Sustainable (IT-) Users

University of Hohenheim | Digital Management28

Sustainable (IT-) Users


Research Question

Baumbach et al. (2018)

What factors influence individuals to

behave in an environmentally sustainable

manner across the different life cycle

stages of information technology (IT)?



Concerns an individual’s behavioral intention to use IT with the aim of increasing sustainability.

An increased sustainability can

either be due to adjusting energy-

saving settings of IT or to buying

"Green-IT".

Focuses on the way IT is disposed.

Intention is described as the

behavioral intention to dispose IT

sustainably.

A consumer’s attention to the production of IT, which can be considered within the IT purchase process.

The stage captures an individual’s

behavioral intention to buy

sustainably manufactured IT.

Manufacturing / Buy of IT Use of IT Disposal of IT

Life Cycle of IT


Ajzen (1985); Ajzen and Fishbein (1980); Baumbach et al. (2018)

EA/EC

Attitude

Social

Norms

Perceived

Behavioral

Control

Behavioral

Intention

GEK PEK

Manufacturing

/Buy

sustainable IT

Use IT

sustainable

Dispose IT

sustainably

Traditional theoretical constructs (Theory of Planned Behavior)

Newly developed constructs

Life cycle stages of IT

EA/EC

ENVIRONMENTAL AWARENESS / ENVIRONMENTAL CONCERN: Concern about the environment.

When I think of the consequences of IT on the

climate, I am very worried.

GENERAL ENVIRONMENTAL KNOWLEDGE: Common

understanding of environmental related issues.

Fossil fuels produce carbon dioxide in the

atmosphere when burned.

PERSONAL ENVIRONMENTAL KNOWLEDGE: Specific personal environmental knowledge and

understanding

I know the meaning of the labels affixed on the

sustainable technologies (e.g., energy-efficient

devices).

General EK

Personal

EK

Analysis

• Development of questionnaire

• Conduction of Online Survey >300 participants

• Application of Structural Equation Modeling

Where in this lifecycle does sustainability play a role from a customer’s perspective?


Managerial Implications

1. Individuals prefer to buy IT which is

sustainably produced sustainable

manufacturing and marketing campaigns

2. Individual’s use IT to behave sustainable IT

may be designed to offer sustainability

attributes during usage (e.g., improving

carbon footprint)

3. Individuals pay attention to the disposal of IT

IT should be designed to offer simple and

sustainable way of recycling


Results Managerial Implications

“Environmental

Factors are positively

related to the

intention of

environmentally

sustainable behavior

across the life cycle

of IT”

Results and Implications


Sustainable (Decision-Making) Citizens


Research Article: Supporting Citizens’ Political Decision-Making Using Information Visualisation


Please prepare the following questions for the Live Session on June, 15

What are the paper‘s key thoughts? (~5 sentences)

Which (self-drawn) figure represents the paper?

What are the most interesting direct quotes? (~3 quotes)

Which references seem to be worth reading next? (~2)

What is most objectionable? (1-2 thoughts)

When / for what will I cite the paper? (1-2 thoughts)

Questions

1

2

3

4

5

6


Motivation| Living up to your Ideals may be challenging!

Everyday life Society at large

At the end of a cold winter day, there is hardly

anything more pleasant than a long hot shower…

…but wasn't I trying to limit my resource

consumption?

Damn, these pandemic measures have left me

isolated for weeks now and I really want to hang

out with my friends again…

…but how would relaxing social distancing

measures affect the overall spread of the virus?


Background | Some Theory on Human Decision-Making

Bordalo et al. (2012, 2013); Kluger and DeNisi (1996); Knobloch-Westerwick et al., (2020); Westerwick et al., (2017) ;

Can an Information-Systems (IS-) based tool influence individuals’ decision-making by

providing immediate feedback on decision consequences?

Salience Theory

Selective Information Search

FeedbackInterventionTheory

• Decision-makers are often found to make sub-optimal and irrational

decisions resulting from limited cognitive resources

• A bias in favor of the salient aspects of a decision leads to an attitude-

behavior gap

• People tend to seek information in ways that are partial towards their

existing beliefs

• This can determine the selective perception of salient decision aspects

• An effective way to overcome salience bias and selective information search

is by making the implications of one’s behavior salient in real time

• Individuals compare elements of a feedback intervention with (internal or

external) standards and adjust their behavior to attain the standard


Application Context | Citizens’ Decisions on Renewable Energy

• Public support for sustainability runs high

in all European countries – see

#FridaysForFuture

• Common mistake to expect citizens to

welcome developments

they claim to support

Acceptance of Renewable EnergyDecision-making in a citizen context

• Serious consequences at all levels of a

society

• Overwhelming complexity of decisions

often involves a multitude of outcomes

• Unlike in organizational and consumer

contexts, little attempts to make a

broad set of information available for

citizens to reflect on decision

implications


Approach | Research Hypothesis & Data Collection

ResearchHypothesis

ResearchModel

Data Collection

Citizens’ decisions on renewable energy change when respective consequences become clear.

Baseline Decision

(Limited) Information Decision

Information Visualization

IS-tool supported Decision

• Participants are requested to decide on

the proportion of coal-fired plants they

would replace with renewable wind

energy – assuming they had free choice

• IS-Tool provides immediate feedback in

terms of visualizing the location of newly

required wind turbines on a map

• Participants can reevaluate their

decision until the decision outcome is in

line with their preference


Results of a Data Collection

• Young, urban, and environmentally aware

citizens are willing to accept a high percentage

of renewable wind energy. This result reflects

trends and socio-economic developments at

the time when the survey was conducted.

• The tool influences citizens’ decision-making.

In particular, we find that all analyzed cross-

sections of citizens (e.g., different age,

different political affinity, different levels of

education) within sample changes the amount

of renewable energy initially desired, after

interacting with our tool.

• Citizens update, however not completely turn

over their preferred level of renewable wind

energy after interaction with the tool.

Results

1

2

3


Sustainable (Grocery Store) Customers


Motivation

• Food system is a major driver of global environmental challenges1

• Everyone can contribute by making sustainable food choices2

• These decisions about food consumption are increasingly made online3

The Need for Ecologically Sustainable Food Consumption

• e-commerce continuously grows4

• Grocery purchases made online as well as different kinds of online food

services are increasing3

• Many advantages, e.g., time savings, convenience, and flexibility,

especially in times of uncertainty like COVID-195

Rising Relevance of Online Food Shopping

Online grocery stores, delivery services, and subscription services represent choice environments in which consumers

decide between different food products. These choice environments can be modified using Digital Nudging Elements

(DNEs).

References: 1) Noleppa (2012); 2) Ferrari et al. (2019) and Mont et al. (2014); 3) Centraal Bureau voor de Statistiek (2019); 4) Wigand (1997); 5) Gassmann (2020); 6) PWC (2018)


Research Gap & Research Questions

• Modifying the choice environment to influence choices1

• Goal: Help making better choices without limiting freedom of

choice or manipulating incentives2

• Especially intuitive decisions are prone to heuristics, leading

to faster, but potentially undesirable decisions starting point

for nudging3

1. Which of the DNEs default rules, simplification, and social norms are effective in online food shopping

contexts regarding the promotion of ecologically sustainable food choices?

2. Do the DNEs differ in their influence on different consumer groups?

Research Questions

Nudging

• Only Demarque et al. (2015) focus on the design possibilities of social norms in online food contexts to promote ecologically

sustainable food choices

• Default rules and simplification are not evaluated in online food shopping contexts yet

• Hence, no comparison of the effects of these DNEs exists so far

Gap

Nudging to Promote Ecologically Sustainable Choices

• Food behavior is highly habitual prone to nudging4

• Lehner et al. (2016) and Ferrari et al. (2019) reviewed

prior research on the effect of nudging to leverage

healthier and ecologically sustainable food choices

= Default rules, changes to physical environment, simplification, and social norms

References: 1) Münscher et al. (2016); 2) Thaler and Sunstein (2009); 3) Kahneman (2011) and Tversky and Kahneman (1974); 4) van’t Riet et al. (2011)


Theoretical Background and Prior Research on Default Rules, Simplification, and Social Norms

Describes a setting in which the

preferred option is pre-selected

and will be maintained if the

person does nothing1

Campbell-Arvai et al. (2014):

default meat-free options

promote vegetarian meals

when eating out

Kallbekken and Sælen (2013)

and Vandenbroele et al.

(2018): default reduced plate

size leads to less food waste

Default Rules

Utilizes the effect of social

pressure & conformity by giving

information about appropriate

behavior within a group3

“70% bought at least one

ecological product”

(Demarque et al. 2015, p.

169)

Linder et al. (2018) and

Kameke and Fischer (2018)

used descriptive norms to

reduce food waste

Social Norm

Represents the transportation of

condensed information about a

complex construct/“Framing” of

information to activate values1,2

Van Gilder Cooke (2012) used

GHG emission labels to

promote environmentally-

friendly burgers

Redesign of menus in

restaurants (Bacon and Krpan,

2018; Kurz 2018)

Simplification

Definition

Prior

Research:

Food Context

Our Implementation

References: 1) Thaler and Sunstein (2009); 2) Sunstein (2014); 3) Aldrovandi et al. (2015) and Kormos et al. (2015)


Research Process

Implement DNEsin online shop

Conduct field experiment with

shopping task

Calculate sustainability

score

RQ1:(non)parametric

tests and multiple regression

RQ2:cluster analysis and

(non)parametric tests


Research Process



shopping task


score

RQ1:(non)parametric





Research Process



shopping task


score

RQ1:(non)parametric




• Structure: introduction, scenario description,

recipe, online shopping task, survey

• Run #1: random assignment to control group or

implementation of one of the three DNEs

• Run #2: repetition during revision for additional

DNE salience

Collect Empirical Data

References: 1) Onwezen et al. (2019) based on Steptoe et al. (1995); 2) Onwezen et al. (2014, 2019)


Research Process



shopping task


score

RQ1:(non)parametric









DNE salience


• Product analysis: identification of most and

least sustainable option for each ingredient

• Assignment: 0, 1, or 2 for least, second most,

and most sustainable ingredient option

• Calculation: sustainability score (SC) for each

participants shopping cart from 0-16

Determine Sustainability of Choices



Research Process



shopping task


score

RQ1:(non)parametric









DNE salience









• Comparison of SCs: ANOVA and Kruskal-Wallis

tests of SCs between control and DNE groups

• Multiple regression analysis: inclusion of

control variables regarding consumption

behaviour and motives (Food Choice Question-

naire FCQ1 and Self-reported Consumption SRC2)

Compare DNEs



Research Process



shopping task


score

RQ1:(non)parametric









DNE salience











• Multiple regression analysis: inclusion of

control variables regarding consumption

behaviour and motives (Food Choice Question-

naire FCQ1 and Self-reported Consumption SRC2)

Compare DNEs

• Clustering of participants: two-step cluster

analysis with hierarchical Ward’s and

partitioning k-means algorithms



within clusters

Compare DNEs within Participant Groups



Results: Comparison of SCs between Control and DNE Groups

Total C DR S SN

N 291 73 74 68 76

Mean 9,35 9,29 9,55 9,53 9,04 C .

Standard deviation 2,92 2,71 3,16 2,97 2,86 S *

Median 10 9 10 10 10 SN .

Interquartile range 4 4 3 4 4

Pair

wis

e p

ost

-hoc

t-te

sts

Pair

wis

e p

ost

-hoc

Mann-W

hit

ney-

U t

est

s

p-value significance codes: *** for < 0.001, ** for < 0.01, * for < 0.05, + for < 0.1Sh

apir

o-W

ilk

norm

ali

ty t

est

Bart

lett

vari

ance

test

AN

OV

A

Kru

skal-

Wall

is t

est

SC


Results: Multiple Regression Analysis with DV SC Including Control Variables FCQ and SRC

Variable Description Estimate p-value

Intercept 6,05 0,000 ***

Group DR Default rules 0,80 0,094 .

Group S Simplification 0,71 0,136

Group SN Social norms 0,07 0,876

FCQ1 Healthy 0,00 0,998

FCQ2 Enables mood monitoring 0,00 0,969

FCQ3 Convenient -0,07 0,568

FCQ4 Provides pleasurable sensations 0,06 0,669

FCQ5 Natural 0,17 0,379

FCQ6 Affordable -0,33 0,005 **

FCQ7 Helps control weight -0,06 0,483

FCQ8 Familiar 0,06 0,574

FCQ9 Environmentally friendly 0,25 0,195

FCQ10 Animal friendly 0,03 0,857

FCQ11 Fairly traded 0,27 0,193

SRC1 Vegetables 0,25 0,120

SRC2 Fruit -0,12 0,330

SRC3 Dairy 0,02 0,785

SRC4 Fish 0,05 0,736

SRC5 Meat -0,18 0,101

p-value significance codes:

*** for < 0.001, ** for < 0.01, * for < 0.05, + for < 0.1


Results: Comparison of SCs between Control and DNE Groups within Clusters of Participants

Total C DR S SN

C1 N 95 32 18 21 24

Mean 10,30 10,00 10,11 11,52 9,75

Standard deviation 2,60 2,89 2,70 1,97 2,42 . . C-S * C-S *

Median 10 10 10 11 10


C2 N 90 16 31 23 20

Mean 8,36 8,50 9,90 8,13 7,65

Standard deviation 2,84 1,75 3,04 3,01 3,03 S *

Median 8 8 9 8 7


C3 N 106 25 25 24 32

Mean 9,34 8,88 9,96 9,13 9,38

Standard deviation 3,00 2,83 3,57 2,80 2,84

Median 10 9 10 10 10


Cluster codes: C1 - environmentally-conscious, C2 - environmentally-unconscious, C3 - pragmatic

p-value significance codes: *** for < 0.001, ** for < 0.01, * for < 0.05, + for < 0.1

Shapir

o-W

ilk

norm

ali

ty t

est

Bart

lett

vari

ance

test

AN

OV

A

Kru

skal-

Wall

is t

est

Pair

wis

e p

ost

-hoc

Mann-W

hit

ney-

U t

est

s

SC

SC

Pair

wis

e p

ost

-hoc

t-te

sts

SC

• Motives: high importance of naturalness,

environmental friendliness, fair trade, …

• Consumption: more plant-based or veggie

Environmentally-conscious

• Motives: low importance of naturalness,

environmental friendliness, fair trade, …

• Consumption: less plant-based or veggie

Environmentally-unconscious

• Motives: high importance of convenience,

price, familiarity, …

• Consumption: in-between

Pragmatic


Contribution

1. Default Rules can be implemented in online

food services that increasingly have the power to influence our food choices

2. Help environmentally-conscious customers

with simplification nudge to transfer their good intentions into concrete choices

3. Consumers could profit from time savings due

to reduced decision-making efforts as well as

support to act on their societal responsibility

4. Customers’ price sensitivity has a negative

influence on SCs; hence, this relationship

needs to be dissolved

Practical Contribution

1. Complementing the research by Demarque et

al. (2015) about the DNEs social norms, we

transferred two additional major NEs from the physical to the digital world

2. We compared different DNEs and shed new

light on possible differences in their impacts

3. We identified three typical consumer types,

which enabled us to examine the

effectiveness of the different DNEs in different consumer groups

4. We found that the DNE simplification proved

to be effective for environmentally-conscious

consumers

Theoretical Contribution

+


Limitations and Further Research

So far, only the three most common (D)NEs in

the food consumption domain have been

considered.

Only one implementation of each DNEs has been

considered yet.

Despite incentive to behave as usual, the

observations base on an artificial field

experiment.

The sample size is limited, especially regarding

within-cluster comparisons.

Limitation

Include further (D)NEs such as feedback,

reminders, and, also, salience.

Consider different implementations and levels

of DNEs.

Partner with online food services to collect

real-life data on consumer behavior.

Collect more data in collaboration with

partners to gain more reliable insights.

Implications for Further Research

DNE Number

DNE Design

Real-world

Observations

Sample Size

#

Sustainable (IT-) Users

Documents

Transcript of Sustainable (IT-) Users