Title: Learning and Technology Adoption: A Structural and Behavioral Analysis
1Learning and Technology AdoptionA Structural
and Behavioral Analysis
- Brent Zenobia Department of Engineering and
Technology Management Portland State
Universitybcapps_at_hevanet.comCharles Weber - Department of Engineering and Technology
Management Portland State Universitywebercm_at_gmail
.com
2Abstract
- What is the relationship between learning and
technology adoption? A case study of alternative
transit mode choice suggests that adoption is a
family of three learning processes which includes
technology selection, evaluation, and
maintenance. Structural and behavioral analysis
is applied to uncover the properties, dimensions,
and dynamic behavior of these processes.
Technology evaluation and selection are
cognitively distinct mental processes with
different information requirements. -
3Adoption as a Learning Process
- Adoption is learning about tools
- Tool use sets us apart from most other primates
- Baboons live in social groups but they dont
use tools,have language, or possess a theory of
mind - Social life drove the evolution of the primate
brain toa certain point, then tool use took over
and spurredthe development of language. This
rewired the humanbrain in important ways (Cheney
Seyfarth, 2007) - Thus, learning about tools is even more
fundamental to human cognition than language. - Adoption is also useful as a substantive domain
to study learning - Theories of learning that are difficult to test
in organizational settingsmay reveal themselves
in high-involvement adoption situations - Such studies will have the character of basic
research
4The Current State of Adoption Theory
Theoretical progress has stagnated the growth
of appropriate theory is at an apparent
standstill. (Katz, 1999, pg. 145) a kind of
sameness...stereotyped approaches. (Rogers,
2003, pg. 40) A reified concept without
behavioral referents (Eveland, 1979, pg. 5)
5The Rogers Adoption Model
OUTDATED
and yet, Rogers himself said this model is wrong.
The supporting evidence is thin, the model is 50
years old, and the stages it predicts often occur
out of order or else are skipped entirely. Rogers
pleaded for decades for someone to apply
qualitative process research methods (as opposed
to surveys) to construct a better model of
technology adoption. No causal, empirically
derived theory of adoption yet exists.
- By far the most widely cited adoption model
Source Rogers (2003)
6Research Objective
To construct a theory of thetechnology adoption
process that is suitable for implementation in
simulated consumer agents.
- Why simulated consumer agents?
- Because agent-based social simulation is the
keyto reviving theoretical progress in adoption
research - Forces parsimony and conceptual clarity
- Permits controlled experimentation
- Enables falsification of hypotheses elimination
of weak theory
Source Goldspink (2002) Andrews, Baptista and
Patton (2005)
7Research Setting
The structural and behavioral aspects of
individual adoption are more easily brought out
into the openin high-involvement decision
settings like commuting
8Process has Structure and Behavior
- Structure key concepts static relationships
- Requires sampling a wide variety of outcomes
- Therefore, outcomes must be known in advance
- Retrospective data is needed
- Behavior patterns in the sequence of events
- Requires watching the process as it unfolds
- Therefore, outcomes cannot be known in advance
- Longitudinal data is needed
Two-stage complementary assistance design Case
1 STRUCTURE behavior Case 2 structure
BEHAVIOR
Sources Morse (1991) Morgan (1998) Van de Ven
and Huber (1990) Leonard-Barton (1990)
9Two-Case Design
- Passport Plus Transit Pass
- Emphasis on structural analysis
- (Strauss and Corbin, 1998)
- Retrospective data collection
- 14 interviews other evidence
- Novice Winter Bike Commuters
- Emphasis on behavioral analysis
- (Miles and Huberman, 1994)
- Longitudinal data collection
- 28 interviews other evidence
Outputs Relationship diagrams decision
diagrams sequence diagrams UML model
10Structural Findings
Beliefs,judgements about cause and effect
A map of how technologiesare linkedto
motivesvia beliefs
Momors,MOtive-to-MOtive Relations
Temors,TEchnology-to-MOtive Relations
influences
Motives,representations of inner mental states
Technologies,tools pertaining to motives
selection
Desires,pursued fortheir own sake
Needs,pursued for the sakeof something else
11Structural Findings
A map of how technologiesare linkedto
motivesvia beliefs
but HOW?
influences
Technologies,tools pertaining to motives
selection
Behavioral analysis is needed!
12Behavioral Findings I
- Adoption is a set of three cognitive
processeswith distinct information requirements. - Selection ? Which technology best suits my
current needs? - ? Uses beliefs
- ? Simple behavior runs until it terminates
- Evaluation ? How well does this technology suit
my needs? - ? Changes beliefs
- ? Event-driven behavior stimulus/response
- Maintenance ? Is this technology ready to use?
- ? Factual
- ? Event-driven behavior stimulus/response
13Behavioral Findings IIThe Selection Process (n
75 episodes)
CantQuit
NoOptions
SensitiveMotive
Quit
ReframeSituationalNeeds
MultipleOptions
MakeTradeoff
Closure
OneOption
Committo Option
TakeAction
FrameSituationalNeeds
PromptingMotive
ActionRequired
Procrastinate
RecallTechnologyOption
AssessOptionalNeeds
ScreenOption
StatusQuo
Framing Stage
Screening Stage
Choice Stage
14Behavioral Findings IIIThe Evaluation Process (n
151 episodes)
- Adoption criteria
- Relevance
- Familiarity
- Value
15Behavioral Findings IVThe Maintenance Process (n
57 episodes)
- Gain access buy, subscribe, renew, borrow.
- Lose access sell, unsubscribe, expire, discard.
- Gain operation repair, resupply.
- Lose operation break, consume.
- Adoption criteria
- Opportunity
- Accessibility
- Operability
16Behavioral Findings V
- Operational linkage between selection
evaluation occurs through the mirror mechanisms
of - Consolidation of needs into an habitual set of
options - and
- Differentation of needs into a sensitized set of
options
17The Seven Criteria for Adoption
- Timing Is my need immediate or in the future?
- Relevance Is this technology relevant to my need?
- Familiarity Have I used this technology before?
- Value Is it positive, negative, mixed, or
unknown? - Opportunity Can I get this technology?
- Accessibility Do I already have this technology?
- Operability Does this technology function?
During future research these criteria will
provide the basisfor a survey instrument to
measure the stage of adoption
18Summary
- Adoption is a set of three cognitive processes
having distinct information requirements - Selection
- Evaluation
- Maintenance
19Main Contributions
- Empirical findings
- Research has established an empirical link
between technology adoption and basic cognitive
processes. - Selection and evaluation are cognitively distinct
mental processes with different information
requirements. - The first empirically derived theory of adoption
- An explanatory theory more than a model
- In the Whitehead tradition
- Explicitly decision-theoretic
Sources Whitehead (1929)
20Questions?
21Bibliography
- Andrews, C. J., Baptista, A. I., Patton, S.
(2005). Grounded theory and multi-agent
simulation for a small firm. In T. Terano H.
Kita T. Kaneda K. Arai H. Deguchi (Eds.),
Agent-Based Simulation From Modeling
Methodologies to Real-World Applications. Tokyo
Springer-Verlag. - Cheney, D. L., Seyfarth, R., M. (2007). Baboon
Metaphysics The Evolution of a Social Mind.
Chicago University of Chicago Press. - Eveland, J. D. (1979). Issues in using the
concept of 'adoption of innovations'. Journal of
Technology Transfer, 4(1), 1-13. - Goldspink, C. (2002). Methodological implications
of complex systems approaches to sociality
Simulation as a foundation for knowledge. Journal
of Artificial Societies and Social Simulation,
5(1). - Katz, E. (1999). Theorizing diffusion Tarde and
Sorokin revisited. Annals of the American Academy
of Political and Social Science, 566, 144-155. - Miles, M. B., Huberman, A. M. (1994).
Qualitative Data Analysis. Newbury Park, CA.
Sage. - Rogers, E. M. (2003). Diffusion of Innovations
(5th ed.). New York Free Press. - Strauss, A. L., Corbin, J. (1998). Basics of
Qualitative Research Techniques and Procedures
for Developing Grounded Theory (2nd ed.).
Thousand Oaks, CA. Sage. - Whitehead, A. N. (1929/1969). Process and
Reality An Essay in Cosmology (paperback ed.).
New York Free Press. - Photo credits Bike Portland, Flexcar, Trimet,
Toyota
22Behavioral Modeling
23Composite Sequence Analysis
- A method for identifying typical process
sequencesand families of sequences (Miles and
Huberman, 1994) - Provided evidence for the agents activity
diagrams and statecharts - What follows is a step-by-step illustration of
behavioral analysis - Interview?Transcript
- Transcript?Chronology
- Chronology?Decision Diagram (Langley, 1999)
- Decision Diagram?Sequence Coding
- Sequence Coding?Dataset
- Dataset?Sequence Diagram
- Sequence Diagram?UML Model
24Interview?Transcript (optional)
Transcript
Paragraph
032
- B How long have you been using TriMet?
- 1 Oh, wellI moved back to Oregon after a
number of years away, a long time, almost 20
years away, in 1998. Ive been using TriMet since
that time. I grew up with itI think it was
called Rose City Transit when I was small. My mom
took me everywhere in Portland on Rose City
Transit. I grew up with riding the bus. Shes
from Germany, they always had decent mass
transit. She didnt know how to drive, so we were
always on the bus somewhere. - time mark 1100
- B If I were to ask you about some of the
feelings or meanings that you have personally
constructed around the bus or mass transit, could
you describe some of those to me? - 1 I think its often very convenient. Its
often slower than if you park your car, but on
the other end you dont have to pay, you dont
have to look for a parking place or pay for it,
so to me it comes out even - even though I can
get downtown quicker from my house in my car than
I can on TriMet. But then I might use my time
hunting for a parking place. I usually enjoy it,
its sort of like a little cultural experience.
Sometimes Im too tired to have the cultural
experience, because you just never know what
youre going to get on the bus. Sometimes its
like walking into a play.
033
034
035
036
25Transcript?Chronology
(or voicerecording index)
Timeline Issue Category Paragraph
Childhood Rides Rose City Transit bus with mother Personal 33
1998 Moves back to Portland after long absence lives in area with limited walking options Functional-Social-Personal 33, 76, 91
1998 Becomes caregiver for elderly mother Social-Personal 76
1998 Unable to find acceptable bus options for visiting mother Functional-Social 78, 80
2001? Working full time downtown Functional 63
Sept 2001 Hired as adjunct faculty for PSU and PCC Functional 13
2001-2006 Busy schedule pressed for time Functional 82, 84, 89
2001-2006 Dissatisfied with cost of campus parking Functional 9, 17, 159
2001-2006 Parking hassles at PSU PCC Functional 9, 17, 29, 159, 165
2001-2003 Bus use increasing buys monthly pass Functional 21, 63, 65-66
2001-2006 Uses personal auto to drive to transit center (w/ trip chaining) Functional 21, 27
2001-2006 Combines Trimet with PCC shuttle bus Functional 31
2001-2006 General satisfaction with Trimet, except for cost, availability, and accessibility issues Functional 36, 221, 227-229, 231, 235
2001-2006 Enjoys using bus ride for teaching prep, rest periods/naps Functional-Personal 9, 13, 17, 217-219
2001-2006 Tired of driving, but thankful to have a car Functional 97, 249
26Chronology?Decision Diagram
- Commuting between home, campus, downtown, mother
- Increasing bus use buys monthly pass
- Utilizes shuttle bus to commute between campuses
- Tries to reduce use of personal vehicle (trip
chaining) - Occasional walking
Hired as adjunct
Working as adjunct faculty at PSU PCC
Parking hassles at PSU PCC
(--)
(--)
(--)
Settles in area with few walking options
Campus parking is expensive
Pressed for time
FUNCTIONAL
Uses bus time for teaching prep
Unable to find acceptable bus options for
visiting mother
Rides Rose City transit with mother
Becomes caregiver for elderly mother
SOCIAL
Caregiver for elderly mother
Moves back to Portland after long absence
Uses bus time to rest and nap
PERSONAL
(-)
(-)
()
()
Commute tradeoff
Childhood
Sept 2001
2003
1998
27Decision Diagram?Sequence Coding
- Commuting between home, campus, downtown, mother
- Increasing bus use buys monthly pass
- Utilizes shuttle bus to commute between campuses
- Tries to reduce use of personal vehicle (trip
chaining) - Occasional walking
Hired as adjunct
Sequence codesare derived fromuse case analysis
Working as adjunct faculty at PSU PCC
Parking hassles at PSU PCC
(--)
(--)
(--)
Settles in area with few walking options
Campus parking is expensive
Pressed for time
FUNCTIONAL
Uses bus time for teaching prep
Unable to find acceptable bus options for
visiting mother
EVTS FSN AR RCL (bus) AON SCO RCL (car) AON SCO
CLO MO MT CMT ACT (bus)
Rides Rose City transit with mother
Becomes caregiver for elderly mother
SOCIAL
Caregiver for elderly mother
Moves back to Portland after long absence
Uses bus time to rest and nap
PERSONAL
(-)
(-)
()
()
Commute tradeoff
Childhood
Sept 2001
2003
1998
28Sequence Coding?Data Set
- 32 interviews yielded 283 sequences
- 75 Selection sequences
- 151 Evaluation sequences
- 57 Maintenance sequences
EVTS FSN AR RCL (bus) AON SCO RCL (car) AON SCO
CLO MO MT CMT ACT (bus)
29Dataset?Sequence Diagram
CantQuit
NoOptions
SensitiveMotive
Quit
ClosureExceeded
MultipleOptions
MakeTradeoff
ReframeSituationalNeeds
OneOption
Committo Option
TakeAction
FrameSituationalNeeds
PromptingMotive
ActionRequired
Procrastinate
RecallTechnologyOption
AssessOptionalNeeds
ScreenOption
StatusQuo
Framing Stage
Screening Stage
Choice Stage
30CantQuit
NoOptions
SensitiveMotive
Quit
ClosureExceeded
MultipleOptions
MakeTradeoff
ReframeSituationalNeeds
OneOption
Committo Option
TakeAction
FrameSituationalNeeds
PromptingMotive
ActionRequired
Procrastinate
RecallTechnologyOption
AssessOptionalNeeds
ScreenOption
StatusQuo
31PromptingMotive
FrameSituationalNeeds
ActionRequired
AssessOptionalNeeds
RecallTechnologyOption
ClosureExceeded
ScreenOption
ReframeSituationalNeeds
NoOptions
MultipleOptions
OneOption
CantQuit
MakeTradeoff
SensitiveMotive
Committo Option
Quit
Procrastinate
TakeAction
StatusQuo
32Sequence Diagram? Activity Diagram
Select Technology ltltpreconditiongtgt usage
situation Prompting Motive Motive ltltpostconditio
ngtgt technology chosen
Info
Help
PromptingMotive
AcquireHelp
AcquireInfo
RecallSituational Need
Situational Context
Check for Closure
else
closure achieved
Check ifAction Required
else
From interviewto UML diagraman unbrokentrail
of evidence
action required
status quo
Reject Technology
Recall Technology
Technology
Technology Set
RecallOptional Need
Optional Need
Check for Closure
else
Check forViolation
closure achieved
technology set lt 1
technology set gt 1
Optional Need Set
else
else
limit exceeded
SensitizeMotive
MakeTradeoff
Check for Closure
else
ReframeContext
closure achieved
ChooseTechnology
SituationalContext
Check toBegin Action
limit exceeded
else
use
procrastinate
Situational Need
Optional Need
Technology
33Bibliography
- Langley, A. (1999). Strategies for theorizing
from process data. The Academy of Management
Review, 24(4), 691-710. - Miles, M. B., Huberman, A. M. (1994).
Qualitative Data Analysis. Newbury Park, CA.
Sage.