
l
o
n
l
y
wh
en
p
r
oj
ect man
a
g
e
r
kn
o
w
s t
h
e rel
a
t
i
ons
hi
p am
ong t
h
ese m
e
t
r
i
c
s. The rel
a
t
i
o
ns
hi
p am
ong t
h
e
s
e
m
e
t
r
i
c
s can be de
ri
ve
d usi
ng t
h
e
hi
st
ori
cal
dat
a
of si
m
i
l
a
r
pr
oject
s o
f
t
h
e or
ga
ni
zat
i
on b
u
t
i
t
i
s
po
ssi
b
l
e
onl
y
w
h
e
n
or
ga
ni
zat
i
on p
r
oces
s
m
a
t
u
ri
t
y
l
e
vel (e.g
. SE
I C
M
M
i
l
e
vel
4 or a
b
o
v
e
)
i
s
hi
g
h
. I
f
or
gan
i
zat
i
on p
r
oces
s l
e
vel
i
s
not
hi
g
h
o
r
or
ga
ni
zat
i
on
d
a
t
a
rep
o
si
t
o
ry
i
s
not
s
u
f
f
i
c
i
e
nt
t
o
dri
v
e
an
y st
atistical relat
i
o
n
s
h
i
p
(o
rg
an
izatio
n
d
a
ta reposito
ry
shoul
d
c
ontain sufficient
data
points
e.g. sa
m
p
le size >
1
0
to
d
r
i
v
e an
y
mean
in
gfu
l
statistical relatio
n
)
t
h
en
rel
a
t
i
ons
hi
p p
u
b
l
i
s
he
d by
var
i
ous
gr
ou
ps I
S
B
S
G [
2
]
a
nd
QSM
[
3
]
can be use
d
. R
e
l
a
t
i
ons
hi
p deri
ved
usi
n
g
or
ga
ni
zat
i
on
hi
st
ori
cal
dat
a
o
f
si
m
i
l
a
r pr
o
j
ect
i
s
al
way
s
bet
t
e
r t
h
a
n
t
h
e i
n
d
u
st
ry
pu
bl
i
s
he
d
dat
a
.
Fol
l
o
wi
ng
s
u
b
section
will explain m
e
thod t
o
drive
re
lationship using hist
orical
data
a
nd
also s
o
m
e
of the industry publishe
d
rel
a
t
i
ons
hi
p
.
If
hi
st
ori
cal
dat
a
i
s
not
avai
l
a
bl
e t
h
en
t
h
e
n
p
r
o
j
ect
m
a
nager can use t
h
e rel
a
t
i
ons
hi
p
pu
bl
i
s
hed
by
vari
ous s
o
ft
wa
re benc
h m
a
rki
ng o
r
ga
ni
zat
i
on
or by
rel
a
t
i
ons
hi
p p
u
b
l
i
s
hed
by
resear
cher
, ho
we
ver
,
t
h
ese
rel
a
t
i
ons
hi
p m
a
y
not
be exact
l
y
appl
i
cabl
e
t
o
t
h
e p
r
o
j
ect
b
u
t
i
t
can defi
ni
t
e
l
y
gi
ve i
d
ea or di
rect
i
on
ho
w t
h
ese
m
e
t
r
i
c
s are rel
a
t
e
d t
o
eac
h
ot
her
.
W
e
are
gi
vi
n
g
h
e
re t
h
e r
e
fere
nce
of
o
u
r
resea
r
c
h
w
o
r
k
t
h
at
has
bee
n
ha
s
b
een pub
lish
e
d
2.
1.
Equat
i
o
n
fo
r
To
ta
l Number of
Def
e
ct
s Estima
t
e
d f
r
o
m
So
ft
wa
re
Size,
Ef
fo
rt
s a
n
d Pro
duct
i
v
i
ty
In
hi
s
pa
per
M
r
i
dul
B
h
ar
d
w
aj
an
d
A
j
ay
R
a
na [
5
]
,
“
I
m
p
act
o
f
si
ze a
n
d
pr
od
uct
i
v
i
t
y
on
t
e
st
i
ng
a
n
d
rew
o
r
k
ef
f
o
rt
s
for
we
b-
base
d p
r
o
j
ect
s”, es
t
a
bl
i
s
hed t
h
e f
o
l
l
o
wi
ng
rel
a
t
i
ons
hi
p bet
w
ee
n n
u
m
b
er of
d
e
fect
s,
size, produ
ctivity an
d
efforts. Th
is
relatio
nsh
i
p
was
e
s
t
a
bl
i
s
he
d by
usi
n
g
st
at
i
s
t
i
cal
techni
que
s u
s
i
n
g t
h
e
i
n
d
u
st
ry
benc
h
m
arki
n
g
dat
a
p
ubl
i
s
hed
by
IS
B
S
G [6]
.
N
u
mber of
def
ect
s = - 2.84
+ 0
.
00
011
4 * ef
fo
rt
s +
0
.
0
290
*
size -
0
.
1
22 *
product
i
v
i
ty
(
1
)
As co
ncl
u
de
d by
M
r
i
d
ul
B
h
ar
dwa
j
an
d A
j
ay
R
a
na [5
] “Co-efficient of size is
m
u
ch higher than t
h
e
co-e
fficient o
f
effo
rt th
at
means size has
m
u
ch signifi
cant im
pact
o
n
th
e to
tal n
u
m
b
er o
f
d
e
fects in
co
m
p
ariso
n
to
effo
rt. C
h
an
g
e
in
so
ftware si
ze will h
a
v
e
b
i
g
g
e
r im
p
act o
n
th
e
nu
m
b
er
o
f
d
e
fects id
entified
,
therefore,
we c
a
n say that
while plan
n
i
ng
th
e sof
t
w
a
r
e
pro
j
ect w
e
sh
ou
ld use app
r
op
r
i
ate
so
f
t
w
a
r
e
esti
mati
on
t
ool
s a
nd t
e
c
h
n
i
que t
o
re
duc
e
t
h
e m
a
rgi
n
o
f
err
o
r i
n
si
ze and we
should
re
-estim
ate the size after e
v
ery
pha
se
to re
-calibrate
ove
rall ef
fo
rts.
Lo
w c
o
nstant
num
ber i
n
eq
uatio
n
si
g
n
i
fies
th
at th
ere is
v
e
ry less en
v
i
ronmen
t
noi
se o
n
t
h
e t
o
t
a
l
num
ber of defect
s an
d i
t
is si
gni
fi
cant
l
y
depe
n
d
s o
n
so
f
t
ware si
ze so err
o
r i
n
si
ze est
i
m
at
i
on
may
led
to
error in
esti
m
a
t
i
o
n
o
f
testin
g
efforts. Resu
lts
also suggest that in web-
base
d p
r
o
j
ect
s t
h
e num
ber o
f
d
e
fects id
en
tified
is d
i
rectly p
r
opo
rtio
n
a
l to
th
e produ
ctiv
ity i.e. h
i
g
h
er
prod
u
c
tiv
ity will l
e
d
to
m
o
re d
e
fects
foun
d
and
lower p
r
od
u
c
tiv
ity will lead
to
fewer d
e
fects fo
u
n
d
,
th
erefor, less testin
g
and rework
effo
rt
will b
e
requ
ired
if we
sp
en
d
m
o
re time o
n
d
e
v
e
lopmen
t (i.e. ti
m
e
sp
end
till co
n
s
tru
c
tion
ph
ase). Th
e to
tal num
b
e
r o
f
d
e
fects
will g
e
t
red
u
c
ed
and
it will d
i
rectly c
o
n
t
ribu
te in
red
u
c
i
n
g th
e
rewo
rk
effo
rts.”
For
t
h
e
no
n
-
w
e
b-
base
d p
r
oje
c
t
s
sim
i
l
a
r rel
a
t
i
ons
hi
p
was
p
r
esent
e
d M
r
i
d
ul
B
h
ar
d
w
a
j
a
nd
A
j
ay
R
a
na
[7]
i
n
IEEE i
n
ternational c
o
nfe
r
e
n
c
e
on “
I
nte
r
nati
onal C
o
nfe
r
en
ce on F
u
turistic Tre
n
ds in Com
putational Analysis
and
Kn
o
w
l
e
d
g
e
M
a
nagem
e
nt
“ hel
d
at
Great
er NO
ID
A, f
o
l
l
o
wi
ng
rel
a
t
i
ons
hi
p bet
w
ee
n n
u
m
b
er of d
e
fect
s,
si
ze, p
r
o
d
u
ct
i
v
i
t
y
and
ef
fo
r
t
s was
prese
n
t
e
d. T
h
i
s
rel
a
t
i
ons
hi
p
was a
l
so est
a
bl
i
s
h
e
d
by
usi
n
g
i
n
dust
r
y
benc
hm
arki
n
g
dat
a
p
u
b
l
i
s
he
d
by
ISB
S
G
[
6
]
.
N
u
mber of
def
ect
s = - 11
.3
- 0
.
00
027
2 * effo
rt
s +
0
.
05
29
*
size + 0.538
*
pro
d
uct
i
v
i
t
y
(2
)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Key Software
Metrics and its Impact
on ea
c
h
other
for Software Development Projects
(
M
ri
dul
Bh
ar
d
w
aj
)
24
6
As con
c
lud
e
d
b
y
Mridu
l
B
h
ardwaj
an
d
Ajay Ran
a
[7
], “h
igh
e
r pro
d
u
c
t
i
v
ity will lead to
t
h
e less
num
ber o
f
de
fect
s so n
o
n
-
web
base
d pr
oject
s
h
o
u
l
d
be pl
an
ne
d w
i
t
h
m
u
ch exp
e
ri
ence t
eam
(hi
g
h
e
r
pr
o
duct
i
v
i
t
y
) so t
h
at
we sho
u
l
d
ha
ve l
e
ss defect
an
d
re
work e
f
fort. It
also suggests
that size has
m
u
ch
si
gni
fi
ca
nt
i
m
pact
on
t
o
t
a
l
nu
m
b
er of
de
fect
i
n
c
o
m
p
ari
s
on
t
o
e
f
f
o
rt
s.
I
n
m
u
lt
i
l
i
n
ear re
gressi
o
n
e
quat
i
on
si
ze
co-e
fficient is m
u
ch highe
r than the
efforts
co-efficient that
m
eans size
has m
u
ch si
gni
fi
cant
im
pact
on t
o
t
a
l
n
u
m
b
e
r of defect in
co
m
p
ariso
n
t
o
efforts.
Ch
ang
e
in
u
n
i
t
size will h
a
v
e
b
i
gg
er im
p
act i
n
co
m
p
arison
t
o
un
it
chan
ge i
n
ef
fo
r
t
. So st
udy
c
o
n
c
l
ude t
h
at
w
h
i
l
e
pl
an
ni
n
g
t
h
e
soft
ware
pr
o
j
e
c
t
we sh
o
u
l
d
u
s
e ap
pr
op
ri
at
e t
ool
s
to red
u
ce the
m
a
rgin o
f
er
ro
r in si
ze estimation and we
shoul
d
re-esti
m
ate the size,
after eve
r
y phase of
d
e
v
e
l
o
p
m
en
t life cycle, to re-calib
rate ov
erall effo
rts an
d to min
i
m
i
ze th
e im
p
act on the
project
plan.”
2.
2.
Rel
a
ti
ons
h
i
p
a
m
on
g Pr
od
uct
i
vi
ty
, E
f
for
t
,
S
o
f
t
w
a
re
Si
z
e
and Pr
ojec
t
Du
rati
on
as
Per I
S
B
S
G
ISBSG in its bo
ok
“Practical
So
ft
ware Proj
ect
Est
i
m
a
t
i
on” [
6
]
has
p
u
b
l
i
s
hed
eq
uat
i
o
n
f
o
r
so
ft
wa
re
si
ze, eff
o
rt
s
,
p
r
oject
du
rat
i
o
n,
pr
o
duct
i
v
i
t
y
an
d m
a
xim
u
m
t
eam
si
ze. These
equat
i
o
ns
hav
e
been
deri
ve
d
aft
e
r
the statistical
analysis of
projects in t
h
e ISBSG re
p
o
sitory
.
ISBS
G study
s
h
o
w
e
d
that so
ftware s
i
ze an
d
max
i
m
u
m
tea
m
size are th
e k
e
y m
e
trics for estim
a
ti
ng p
r
o
j
ect
du
ra
t
i
on an
d e
f
f
o
r
t
s. So
ft
war
e
s
i
ze and
m
a
xim
u
m
t
e
am
si
ze are t
h
e i
nde
pen
d
e
n
t
m
e
t
r
i
c
s and
ot
he
r m
e
t
r
i
c
s prod
uct
i
v
i
t
y
, eff
o
rt
s and
p
r
o
j
ect
d
u
rat
i
o
n
depe
n
d
s
o
n
t
h
e
s
e t
w
o.
We
ha
ve cl
assi
fi
e
d
t
h
e eq
uat
i
o
n
i
n
f
o
l
l
o
wi
ng
2
g
r
o
ups
.
Grou
p 1
-
Eq
uat
i
o
n
s
t
o
sh
o
w
ho
w
pr
o
duc
t
i
v
i
t
y
, effo
rt
s a
n
d
p
r
o
j
ect
du
r
a
t
i
on
depe
n
d
s
on
so
ft
wa
re si
ze an
d
m
a
xim
u
m
t
e
am
si
ze (sect
i
on 1.
2.
1 t
o
sect
i
o
n
1.
2.
3)
Grou
p 2
-
E
quat
i
o
ns
t
o
s
h
ow
h
o
w
p
r
od
uct
i
v
i
t
y
, eff
o
rt
s an
d
pr
o
j
ect
du
rat
i
o
n
depe
nds
o
n
s
o
ft
wa
re si
ze
(
s
ectio
n 1.2.1 to
section
1
.
2
.
3)
2.
2.
1.
Equa
tion
f
o
r Prod
uctivi
ty
, Estima
ted fro
m
Softw
are
Siz
e
and
Maxim
u
m Team Siz
e
As pe
r ISB
S
G
[7]
for
new
d
e
vel
o
pm
ent
project
s
,
pr
o
duct
i
vi
t
y
can be est
i
m
a
t
e
d usi
n
g soft
ware si
ze
an
d m
a
x
i
m
u
m
tea
m
size as p
e
r th
e fo
llo
wi
n
g
equ
a
tio
n.
Prod
ucti
vi
ty
= 37
.4
8 *
Siz
e
(-0.496)
*
Tea
m
Size
(0
.759)
Whe
r
e: Size
=
Softwa
re size
in function poi
n
ts
Team
Size = Maxim
u
m
tea
m
s
i
ze
2.
2.
2.
Equa
tion
f
o
r
Eff
o
rts
,
Estim
a
te
d fr
om
Software
Siz
e
and Maximum
Team Siz
e
As
per
ISB
S
G [
7
]
,
f
o
r ne
w de
vel
opm
ent
pr
oject
s
,
t
o
t
a
l
pr
oject
e
f
f
o
rt
s can
be est
i
m
a
t
e
d usi
n
g
soft
ware
size a
n
d m
a
xim
u
m
tea
m
size as pe
r the
following
equation.
Effo
rts
= 37
.4
8 *
Siz
e
(.504)
*
Tea
m
Size
(0.759)
Whe
r
e: Size =
Soft
ware
size i
n
function
poi
nts
Team
Size = Maxim
u
m
tea
m
size
2.
2.
3.
Equa
tion
f
o
r
Projec
t
Dur
a
t
i
on, Es
tima
te
d fr
om
Softw
are Siz
e
and
Maxim
u
m Te
am Siz
e
As per ISB
S
G
[7]
,
f
o
r ne
w d
e
vel
o
pm
ent
pr
oject
s
,
n
o
sui
t
a
bl
e
eq
uat
i
o
nca
n
be deri
ve
d
a
s
m
a
xim
u
m
t
e
am
si
ze does
not
m
a
ke si
g
n
i
f
i
cant
i
m
pact
on
pr
o
j
ect
d
u
rat
i
on
.
2.
2.
4.
E
qua
ti
on
f
o
r Prod
ucti
vi
ty
, E
s
tima
ted fro
m
So
ftw
are Siz
e
As pe
r ISB
S
G
[7]
,
f
o
r
ne
w de
vel
o
pm
ent
pro
j
ect
s, no s
u
i
t
a
bl
e equat
i
o
n can
be de
ri
ve
d as s
o
ft
ware si
ze
doe
s
not
m
a
ke
si
gni
fi
ca
nt
i
m
pact
on
t
h
e
p
r
o
d
u
ct
i
v
i
t
y
.
2.
2.
5.
Equa
tion
f
o
r
Eff
o
rts
,
Estim
a
te
d fr
om
S
o
f
t
w
a
re
Siz
e
As
per
ISB
S
G
[7]
,
fo
r
ne
w de
vel
o
pm
ent
pr
oj
ect
s, t
o
t
a
l
p
r
o
j
ect
eff
o
rt
s ca
n
be est
i
m
at
ed usi
ng
so
ft
w
a
r
e
size
Effo
rts
= 23
.2
5 *
Siz
e
(.814)
Whe
r
e: Size =
Soft
ware
size i
n
function
poi
nts
Team
Size = Maxim
u
m
tea
m
size
2.
2.
6.
Equa
tion
f
o
r
Projec
t
Dur
a
t
i
on, Es
tima
te
d fr
om
So
ftw
a
re Siz
e
As
per
ISB
S
G
[7]
,
f
o
r
ne
w
de
vel
o
pm
ent
pr
oj
ect
s, t
o
t
a
l
p
r
oje
c
t
du
rat
i
o
n ca
n
be e
s
t
i
m
a
t
e
d u
s
i
n
g
so
ft
wa
re s
i
ze
Projec
t dur
ati
o
n
= 0.543
*
Size
(.
408)
Whe
r
e: Size =
Soft
ware
size i
n
function
poi
nts
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
IJEC
E V
o
l
.
6, No
. 1, Feb
r
uar
y
20
1
6
:
24
2 – 24
8
24
7
3.
RESULTS
A
N
D
DI
SS
CU
S
I
ON
For
we
b ba
se
d p
r
o
j
ect
, i
f
w
e
com
b
i
n
e t
h
e equat
i
o
n
(1
)
of sect
i
o
n
1.
1
and e
q
uat
i
o
n
s
descri
be i
n
sect
i
on
1.
2,
we
get
t
h
e
f
o
l
l
o
wi
ng
de
fect
e
quat
i
on i
n
t
e
rm
s of
soft
ware
si
ze a
n
d
m
a
xim
u
m
team
si
ze.
Number
of
D
e
fects
= -
2.
84
+ T
e
am
Si
z
e
(
0
.7
59)
(0
.00
42
*
Siz
e
(
.
50
4)
- 4
.
57
25
*
Siz
e
(
-
0
.
49
6)
)
+
0.
02
9
0
*
Si
z
e
Eq
uat
i
ons
des
c
ri
be
d i
n
sect
i
on
1.
2.
1 t
o
1..
26 s
u
g
g
est
t
h
a
t
out
of t
h
e f
o
u
r
key
so
ft
w
a
re
m
e
t
r
i
c
s
soft
ware
size signi
ficantly im
pact
the
othe
r three metrics(project
ef
fo
r
t
, du
r
a
tion
an
d pr
odu
ctiv
ity)
.
Produ
ctiv
ity do
es
no
t sign
ifi
can
tly d
e
p
e
nd
o
n
th
e soft
ware size bu
t it
rep
r
esen
ts t
h
e
no
n
lin
ear
relatio
n
s
h
i
p
wi
t
h
s
o
ft
ware
si
ze an
d m
a
xi
m
u
m
t
e
am
si
ze, h
e
nce
,
It
i
s
r
ecom
m
e
nde
d
n
o
t
t
o
ha
ve
a
ve
ry
bi
g t
eam
si
ze as i
t
m
i
ght
im
pact
the o
v
eral
l
pr
o
duct
i
v
i
t
y
. Tot
a
l
pro
j
ect
du
rat
i
on
onl
y
depe
n
d
s o
n
t
h
e soft
ware si
ze and
i
t
does
not
de
pe
nd
on
t
h
e
m
a
xim
u
m
t
e
am
si
ze. It
im
pli
e
s t
h
at
we cann
o
t
red
u
ce
pro
j
ect
d
u
rat
i
on
by
i
n
creasi
ng t
h
e
tea
m
s
i
ze. Th
is fact is co
n
t
rary to
th
e p
e
rcep
tio
n
th
at we
can reduce the
project
d
u
rat
i
on
by
i
n
creasi
ng t
h
e
project team
size.
We ca
n c
onclude
that s
o
ft
ware
size is
t
h
e im
p
o
r
tan
t
m
e
trics and
a significan
t effort mu
st
b
e
p
u
t
during
proj
ect in
itiatio
n
p
h
a
ses to
estimate th
e p
r
oj
ec
t size. As so
ft
ware size will h
e
lp
in
estim
at
in
g
th
e
p
r
oj
ect du
ration
and
pro
j
ect effo
rts so
error in
esti
m
a
ti
n
g
th
e so
ft
ware size will h
a
v
e
significan
t i
m
p
act o
n
t
h
e
accuracy
of project
duration and e
f
fo
rt. All these key
metrics
m
u
st be
re-calibrated
during t
h
e
project
devel
opm
ent life cycle
In his book “
F
ive C
o
re
Me
tric
s
–
The Intelligence be
hind Succes
sful Soft
ware
Managem
e
nt
Project”, Put
n
a
m
,
Lawrence H., and Ware [1]
desc
ribe
t
h
e
fiv
e
co
re m
e
tr
ics th
at can
b
e
h
e
lpfu
l to m
a
n
a
g
e
th
e
soft
ware
de
vel
opm
ent
pr
o
j
ec
t
execut
i
o
n.
We
have t
r
i
e
d
t
o
est
a
bl
i
s
h
t
h
e rel
a
t
i
o
ns
hi
p
am
ong t
h
ese
m
e
t
r
i
c
s
usi
n
g t
h
e
benc
hm
arki
ng
dat
a
pu
bl
i
s
he
d by
Int
e
r
n
at
i
o
na
l
S
o
ft
ware B
e
nc
h
m
arki
n
g
Ser
v
i
ce Gr
ou
p,
o
u
t
of t
h
e
fiv
e
m
e
trics, we
h
a
v
e
ex
cl
u
d
e
d
t
h
e “Reliab
ility”
metr
ic as it can
o
n
l
y
b
e
m
eas
u
r
ed
o
n
c
e software
d
e
v
e
l
o
p
m
en
t is co
m
p
leted
an
d
it will n
o
t
h
e
lp
in
p
r
oj
ect
ex
ecu
tion
.
Our stud
y sho
w
s
th
at p
r
oj
ect duratio
n
depe
n
d
s
on s
o
ft
ware
si
ze an
d n
o
t
o
n
t
eam
si
ze, o
u
r st
udy
also
estab
lish th
e fact th
at
produ
ctiv
ity d
o
es no
t
depe
n
d
on
s
o
f
t
ware si
ze
b
u
t
o
n
t
eam
si
ze. O
u
r
fi
n
d
i
n
gs
are i
n
l
i
n
e
wi
t
h
t
h
e
fi
ndi
ng
pu
bl
i
s
he
d
by
“QSM
Soft
ware
Almanac” [3] in hi
s researc
h
edition 2014. Ou
r study shows that Softwa
re
Size influences all the
co
r
e
m
e
tr
ics a
n
d
h
e
n
ce industr
y b
e
st p
r
actices (
e
.g. f
u
n
c
tio
n
p
o
i
n
t
)
m
u
st
b
e
u
s
ed
to
measu
r
e th
e So
ftw
a
re
Size. Our find
in
g
also
in
lin
e with
o
u
r earlier res
earc
h
w
o
rk p
u
b
l
i
s
he
d [
5
]
i
n
AC
M
SIGS
OFT M
a
rc
h
20
15
,
wh
ere
we estab
lish
e
d
th
at to p
r
ed
ict th
e nu
m
b
er o
f
d
e
fects, Software
Size is th
e mo
st in
fl
u
e
n
c
es
metric
a
m
o
n
g
th
e
fiv
e
core m
e
trics.
4.
CO
NCL
USI
O
N
We can concl
ude that s
o
ftware size is the im
por
t
a
nt
m
e
t
r
i
c
sand a si
g
n
i
fi
cant
eff
o
rt
m
u
st
be put
d
u
ring
p
r
oj
ect in
itiatio
n
p
h
a
ses to
estim
a
t
e th
e p
r
oj
ect size.
As so
ft
ware si
ze willh
elp
i
n
esti
m
a
t
i
n
g
th
e
p
r
oj
ect
d
u
ration
and
p
r
oj
ect effo
rts so
error in
esti
m
a
tin
g
th
e so
ft
ware size will h
a
v
e
sig
n
ifican
ti
m
p
act
o
n
t
h
e
accuracy of proj
ect duration and effo
rt. Allthese key metrics
m
u
st
be
re-calibrate
d during the
proj
ect
devel
opm
ent life cycle.
REFERE
NC
ES
[1]
Putnam
, Lawre
n
ce H.
, and W
a
re M
y
ers, “
F
i
v
e Core Me
tric
s—The Inte
llig
e
n
ce beh
i
nd Suc
cessful Softwar
e
Management”, New
York:
Dorset House Publishi
ng Compan
y
,
In
c., 2002.
[2]
ISBSG “The Benchmark data fo
r Soft
ware estim
ation
”
, Release 1
0
(2011)
[3]
QSM Software Almanac,
Application Developm
ent Series
. 2014
Research
Ed
itio
n.
[4]
Mridul Bhardwaj and Ajay
R
a
na, 2014
, Estimate Soft
ware Functional Size before
Requir
e
ment phase of
Developm
ent L
i
fe C
y
cl
e,
International Journal
of Innovations
&
A
d
vancement in Computer Science
, vol. 3 Issue 4
June-2014, pp
7
9
-83
[5]
Mridul Bhardwaj and
Ajay
R
a
na, 2015, Impact
of Size
and Prod
uctivity
on
Testing
and R
e
work
Efforts for Web-
based Dev
e
lopment Projects
,
AC
M SIGSOFT Software Eng
i
neering Notes
, vol. 40
Number 2, M
a
r
c
h-2015, pp
1-4
[6]
Book “
Practical Software
Projec
t Estimation
”, b
y
International S
o
ftware Be
n
c
hmarking Standards
Group, Page no
246-248
[7]
Mridul Bhardwaj and
Ajay
R
a
n
a
, 2015
,
Estimation of
Te
sting
and Rework Effo
rts
for non-Web
-
based Softwar
e
Developm
ent P
r
ojec
ts, I
EEE
i
n
terna
ti
onal
con
f
erence on
“Futuristic Tr
e
nds
in Computation
a
l Analy
s
is and
Knowledge Man
a
gement
” at Greater NOIDA,
25-
27 Februar
y
201
5.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Key Software
Metrics and its Impact
on ea
c
h
other
for Software Development Projects
(
M
ri
dul
Bh
ar
d
w
aj
)
24
8
BIOGRAP
HI
ES OF
AUTH
ORS
Mridul Bhardw
aj is h
a
ving
more th
an 17
years’ exp
e
rience of I
T
project and p
r
oduct
development
in
various techno
logies. Curren
t
ly,
He is senior
project manager
in one of the
lead
ing IT com
p
an
y
in INDIA and als
o
doing re
s
ear
ch in Software quality
metrics for software
development pro
j
ect. He h
a
s published 5 p
a
pe
rs in
reputed intern
ational
journal.
Prof (Dr.) Ajay
Rana is hav
i
ng a rich exp
e
rien
ce of
Industr
y
and Academia
of aro
und
15 y
ears.
He is Founder Director /Group Dire
ctor / Directo
r
Professor / Me
nt
or of more than 27 different
Institutions and
Innovativ
e Program
s at Am
it
y
Group. He obtai
ned Ph.D. in Com
puter Scienc
e
and Eng
i
neering
,
M.Tech
(Master of
Technolog
y
)
in Computer
Scien
c
e and
Engineer
ing and
MBA (Master o
f
Business Ad
minist
ration) He h
a
s published more than 177 Res
earch Pap
e
rs in
reputed J
ourna
ls
and P
r
oceed
ing
s
of Internat
iona
l and Nat
i
onal C
onferenc
e
s
.
He h
a
s
co-author
ed
05 Books and co-edited 36 Con
f
erence Proceedings.
He has delivered Inv
ited
lectures in
more
than 36 Technical and Man
a
gement Workshop / C
onferences programs in India and abroad. H
e
is
a m
e
m
b
er of
Board of Govern
ance
(BOG), Ad
vis
o
r
y
Coun
cil
(
A
C), Acad
em
ic
Execu
tive
(AE)
M
e
m
b
er, Board
of S
t
udies
(BOS
) and S
p
ec
ia
l M
e
m
b
er of m
a
n
y
I
ndian and
F
o
rei
gn Univers
iti
es
as well
as Indu
stries. He
is
Ed
itor
in Chi
e
f,
T
echni
cal
Com
m
itte
e Mem
b
er,
A
dvisor
y
Board
Member for 18
Plus Techn
i
cal J
ournals and
Con
f
erenc
e
s a
t
Na
tio
nal
and In
terna
t
i
onal
Leve
ls.
Evaluation Warning : The document was created with Spire.PDF for Python.