B588 Online Surveys Of User And Employee Data; Tolaba

B588 Online Surveys of User and Employee Data; Tolaba B588 Online Surveys of User and Employee Data; Tolaba

User Manual: B588 Online Surveys of User and Employee Data; Tolaba

Open the PDF directly: View PDF PDF.
Page Count: 12

iSHAREi:
SHARE
SESSION
REPORT
M
en
-.J
=
Go,
B588
Rg~lg~e~usX~r~f~~t~~~r
and '2.5"
SHARE
NO.
SESSION
NO.
SESSION
TITLE
ATTENDANCE
CGS
VM
Systems
Management
Terrance
Baughman
PROJECT
SESSION
CHAIRMAN
Columbia
GAS
System,
1600
Dublin
Road, Columbus, Ohio 43017
SESSION
CHAIRMAN'S
COMPANY,
ADDRESS,
AND
PHONE
NUMBER
INST.
CODE
(614) 486-3681
ON-LINE
SURVEYS OF USER AND EMPLOYEE SATISFACTION
1. Abstract
On-line
survey
facilities
have
been
developed
at
the
T J Watson
Research
Center
(IBM,
Yorktown
Heights,
N.Y.)
Their
purposes
are
to
evaluate
user
satisfaction
with
computer
services
and
employee
satisfaction
with
the
job.
These
facilities
are:
Surveys
at
logoff/dis
conn
time
about
system
response
time
On-line
opinions
surveys
of
user
and
employee
satisfaction
A
general
description
of
the
facilities,
implementation,
advantages,
human
factors)
security
considerations,
and
achieved
objectives
are
out-
lined
in
this
paper.
Author
Jorge
H
Tolaba
Thomas
J.
Watson
Research
Center
of
IBM
P.O.Box 218
Yorktown
Heights,
N.Y. 10598
l/a/pal/l
2.
SURVEY CONDUCTED
AT
LOGOFF/DISCONN TIME
2.1
Introduction
The
LOGOFFs
and
DISCONNs
commands
are
intercepted
during
the
last
part
of
the
working
day
and
the
users
are
asked
to
rate
the
system
response
time.
Their
responses
are
transmitted
in
the
form
of
messages
to
a
receiving
virtual
machine. The
receiving
virtual
machine
captures
these
messages
into
a
CMS
file,
and
·after
completion
of
the
daily
survey
it
computes
sta-
tistics
and
generates
reports
using
the
collected
data.
2.2
Objectives
To
obtain
from
the
users
an
evaluation
of
prime
shift
system
response
time
for
their
terminal
session.
To
determine
level
of
service
wh.ich
satisfy
users
and which
users
need
help.
To
generate
statistical
reports
and
maintain
a
history
of
it.
2.3
Design Considerations
The
interception
of
the
LOGOFFs
and
DISCONNs
is
done by
placing
CMS
execs
with
these
names and
their
valid
truncations
in
the
system.
This
has
the
disadvantage
that
the
LOGOFFs
and
DISCONNs
exercised
by
the
user
at
the
CP
level
are
not
intercepted.
However
this
route
was
chosen
since
it
was
less
practical
to
modify
the
pertinent
system
modules.
It
also
offers
the
users
the
choice
of
bypassing
entirely
the
survey
by
simply
typing:
'CP
LOGOFF'
or
'CP
DISCONN'.
This
is
desirable
specially
since
the
survey
is
conducted
daily.
Naive
users
The
survey
is
carried
out
in
a
conversational
way
and
without
resorting
to
full
screen
displays.
This
avoids
situations
where
users
not
familiar
with
full
screen
mode
may
not
know
how
to
proceed.
It
also
eliminates
the
prob-
lem
of
having
to
take
into
separate
consideration
typewriter
type
termi-
nals
or
certain
other
terminals
(like
Textronix
graphic)
that
are
treated
as
2741
typewriter
terminals
by
the
system.
Burden on
the
users
To
avoid
burdening
the
users
the
LOGOFFs,
DISCONNs
interception
takes
place
at
random
only
in
one
out
of
four
cases,
on
business
days
from
3:30PM
to
6:00PM
and
is
limited
to
one
per
user
per
day.
~
:(,,""1
CO
Syntax
equivalent
to
CP
LOGOFF
and
CP
DISCONN
Any
permissible
syntax
for
the
CP
logoff
and
dis
conn
commands
can
be
exer-
cised
when
the
logoff/dis
conn
execs
are
invoked.
In
addition
they
are
operational
in
any
environment
where
the
CP
commands
are
valid
(Editors,
Dial
PVM
etc.
)
Speed
of
execution
The
design
results
in
a a
prompt
logoff
and
dis
conn
when
the
survey
is
not
taken
place
(Before
3:30PM
and
after
6:00PM).
This
is
particularly
impor-
tant
because
&ly
noticeable
increase
of
waiting
time
when
logging
off
or
disconnecting
would
be
a
frustration
unacceptable
to
the
user.
Security
A
user
is
disconnected
or
logged
off
if
he
does
not
answer
within
a
cer-
tain
time
the
question
raised
by
the
execs
when
interception
and
surveying
is
taking
place.
This
is
done
for
security
reasons
to
take
into
account
the
case
of
somebody
logging
off
(or
disconnecting)
and
walking
immediate-
ly
away from
the
terminal.
Users
running
disconnected
For
obvious
reasons
the
survey
is
bypassed
for
the
case
of
users
running
disconnected.
2.4
General
description
There
are
two
main
components
to
the
facility,
namely:
a
collection
of
CMS
system
execs
and
a
receiving
virtual
machine.
The
first
component
intercepts
the
LOGOFFs
and
DISCONNs
commands,
asks
the
user
to
enter
a
rating
of
system
response
time
and
then
proceeds
to
com-
plete
the
command's
intended
action.
The
second
campon,ent
captures
the
responses
of
the
users
and
produces
sta-
tistical
reports.
2.4.1
CMS
execs
that
intercept
the
LOGOFFs
and
DISCONNs
These
execs
logoff
and
disconnect
a
user
after
asking
the
following
ques-
tion:
"How
do
you
rate
today's
system
response
for
processing
your
prime
shift
work? Type 1
for
very
good,
2
for
.good, 3
for
fair,
4
for
poor,
5
for
very
poor,
or
0
for
no
opinion."
They
do
not
survey
the
user
if
one
of
the
following
conditions
occur:
1)
The
day
of
the
week
is
not
a
business
day
2)
The
time
of
the
day
is
not
in
the
interval
(15:30
to
18:00)
3)
The
user
is
running
disconnected
4)
The
user
does
not
have
a
R/W
minidisk
accessed
as
A
5)
The
user
has
already
participated
in
the
survey
during
that
day
The
following
additional
precaution
is
taken:
6)
The
user
is
logged
off
or
disconnected
if
he
fails
to
response
within
60
seconds
2.4.2
Receiving
virtual
machine
The
following
are
the
main
functions
of
this
receiving
station:
a)
To
capture
any
messages
send
to
it
by
the
users
during
the
time
the
survey
is
conducted
b)
To
compute
statistics
on
the
users
responses
at
the
conclusion
of
the
survey
c)
Maintain
on
line
only
information
for
the
last
20
business
days
Daily,
on
business
days,
the
rece1V1ng
virtual
machine
wakes
up
at
15:25
and
places
itself
into
a
receive
state
during
the
survey's
time
(15:25
to
18:00).
It
then
captures
the
incoming
users
responses
and
places
them
in
a
CMS
file.
At
18:00
it
leaves
the
receiving
state
and
generates
or
modi-
fies
the
following
files:
RESPONSE
FILE,
contains
statistics
(including
the
average
rating
of
response
time)
for:
the
last
business
day,
accumulative
month
to
date,
for
the
last
20
business
days.
This
file
is
shown
in
figure
1
of
Sec-
tion
2.6
COMPLAIN
FILE,
contains
a
history
of
the
users
ratings
for
the
last
20
business
days
sorted
in
descending
order
from
the
lowest
to
the
highest
rating
and
subsequently
by
frequency.
RESPONSE
HISTORY,
contains
a
record
of
the
RESPONSE
FILE
for
the
last
20
business
days.
RESPONSE
LOG,
contains
a
record
of
the
users
ratings
for
the
last
20
business
days.
Extraneous
messages,
no
opinion
ratings,
and
duplicate
entries
deleted.
A
duplicate
entry
is
a
record
with
the
same
llserid,
the
same
rating
for
the
same
date.
After
generating
these
files
the
virtual
machine
places
itself
into
a
sleep
state
until
15:25
of
the
next
business
day.
~
c.n
CCl
2.5
Graphics obtained
by
using
the
collected data
The
average
daily
user
rating
of
response
time
was
plotted
versus
some
of
the
recorded
system
measurements.
Figures
2 and
3,
in
Section
2.6,
show
the
plots
of
the
average
response
time
rating
versus
the
daily
average
response
time
and
the
daily
average
expansion
factor
respectively,
for
the
VM
V
system
at
Yorktown, and
for
the
interval
November
18,
1981
to
Febru-
ary
2,
1982.
Figures
4,
5 and
6,
in
Section
2.6,
show
the
daily
user
rating
of
response
time
of
the
3
VM
systems
at
Yorktown, on a monthly
average
basis,
from
June
82
to
June
83
together
with
the
hourly
response
time,
hourly
expan-
sion
factor
and downs
per
day.
2.6.
Figures
.for
generated
report
and graphs
FIGURE
1
DATE:
83/07/20
RESPONSES
(INCLUDING
NO
OPINION):
82
NO
OPINION:
17
21%
VERY
VERY
NUMB
AVG
SERVICE
GOOD GOOD
FAIR
POOR
POOR
RESP
RATING
---------
...
--
RESP.
T.
17%
11
28%
18
35%
23
9%
6
11%
7
65
2.69
RESP.
T. P>50
0%
0
0%
0
0%
0
0%
0
100%
1 1
5.00
DATE:
83/07/01
TO
83/07/20
RESPONSES
(INCLUDING
NO
OPINION): 917
NO
OPINION:
226
25%
VERY VERY
NUMB
AVG
SERVICE
GOOD
GOOD
FAIR
POOR
POOR
RESP
RATING
------------
RESP.
T.
13%
88
28%
193
31%
214
13%
93
15%
103 691
2.90
RESP.
T.
P>50
8%
2
15%
4
46%
12
8%
2
23%
6
26
3.23
DATE:
83/06/22
TO
83/07/20
RESPONSES
(INCLUDING
NO
OPINION): 1392
NO
OPINION:
347
25%
VERY
VERY
NUMB
AVG
SERVICE
GOOD GOOD
FAIR
POOR POOR
RESP
RATING
------------
RESP.
T.
12%
130
28%
297
29%
305
13%
139
17%
174 1045
2.93
RESP.
T. P>50
10%
3
17%
5
47%
14
7%
2
20%
6
30
3.10
Q)
E
;:
3.6
.
Co
II)
Q)
...
3.2
~
0
C)
C
;:
2.8
0
...
...
Q)
~
2
.....
.
C)
>
Figure
2
Prime shift
YKTVMV
Time interval:
(81/11/18
to
82/02/05)
J 0
<C
2.0
L'--~'~I--
__
~
__
-L
____
~
__
-L
____
~
__
~
____
~
__
~
__
~
0.3
0.....
o.S
Daily average
LInear
fit
coefficient.
Ordinate
to
the
origin:
.70
Slope: 3.86
R Squared factor. .671
Dally
avg. re.p
time
fO.r
a
2.5
avg.
0.6 0.7
0.8
response time
".er
rating of r
••
p. time: .47
Figure
3
PrIme shIft
YKTVMV
Time interval:
(81/11/18
to
82/02/05)
CD
E
:;:
3.6
.
0..
II)
CD
...
3.2
....
0
0
0
00
/'
NO)
c:1).=
2.8
=
.....
0
...
...
CD
10
o 0 0
o
o _
.;....-"
0
Yo
8 0
o 0
II)
2.4
:J
8 § 0
. o
0)
> o
< 2.0
2.5 3.0 3.5 4.0
4.5
5.0
Daily
average expansion factor
Lln.a~
fit co.fflclents
Ordinate
to the origin: .35
Slop.: .67
R Squar.d factor: .776
Dally
avg.
expo
factor for a 2.5 avg.
u.er
rating of resp. time: 3,22
oj
V
~
T
x
1.0
!:l
0.8
SERVICE OBJECTIVES -
prime
shift
HOURLY RESPONSE
TIME
(seconds),
90th
percentile
and
USER'S
RATING
of
response
time
z
0.6
:=C""_c~"_~".c,u~_~_~~-~
___
~~---~~""";----;
;
__
ee'C
~
-=
1:
:
__
~~~MJ~~~~~~~~~_____________
20
---
-......--------
-......--+--~.
-
-:--
0
0-
~
0.4
---.
•
RESPONSE
TIME 1
Ij
• USER'S RATING
0.2
L ' , , "
10
1.0
'"
~
o.e
-===~:::::=----::-.B!~tp!=~:~-.,..-=-~~=::::~-:=-:.::.-..::.~=-==-:::":~::'-.:::-.~======--:_-:_-:::~::--_~
-~-..:.--=.--::
0.6_.:::..::::...~~~ID:J~E-(LOAlJ_0~~_
_______
________
~_~----~~:-
.
----.
0.4
• RESPONSE TINF.
• USER'S RATING
0.2L-
__
~
____
-L
____
~
__
~
____
-L
____
~
____
~
__
~
____
-L
____
L-
__
~
____
~
I.°lr-----r----,-----r----,r---~----_,----_r----~----._----r_--_,----_,
O.B
_-:-:=------"=:'"-:::.B!P..~=:;2~.::-~--:v=---=-----..-==":.:":.!I-....::.~--=-=:.=:.::.:e--=-........-::..-I!=~-
...
~-~----:...-:...-:-~
0.6
__
~~~~T
E.(LOALJ_O~~___________
_
_____
_
0.4
• RESPONSE TIME
• USER'S RATING
30
lj
2.5 "
'<
2.0
0:::
rn
1.5
~
=>
1.0
3.0
lj
2.5
"
2.0 5
!"
1.5
~
:0
0.2 1.0
JUN
JUL
AUG
SEP
OCT
NOV
DEC
JAN
FEB
MAR
APR
MAY
JUN
~ ~
V:
YORKTOWN
VM
V
T:
YORKTOWN
VM
T
X:
YORKTOWN
VM
X
Figure
4
0:
0
...
u
<
..
z
V
~
"
~
1.1
NI5
~
--
"-
T
~
§
0:
0
t
;::
z
X
~
..,
~
SERVICE OBJECTIVES -
prime
shift
HOURLY EXPANSION FACTOR.
90th
percentile
and
USER'S RATING
of
response
lime
8 , I I
I-~
-_
• EXPANSION FACTOR ____ • USER'S RATING ____ 2 B
:
-:-:~:~:::"'---7!~~~"'~=--=--:ii:---=---~--""-=-.,,.--""I!!:::-':-'::'-'-::::::""'----.-
"'--::.:~
•
.:.--.:.
- - 2 4
EXPANSION FACTOR GOAL
~
5 0
5
-----------------_~~~~~1'~!9!!~----------
20
3'
-
~
~~
I I I
-~--~~
16
I
"
z
~
~V
0:
OJ
UJ
::>
8 • EXPANSION fACTOR • USER'S RATING
__
:-===~::~~~~~~~~~5~-~-~~~~~~~~-:=~~::~~===~~~~~~~~~--
2.fJ
~
;:
EXPANSION FACTOR GOAL
~
5.0 EXPANSION
FACTOR
GOAL
~
4-!...c:::-':::'
5
-------------___
_
_________
~~~
4
3~'----~----~----L---~
____
~
____
_L
____
_L~
__
~
____
L_
__
~L_
__
_J
____
~
2,4
~
'r
2.0
f3
~
1.6
8rl----.----.,----.----.-----r----,-----.----.----,-----r----.----.
SER'S RATING
_---
•
EXPANSION
FACTOR
• U
.:::...t!:~-::-_';::'_-.::..-:..-::.-_-
2.8
~
!;i
4
7r--~~~~E~_~~~~=~~---~-==~~~--~~~~~~~~
_~
6
---
-;~-;;SION
FACTOR GOAL
~
5_0
E!!'~.§.I~..1'~Cl'O.!!.QCMk..t.L4
_____
__==_:;=___,
5
--------------------------~
,
~
1.6
:
1'::-:::---.
L
~_
, ,
3
JUN
2.4
~
X
2.0
~
UJ
::>
JUL
AUG
82
V: YORKTOWN
VM
V
T: YORKTOWN
VM
T
X: YORKTOWN
VM
X
SEP
OCT
NOV
DEC
Figure
5
JAN
B3
fEB
MAR
APR
MAY
JUN
SERVICE OBJECTIVES -
prime
shift
DOWNS/DAY
and
USER'S RATING
of
response
lime
2,Or'----~-----,r-----r-----,------r-----r-----,------,-----.-----.------r-----,
>-
---
...
_--.
~--
....
--- ---
....
-
..
---.-----
~
• DOWNS/DAY • U.§f:R'S RATING
---t
2.B "
< 1.5 t
"'-...
RATING.:
2~
_______
..
__
---
__
=---------~
__
..
--:.::.--_-_-
~
€.
F-----
....
--::'~7;:::::--:tJ-------
... ---- ---
..
--
-
2.4
~
~
1.0
~
g
0.5
_~~~~~U~L@~~~~----------------
~-:::~=-=
01
~
~~,
~
__
J_
____
~
__
_J
____
_L
_____
L_
__
~
____
~
2.0r'-----.-----,------.-----.------r-----r-----,------.-----.-----,------.-----,
~
1.5
'"
~
1.0
~
g
0.5
0
2.0
~
1.5
~
1.0
~
g 0.5
• DOWNS/DAY • USER'S RATING
__
--
,-
==::
:~:.:::.:.
..
-~
..
~~~:_~_;5::-~-~-.:.-:.-:.-!:"::===-=-=e.:.-.=.--=--:~::~~.::=.::=.;:--:--:---=--;.::
;:::--a::'::
--
_~~~~U~L@~~~_~---------~------------------
~--
~-------
---------.----------
~~-----.
• DOWNS/DAY • USER'S RATING • ____
---
~~~~~
...
~~~~~~~~~~=-~~~=~~~--~~~~~~~~~~~.~-~~----
_____
~~~U~L@~~~~
_______
_
v,
2.0
e3
:n
'-'
1.6
u"
~
;:
U~
~
UI
I.
-
2.
~
~
;:
U~
UJ
U~
:n
o t
=-:
"--------=
~
d
1.6
::>
JUN
JUL
AUG
SEP
OCT
NOV
DEC
JAN
FEB
MAR
APR
MAY
JUN
~
~
V:
YORKTOWN
VM
V
T: YORKTOWN
VM
T
X: YORKTOWN VM X
Figure
6
N>
<=
N
3.
ON-LINE
OPINION SURVEYS OF USER AND EMPLOYEE
SATISFAC-
TION
This
facility
has
been
used
to
conduct
the
Annual
Survey
of
Yorktown
Com-
puter
Services
and
the
IBM
Research
Division
Opinion
Survey.
3.1
Introduction
In
response
to
a
system
command
SURVEY
the
user
is
presented,
in
full
screen
mode,
with
the
pertinent
survey
questions
and
offered
the
opportu-
nity
to
enter
related
comments. At
the
end
of
the
session
the
user's
replies
are
transmitted
to
a
collecting
virtual
machine
for
tabulation
and
report
generation.
3.2
Objectives
The
main
objectives
of
on-line
surveys
are
the
capability
of
conducting
surveys
on
a
very
short
notice,
without
all
the
hassle
of
related
paper-
work
and
mailing,
and
the
ability
to
generate
automatic
summary
reports
of
answers
either
on a
partial
progressive
or
final
basis.
These
objectives
have
been
achieved
only
partially
since
in
most
cases
not
all
the
survey
participants
have
access
to
a
computer
terminal,
are
familiar
with
the
pertinent
techniques,
or
are
willing
to
take
a
survey
on
line
preferring
instead
the
conventional
way
of
filling
a
paper
form.
Reduction
in
the
use
of
paper,
greater
flexibility
for
the
user
to
file
his
survey,
improved
security,
elimination
of
ballot
stuffing
and
greater
appeal
to
certain
users
are
some
of
the
additional
advantages
of
on-line
surveys.
3.3
General
description
The
facilities
covered
in
this
paper
operate
under
VM/SP
and
are
func-
tional
in
tube
type
terminals
only.
The
different
components
were
written
in
CMS
EXEC2
language.
Full
screen
support
is
provided
by
1053270 (The
Display
Input/
Facility).
There
are
three
main components
to
the
faciiity,
namely:
A
system
command, named
SURVEY,
to
be
invoked
by
the
user
in
order
to
take
the
survey.
A
receiving
station
or
collector
where
the
surveys
are
transmitted
to
and
collected.
Postprocessing
programs
for
tabulation.
3.4
Flexibility
to
the
user
and human
factors
The
questions
of
the
survey
may
be
answered
by
the
user
in
any
order.
An
index
permits
the
selection
of
any
desired
question.
Any
or
all
the
answers
may
be
reviewed
at
any
time
and/or
before
transmission.
The
sur-
vey
may
be
ended
by
the
user
at
any
time
with
the
option
of
continuing
lat-
er.
This
enables
him
to
take
the
survey
in
several
sessions
if
he
so
desires
or
finds
it
necessary.
However
this
feature
is
not
available
in
surveys
about
very
sensitive
or
confidential
matters
for
security
reasons.
Comments
may
be
entered
by
means
of
the
participant's
preferred
CMS
editor.
After
the
invocation
of
the
command
the
user
is
presented
with
an
intro-
duction
that
contains
instructions
to
the
survey.
The
survey
itself
con-
sists
of
the
displays
of
a
certain
number
of
frames
containing
the
survey
questions
and
instructions
to
the
user
on
how
to
proceed.
Defined
PF
keys
permit
him
to
slide
from
display
to
display.
There
is
little
room
for
operational
errors
since
input
fields
are
the
only
ones
that
are
unpro-
tected,
furthermore
the
autoskip
feature
of
full
screen
mode
is
in
effect
throughout
the
survey.
3.5
System command SURVEY
SURVEY
is
the
CMS
system
command
that
initiates
the
survey.
It
is
opera-
tional
only
during
the
duration
of
the
survey.
The
users'
replies
and
comments
are
written
in
a
CMS
file
in
one
of
his
minidisks.
At
the
end
of
the
survey
the
users'
answ,ers
are
transmitted
to
a
receiving
virtual
machine
for
collection.
The
conduction
of
a new
survey
requires
the
change
of
the
input
data
file
containing
the
questions
and
1053270
full
screen
control
characters.
In
the
case
of
confidential
surveys
and
for
security
reasons
the
fo11owing
are
additional
restrictions:
The
availability
of
the
command
SURVEY
is
restricted
to
business
hours.
During
the
users'
session
his
answers
are
kept
in
temporary
disk
space.
The
advantages
are
that
the
information
is
not
accessible)
and
it
is
erased
in
case
of
a
system
crash.
The
obvious
disadvantage
is
that
the
information
is
lost
if
the
system
crashes
with
the
result
of
the
user
having
wasted
his
effort.
3.6
The
collector
The
survey
files
are
read
as
soon
as
they
arrive
to
the
collector
virtual
reader.
The
originating
user
id
is
detected
at
this
time
and
recorded
with
the
aim
of
precluding
ballot
stuffing.
In
the
case
of
confidential
sur-
veys
and
for
security
reasons
the
following
additional
features
are
imple-
mented:
The
user
id
undergoes
checks
for
eligibility
(i.e
if
it
is
a
regular
IBM
employee)
and
for
previous
participation.
Despite
these
checks
the
par-
ticipant
remains
anonymous. .
Rejected
files
are
discarded
and
the
originator
is
notified
electron-
ically
of
the
reasons
for
the
rejection.
Accepted
survey
files
are
assigned
a
sequential
number
and
encrypted
pro-
vided
the
virtual
machine
is
operating
in
enciphering
mode.
Encryption
is
done
by
means
of
IPS/CMS
Cryptographic
Program
(IBM
IUP
5796-PPK). The
originator
is
notified
by
electronic
mail
about
his
survey's
acceptance.
An
enciphering
key
is
used
for
encryption
of
the
files.
Enciphering
mode
is
established
at
the
time
the
enciphering
key
is
entered
manually
by
the
survey
administrator.
The
enciphering
key
resides
in
virtual
memory
only
for
security
reasons.
The
key
is
tested
to
determine
if
it
is
the
right
key.
Wrong
keys
are
rejected.
The
right
key
is
the
key
entered
by
the
sur-
vey
administrator
at
key
establishment.
Key
establishment
is
done
only
once
by
means
of
a
key
generation
program.
No
record
of
the
key
remains
in
disk
storage.
The
collector
notifies
the
survey
administrator
when
is
not
operating
in
enciphering
mode
(i.e.
every
time
the
system
is
brought
up).
Multiple
checks
exclude
the
possibility
of
a
survey
file
being
lost.
In
this
regard
the
collector
notifies
the
survey
administrator
and
the
pro-
grammer
about
the
nature
of
any
malfunction
and
then
logs
itself
off.
Access
to
the
receiving
virtual
machine
is
restricted
to
the
survey
admin-
istrator
and
authorized
employees.
N
3.7
The
postprocessing
programs
~
c...::l
This
component
completes
the
cycle
of
automating
the
conduction
of
surveys.
There
are
currently
two
postprocessing
programs.
One
for
confidential
surveys
(i.
e.
an
IBM
Opinion
SURVEY),
and
another
for
surveys
where
confidentiality
is
not
so
critical.
In
either
case
the
programs
require
information
about
the
structure
and
format
of
the
survey.
The
final
output
consists
of
summary
reports
containing
statistical
tabulations
of
the
survey's
answers,
plus
files
where
the
users'
comments
are
classified
and
appended.
Postprocessing
may
be
done
on
several
sessions
as
the
survey
progresses
or
at
its
conclusion.
In
the
case
of
the
IBM
Research
Division
Opinion
Survey
the
postprocessing
component
is
an
interface
to
TOSS
(Total
Opinion
Survey
System)
developed
by
IBM
Corporate.
The
following
steps
and
considerations
apply
to
the
case
of
confidential
surveys:
The
survey
files
are
deciphered.
The
cipher
key
is
entered
manually
by
the
survey
administrator
and
it
resides
in
virtual
memory
only.
Generated
reports
are
encrypted
and
all
deciphered
files
are
erased
before
the
ses-
sion
is
ended.
Access
to
the
postprocessing
programs
is
restricted
to
the
survey
administrator
and
authorized
employees.
3.8
Usage
The
survey
facility
has
been
used
at
the
Thomas
J.
Watson
Research
Center
of
IBM
since
1981
to
conduct
the
Annual
Survey
of
Yorktown Computing
Ser-
vices.
It
was
also
used
to
conduct
the
IBM
Research
Division
Opinion
Sur-
vey,
Fall
1982.
These
surveys
were
also
available
in
paper
form.
1100 employees
participated
in
the
IBM
Research
Division
Opinion
Survey.
Of
this
517
took
the
survey
on
line
(47%
of
the
total).
356
users
took
the
1982 Annual
Survey
of
Yorktown Computing
Systems.
Approximately
half
of
them
(177)
participated
on
line.
3.9
Demonstration
The
following
example
demonstrates
through
captured
spooled
console
images
the
case
when a
user
invokes
the
SURVEY
command,
replies
to
questions
1 and
2,
selects
from
the
index
question
9,
answers
question
9,
enters
comments
to
question
9,
answers
question
10,
ends
the
survey
and
transmits
his
answers.
N)
CO)
.r;;..
The
survey
introduction
is
displayed
following
the
invocation
of
the
system
command
SURVEY.
In
this
case
the
introduction
extends
over
several
frames.
Only
the
first
frame
is
displayed.
ANNUAL
SURVEY OF YORKTOWN COMPUTING SERVICES
In
order
to
determine
user
satisfaction
with
our
services,
we
are
conducting
this
survey.
Please
base
your
answers
only
upon
your
experiences
during
the
past
three
months.
Do
not
answer
questions
which
do
not
apply
to
you
or
for
which
you
have
no
opinion.
This
survey
consists
of
39
questions.
You can
proceed
to·
the
following
question
by
pressing
the
ENTER
key
or
go
back
to
the
preceding
question
by
pressing
PF7.
You
are
not
limited
to
this
sequential
mode
but
may move
to
any
question
you
wish
by
pressing
PF2
and
making
a
question
number
selection
from
a
question
index.
You may
go
back
to
enter
your
response
to
a
previous
question
that
was
left
unanswered,
or
to
change
a
previous
response.
Whenever
a
question
applies
to
more
than
one
service,
the
Press
'ENTER
or
PFS'
to
CONTINUE
(Next
franie);
PF3
(Quit);
The
first
question
is
displayed
after
the
introduction.
The
cursor
is
positioned
at
the
first
input
field.
QUESTION 1 OF 39
Which one
of
these
computers
have
you
used
the
most
during
the
past
three
months?
(Enter
an x
in
appropriate
field.)
==>
VM-V
==>
VM-T
==>
VM-X
==>
MVS/TSO
==>
IBM
PC
==>
IBM
PC
connected
to
VM
PF2
(ForquestiOnindexT;Press'ENTER'or-PFS
to
CONTINUE;
PF3
(End);
PF4
(To
enter
comments);
PF7
(To
view
Introduction);
N
en
en
After
a
reply
has been
entered
in
the
first
field
the
cursor
moves
to
the
next
input
field
(autoskip
feature).
QUESTION 1 OF 39
Which one
of
these
computers
have
you
used
the
most
during
the
past
three
months?
(Enter
an x
in
appropriate
field.)
==>
x
VM-V
==>
==>
==>
VM-T
VM-X
MVS/TSO
==>
IBM
PC
==>
IBM
PC
connected
to
VM
PF2
(For
question
indexr;--Press
'ENTER'or
PF8
to
CONTlllfUE;
PF3
(End);
PF4
(To
enter
comments);
PF7
(To
view
Introduction);
Question
2 is
displayed
next
following
the
replies
to
question
1.
QUESTION 2 OF 39
Where
are
you
located?
(Enter
an x
in
appropriate
field.)
==>
x
Yorktown
==>
Eastview
==>
Another
IBM
location
(specify):
PF2
(For
question
index);
Press
'ENTER'or
PFS
to-COI\lTTNUE;
PF3
(End);
PF4
(To
enter
comments);
PF7
(Previous
question);
HI
c=
~
The
survey
index
is
displayed
after
pressing
the
PF4
key.
The
index
extends
over
several
frames.
Only
the
first
frame
is
displayed.
Question
9 is
then
selected.
1-
2.
3.
4.
5.
6.
7.
S.
9.
10.
11.
SURVEY INDEX
Which
one
of
these
computers
have
you
used
the
most
during
the
past
three
months?
Where
are
you
located?
For
what
purposes
do
you
use
Yorktown's
VM
and/or
MVS
computing
systems?
For
what
purposes
do
you
use
the
IBM
PC?
To
what
extent
has
your
PC
usage
affected
your
usage
of
Yorktown's
VM
and
MVS
computing
systems?
How
do
you
rate
average
system
response
for
most
of
the
work
that
you
do
on
prime
shift?
How
do
you
rate
average
system
response
for
your
prime
shift
interactive
work
that
requires
moderate
to
large
amounts
of
computer
resources
(e.g.,
system
responses
exceeding
several
seconds)?
How
do
you
rate
turnaround
time
for
your
prime
shift
batch
work?
How
do
you
rate
turnaround
time
for
your
printouts?
How
do
you
rate
system
reliability
(number
of
times
the
system
goes
down
during
your
working
session)?
How
do
you
rate
system
availability
(fraction
of
time
system
is
available
when
you
need
to
work)?
ENTER QUESTION
NUMBER
===>
9
AND
PRESS ENTER
PF3
(Quit
Index);
PF7-
(Previous
frame);
PFS
(Next
frame);
Next
the
user
enters
his
replies
to
question
9
and
then
decides
to
enter
some
comments.
To
do
this
he
presses
the
PF4
key.
QUESTION 9 OF 39
How
do
you
rate
turnaround
time
for
your
printouts?
(Enter:
1
for
very
good,
2
for
good,
3
for
fair,
4
for
poor,
5
for
very
poor.
Leave
blank
for
no
opinion.)
==>
2
Versatec/Benson
(DRAFT)
printers
==>
2
APS-5
photocomposer
==>
1
3SOO
==>
1
Self-service
printers
at
Yorktown
==>
2 6670
printers
at
Yorktown
==>
3
Confidential
output
at
Yorktown
==>
At
Eastview
==>
At
other
IBM
locations
over
network
==>
IBM
PC
PF2
(For
question
index)
;Press-'ENTER'
or
PFS
to
CONTINUE;
PF3
(End);
PF4
(To
enter
comments);
PF7
(Previous
question);
NI
=
-..l
The
intention
to
enter
comments was
signaled
at
question
9
by
pressing
the
PF4
key.
the
user
enters
now
his
comments,
and
files.
He is
then
returned
to
the
survey
questionnaire.
C$9$
XEDIT:
MEMO
A 1 F 80 TRUNC=80 SIZE=3 LlNE=3
COL=l
AL
T=2
00000 * * * TOP
OF
FILE
* * *
00001
COMMENTS
TO
9:
How
do
you
rate
turnaround
time
for
your
printouts?
00002
00003
Printouts
are
sometimes misplaced
by
the
operator
I
...
+
••••
1
••••
+
••••
2
••••
+
••••
3
••••
+
••••
4
••••
+
••••
5
•.••
+
••••
S
••••
+
••••
7
•••
00004 * * * END OF
FILE
* * *
===>
file
XEDIT/ykt
1
FILE
After
answering
question
10
the
user
decides
to
end
the
survey
by
pressing
the
PF3
key.
QUESTION 10 OF 39
How
do
you
rate
system
reliability
(number
of
times
the
system
goes
down
during
your
working
session)?
(Enter:
1
for
very
good,
2
for
good,
3
for
fair,
4
for
poor,S
for
very
poor.
Leave
blank
for
no
opinion.)
==>
2
VM-V
==>
2
VM-T
==>
2
VM-X
==>
1
MVS/TSO
==>
2
Network
-among
Yorktown
systems
==>
2
Network
-
between
Yorktown
and
other
IBM
locations
==>
3
MSS
==>
1 IBM
PC
PF2
(For
question
index);
Press
'ENTER'
or
PF8
to
CONTINUE;
PF3
(End);
PF4
(To
enter
comments);
PF7
(Previous
question);
N
CI)
00
The
user
is
presented
with
a
multiple
choice
menu.
He
then
chooses
option
1.
Your
replies
to
this
survey
are
stored
only
on
your
disk.
What would
you
like
to
do
with
them?
1.
Transmit
them
(for
tabulation
with
other
users'
replies)
and
erase
them
from my
disk.
2.
Do
not
transmit
them,
but
keep
these
replies
on my
disk
so
that
I
may
continue
the
survey
at
some
other
time.
3.
Do
not
transmit
my
replies
and
erase
them
from my
disk.
4.
Do
not
transmit
my
replies.
Let
me
resume
taking
the
survey
at
the
point
where
I
left.
5.
Do
not
transmit
my
replies.
Let
me
review
the
survey
starting
at
question
1.
ENTER SELECTION
NUMBER
===>
AND
PRESS ENTER
3.91
Format
of
the
survey
replies
in
the
Collector.
The
following
data,
transmitted
to
the
receiving
virtual
machine
for
tabu-
lation,
corresponds
to
the
example
in
the
demonstration
in
section
3.9.
The
data
has
undergone
some
preliminary
processing
and
is
ready
for
postprocessing.
YKTVMT
TOLABA
TUESDAY
07/26/83
13:19:48 13:22:33
ANSWERS
TO
QUESTION
1
X
ANSWERS
TO
QUESTION
2
X
COMMENTS
TO
QUESTION
3
COMMENTS
TO
3:
For
what
purposes
do you
use
Yorktown's
VM
and/or
MVS
computing
systems?
ANSWERS
TO
QUESTION
9
2 2 1
123
COMMENTS
TO
QUESTION
9
COMMENTS
TO
9:
How
do you
rate
turnaround
time
for
your
printouts?
Printouts
are
sometimes
misplaced
by
the
operator
ANSWERS
TO
QUESTION
10
222
1 2
231

Navigation menu