Beyond Technology: Improving Energy
Efficiency through SocialPsychological Approaches
Chien-fei Chen, Ph.D.
Research Professor & Director of Education and
Diversity
NSF-DOE Engineering Research Center, CURENT
University of Tennessee, Knoxville
Focus of Environmental Sociology
1-2
Research Methodology
• Qualitative vs. Quantitative (focus group, interviews)
• Survey technique asks questions of large numbers of
persons to gain information on attitudes and behaviors
 Reliability and validity:
• Test – retest, internal, inter-rater, …
• Face, convergent & discriminant, predictive…
 Advantages of survey approach:
• Cheap and fast
• Go beyond behavioral level
 Common issues:
• Sample representativeness
• Low response rate or careless answers
• Social desirability
• Basically a correlational study
The Experimental Research Process
Consider and control
confounding variables
© 1999 John Wiley and Sons, Inc.
Summary of main research methods in environmental psychology
SETTING
METHOD
STRENGTHS
WEAKNESSES
USE
Environment
independent
setting
Questionnaire
studies
High external
validity Costeffective
method for
reaching large
populations
No
manipulation of
variables
No causal
inferences
Describing
Artificial setting
Laboratory
experiments
High internal
validity
Control of
variables
Low external
validity
Artificiality
Testing theories
or hypotheses
Identifying
causal
relationships
Simulation
studies
Good balance
between
external/
internal validity
Realistic
visualisation
Requires
advanced skills
and equipment
Often
perceived as
‘fictitious’
Study complex
humanenvironment
dynamics
Visualise and
evaluate future
developments
Case studies
High external
Low internal
Descriptions
Real setting
populations or
practices
Studying
relationships
among
variables
Social Psychological Factors and
Energy Behaviors
1. Why behaviors
matter?
2. Decisionmaking models
3. Factors influencing
public acceptance of
energy-efficiency
measures & renewables
6. Energy efficiency in
office buildings
Content
5. Promoting
Demand Response
4. Interventions
(feedback, norms,
framing, etc.)
6
U.S. Energy Consumption in 2010
Transportation
-other, 16.90%
Industry,
29.80%
Residential &
personal
transport
account for
34%
Residential,
22.40%
Personal
Transport,
11.40%
Commercial,
19.40%
Source: EIA
7
Traditional Economic & Engineering Models
Typical responses to energy crisis:
• Find new energy resources
• Develop technologies – engineers’ job
• Provide financial incentives for people to reduce
consumption, to adopt more efficient technologies
Assumption of rationality
 People are instrumental and self-interested, consistent,
cost-benefit based.
But it is often not true.
8
Creating an Energy Revolution
“A revolution doesn’t happen when society adopts new
tools, it happens when society adopts new behaviors” (Glay
Shirky, Digital Guru).
“Mitigate future climate change will be made by energy
consumers, rather than suppliers… not a straightforward
and easily achievable goal” (EIA, 2009).
9
Your Behaviors Matter
Behavioral Change
Potential Savings:
Various types of
household behaviors
22% reduction in household & personal
transportation energy use over a 5-8 year
period (Laitner & Ehrhardt-Martinez, 2009)
In crisis situations
Community-level electricity savings could
be 25% in 6 weeks & post-crisis savings of
8-10% (Leighty & Meier, 2010)
Feedback programs &
devices
Save electricity 4-12%
(Ehrhardt-Martinez, et al. 2010)
Real-time feedback
with smart programs +
social science insights
Save electricity 20-35%
10
Social Psychological Factors and
Energy Behaviors
1. Why behaviors
matter?
2. Decisionmaking models
3. Factors influencing
public acceptance of
energy-efficiency
measures & renewables
6. Energy efficiency in
office buildings
Content
5. Promoting
Demand Response
4. Interventions
(feedback, norms,
framing, etc.)
11
Assumptions of Human Behaviors
Without Deliberation
• In-output model
• Behaviors controlled by
environment
Not Empathizing:
• Internal judgments
• Cognitive process
• Interpersonal relationships
With Deliberation
• Emphasize human
agency
• Mindful or cognitive
process
Empathizing:
• Attitudes
• Perceptions
• Motivations
(Asch, 1951, Sherif, 1935)
Actors are mindless robots
Actors are mindful
12
Theory of Planned Behavior
(Icek Ajzen, 1991)
13
Norm Activation Model (Schwartz, 1977)
Awareness of
consequences
Ascription of
responsibility
Personal
Norms
Pro-social or
environmental
behaviors
14
Social Psychological Factors and
Energy Behaviors
1. Why behaviors
matter?
2. Decisionmaking models
3. Factors influencing
public acceptance of
energy-efficiency
measures & renewables
6. Energy efficiency in
office buildings
Content
5. Promoting
Demand Response
4. Interventions
(feedback, norms,
framing, etc.)
15
Social-psychological Factors Influencing Energy
Saving at Workplace
• Sampled 584 employees from 9 electricity companies in
Jiangsu Province, China.
• Investigated the relationships among social norms, behavioral
control, attitudes, energy concern influence energy saving
intention and energy behavior at workplace.
• Based on the Theory of Planned Behavior.
Chen, C.F., & Knight, K. (2014). Energy at work: social psychological factors affecting
energy conservation intentions within Chinese electric power companies. Energy
Research & Social Science, 4, 23-31.
16
Social-Psychological Factors Affecting
Electricity Saving Behaviors at Work (China)
Attitudes
.35***
.47***
Energy Concern
.04
.38***
Behavioral
Control
.61***
Intention to Save
.43***
Group Norms
17
Social Psychological Factors Affecting Support
of Renewable Energy (U.S.)
Electricity saving behaviors (-0.22)**
Energy conservation attitudes (0.18)*
Energy concern (0.24)*
Global warming consequences(0.18)*
Economic benefits (-0.10)
Support of
Renewable
Energy
Personal norms (0.34)
Social rewards (-0.08)
Familiarity with renewable(0.05)
Political orientation (0.20) *
* P<0.05; ** p<0.01, green boxes on the left were not significant
18
Social Psychological Factors Affecting Intention to Use
SWH and AFV (China)
* p < 0.05; ** p < 0.01, *** p < 0.001
19
Factors Influencing
Acceptance of Smart Meter
Useful1
Useful2
Useful3
Useful4
Useful5
Useful6
.89
Priv4R
.51
Money2
Money6
Environ1
Environ2
Environ3
Environ4
.91
Priv7R
Price2
.94
Price3
.93
.85
Price
privacy
.52, p < .001
.88
.84
-.14, p < .001
-.093, p = .001
.76
.61
.70
.56
-.059
HabitE
1
Money
Money3
Money4
.92
Usefulness
Useful8
Money1
Priv6R
.74
.89
.87
.80
Priv5R
-.018
HabitC
1
-.007
.091, p = .01
.83
.014
.035
.83
HabitCH
.80
.84
.87
Support
Environ
.73
.68
Habit7
Habit6
.95
.95
Trust1
•
•
Trust
.89
Trust2
Sampled 817 U.S. residents across the U.S.
Based on the Extended Technology Acceptance Theory.
.83
Trust3
.92
Trust4
20
Social Psychological Factors and
Energy Behaviors
1. Why behaviors
matter?
2. Decisionmaking models
3. Factors influencing
public acceptance of
energy-efficiency
measures & renewables
6. Energy efficiency in
office buildings
Content
5. Promoting
Demand Response
4. Interventions
(feedback, norms,
framing, etc.)
21
Intervention/Strategies to Change Behaviors
• Feedback programs (direct vs. indirect)
• Social norms: descriptive and injunctive norms shape
people’s behaviors; developing strategies in a social
context
• Goal setting: define what people are trying to attain and
be able to evaluate their progress
• Message framing: emphasizing a particular aspect of an
object/activity while limiting emphasis on other aspects
• Commitments: help people to sure their actions are
consistent with their ideals
22
Average Household Electricity Savings of
Historical Program by Feedback Type
Potential
Resource
Savings:
20-35%
Real-Time
Plus
Feedback
w/Smart
Program
Plus Smart
Application of S.S.
Insights
23
Social Norms Approach: Opower
24
25
Social Psychological Factors and
Energy Behaviors
1. Why behaviors
matter?
2. Decisionmaking models
3. Factors influencing
public acceptance of
energy-efficiency
measures & renewables
6. Energy efficiency in
office buildings
Content
5. Promoting
Demand Response
4. Interventions
(feedback, norms,
framing, etc.)
26
Message Framing and Electricity Saving
• RQ1: Which type of messages are effective to 1) change
attitudes toward electricity saving? 2) to boost perceived
efficacy (“whether can I make a difference”) on saving
electricity?
• RQ2: Do individuals with difference on environmental
concern react to messages differently?
• Design: Four (2 x 2) manipulations on benefit message
framing of saving electricity
Benefit framing: Environmental vs. Financial
Temporal framing: Long-term vs. Short-term
27
Condition 1/2: Financial, Long-term/Short-term
28
Condition 3/4: Environmental, Long-term/Short-term
29
29
Results of Message Framing
• 292 US residents
• Environmental messages, in general, are more
effective than the financial benefits in producing
positive attitudes toward energy saving.
• Short-term benefits boost perceived efficacy among
people with lower environmental concern.
6.5
6.04
5.88
6
5.5
5.13
Low NEP
High NEP
5
4.5
4.36
4
3.5
Long-term
Short-term
30
Recycling Towels in Hotels:
Evidence of Descriptive Norms
Below “Please reuse your towels”
Control: HELP SAVE THE
ENVIRONMENT. You can show your
respect for nature and help save the
environment by reusing your towels
during your stay
Social Norm: JOIN YOUR FELLOW
GUESTS IN HELPING TO SAVE THE
ENVIRONMENT. Almost 75% of
guests who are asked to participate in
our new resource savings program do
help by using their towels more than
once. You can join your fellow guests
in this program to help save the
environment by reusing your towels
during your stay.
Goldstein, Cialdini, & Griskevicius (2008, J of Consumer Research)
31
Financial Incentives -> Customer Segmentation
• Traditional approach to promoting demand response
(DR): economic approach – peak & off-peak pricing,
dynamic pricing, additional financial incentives, etc.
Will desired behavioral change disappear after
incentives are removed?
• To what extent financial incentives help customers
accept DR and energy saving programs? Same for
everyone?
• Money is often less of a motivator than other social
psychological variables (Stern et al., 1986).
32
• RQ: Are the requested incentives for different DR behaviors
predictable by energy use, demographic variables, and socialpsychological variables?
• Impact: more accurate estimation of adjustable loads as a function of
financial incentives
• Method: Online survey across 48 states, 754 valid responses
collected
How much monetary rewards do you expect in
exchange for… ?
•
•
•
•
Adjusting AC setting by 2-3oF when at home
Adjusting heater by 2-3oF when at home
Adjusting AC setting by > 5oF or shutting AC down
before leaving home
Adjusting heater setting by > 5oF or shutting heater down
before leaving home
33
Factors Affecting Requested Incentives
Energy Use Info
Demographics
Social-Psychological
Monthly Bill_Ave
Age
Energy Concern
Stay Home (9-5)
Gender
Bill Consciousness
Light Use
Income
Frugality
Computer Use
Education
Need for Comfort
Thermostat Settings
Political Orientation
Need for Convenience
Night Adjustments
House Sqft
Need for Control
Household Size
Trust
Weather Region
Subjective Norm
Perceived Control
Block 1
Block 2
Block 3
34
Financial Incentives and HVAC Behaviors
25.00%
no incentive
Acceptance rates in respondents
<5% of montly bill
20.70%
5%-10% of monthly bill
20.00%
18.30%
17.20%
would not do it
15.00%
14.10%
13.30%
12.60%
10.00%
9.90%
11.30%
11.70% 12.10%
10.50%
11.00%
10.70%
9.80%
9.10%
8.00%
7.80%
12.30%
11.00%
9.00%
7.10%
5.30%
6.00%
5.80%
5.00%
0.00%
Adjust AC for 23°F_at home
Adjust heater for 23°F_at home
Adjust AC for
>5°F_away
Adjust heater for
>5°F_away
Automation_AC
Automation_Heater
35
Financial Incentives and HVAC Behaviors
• Finding 1: The amount of the incentives requested does
not have a linear relationship with some factors

E.g. Small homes -> smaller amount of requested incentives to lower
thermostat setting in winter, large homes is not associated with higher
requested incentives.
• Finding 2: Social-psychological variables enhance the
predictive power of the model
Heater_Home
Block 1
Block 2
Block 3
No rewards
χ2(7)=9.23
(p=.236)
χ2(16)=35.01
(p=.004)
χ2(29)=58.37
(p=.001)
Would not do
χ2(7)=19.53
(p=.007)
χ2(16)=19.60
(p=.020)
χ2(29)=64.06
(p<.001)
Total
ΔF=2.75
(p=.008)
ΔF=1.88
(p=.060)
ΔF=2.53
(p=.002)
36
Factors Affecting Amount of Requested
Incentives (Adjust 2-3oF Heater at Home)
Energy Use Info
Demographics
Social-Psychological
Age
Energy Concern
Monthly Bill Ave
Gender
Bill Consciousness
Stay Home (9-5)
Income
Frugality
Lights Use
Education
Need for Comfort
Computer Use
Political Orientation
Need for Convenience
Thermostat Settings
House Sqft
Need for Control
Night Adjustments
Household Size
Trust
Weather Region
Subjective Norm
Perceived Control
Block 1
Block 2
Block 3
37
Analysis of Customer Segmentation (1)
• Raising AC thermostat setting for 2-3oF in summer
Features of the most
cooperative group
(no incentives)
•
•
•
•
•
Smaller homes
Lower electricity bills
More frugal
Lower comfort needs
Lower needs for
convenience
Features of the least
cooperative group
(won’t do it anyway)
• Won’t put PC in
sleep
• Lower thermostat
settings generally
• Older
• Higher comfort needs
• Poorer energy-saving
attitudes
• Lower level of
subjective norms
38
Analysis of Customer Segmentation (2)
• Accepting automation at home in summer
Features of the most
cooperative group
• Higher thermostat
settings generally
• Lower electricity bills
• More Democrats
• Higher energy concerns
• More frugal
• Lower comfort needs
• Lower needs for
convenience
• Positive energy-saving
attitudes
Features of the least
cooperative group
• Lower thermostat
settings generally
• Higher income
• Lower energy concerns
• Lower money
consciousness
• Lower trust
• Poorer energy-saving
attitudes
39
• How attractive are current DR programs in the market?
 Automatic A/C cycling
 Automatic A/C shut down
 Automatic thermostat adjustment
 Voluntary A/C adjustment
• How specific features of DR programs affect attractiveness?
 Frequency, duration, etc.
 Level of control
 Amount of rewards
• Predictable by energy use habits, demographics and socialpsychological variables?
40
41
Social Psychological Factors and
Energy Behaviors
1. Why behaviors
matter?
2. Decisionmaking models
3. Factors influencing
public acceptance of
energy-efficiency
measures & renewables
6. Energy efficiency in
office buildings
Content
5. Promoting
Demand Response
4. Interventions
(feedback, norms,
framing, etc.)
42
Future Work: Triple Levels of Integration
Technology
• Design of building
office setting
•Visibility of behaviors, human
interaction & control
Structural
Individual
• Company culture& policy
• Personality & Profiles
•Leadership
•Attitudes (Frugality)
•Group norms
•Self-efficacy
•Reward systems
•Habits
43
Energy Use in Office Buildings
Social Dilemmas
• Social dilemmas: a collective action dilemma, where
individual interests and collective interests are at odds
(Dawes, 1980; Van Lange, et al., 2013).
• Public goods dilemma: each member must decide whether
or not to contribute to a public good (or a resource) from
which all may benefit regardless of whether or not they
contributed.
• Diversity of social dilemmas: Cooperation forms (giving or
contributing vs. not taking or consuming), time span (short
vs. long), or scale (small vs. large) (Van Lange et al.,
2013).
44
Mechanisms of Increasing Cooperation
Potential Effects
Influencing Payoff Structure
in Social Dilemmas
Making Free-riding Less
Appealing
Reducing Fear of
Free-Ridership
Increasing Sense of
Responsibility
Factors
Personality (Pro-social vs. Pro-self)
Social rewards/punishment
Emotion (shame vs. guilt)
Group size
Group member’s identity
Group cohesion
Cooperative culture
Trust
Communication
Self-efficacy
Group identity
Other members’ decision to cooperate
(may diffuse responsibility)
Building Technology
45
Conclusions
• Human beings are not always rational.
• Information programs may be effective in changing attitudes but may
not be effective in changing behaviors in specific situations.
• There is potential in applying social norm and framing approaches.
• Energy use in public domain relates largely to social matters, not
individual ones.
• Social-technological approaches are needed for persistent
behavioral change and reducing free-riding:
• Making cooperation in public domain easier
• Changing the payoff structure of social dilemmas (in a more
friendly way)
• Increasing the transparency for the situation
46