An Analysis of Email Response Policies under Different Arrival Patterns By Ashish Gupta Ramesh...
-
Upload
roland-farmer -
Category
Documents
-
view
217 -
download
3
Transcript of An Analysis of Email Response Policies under Different Arrival Patterns By Ashish Gupta Ramesh...
An Analysis of Email Response Policies under An Analysis of Email Response Policies under Different Arrival PatternsDifferent Arrival Patterns
By By
Ashish GuptaAshish Gupta Doctoral Student, Department of Management Science & Information Systems,
Oklahoma State University, Stillwater.
Ramesh Sharda Ramesh Sharda Regents Professor of Management Science & Information Systems,
Director, Institute for Research in Information Systems, Oklahoma State University, Stillwater.
01/06/0501/06/05 ICS-2005ICS-2005 22
Objective of the studyObjective of the study
To improve individual knowledge worker To improve individual knowledge worker performance by identifying policies that will :-performance by identifying policies that will :-
To model email work environment by considering To model email work environment by considering various email characteristics.various email characteristics.
Improve response time of emails and primary task Improve response time of emails and primary task completion timecompletion time
Reduce number of interruptions Reduce number of interruptions
Validate the results of prior research.Validate the results of prior research.
01/06/0501/06/05 ICS-2005ICS-2005 33
Problem significanceProblem significance
20042004 AMA Research on w AMA Research on workplace E-Mail & Productivityorkplace E-Mail & Productivity On a typical workday, time is spent on e-mail is ?????On a typical workday, time is spent on e-mail is ?????
0–59 minutes 77.9% 0–59 minutes 77.9% 90 minutes–2 hours 18%90 minutes–2 hours 18% 2–3 hours 2%2–3 hours 2% 3–4 hours 2.5%3–4 hours 2.5%
Osterman Research-Osterman Research- How often do you How often do you
check your E-mail for new messages check your E-mail for new messages
when at work?when at work?
01/06/0501/06/05 ICS-2005ICS-2005 44
Problem significanceProblem significance
E-Policy Institute (2004)E-Policy Institute (2004) Annual Email growth rate= 66 %Annual Email growth rate= 66 %
Corporate ResearchCorporate Research IBM, Microsoft, Xerox, Ferris, Radicati, etc.IBM, Microsoft, Xerox, Ferris, Radicati, etc.
Need for more research in MS/IS thatNeed for more research in MS/IS that Looks at the problem of information overload and Looks at the problem of information overload and
interruptions simultaneously.interruptions simultaneously.
01/06/0501/06/05 ICS-2005ICS-2005 55
Extant ResearchExtant Research
Overload due to emails-Overload due to emails- First reportedFirst reported byby Peter Denning Peter Denning (1982). (1982).
Most recently reported byMost recently reported by Ron Weber (MISQ, Ron Weber (MISQ, Editor-in-Chief 2004)Editor-in-Chief 2004)
Interruptions due to emails-Interruptions due to emails-Reported by someReported by some- Speier,et.al.1999, Jackson, et.al., - Speier,et.al.1999, Jackson, et.al., 2003, 2002, 2001), Venolia et.al. (2003) 2003, 2002, 2001), Venolia et.al. (2003)
01/06/0501/06/05 ICS-2005ICS-2005 66
Extant ResearchExtant Research
““The nature of managerial work”, Mintzberg (1976)The nature of managerial work”, Mintzberg (1976) ““Managerial communication pattern”, Ray Panko (1992)Managerial communication pattern”, Ray Panko (1992) ““Email as a medium of managerial choice”, M. Markus Email as a medium of managerial choice”, M. Markus
(1994)(1994) ““You have got (Lots and Lots) of mail” in “The Attention You have got (Lots and Lots) of mail” in “The Attention
Economy” by Davenport (2001)Economy” by Davenport (2001) ““The Time Famine: Towards a Sociology of Work Time”, The Time Famine: Towards a Sociology of Work Time”,
Leslie Perlow (1999)Leslie Perlow (1999)
01/06/0501/06/05 ICS-2005ICS-2005 77
Phenomenon of InterruptionPhenomenon of Interruption
Interrupt arrives
IL + Interrupt processing
Interrupt departs
Recall time- RLPre-processing Post-processing
Interruptions-Interruptions- According to According to distraction theorydistraction theory, interruption is , interruption is “an “an externally generated, randomly occurring, discrete eventexternally generated, randomly occurring, discrete event that breaks continuity of cognitive focus on a primary task“ that breaks continuity of cognitive focus on a primary task“ (Corragio, 1990; Tétard F. 2000).(Corragio, 1990; Tétard F. 2000).
01/06/0501/06/05 ICS-2005ICS-2005 88
Previous Research Model Previous Research Model
Performance Measures1. % Increase in utilization2. Number of interruptions per task. 3. Primary task completion time4. Email response time.
Task complexity(Pure simple) vs. (more-simple & less-complex) vs. (equal-simple & complex) vs. (less-simple & more-complex) vs. (pure complex)
Workload LevelLow vs. Medium vs. High
Email PolicyFlow vs.
Scheduled vs.Triage
Only “high” dependency on email communication (3 hrs) with exponential email arrivals was studied
01/06/0501/06/05 ICS-2005ICS-2005 99
Detailed Research model
Performance variables
(a) % increase in Utilization(b) Time spent due to interruptions(c) Average response time of emails(d) Average completion time of primary task.(e) Total no. of interruptions/ day
Email processing strategies(C1, C2, C4, C8, C)
Email characteristicsProcessing Time*
(Large, Small)
Arrival Rate(V. Low, Low, High, V. High)
Dependency on email communication
(Very Low, Low, High, Very High)
Email arrival pattern(Expo, NSPS)
Work Environment
* Processing time is based on email category
01/06/0501/06/05 ICS-2005ICS-2005 1010
Email typesEmail types
Emails differentiated on the basis of its ‘content’ Emails differentiated on the basis of its ‘content’ or the ‘action required by the user’or the ‘action required by the user’Notation Email type Discrete arrival
percentage
1 Priority email 5%
2 Spam 5%
3 Informative email 20%
4 Email with non-diminishingservice time
55%
5 Email with diminishingservice time
15%
01/06/0501/06/05 ICS-2005ICS-2005 1111
Email PoliciesEmail Policies
Dependency on Email Communication
Policy type Very Low(1 hr)
Low (2 hrs)
High (3 hrs)
Very High (4 hrs)
Notation # of Emailhour- slots
Triage 8am-9am 8am-10am 8am-11am 8am -12 noon C1 1
Schedule 8am-8:30am4:30pm- 5pm
8am-9am4pm-5pm
8am-9:30am 3:30 am to 5:00
pm
8am-10am3pm- 5pm
C2 2
Schedule 8am-8:15am,11am-11:15am1pm-1:15pm4:45pm- 5pm
8am-8:30am,11am-11;30am1pm-1:30pm4:30pm- 5pm
8am-8:45 am, 11am-11:45am,1 pm - 1:45 pm, 4:15 pm - 5:00
pm
8am-9am11am - 121pm- 2pm4pm- 5pm
C4 4
Schedule 8am-8:08am9- 9:08amand so on
8-8:15am9-9:15am
10-10:15amand so on
8-8:23am9-9:23am
10-10:23am and so on
8- 8:30am9- 9:30pm
10- 10:30pmand so on
C8 8
Flow Processed as soon
as emails arrive
Processed as soon
as emails arrive
Processed as soon
as emails arrive
Processed as soon
as emails arrive
C NotApplicable
01/06/0501/06/05 ICS-2005ICS-2005 1212
MethodologyMethodology
Discrete event simulation using Arena 8.01Discrete event simulation using Arena 8.01
Model Run length= Model Run length= 500500 days days
Model Warm-up time= Model Warm-up time= 5050 days days
No. of replications of each model= No. of replications of each model= 2020
1616 scenarios evaluated for scenarios evaluated for 55 different policies. different policies.
Thus, Total number of simulations models= Thus, Total number of simulations models= 16 x 5= 8016 x 5= 80
Total number of data points generated
= 80 x 20 = = 80 x 20 = 16001600
01/06/0501/06/05 ICS-2005ICS-2005 1313
ScenariosScenarios
Scenarios Email (E) dependency E Arrival pattern E processing time
1 Very low Time stationary Expo Small
2 Very low Time stationary Expo Large
3 Very low Non-Stationary Expo Small
4 Very low Non-Stationary Expo Large
5 Low Time stationary Expo Small
6 Low Time stationary Expo Large
7 Low Non-Stationary Expo Small
8 Low Non-Stationary Expo Large
9 High Time stationary Expo Small
10 High Time stationary Expo Large
11 High Non-Stationary Expo Small
12 High Non-Stationary Expo Large
13 Very High Time stationary Expo Small
14 Very High Time stationary Expo Large
15 Very High Non-Stationary Expo Small
16 Very High Non-Stationary Expo Large
01/06/0501/06/05 ICS-2005ICS-2005 1414
ParametersParameters
S #
Type 4 email (E)
Processing
time (PT)
Type 5E PT(min)
Total Email PT per day
Avg. EmailArrival Rate
Primary Task
(P) ArrivalRate /day E Util P Util
Min(E+P)Util
1 5 5 1 12 62 0.125 0.775 0.9
2 15 15 1 5 62 0.125 0.775 0.9
3 5 5 1 12 62 0.125 0.775 0.9
4 15 15 1 5 62 0.125 0.775 0.9
5 5 5 2 24 52 0.25 0.65 0.9
6 15 15 2 10 52 0.25 0.65 0.9
7 5 5 2 24 52 0.25 0.65 0.9
8 15 15 2 10 52 0.25 0.65 0.9
9 5 5 3 36 42 0.375 0.525 0.9
10 15 15 3 15 42 0.375 0.525 0.9
11 5 5 3 36 42 0.375 0.525 0.9
12 15 15 3 15 42 0.375 0.525 0.9
13 5 5 4 48 32 0.5 0.4 0.9
14 15 15 4 20 32 0.5 0.4 0.9
15 5 5 4 48 32 0.5 0.4 0.9
16 15 15 4 20 32 0.5 0.4 0.9
Processing time of(a) Type 1 email- Expo(10 min)
(b) Type 2 email- Expo (0.5 min)
(c) Type 3 email- Expo (5 min)
(d) Primary task- Expo(6 min)
01/06/0501/06/05 ICS-2005ICS-2005 1515
Bird’s Eye view of Entire Bird’s Eye view of Entire model built using Arenamodel built using Arena
f or t ypes 2 3 4em ail s t r at egy
im plem ent ing
k W
Pr eem pt1 2 3 4
Pr o c e s s em a il
t im ee m ail e f f ec t iv e
nonof f ice hour shold dur ing
Emai l hours Non Emai l hours Knowledge Worker (Busy- green, Idle- whi te, Inactive- hatched)
t y p eAs s ign e m a ile m ails
E ls e
e m a il t y p e = = 1e m a il t y p e < = 4
t y p e o f em a il
f or t ype 5em ail s t r at egy
im plem ent ing
k W
Pr eem pt e m ail 5 in 1T r u e
F a ls e
t y p e 5d e c id in g s er v ic e t im e o f
s p e nte m ail t y p e 5 t im e
E ls e
e m a il t y p e = = 1e m a il t y p e = = 2e m a il t y p e = = 3
s p lit e m a il t y p e
s p e nte m ail t y p e 2 t im e
s p e nte m ail t y p e 3 t im e
s p e nte m ail t y p e 4 t im e
s p e nte m ail t y p e 1 t im e
s im plep la n n in g s im p le p la n n in g o r n o t s im p le
I n t er r u p t e d d u r ingT r u e
F a ls e
e x e c u t io n s im p le
s im pleRL p la n n in g
T r u e
F a ls e
e x e c u t io n o r n o t s im p leI n t er r u p t e d d u r ing
s im plee v a lua t io n
T r u e
F a ls e
e v a lua t io n o r n o t s im p leI n t er r u p t e d d u r ing
Dis p os e s im p le
s im pleRL ex e c u t in g
s im pleRl e v a lu a t ing
Hold s im ple
s im pleI L p la n n in g
s im pleI L ex e c u t in g
s im pleI L ev a lu a t in g
As s ign 1 9 t im et a s k c o m p le t io na v e r a g e s im p le
s e t p la n n ing RT0
s e t e x e c ut io n RT0RT0
s e t e v a lu at io n
As s ign 2 3
As s ign 2 4As s ign 2 5
k W
Pr eem pte m ail 5 in 2
wo r k d a yin t e r r u p t ion s p e r
Av e r a g e
in t e r r u p t ion s s pn o o f e x e s
in t e r r u p t ion sn o o f s e v a l
in t e r r u p t ion sn o o f
r e s po n s e t im ea v e r a g e em a il Dis p os e e m a il
r e s po n s e t im em a x im u m e m a il
r e s po n s e t im em in im u m e m a il
r e s po s e t im ed e v ia t io n in e m a il
s t a nd a r d
00:00:00
0
Jan u ary 1 , 2 0 0 5
0
0
0
0
0 0
0
0 0
0
0
0 0
0
00
0 0
0
0 0
0
0
Zoom in follows….
Arena Email flow SnapshotArena Email flow Snapshot
1
2
for types 2 3 4email strategyimplementing
kW
Preempt1234
Process email
Else
email type==1email type==2email type==3
split email ty pe
0
1
2
Preempts the KW when an email of type 1 arrives during email hrs . Stores remaining processing time in an attribute ‘RT’
timeemail effective
nonoffice hourshold duringtype
Assign emailemails
Else
email type==1email type<=4
type of email
0
3
for type 5email strategyimplementing
kW
Preempt email 5 in 1True
False
type 5deciding service time of
spentemail type 5 time
kW
Preemptemail 5 in 2
0
0
0
0
3
Releases emails of type 2,3,4 on the basis of policy
Emails created based on different schedules that determines whether it is Expo or Non-
Stationary Expo and at what rate
To record output statistics of each
email type separately
Checks if email has been in system for > or < than 24 hrs
01/06/0501/06/05 ICS-2005ICS-2005 1717
Arena Primary Task SnapshotArena Primary Task Snapshot
simpleplanning s imple planning or not s imple
Interrupted duringTrue
False
simpleRL planning
Hold simple
simpleIL planning
Assign 19
set planning RT0Assign 23 interruptions sp
no of
0 0
0
0
0 0
Attribute RT is reset to 0 to erase the memory. This makes the attribute RT reusable for recording remaining time
interrupted primary task in future.
Checks to see if RT>0. If yes, RL and IL are added If no, Primary task is sent next processing stage
01/06/0501/06/05 ICS-2005ICS-2005 1919
Model LogicModel LogicNew email arrival Ei occurs at time T0, for all i ={n : n = 1 . . 5}If i = 1,
Step1. Email released at T0. Step2. If STATE (KW) == IDLE & E1.WIP=0
KW seized; Than, Set RT = Ta = 0;
IL = 0, RL = 0;Process E1;
Release KW;If STATE (KW) == BUSY & E1.WIP=0 Seize KW;
Than, Set RT = Ta; Record IL = Tria (a, b, c), Tb;
Process E1;Release KW;
Calculate; χ = Tb /( Ta + Tb) for all 0 ≤ χ ≤ 1Calculate;RL = {RT * [χ * *( K-1)] * [ (1- χ)* * ( L-1 ) } / Beta (K,L)
01/06/0501/06/05 ICS-2005ICS-2005 2020
Model LogicModel Logic For K = 2, L = 1;For K = 2, L = 1; Calculate;Calculate;
T1 = IL + Tb + RL; T1 = IL + Tb + RL; Seize KW for time T1;Seize KW for time T1;
Process PiProcess PiSet RT=0;Set RT=0;
Release KW;Release KW;If If ii = = 2 || 3 || 4 || 5,2 || 3 || 4 || 5,
Step.3 Step.3 Release Release Ei, Ei, ifif{(STATE(dummy) == IDLE_RES && {(STATE(dummy) == IDLE_RES && Process email 1234.WIP == 0 && Process email 1234.WIP == 0 && email 5 in 1.WIP == 0 && email 5 in 1.WIP == 0 && email 5 in 2.WIP == 0 ) || email 5 in 2.WIP == 0 ) || ( STATE(anti dummy) == IDLE_RES && ( STATE(anti dummy) == IDLE_RES && Primary.WIP == 0 && Primary.WIP == 0 && NQ(Hold primary.Queue) == 0 && NQ(Hold primary.Queue) == 0 && IL Primary .WIP=0 && IL Primary .WIP=0 && RL primary.WIP == 0 ) } =RL primary.WIP == 0 ) } = TRUETRUE
Else Hold;Else Hold;
01/06/0501/06/05 ICS-2005ICS-2005 2121
Model logic- commentsModel logic- commentsIf New arrival = PnIf New arrival = Pn
Step4.Step4. Release if, Release if,STATE(kW) == IDLE_RES;STATE(kW) == IDLE_RES;
Else Hold;Else Hold;
//*****//*****Tb- Value added time spent on the task Tb- Value added time spent on the task BBeforeefore interruption interruptionTa- Value added time spent on the task Ta- Value added time spent on the task AAfterfter interruption interruption χ - Fraction of task completed before interruption occurredχ - Fraction of task completed before interruption occurredIL – Interruption LagIL – Interruption LagRL – Resumption LagRL – Resumption LagPi – interrupted primary taskPi – interrupted primary taskDummy resource- implements email hoursDummy resource- implements email hoursAnti-dummy resource – implements non- email hoursAnti-dummy resource – implements non- email hours *****//*****//
Stop;Stop;Stop;
ResultsResults
(a) Percent Increase in Utilization
% increase in utilization (base value=0.9)
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
9
8
7
6
5
4
3
2
1
Email dependency
high
low
very high
very low
% increase in utilization (base value=0.9)
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
8
7
6
5
4
3
2
1
Email Arriv. Pattern
Expo
Non Stationary Expo
% increase in utilization (base value=0.9)
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
9
8
7
6
5
4
3
2
EPT
large
Small
ResultsResults
(b) Additional Time (min) spent per day due to interruptions
Additional Time spent / day due to interruptions
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
40
30
20
10
0
Email Dependency
high
low
very high
very low
Additional Time spent / day due to interruptions
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
40
30
20
10
0
Email Arriv. Pattern
Expo
Non-Stationary Expo
Additional Time spent / day due to interruptions
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
40
30
20
10
Email Processing Tim
large
Small
01/06/0501/06/05 ICS-2005ICS-2005 2424
Response time resultsResponse time results
Avg. Email Response Time
= Avg. Email processing time (Value added)
+ Avg. Email wait (Queue) time [fig. c]
Avg. Primary Task (PT) Completion Time [fig. d.3]
= Avg. PT value added processing time
+ Avg. PT non-value added processing time due recalling & switching [fig. d.1]
+ Avg. PT wait (Queue) time [fig. d.2]
ResultsResultsEmail wait time
POLICY
CC8C4C2C1
Est
imat
ed M
argi
nal M
eans
400
300
200
100
0
Email Dependency
high
low
very high
very low
Email wait time
POLICY
CC8C4C2C1
Estim
ate
d M
arg
ina
l M
ea
ns
400
300
200
100
0
Email Pattern
EA
NSEA
Email wait time
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
400
300
200
100
0
EPT
large
Small
(c) Email Wait time i.e. inbox queue and holdup
time
ResultsResultsNon-Value Added time (RL+ IL) spent
per Primary Task (min)
POLICY
CC8C4C2C1
Est
imat
ed M
argi
nal M
eans
1.0
.8
.6
.4
.2
0.0
Email Dependency
high
low
very high
very low
Non-Value Added time (RL+ IL) spent
per Primary Task (min)
POLICY
CC8C4C2C1
Estim
ate
d M
arg
ina
l M
ea
ns
.8
.7
.6
.5
.4
.3
.2
.1
Email Arriv. Pattern
Expo
Non-Stationary Expo
Non-Value Added time (RL+ IL) spent
per Primary Task (min)
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
1.0
.8
.6
.4
.2
0.0
Email Processing tim
large
Small
(d.1) Avg. Additional time spent (wasted) in recalling and
switching for processing one primary task
ResultsResults
(d.2) Average Primary Task Wait Time
Avg Primary Task Wait time (min)
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
2000
1000
0
Email Dependency
high
low
very high
very low
Avg. Primary Task Wait Time (min)
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
1600
1400
1200
1000
800
600
400
200
0
Email Arriv. Pattern
Expo
Non-Stationary Expo
Avg Primary Task Wait time (min)
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
2000
1000
0
Email Processing Tim
large
Small
ResultsResultsAvg Primary Task Completion time
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
2000
1000
0
Email Dependency
high
low
very high
very low
Avg Primary Task Completion time
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
1800
1600
1400
1200
1000
800
600
400
200
0
Email Arriv. Pattern
EA
NSEA
(d.3) Average Primary Task Completion Time
Avg Primary Task Completion time
POLICY
CC8C4C2C1
Est
ima
ted
Ma
rgin
al M
ea
ns
2000
1000
0
Email Processing Tim
large
Small
01/06/0501/06/05 ICS-2005ICS-2005 2929
Optimal Policy ??Optimal Policy ??
Previous research found C4 as the optimal policy (no Previous research found C4 as the optimal policy (no consideration was given to email arrival pattern and consideration was given to email arrival pattern and characteristics).characteristics).
Current Research found under varying email arrival Current Research found under varying email arrival characteristics-characteristics- Optimal policy for primary task completion time - C1 & Optimal policy for primary task completion time - C1 &
C2 closely followed by C4.C2 closely followed by C4. Optimal policy for email response time – C Optimal policy for email response time – C Optimal policy for reducing interruptions- C1& C4 closely Optimal policy for reducing interruptions- C1& C4 closely
followed by C2followed by C2
01/06/0501/06/05 ICS-2005ICS-2005 3030
Limitations of the modelLimitations of the model
Assumptions of the model are its limitationsAssumptions of the model are its limitations Knowledge worker works strictly according from 8 to 12 Knowledge worker works strictly according from 8 to 12
and then from 1 to 5pm. Need for relaxing the work-hrs.and then from 1 to 5pm. Need for relaxing the work-hrs. Knowledge worker is busy only 90% of the time in a given Knowledge worker is busy only 90% of the time in a given
workday.workday. KW is working on interruptible primary task. In reality, not KW is working on interruptible primary task. In reality, not
all primary tasks are interruptible. For e.g. group meetingsall primary tasks are interruptible. For e.g. group meetings Primary task modeled is interruptible only 3 times.Primary task modeled is interruptible only 3 times. Emails are not interrupted.Emails are not interrupted.
01/06/0501/06/05 ICS-2005ICS-2005 3131
Limitations & future Limitations & future researchresearch
Perform the study in field or experimental Perform the study in field or experimental settings.settings.
Modeling utility/ life of an email.Modeling utility/ life of an email. Modeling group knowledge network and at Modeling group knowledge network and at
organizational level. organizational level. Modeling by incorporating more doses of Modeling by incorporating more doses of
reality. Considering other communication media reality. Considering other communication media along with email.along with email.
http://iris.okstate.edu/rems/http://iris.okstate.edu/rems/Suggestions or comments or Questions????Suggestions or comments or Questions????