Usefulness of learning policy lessons : and the implications for policy transfer

Margarita Yerastova // 11330627

Thesis Advisor: Dr. Marco te Brömmelstroet

Second Reader: Dr. Mendel Giezen

Masters of Urban and Regional Planning

ladyenoki@gmail.com

UNIVERSITY OF AMSTERDAM

GSSS

GRADUATE SCHOOL OF SOCIAL SCIENCES

MASTER URBAN AND REGIONAL PLANNING

Usefulness of Learning Policy Lessons

and the Implications for Policy Transfer

Master’s Thesis

Margarita Yerastova

Student Number: 11330627

11 June, 2018

Thesis Supervisor:

Dr. Marco te Brömmelstroet

Second Reader:

Dr. Mendel Giezen

Yerastova

Page 2 of 93

Acknowledgements

I would like to express my gratitude to my thesis advisor, Dr. Marco te Brömmelstroet, with input from

Dr. David Evers, and to Meredith Glaser at the University of Amsterdam for their invaluable guidance and

feedback, and for providing me access to the data. Their advice and expertise has steered this thesis in a

socially and scientifically-relevant direction. I would also like to extend my gratitude to the second reader,

Dr. Mendel Giezen, and to the interviewees without whom this research would not be possible.

Furthermore, I would like to express my gratitude to my parents who have always pushed me to

pursue my interests no matter how far they took me and to thank all my colleagues, mentors, and friends

who have believed in me and in a brighter future, who continue to push the envelope on what is possible.

Yerastova

Page 3 of 93

Abstract

In recent years, cycling has seen a renaissance as part of a human-scale strategy to address the

mobility dilemma and improve the quality of life in cities. The successful integration of cycling into the

mainstream transport system has inspired many to come to the Netherlands to engage in ‘in situ’

experiential learning with study tours in search of inspiration inform their local policy. While ‘usefulness’

has been explored through literature as a metric in addressing gaps in the acceptance and usability of

tools, it has not yet been applied to learning as a strategic tool in itself to evaluate its efficacy for policy

transfer. This thesis aimed to bridge that knowledge gap by testing the application of the metric to see if

the Perceived Usefulness (PU) of the lessons learned translated into Actual Usefulness (AU) through the

case study of the 2017 Velo-city conference that was held in the Netherlands, as compared with the study

tours, that provided an interesting opportunity to study ways of learning ‘in situ’ about cycling policies and

practice. A survey questionnaire was then analyzed and interviews were conducted with the study

ultimately finding that, out of the identified ‘ways of learning’, actual usefulness best produces actionable

results with the experiential learning offered at conferences and even more so with the deeper-focused

study tours.

Yerastova

Page 4 of 93

Table of Contents

Acknowledgements ... 2

Abstract ... 3

Table of Contents ... 4

Table of Figures ... 5

List of Tables ... 5

1. Introduction ... 7

1.1 Research Aims ... 7

1.2 Research Questions ... 8

1.3 Limitations ... 8

1.4 Organization of the Report ... 9

2. Literature Research ... 11

2.1 Policy Transfer & Policy Learning ... 11

2.2 Ways of Learning ... 12

2.2.1 Organizational and Institutional Learning ... 12

2.2.2 Educational Learning ... 13

2.2.3 Experiential and Social Learning ... 13

Conferences ... 13

Site Visits and Study Tours ... 14

2.3 Usefulness as a Metric ... 15

2.4 Concluding Remarks ... 16

3. Research Methodology ... 18

3.1 Research Design ... 18

3.2 Data Collection ... 19

3.3 Survey Questionnaire Methodology ... 19

3.4 Perceived Usefulness (PU) Metric Methodology ... 19

3.5 Actual Usefulness (AU) Metric Methodology ... 22

3.6 Semi-structured Interview Questionnaire Methodology ... 23

4. Case Study Results ... 25

4.1 Engagement with Different Ways of Learning and Methods ... 25

4.2 Perceived Usefulness (PU) ... 34

4.3 Actual Usefulness (AU) ... 37

4.4 Further Unpacking Usefulness ... 43

5. Reflection, Conclusion, and Recommendations ... 46

5.1 Reflection ... 46

5.2 Conclusion ... 47

5.3 Recommendations ... 48

6. References ... 50

7. Appendix ... 54

7.1 Velo-city 2017 Survey Questionnaire ... 54

7.2 Semi-structured Interview Questions for Velo-city 2017 Attendees ... 62

7.3 Semi-structured Interview Questions for People for Bikes Study Tours 2010-15 ... 63

7.4 Interview Consent Letter ... 64

7.5 Thematic Analysis Codes ... 65

Yerastova

Page 5 of 93

Table of Figures

Figure 1. Kolb's experiential learning cycle (Bracic, 2017, p. 10) ... 14

Figure 2. Outcomes and Process Framework, (te Brömmelstroet et al., 2016, p. 302) ... 15

Figure 3. Velo-city Learning Methodology Popularity – Sample A ... 27

Figure 4. In-conference Platform Popularity – Sample A ... 28

Figure 5. Conference Profession Demographics – Sample A (left) and Sample A-A (right) ... 29

Figure 6. Velo-city Learning Methodology Popularity – Sample A-A ... 30

Figure 7. In-conference Platform Popularity – Sample A-A ... 31

Figure 8. Demographic Snapshots of Learning Methodology Popularity – Sample A ... 32

Figure 9. Demographic Snapshots of Learning Methodology Popularity – Sample A-A ... 33

Figure 10. Survey Questionnaire Usefulness for Sample A-A ... 36

Figure 11. Sample A-B Interviewee List ... 37

List of Tables

Table 1: Metric for Perceived Usefulness (PU) on the process of learning ... 21

Table 2: Metric for Actual Usefulness to the outcome of learning ... 23

Table 3. Velo-city Survey: Methodologies for learning (more about cycling) ... 25

Table 4. Velo-city Learning Methodology Popularity – Sample A ... 26

Table 5. PU Metric for Sample A-A ... 35

Table 6. Frequency Table – AU Metric for Sample A-B // Conference Attendees ... 38

Table 7. Frequency Table Analysis – AU Metric for Sample A-B // Conference Attendees ... 38

Table 8. AU Metric for Sample A-B // Conference Attendees ... 39

Table 9. Frequency Table – AU Metric for Sample A-B // Study Tour Participants ... 41

Table 10. Frequency Table Analysis – AU Metric for Sample A-B // Study Tour Participants ... 41

Yerastova

Page 6 of 93

Chapter 1

Introduction

Yerastova

Page 7 of 93

1. Introduction

Cycling has seen a recent renaissance, with a renewed interest from planning, as part of a humanistic

strategy to address the mobility dilemma and improve the quality of life in cities. The Netherlands is a

country regarded as the “gold standard” for their sustainable planning achievements with both academics

and practitioners, and more specifically within transportation planning for achieving cycling “best

practices”. The Netherlands has a national total modal split for bicycles at 27%, up to 35-40% in the top

municipalities, with only 181 deaths out of 14.4 billion kilometres traveled by bicycle reported in 2005

. In

an updated report from 2016, the use of both bikes and e-bikes collectively in kilometres travelled has

gone up by a further 11%

. The successful integration of cycling into the mainstream transport system

has inspired hundreds of individuals and groups to come to the Netherlands to engage in ‘in situ’

experiential learning through study tours in search of inspiration and ideation to take back to their home

countries to inspire and inform their local policy. Furthermore, the international cycling conference

Velo-city came back to the Netherlands in 2017, providing another interesting opportunity to study ways of

learning ‘in situ’ about cycling policies and practice. This then begs the question: how actually useful are

these visits?

It is from the desire for new and innovative ideas that policymakers look to approach learning as a

powerful tool for the creation and dissemination of knowledge to impact policy decisions (Stone, 1999).

Much literature has been written about policy transfer, policy mobility, and policy learning. However,

learning remains a black box in the policy transfer process in transportation as the transition towards

sustainability increasingly becomes a tenet in urban planning. The decision-making process around which

policy measure or approach would be the most effective is naturally based on some insight into its

usefulness. Thus, utilizing this approach to analyze the lessons learned can be helpful to evaluate

strategies for learning, their implications, and further guide policy transfer research.

1.1 Research Aims

The industry that organizes and leads the groups and delegations coming to the Netherlands to be

immersed in its cycling culture and engage in “in situ” experiential learning is relatively undeveloped,

informal, and lacks systems or records for follow-up. Therefore, research into the industry represents an

opportunity to delve into these visits to fill in the knowledge gap on their outcomes. Thus, the aim of this

research is to better understand the outcomes of these visits and learning experiences through the metric

1

Source: http://www.fietsberaad.nl/library/repository/bestanden/CyclingintheNetherlands2009.pdf

2

_{The report acknowledges this may be due to the growing popularity of e-bikes and people moving further away. Source:}

Yerastova

Page 8 of 93

of usefulness. Usefulness has been explored through literature in addressing gaps in the acceptance and

usability of tools, both in the planning field and in informational technology (F. D. Davis, 1989; Karahanna

& Straub, 1999; te Brömmelstroet, Curtis, Larsson, & Milakis, 2016; Wechsung, 2014). As we can

consider ways of learning to be strategic tools that can be employed, the metric of usefulness has not yet

been applied to evaluate them or their effect on policy transfer. Therefore, the research also aims to both

test the application of this metric and reflect on its validity and merit, as well as to add to the literary

discourse by building on the theoretical framework about policy learning. The scientific relevance of the

thesis is that it operationalizes an approach to analyzing learning in an interesting direction. The societal

relevance is that it offers practitioners and policymakers greater insight by understanding the role of the

home context in either hindering or supporting actual policy transfer by validating if the “perceived

usefulness” of the lessons turned into “actual usefulness”.

1.2 Research Questions

The above research aims are translated into the main research question below, followed by the

sub-questions identified as necessary to answer it:

How do the different ways of learning affect both their perceived and actual usefulness to

influence policy transfer?

1. What ways of learning and methods do actors involved in policy transfer engage in?

2. How do the different ways of learning correspond with their perceived usefulness?

3. To what extent does the perceived usefulness correspond with actual usefulness in

influencing policy transfer?

1.3 Limitations

First, the biases inherent with qualitative analyses from the views and interests of the researcher will

always affect the results of the research where it may not be replicable. That being said, the research was

also subject to the questions and data from a survey questionnaire already formulated without input from

the researcher; therefore, the subsequent metrics and analyses were constrained by the availability of the

data, its logic, and its language. Secondly, the time limitations for the thesis and research participation

have had to significantly simplify the scope of the research by narrowing it further to the data available to

only really analyze two of the learning platforms: Study tours and Conferences. The study additionally

limited the research to the geographic boundary of participants from the United States (US) and Canada,

both per the linguistic familiarity to the researcher as well per the study tours from People for Bikes, which

Yerastova

Page 9 of 93

is an organization based in the US. Third, the field of policy and knowledge transfer involves many

different stakeholders, transfer agents, and actors. While the participants varied, the sample is not

inclusive of the entire broad spectrum that has involvement in or effect upon cycling policy transfer. In fact

with conferences, most demand travel and tend to draw the same familiar crowd, so there’s an inherent

bias with the participation.

Furthermore, the limitations of the subjective nature of creating a metric for Perceived Usefulness (PU)

and Actual Usefulness (AU) do not set out to prove or disprove anything - they only add another unique

approach to the literature for further research. The metrics chosen to operationalize PU and AU do

constitute a good approach, but there could be other, more appropriate dimensions as well as other

influencing factors, such as personal dynamics and other institutional and contextual barriers that are

going to be limiting (te Brömmelstroet, 2013).

Finally, as many studies also state, deeper longitudinal studies would increase the knowledge gained

about the topic as the limitations of this thesis do not allow for comprehensiveness.

1.4 Organization of the Report

This thesis is structured into Chapters to organize and present the research in a logical and consistent

manner to answer the research questions. The research itself is divided into two phases: the first phase is

based on the survey results from the case study and an evaluation analyzed with the Perceived

Usefulness metric, while the second phase is based on follow-up interview results and analyzed by the

Actual Usefulness metric.

Chapter 2 begins by providing the necessary content on relevant

case-independent literature in order to explore the theory on policy tourism and transfer, the different ways of

learning, and generalizing the concept of usefulness. Chapter 3 introduces the research methodology

with the case studies, data collection, and the metric methodology used in the analysis. Chapter 4 then

discusses the case study results. Chapter 5 concludes the report in a reflection on the research, draws

conclusions, and makes recommendations for future research. Chapters 6 and 7 list the references and

supporting appendix documents.

Yerastova

Page 10 of 93

Chapter 2

Yerastova

Page 11 of 93

2. Literature Research

2.1 Policy Transfer & Policy Learning

The quest for learning can be triggered by the need to adapt to changing global trends through the

creation and transfer of knowledge, creating vast networks of epistemic communities that facilitate this

(Evans, 2009, p. 252). Policy transfer, also known as “lesson-drawing”, is the process of transferring

knowledge about policies from one political system for use in the development of another with the

aspiration of policy change (Dolowitz & Marsh, 2000; Evans, 2009; Wolman & Page, 2012). However,

though the term ‘transfer’ implies linearity in literalist terms of clear lessons translated into full policy

transplant from one political context to another, it’s a much more generalized process of on-going policy

learning (Hudson & Kim, 2014; Marsh and Evans, 2012; McCann, 2012). Policy learning, therefore, is one

of many influences on policy transfer outcomes that can be better defined as complex and dynamic

assemblages of ever-evolving ideas from multiple institutions that evolve the decision-making process

(Hudson & Kim, 2014). Policy transfer requires not only acquiring knowledge, but understanding the local

context in which it will be applied (Wolman & Page, 2012).

Policy lessons can be drawn from any level, though local actors draw from more regional contexts

while national actors take on a broader international approach (Marsden & Stead, 2011, p. 493). The

roles of these actors differentiate policy transfer from policy diffusion, where the first defines a state-led

role and the latter implies the transfer of information from social systems (Marsden & Stead, 2011).

Identifying these actors and their roles in the transfer of knowledge is crucial to facilitating analysis of how

policies are mobilized (McCann, 2011). Actors who traditionally facilitate policy learning and transfer are

bureaucrats, politicians, and other government experts, though the broader epistemic community

identifies consultancy firms, law firms, non-governmental organizations (NGOs), think-tanks, banks, and

transnational advocacy networks who have considerable influence and leverage on agenda-setting

(Evans, 2009; Stone, 2004). These actors highlight the multilateral dynamic environment of policy transfer

and how they, as key players, can impact domestic politics through the creation and dissemination of

knowledge (Stone, 2004). The interactions between state and non-state actors thus create a “shared

experience of learning” and a common perspective (Stone, 2004).

Policy travel or policy tourism, as McCann (2011) defines, is a set of activities that specifically aims to

inform policy learning and broaden the scope of knowledge, such as site visits and conference

attendance (p. 117), that can make those actors participating in the travel effectively global transfer

agents (Stone, 2004). Though policymakers look abroad for solutions, the heterogeneity and potential

information overload can generate uncertainty and support a bounded rational approach to policy

Yerastova

Page 12 of 93

transfer, limiting the knowledge the transfer agent can apply back home (Ettelt, Mays, & Nolte, 2012;

Marsden, Frick, May, & Deakin, 2012).

In this regard, the bounded rationality explanation suggests that actor preferences are shaped and

bounded by institutional constraints with structural, social, cultural, and individual factors (Marsden et al.,

2012). Furthermore, actors are bounded by the learning heuristics they use to narrow down the

overwhelming amount of information, either through strictly sticking to “best practice” parameters, or

through relying on trusted peer experience as a strategy (Dunlop & Radaelli, 2017; Macmillen & Stead,

2014; Wolman & Page, 2002).

2.2 Ways of Learning

So how do policy transfer agents and decision-makers learn to learn? Different “ways of” learning can

facilitate ideation and knowledge transfer, with both transformative and restrictive influences on

decision-making processes. Acknowledging how ways of learning and learning heuristics are naturally predisposed

to the preferences of their own bounded rationalities can create opportunities for effective policy transfer

results. For example, though trusted peer experience appears to be limited by bounded rationality, it is

also supplemented by social learning theory that encourages growth of peer networks.

2.2.1 Organizational and Institutional Learning

Organizations play a key role in policy learning and transfer, through emulation and diffusion where

they facilitate the exchange of information (Newmark, 2005). Institutions are the “formal or informal

procedures, routines, and conventions embedded in the organizational structure” (Gonzalez, Healey,

Healey, & Gonza, 2005; Pojani & Stead, 2014). Institutions can seem inflexible, but they can also allow

actors to make efficient decisions within the stability of embedded structures (de Jong, 2016; Pojani &

Stead, 2014). Organizational learning is shaped by the attitude of key senior staff who can be active or

passive in ideation (Marsden, Frick, May, & Deakin, 2010). It’s a strategy based on the benefits of the

collective learning power of the organization (Dunlop & Radaelli, 2017; Evans, 2009).

The “bounded rationality” of the organization or institution predisposes lesson-drawing and policy

choices to certain preferences through the emulation of culturally-similar contexts (Macmillen & Stead,

2014). Organizations can filter and shape the processes of learning by influencing who learns what and

can hinder it through non-decisions and biases (Dunlop & Radaelli, 2017).

Yerastova

Page 13 of 93

2.2.2 Educational Learning

Learning at its core is part of human development as an interaction between the individual and the

outside world, both bounded by socio-cultural structures (Illeris, 2007). Learning is organized in an

institutionalized setting that also itself legitimizes those structures and ideologies (Illeris, 2007). As adults

are more engaged and active participants in their learning process rather than passive, higher

educational learning can be seen as a transformative process (Dirkx, 1998). This transformative learning

is also used as a strategy to promote participatory, social learning approaches that build mutual trust

(Kolb, 2015; Taylor, 2007).

2.2.3 Experiential and Social Learning

The prima facie relationship between policy transfer and social learning is made explicit by the roles

played by the epistemic transfer community (Evans, 2009). McCann (2011) calls the creation of social

learning through this network a “global consultocracy” where policy mobilities are facilitated by vast

informational infrastructures of actors that package and sell information (p. 114). This kind of social

learning is shaped by “individuals, organizations, the ‘cobweb of interaction’ and the feedback effects of

previous policy choices” (Dunlop & Radaelli, 2017, p. 308).

The interaction provided by group environments can offer deeper learning and reflection opportunities

(Kolb, 2015). Environments that foster this collaboration also improve the learning of complex information,

as the theory of cognitive load supports simplifying education to allow better experiential learning (Bracic,

2017, p. 12). Bracic (2017) outlines Merrienboer and Sweller’s (2005) four aspects of experiential learning

on cognitive load: 1) acting by making immediate decisions in the now, 2) learning new concepts, 3)

reflecting and reviewing to process the new information, and 4) acting again by transferring the ideas

back home (p. 12). This is further supported and argued by David Kolb that ideas, and therefore learning,

are “formed and re-formed through experience”, and he subsequently devised a learning model with four

elements conceptualized in Figure 1(Bracic, 2017; Montero, 2017a). Knowledge, to Kolb, is therefore the

product of the dynamic process of transformation of growth-producing experiences (Kolb, 2015).

Therefore, two dynamic instruments of social and experiential learning are conferences and study tours

because both provide this interactive group learning environment.

Conferences - Conferences provide a unique platform for hosting social learning experiences for

people looking for new and innovative ideas to bring home, and are a key for creating connections for

later contact (Marsden et al., 2012). They allow opportunities for policy actors to benefit from “global

circuits of policy knowledge” (McCann, 2011, p. 120). McCann (2011) additionally deems conferences as

“global microspaces” where knowledge is shared about best practices and trust and alliances are

Yerastova

Page 14 of 93

developed among attendees, thus connecting otherwise isolated parties (p. 119). These alliances then

create transnational communities of experts (Stone, 2004).

Site Visits and Study Tours - Site visits, as part of ‘in situ’ experiential learning, are also part of

“fact-finding trips” and policy travel whose purpose is to allow first-hand learning (McCann, 2011). These

experiences provide functional examples (Mild & Schlossberg, 2014) and thus fit within ‘evidence-based’

policymaking (Ettelt et al., 2012). Study tours go further: they are organized trips with an express

educational purpose to learn from the local host agents and build relationships (Montero, 2017a; Pojani &

Stead, 2015; Wood, 2016). Arguments are made that study tours and site visits are the ‘golden standard’

and most effective tool for deeper learning because they provide opportunities for validating and

challenging the knowledge (Ettelt et al., 2012; Hudson & Kim, 2014). This also supports the notion that, in

respect to learning, these fact-finding trips are also political situations where “negotiation and persuasion

occur” and where decision-making can occur (McCann & Ward, 2013). They thus can also be used as

political mechanisms to mobilize support for policy change by utilizing existing examples to create

legitimacy (Hudson & Kim, 2014; Montero, 2017a).

Yerastova

Page 15 of 93

2.3 Usefulness as a Metric

Usefulness, and particularly perceived usefulness (PU), is part of the larger overarching process of

adoption, acceptance, and usage, where the term implies a potential added value (te Brömmelstroet,

Curtis, Larsson, & Milakis, 2016, p. 1180). This approach to learning can help identify common

characteristics, values, and qualities to better understand what knowledge is transferred and why.

Therefore, it has practical implications for providing further insight into how it can influence desirable

directions for acceptance and usage of the knowledge gained from policy lessons.

Usefulness has been explored through literature in addressing the gaps of policy acceptance and

usability, both in the planning field and in informational technology, when it comes to researching the

usage or lack thereof of tools and instruments (F. D. Davis, 1989; Karahanna & Straub, 1999; te

Brömmelstroet et al., 2016; Wechsung, 2014). From a socio-psychological perspective, the theory of

technology acceptance and its model posit that usage and acceptance depend upon beliefs held about

perceived usefulness (PU) and perceived ease-of-use (PEU), where PU is defined as beliefs a person

has about the system being useful to them and PEU is defined as “the belief that its use would be

effortless” (Karahanna & Straub, 1999, p. 238). These beliefs lead to 1) positive attitudes, 2) intentions of

use, and 3) actual usability and therefore, Actual Usefulness (AU) (Karahanna & Straub, 1999, p. 239).

Davis’s 1989 study on PU and PEU highlights the strong correlation between PU and acceptance, with

these beliefs seen as variables and as behavioral determinants (p. 335). Research on usability in the

planning field comes from evaluating an implementation gap in the use of certain tools and instruments,

where the goal is to make these instruments ‘useful’ and accepted, and ultimately improve the quality of

the planning process (te Brömmelstroet et al., 2016). The variables used can be quite subjective, with

Yerastova

Page 16 of 93

literature studying both the process and hedonic content qualities, such as shared insights and “better”

strategies (te Brömmelstroet, 2016). Other variables, such as total system cost and other technicalities

and pragmatic values, can add further to the picture of PU (Wechsung, 2014). The framework identified

by te Brömmelstroet et al. (2016), as shown in Figure 2, are used in this study for operationalizing

‘usefulness’ to evaluate perceived usefulness through the process characteristics and actual usefulness

through the outcome characteristics.

2.4 Concluding Remarks

The literature in both in planning and information technology uses usefulness as a metric to analyze

tool performance to achieve better outcomes and greater tool acceptance. However, this metric needs to

be unpacked depending on the research because it can be illustrated by various process and outcome

qualities. Basing these metrics on the quality of ideas rather than quantity, PU can be operationalized and

a framework can be established. If learning is perceived as a tool as well, then the same metrics can be

applied to examine and ultimately improve the quality of the policy learning process, subsequent

outcome, and finally AU. It can add to the literary discourse by identifying ways of learning with the

greatest added value and positive effects.

Yerastova

Page 17 of 93

Chapter 3

Yerastova

Page 18 of 93

3. Research Methodology

The research methodology uses a case comparison between the ways of learning in an international

conference and on study tours to unravel the perceived and actual usefulness of the learning experiences

in order to make inferences and draw some conclusions. Therefore, the units of analysis are participants

from both cases, who are essentially policy actors.

People for Bikes is an industry coalition that connects the bicycle supply and retail industry with

advocacy efforts and sponsorships to improve bicycling conditions in the United States. Through these

efforts, they were able to sponsor the 5-year Green Lane Project, from 2010 to 2015, to promote

protected cycling infrastructure uptake by sending study tours of select policy actors from some chosen

cities to the Netherlands with the goal of inspiring policy transfer. This organization works with various

host cities and collaborates with researchers from the University of Amsterdam. Additionally in 2017, the

Netherlands also took center stage for cycling policy tourism when it hosted the international cycling

conference Velo-city in Arnhem and Nijmegen, which promoted knowledge-sharing through the

immersive experiential learning environment. The conference brought together an international medley of

consultants, civil servants, policymakers, and students from 40 countries to experience Dutch cycling

firsthand, with high hopes of transferring their experiences back to their home contexts.

3.1 Research Design

The research design follows a testing method that employs both a quantitative and qualitative

approach using case study design for a comparative analysis to answer the research sub-questions, and

through them the main research question. The first sub-question is answered by analyzing the data

collected through the baseline survey questionnaire that identifed themes, patterns, and relevant data to

establish the Perceived Usefulness (PU) metric. The second sub-question is answered through the

application of the PU metric on the identified relevant survey questionnaire responses, which in turn

inform the Actual Usefulness (AU) metric. The third and final sub-question is then answered through the

application of the AU metric on semi-structured interview responses. Using Microsoft’s Excel software,

basic quantitative analyses were done for the survey data and compiled visually into various figures. The

semi-structured interviews were analyzed using a thematic analysis to make inferences, available in

Appendix 7.6, and then a frequency analysis was employed for quantitative generalizations.

The overarching generalization, and thus the tenet of this research, is that actual usefulness should

translate to an action/policy transfer to some capacity, based on the perceived usefulness of the

application of that knowledge that was inspired by the learning the actors engaged in.

Yerastova

Page 19 of 93

3.2 Data Collection

The research data collection consists of two parts: 1) respondent survey data from the Velo-city 2017

conference and 2) semi-structured interview data collected from both conference and study tour

participants from the United States and Canada. The primary survey data is a convenience sample of all

online and paper survey forms submitted at the end of the conference (“Sample A”), from which a smaller

subset was derived of attendees from the geographical boundary of the United States and Canada

(“Sample A-A”). Furthermore, an even smaller convenience sample subset (“Sample A-B”) was derived

from which the interviewees were chosen to contact. This sample set had to fulfill the following

characteristics where they: 1) answered that they attended a study tour at least once in their career on

the survey (see Appendix 7.1 for survey questionnaire) and 2) supplied a valid e-mail address. Due to a

large bias in the sample with many well-educated and/or academic respondents, Sample A-B was further

reduced to focus more on professionals from more recognized organizations to be able to compare them

to the study tour participants. This subset group was then added to the list of individuals who had

completed an interview with University of Amsterdam’s Ph.D. candidate Meredith Glaser prior to the

conference, together with a list of People for Bikes study tour attendees from 2010-2015 also supplied

from the University of Amsterdam’s archives who were noted to be willing to participate in further

research. Emails were sent out from which 15 interviews were conducted and the final list of participants

is shown in Table 6. The semi-structured interviews were conducted over a period of a month using the

audio/video calling applications WhatsApp, FaceTime, and Skype and recorded as audio files using the

software Quicktime on an Apple MacBook Pro laptop. The audio recording transcriptions are not included

in this report but the audio files themselves may be requested from the University archives.

3.3 Survey Questionnaire Methodology

The full survey questionnaire can be found in Appendix 7.1 of this report. The questions were split

between demographic information, general conference evaluation, and questions designed to ascertain

the degree of learning, knowledge sharing, and application of knowledge (which both the metrics of PU

and AU have included). The evaluation of these survey questions and their answers was used to answer

the first research sub-question.

3.4 Perceived Usefulness (PU) Metric Methodology

To establish criteria to ultimately judge policy success or failure, the lens of the PU of policy lessons

may bring insight into the methods of learning offered. PU is contextualized in this study as an innovation

to think about learning and policy transfer, using the construct of added value on quality, where its use is

Yerastova

Page 20 of 93

associated with positive effects (Hazen, Overstreet, & Wang, 2015; Karahanna & Straub, 1999; te

Brömmelstroet, 2013). The ways of learning offered may or may not match up to the PU of the ways of

learning, but nevertheless the survey forms the baseline for qualifying the PU metric and qualifying

learning itself as a tool. The usefulness metric and characteristics identified by te Brömmelstroet et al.

(2013) for improving planning tools form the basis of the PU metric for this study. Their metrics are split

between two frameworks identified earlier in Figure 2, where learning is inherently a continuous process

with the expected value placed on the resulting outcomes. It’s also a process based on the internalization

of explicit information to improve the quality of those outcomes (te Brömmelstroet, 2013, p. 302).

Additionally, the expectation of learning from ‘best practice’ examples is valued high enough to command

the premium price of travel for the ‘in situ’ experience.

Per te Brömmelstroet et al. (2013), learning as a process can also be abstracted into its

multidimensional nature as both an individual and a group process where various actors work together to

maximize their individually-learned assets towards a strategically shared outcome. Thus, this group

outcome is also a means to achieve the individual outcome. Following this logic, the PU of the learning

process can be operationalized using the same metric as the performance of PSS, which also follows the

field of Group Model Building (te Brömmelstroet, 2013, p. 301). First, the individuals engaged in learning

come from different backgrounds with different goals, therefore learning should culminate into a positive

‘reaction’ where enthusiasm, satisfaction, and credibility are collectively achieved to improve

performance. Second, the improvement of ‘insight’ into both the problems at hand and individual

assumptions should lead to better professional behavior. Third, the learning process would focus on

improving the ‘communication’ of the perspectives of both the group members in the same field and

bridging the gap with other fields with the ability to establish a ‘shared language’, as well as coming to a

‘consensus’ on problem definition, goals, and strategies. Finally, the last two goals of the learning process

would be to establish ‘group cohesion’, where individuals can relate to a group, and ‘effectiveness’ that

measures gains in abilities in relation to efforts (te Brömmelstroet, 2013).

Based on te Brömmelstroet’s framework identified in Figure 2 and the survey questions, the PU metric

methodology, shown in Table 1 on the next page, was developed and applied to Sample A-A responses.

The measurable responses to the survey questions used in the metric were ranked on a Likert Scale, with

choices ranging from Not at all, Somewhat, Quite a lot, and Very much. Other measurable responses

used in the metric are analyzed with a simple frequency analysis, such as “Top 3 Choices”.

Yerastova

Page 21 of 93

Individual Reaction (Enthusiasm, Satisfaction, Credibility)

1.

The learning experience matched my expectations (Lectures, Round table discussions, panel

discussions, Speed dating, Pecha Kucha, Master classes, Outdoor).

2.

The indoor sessions were insightful

3.

The bike tours were insightful.

4.

The Velo-city 2017 experience changed my view of cycling.

5.

I had my values or belief systems challenged.

6.

I questioned my personal habits.

7.

I thought about acting differently than I usually do.

Insight (into the problem and assumptions)

8.

I felt inspired to take action on something specific.

9.

I decided to take action on something specific.

10. I had enough time to process information.

11. I gathered information I needed to adopt new ideas or new ways of acting.

12. The most challenging aspects of cycling to transfer to the home context are (top 3 of

(Education/outreach strategies, Communicating about cycling, Policies and governance, Infrastructure

& design, Design guidelines, Investment strategies, Other):

Communication

13. The nature of my conversations with others during Velo-city 2017 were (top 3 of Trivial, Casual, Social,

Strategic, Engaging, Pleasant, Discouraging, Thoughtful, Deeply Reflective)

14. I will most likely share the information (top 3) in these ways (At other conferences, Webinars, Meetings

with colleagues/expert groups, Presentations, Network Groups, Online sources, Audiovisual, Other)

Consensus (about the problem, goals, and strategies)

15. I discussed plans to take action on information gathered at Velo-city.

Development of a shared language

16. It matters to my work that I learn from Dutch examples (compared with from other countries).

17. I plan to share what I’ve gathered from Velo-city 2017 with (top 3 of Colleagues, Professionals in my

network, Politicians = Local/Regional/State/National, Advocacy groups, Civil servants, Friends and

family, I have no plans for sharing information, Other)

Cohesion

18. I felt connected to people in new ways.

Efficiency

19. I expect my Velo-city experience will influence my work in the next year.

20. I expect Dutch cycling principles to influence my work in the next year.

Yerastova

Page 22 of 93

3.5 Actual Usefulness (AU) Metric Methodology

Using the identified framework in Figure 2, AU can be gleaned from the outcome of the learning

process and compared with its perceived usefulness. Understanding the local context into which the

knowledge will be applied is crucial for policy transfer because policymaking is based on local

problem-defining and problem-framing (Wolman & Page, 2002, p. 481). This is crucial for policy relevance and

compromise that would increase the chances of success (Park, Wilding, & Chung, 2014, p. 410).

However, even policy failure can induce social learning and policy change, but only if the political actors

perceive it as so (Skogstad, 2007) by rationalizing that efforts need to be stepped up (Peck, 2011, p.

782). This presents opportunities for non-state actors to re-evaluate their initial beliefs and preferences

(Skogstad, 2007).

As listed in Figure 2, first, ‘novelty’ describes the originality of the idea and its ability to both relate to

the existing paradigm in which it was conceived, as well as its ability to overcome that paradigm. Second,

the idea must be ‘workable’ in that it must be implementable and acceptable legally, socially, and

politically in the context it will be used. Third, the ‘relevance’ of the idea is related to its applicability to the

problem at hand and its effectiveness of solving that problem. And lastly, ‘specificity’ relates to how

complete, explicit, and clear the idea is, as well as the degree of clarity in its communication and

relationship to the expected outcome (p. 302). These metrics also fit well within the ‘bounded rationality’

nature of ways of learning, as they are still bounded by the ‘acceptance’ of the idea which requires an

experience of positive consequences (Schuitema, Steg, & Forward, 2010).

Therefore, as with the PU metric, a similar metric was developed for AU, as shown in Table 2 on the

next page, for Sample A-B. This metric was based on both the outcome framework as shown in Figure 2

and the original survey questions. The metric also guided the development of the interview questions for

both conference and study tour attendees.

Yerastova

Page 23 of 93

3.6 Semi-structured Interview Questionnaire Methodology

The interviews for Sample A-B were necessary in order to make comparisons of learning experiences

between the study tours and conferences, as well as discern their value based on the home context,

which in this case is the North American boundary of the US and Canada. The main questions of the

semi-structured interviews can be found in Appendix of this report, divided into questions for the

conference participants (see Appendix 7.3) and study tour participants (see Appendix 7.4). Altogether, the

questions are poised to reflect the AU metric.

The interviews were all conducted in a semi-structured, open-ended manner, with some questions

being different depending on the interviewee’s background. This allowed for structured questions on the

main topic, but with options to go more in-depth. Various synonyms pertaining to usefulness were used in

the interviews, such as beneficial, valuable, appropriate, convenient, effective, favorable, helpful, and

pragmatic (Source: http://www.thesaurus.com/browse/useful). Other keywords used in the questions that

are found in the original survey that are also reflected in the AU metric that can allude to ‘usefulness’ are

insightful, influencing, and inspiring.

Novelty (Original, Paradigm Relatedness)

1.

I adopted an idea/vision inspired by the information gathered.

2.

I took action on an idea inspired by the information learned in my home context.

3.

The most insightful takeaway from the conference/study tour was:

Workability (Implementability, Acceptability)

4.

What I learned was socially, legally, and politically acceptable.

5.

What I learned was beneficial/valuable/helpful/appropriate/pragmatic/effective.

Relevance (Applicability, Effectiveness)

6.

The Netherlands experience has validated what I learned and made my work more effective.

7.

The knowledge gained improved plans/strategies by making them more novel/relevant/acceptable.

8.

Dutch cycling principles were insightful/beneficial/helpful and have influenced my work.

9.

I was able to overcome challenges/barriers as a result.

Specificity (Completeness, Implicational explicitness, Clarity)

10. The experiential learning experience has changed my view on cycling for improving the quality of life

back in my home context.

11. I was able to adapt what I learned to my home context.

12. I overcame personal beliefs and biases through the learning experience.

13. I can clearly communicate my ideas and strategies as a result of what I learned.

14. I built lasting relationships and/or collaborative partnerships with peers.

Yerastova

Page 24 of 93

Chapter 4

Yerastova

Page 25 of 93

4. Case Study Results

4.1 Engagement with Different Ways of Learning and Methods

As mentioned in the theoretical framework, different “ways of learning” can facilitate ideation and

knowledge transfer through various methods and some may work better than others depending on the

underlying learning approach. As a case study, the conference provided an opportunity to study the

effects of various learning methodology choices, as well as the preferences of adult learners and how

they differ between actors and their professions. In the survey questionnaire, (Sample A) respondents

were asked to identify the frequency with which they engaged with a list of choice methodologies for

learning more about cycling, shown in Table 3 below as Question #14. These methods have been

grouped into the types of learning approaches identified in the theoretical framework.

Table 3. Velo-city Survey: Methodologies for learning (more about cycling)

Educational Learning Question #14 Knowledge Formats Question #7 Conference Platforms Likert Scale Code Question #14

In the last year, how often have you used any of these methods?

Question #7 How well did they match expectations? Academic Courses Professional Courses Master Classes Lectures Master Classes Pecha Kucha Panel Discussions

1 Never in my career Did not participate 2 Not in the last year, but

previously

Much less than expected

3 Once in the last year Less than expected

Organizational/ Institutional Learning Professional Manuals/Books Academic Articles Online Sources (Blogs, Websites) Webinars Audiovisual Material (videos) Network Groups

4 A couple times in the last

year Matched expectations

5 Monthly basis Exceeded

expectations 6 Weekly or more often Greatly exceeded

expectations Experiential/ Social Learning Conferences (other than Velo-city) Study Tours/visits Roundtable Discussions Speed Dating Outdoor Excursions

Yerastova

Page 26 of 93

Table 4. Velo-city Learning Methodology Popularity – Sample A

Table 4 shows the added percentage of respondents answering Likert Scale code numbers 3-6 (listed

in Table 3) in the first column, code numbers 2-6 in the second, and only code number 1 in the third. A

graphic representation of this table is in Figure 3 where the differences are more striking. First, the

methods identified under the Educational Learning Approach are more traditional and formal, such as the

Academic and Professional Courses, as well as Master Classes. Adult learning can be both

transformative, where adults are actively engaged in their learning (Dirkx, 1998), but yet also restrictive,

where the difficulty of the formality of going back to school leaves little desire to meet academic demands

to learn things they may also see little meaning in (Illeris, 2007, p. 208). Table 4 reflects this trend as

many may not have engaged much in formal learning since receiving their degrees.

Second, the methods utilized through an Organizational/Institutional Learning Approach involve less

formal information-seeking practices, such as reading various Professional manuals/books, Academic

Articles, and Online Sources Blogs, or watching Webinars or Audiovisual Materials. Actors may have

incentive and encouragement through their organization’s culture to be pro-active information seekers

(Marsden et al., 2010, p. 9) and many of these sources may owe their popularity to their ubiquity for

disseminating information to a wider global audience, referred to by McCann (2011) as a discourse

analysis as “an attempt to develop a global ethnography” (p. 113). Blogs and other online media can be

transformative on the basis that they form a contemporary transnational advocacy based on the “viral

spread of sustainable transportation best practices” (Montero, 2017b, p. 125-126). Additionally, Network

Groups, as organizations, could be made up of peers or advocates who use digital technologies and

“virtual infrastructure” to effect policy change in “many cities at once”, making them a great method for

Educational Learning

Question #14 Knowledge Formats

% at least once or more in the last year *

% with longer than a year ago added*

% never in their careers* Academic Courses Professional Courses Master Classes 25 67 34 45 75 24 28 59 40 Organizational/ Institutional Learning Professional Manuals/Books Academic Articles

Online Sources (Blogs, Websites) Webinars

Audiovisual Material (videos) Network Groups 79 91 9 81 91 9 97 99 2 46 66 34 86 93 7 81 90 9 Experiential/ Social Learning

Conferences (other than Velo-city) Study Tours/visits

77 91 9

77 91 9

Yerastova

Page 27 of 93

organizational learning (Montero, 2017b, p. 126). The data in Table 4 does agree, but an additional

insight when compared with Figure 3 shows that this could also be due to the sheer frequency because of

the ease with which these learning approaches are accessed and utilized.

Finally, Figure 3 presents more insight on the Experiential Learning Approach through Conferences

and Study Tours, showing they have the highest percentage for the Likert Scale code number 3 (Once in

the last year and 4 (A couple times in the last year). This makes sense given the possible frequency of

attendance. It can be inferred that social learning and participation provided by group environments

support these as popular choices for learning about cycling because they allow informal opportunities for

experiential learning that favor “action-learning techniques” and the ‘in-situ’ experience adult learners

desire (F. D. Davis, 1989; Ettelt et al., 2012; McCann, 2011; Mild & Schlossberg, 2014; Montero, 2017a;

Tiedeman & Knowles, 1979). Planned conferences provide opportunities to bring the “physical, mental,

and social sites of learning together” (Wood, 2016, p. 401) and have the most positive effects when

paired with “active learning opportunities, learning delivered in a longitudinal manner, and when methods

are enabled that facilitate implementation in practice” (D. Davis et al., 1999, p. 873).

Figure 3. Velo-city Learning Methodology Popularity – Sample A

Yerastova

Page 28 of 93

Planned conferences, like Velo-city, offer a variety of learning activities that reflect the overarching

learning approaches, such as workshops, lectures, and other interactive learning programs (D. Davis et

al., 1999, p. 868) to “form common patterns of understanding regarding policy” (Stone, 2004, p. 559).

These activities are found listed in Table 3 under Question #7, as well as below in Figure 4 which reflect

the previously-discussed formal and less formal learning methodologies and support adult learning by

providing interactive and engaging activities, some of which may be more personally meaningful (and

thus insightful) than others (D. Davis et al., 1999; Illeris, 2007). Figure 4 further illustrates these platforms

from which several inferences can be made. First, by adding up the percentages per format by the Likert

Scale code numbers 4-6 (positive answers to expectations) in Table 3, the most popular platforms are

shown to be Lectures, followed by Panel Discussions and Outdoor Sessions. Lectures and Panel

Discussions, as part of the Educational Learning Approach, may be favored because they condense a

wealth of information into a palatable presentation, followed by some time for questions and engagement.

Outdoor Sessions, though third, has the highest percentage of respondents answering that it “greatly

Figure 4. In-conference Platform Popularity – Sample A

Yerastova

Page 29 of 93

exceeded expectations” and with the second highest percentage answering “exceeded expectations”.

This result is in line with the popularity of study tours within the Experiential/Social Learning Approach.

Interestingly, many of the activities reported high levels of “did not participate” responses, which can

be a result of too many sessions programmed simultaneously into the conference’s schedule rather than

simply disinterest. Additionally, it is interesting to note that Panel Discussions came in at 11% for not

matching expectations (addition of 2 and 3 on the Likert Scale), though ranking second favorite. While

they include a panelist discussion group that engages both each other’s experiences, as well as that of

the audience, perhaps they weren’t as much of an engaging active learning environment as desired. This

is in contrast with Lectures, which reported the lowest “did not participate” response rate, and this may

indicate its popularity is due to its sheer familiarity as a learning platform.

Furthermore, it is interesting to examine the demographic differences between conference

participants. Figure 5 shows that the majority of all respondents (Sample A) were Civil Servants, followed

by Consultants/Advisors and Advocates. However, in contrast, Sample A-A shows that specifically of the

US and Canada demographic, academics made up 34% of the respondents, followed by a more even

distribution of the other professions, which infers a population sample bias. Unfortunately, the 22 total

survey respondents who noted a US/Canada origin are hardly a representative population sample.

Nonetheless, it illustrates some interesting differences and similarities between Sample A-A and Sample

A as noted next.

Yerastova

Page 30 of 93

Despite the higher academic prevalence, almost half of the Sample A-A respondents on Figure 6 claim

to have never taken Academic Courses and more than half have never done a Master Class. Additionally,

it appears that all have engaged with Online Sources and Conferences for information on cycling. Though

the sample size is much smaller and much less representative, it still more or less follows a similar pattern

to the Sample A responses. There seems to also be more engagement with Academic articles and

Professional manuals/books, highlighting greater engagement with the Organizational/Institutional

Learning Approach.

Additionally, in contrast with Sample A in Figure 3, while Conferences still rank highest, Study Tours

do not. Possible explanations for this could be that the US/Canada demographic treats conferences

essentially as study tours or simply doesn’t go on many study tours, as indicated by the 29% response

rate under “Never in my career”.

Likewise, the numbers for Sample A-A on Figure 7 follow the same procedure as for Figure 4, and

seem to follow a similar trend, but with greater exaggeration due to the smaller sample size. It shows

Lectures, Panel Discussions, and Outdoor Sessions as still the top 3 favorite learning platforms at the

Figure 6. Velo-city Learning Methodology Popularity – Sample A-A

Yerastova

Page 31 of 93

conference and shows Outdoor Sessions still receiving the highest percentage answering that it “greatly

exceeded expectations”. Speed Dating had the exaggeratingly lowest participation.

Finally, breaking down Figure 3 further by showing respondents’ choices in relation to their professions

presents a more in-depth view on actor preferences. The four professions shown on Figure 8 on the next

page also reflect the main demographic split of Figure 5 for Sample A. While the numbers run fairly

evenly and follow predictable expectations, there are also subtle differences. Consultants put the highest

value on the Network Groups knowledge platform, as well as on Online Sources and attending

Conferences. This may be due to having more flexibility on the private side for learning than on the public

side and there is more incentive to stay abreast of the latest developments in the field. Academic (staff)

had the lowest continuation of Academic and Professional Coursework, but also the lowest response to

“Never in my career”. It could be inferred that academic staff are at the top of their professions within

academia in their fields and/or are continuously being engaged in research themselves. They may also

be more skeptical and critical of Online Sources than the other professions and thus also engage with

Figure 7. In-conference Platform Popularity – Sample A-A

Yerastova

Page 32 of 93

more Academic Articles (in addition to their use in research) that are peer-reviewed and vetted.

Additionally, Advocates answer the highest “Never in my career” to methods for learning about cycling,

such as Academic Courses and Professional Courses, which makes sense if they never formally studied

Figure 8. Demographic Snapshots of Learning Methodology Popularity – Sample A

Yerastova

Page 33 of 93

the subject but got involved through other means.

An analysis was also done on actor preferences Sample A-A, but the sample size is far too small to

yield results this in detail to make any meaningful inferences from but they are included below in Figure 9.

All numbers that were zeros have been removed for an easier visual.

*Numbers represent how many of each profession responded to each knowledge platform

Yerastova

Page 34 of 93

4.2 Perceived Usefulness (PU)

As mentioned in the theoretical framework and established by the framework in Chapter 3.3, Table 5

on the following page summarizes the results of the metric for Sample A-A using associations from the

Likert Scale measured by positive answers (Likert Scale 4-6). As the metrics use the same survey

questions, some of the results are already known, such as that Lectures, Panel Discussions, and Outdoor

Excursions rank the highest in matching or exceeding expectations. Other questions are divided into two

categories where the first shows the percentage answering the ambiguous scale of “Somewhat” and the

second showing the sum of the stronger responses of “Quite a lot” and “Very much”. As the interpretation

of “Somewhat” may border on a more neutral or unconfident position, it has been separated to highlight

this and quite a few of the metrics show 50% or more for this answer (4, 5, 6, 10, 11, 15, 16, 19, 20).

These unconfident responses may be due to the strong wording of the survey questions/conclusions

reflected in the metric and may also reflect a need to unpack and divide this scale in two. Additionally,

since many conferences tend to draw the same traveling audience, many may have confirmation bias so

few may attend who would truly have a transformative experience, supported by such conflicted answers.

Two of the metrics don’t break the 50% threshold at all (7, 9) which also reflect not having strong

experiences. The metrics that do respond highly positively (2, 3, 8, 18) reflect insight, inspiration, and

feelings rather than a changed outlook. This would support the notion that the attendees may have

confirmation bias but were looking more for new information to support what they already know, rather

than an “aha” moment. The preference for interaction and sharing knowledge with colleagues and

network groups (14, 17) is also supported by the social conversational environment (13) as attendees

look to find this new information they seek to overcome challenges they identified (12).

Additionally, the mean of each metric is included for insight into how the “Not at all” answers alter the

metric. For example, though 76% of respondents reported that the bike tours (3) were Quite a lot or Very

insightful, the mean is 2.9 (out of the 1-4 Likert Scale) because there were less respondents, as

compared with a mean of 3.4 for inside sessions (2) with 22 respondents. Furthermore, two

interpretations of the mean have been supplied for the first metric because of the way the “Did not

participate” response (reported in the Mean as Ex(1) as “excluding the Likert Scale 1”) can alter the small

sample being evaluated. There was only one reponse for Speed Dating other than “Did not participate”,

and so the mean can show that the learning platform was either very useful or not at all useful.

Overall, it can be interpreted that on a Likert Scale of 1-4, a mean above 2 could be considered useful

and above 2.5 for the Likert Scale of 1-5. Based on this, it seems metrics 5, 6, 7 show little use to the

conference attendees, and going by the inclusive mean number, Speed Dating, Pecha Kucha, and

Master Classes also ranked poorly for usefulness.

Yerastova

Page 35 of 93

Table 5. PU Metric for Sample A-A

Reaction

1 *Likert Scale: (1) Did not, (2) Much less than, (3) Less than, (4) Matched, (5) Exceeded, (6) Greatly

Lectures Roundtable Discussions Panel Discussions Speed Dating Pecha Kucha Master Classes Outdoor Excursions Likert 4-6 21 (n=22) Likert 4-6 11 (n=21) Likert 4-6 15 (n=22) Likert 4-6 1 (n=21) Likert 4-6 3 (n=21) Likert 4-6 5 (n=22) Likert 4-6 11 (n=20) 95% 52% 68% 5% 14% 23% 55% Mean: 4.2 Ex(1): 4.3 Mean: 2.9 Ex(1): 4.5 Mean: 3.7 Ex(1): 4.3 Mean: 1.2 Ex(1): 5 Mean: 1.6 Ex(1): 4.3 Mean: 1.8 Ex(1): 4.6 Mean: 3.3 Ex(1): 4.5

2 *Likert Scale: (1) Not at all, (2) Somewhat insightful, (3) Quite insightful, (4) Very insightful

Likert 2: 8 (n=22) 36% Likert 3-4: 14 (n=22) 64% Mean: 3.4

3 Likert 2: 4 (n=17) 24% Likert 3-4: 13 (n=17) 76% Mean: 2.9

4 *Likert Scale: (1) Not at all, (2) Somewhat, (3) Quite a lot, (4) Very much

Likert 2: 12 (n=22) 55% Likert 3-4: 5 (n=22) 23% Mean: 2.1

5 Likert 2: 11 (n=22) 50% Likert 3-4: 3 (n=22) 14% Mean: 1.8

6 Likert 2: 12 (n=22 55% Likert 3-4: 2 (n=22) 9% Mean: 1.7

7 Likert 2: 8 (n=21) 38% Likert 3-4: 2 (n=21) 10% Mean: 1.6

Insight 8 Likert 2: 8 (n=21) 38% Likert 3-4: 12 (n=21) 57% Mean: 2.8

9 Likert 2: 8 (n=21) 38% Likert 3-4: 9 (n=21) 43% Mean: 2.5

10 Likert 2: 14 (n=21) 67% Likert 3-4: 5 (n=21) 24% Mean: 2.5 11 Likert 2: 13 (n=22) 59% Likert 3-4: 8 (n=22) 37% Mean: 2.3

12 First Choice (n=15) Second Choice (n=12) Third Choice (n=12)

Infrastructure & design 3 / 20% 3 / 20% 0 / 0%

Investment strategies 3 / 20% 4 / 33% 2 / 17%

Policies & governance 6 / 40% 1 / 8% 2 / 17%

Communicating about cycling 1 / 7% 1 / 7% 1 / 8%

Design guidelines 2 / 13% 0 / 0% 2 / 17% Education/outreach 0 / 0% 2 / 17% 3 / 25% Other 0 / 0% 1 / 8% 2 / 17% Communication 13 (n=17) (n=14) (n=12) Trivial 0 / 0% 0 / 0% 0 / 0% Casual 4 / 24% 1 / 7% 0 / 0% Social 5 / 29% 1 / 7% 2 / 17% Strategic 1 / 6% 2 / 14% 2 / 17% Engaging 5 / 29% 4 / 29% 2 / 17% Pleasant 1 / 6% 2 / 14% 2 / 17% Discouraging 0 / 0% 0 / 0% 0 / 0% Thoughtful 1 / 6% 2 / 14% 4 / 33% Deeply reflective 0 / 0% 2 / 14% 0 / 0% 14 (n=15) (n=11) (n=10) At other conferences 0 / 0% 0 / 0% 2 / 20% Webinars 0 / 0% 0 / 0% 0 / 0%

Meetings with colleagues 11 / 73% 2 / 18% 0 / 0%

Meetings with expert groups 0 / 0% 1 / 9% 1 / 10%

Presentations 2 / 13% 3 / 27% 2 / 20%

Network groups 1 / 7% 4 / 36% 3 / 30%

Online sources 1 / 7% 1 / 9% 1 / 10%

Audiovisual 0 / 0% 0 / 0% 0 / 0%