“Change the Changeable” Framework

for Implementation Research in Health

Mikiko Oono

1,2

, Yoshifumi Nishida

1,2

, Koji Kitamura

1,2

, Asako Kanezaki

and Tatsuhiro Yamanaka

3,2

Artificial Intelligence Research Center, National Institute of Advanced Industrial Science and Technology, Tokyo, Japan

Safe Kids Japan, Tokyo, Japan

Ryokuen Children’s Clinic, Kanagawa, Japan

Keywords: Health Education, Active Learning, Participation, Injury Prevention.

Abstract: In recent years, a “participatory” system of community members and researchers has become a key factor to

obtaining the best outcomes to create a better world. The authors previously suggested the problem structure

change theory to find how to address social issues. In this paper, we suggest the “change the changeable

framework” with three approaches for community-based program implementation: 1) Share the value of the

change the changeable, 2) Systematize the process of a problem structure change, and 3) Build a sustainable

system and discuss a case study based on the framework. In the case study, we conducted a safety education

curriculum for injury prevention that consisted of what is injury prevention, playground safety, indoor safety,

and photovoice project. We introduce an injury surveillance system and a risk recognition system as new tools

for the collaboration of human intelligence and artificial intelligence.

1 INTRODUCTION

In recent years, a “participatory” system of

community members and researchers has become a

key factor to obtaining the best outcomes to create a

better world. For example, in the field of service

design, designers have taken a co-design or co-

creation approach to develop new products based on

people’s needs. Co-design is defined as the creativity

of designers and non-designers working together

from the beginning of the development process

(Sanders, 2008). In the field of engineering, a living

lab is a new research approach for making

innovations happen. In living labs, users are involved

in developing, creating, prototyping, validating, and

testing new products, services, or technologies in a

real-life setting (Schuurman, 2011, 2016). In the

public health field, the community-based

participatory research (CBPR) approach has been

widely used to eliminate health disparities. In CBPR,

community members affected by a health issue are

actively involved as a partner in all phases of research

(Minkler, 2004; Wallerstein, 2010).

Co-design, living labs, and CBPR might seem

unrelated, but they use the same approach to reach

their goals. To the authors, it seems that a boundary

between professionals and nonprofessionals has

become unclear in many fields. The key is that all

powers, which include knowledge, skills, and life

experiences that people have regardless of their

expertise, should be recognized and wisely used to

make innovations happen or to find how to address

social issues. In this study, we defined a participatory

system as a group of assets that functions together to

achieve particular goals. The examples of these assets

are not only people who are multidisciplinary

professionals and nonprofessionals, including parents

and children, but also community resources such as

hospitals, schools, and daycare centers.

The authors previously suggested the problem

structure change theory to find how to address social

issues, as shown in Figure 1 (Nishida, 2017). This

theory has three variables: variable A, which is what

we want to change, variable B, which is what we can

change, and variable C, which is what we cannot

change or is difficult to change. Social problems often

seem rigid and unchangeable because they are

discussed based on the relation between variable A

and variable C. By adding operational variable B

(changeable things) to a problem structure, the

structure is transformed to one that can be changed.

Oono, M., Nishida, Y., Kitamura, K., Kanezaki, A. and Yamanaka, T.

“Change the Changeable” Framework for Implementation Research in Health.

DOI: 10.5220/0006691303610368

In Proceedings of the 10th International Conference on Computer Supported Education (CSEDU 2018), pages 361-368

ISBN: 978-989-758-291-2

361

In this paper, we apply this structure to a particular

social issue, which is improving people’s health by

preventing injuries.

Figure 1: Problem structure change theory.

From the perspective of the problem structure

change theory, co-design, living lab, and CBPR fields

face unique limitations for addressing social issues.

In the field of co-design and living labs, B variables

are often already fixed in the first place because most

professional people are knowledgeable about the

latest technologies. Their primary interest is to

implement such technologies, but they often lack an

impact evaluation of how a society changes after the

technology is implemented in a community. On the

other hand, in the field of CBPR, public health

professionals always start with health issues, which

are the A variables (things to be changed), and they

try to translate what science says about appropriate

behaviours through education. In comparison with

engineers, health professionals are generally not

eager to apply new technologies, such as artificial

intelligence (AI) and the Internet of Things (IoT), and

they tend to practice health education in the

traditional way. It seems that researchers in co-design,

living lab, and CBPR fields have suffered from the A-

C structure on the left side of Figure 1 to obtain

ultimate outcomes.

To change the A-C structure, it is necessary 1) to

identify all the possible B variables and to understand

a precise problem structure by collaborating with

people across many different fields, and more

importantly, 2) to gain the power of B identification

by community members so that they can sustain a

healthy community by themselves.

Recently, information communication technology

(ICT) in education has rapidly become popular and

advanced. Rodriguez et al. used ICT as a tool for

elementary school education and implemented a

system to create a school network at the national level

in Panama (Rodriguez, 2009). Terton and White

developed a computer-based educational game for

children to promote physical and social engagement

and discussed the benefits of computer-based

educational gaming (Terton, 2014). In addition,

advanced technologies including IoT and AI are

currently available for community-based programs.

Those technologies can be used for changing A-C

structure and empowering community members.

In this paper, we first propose the “change the

changeable” framework by expanding the concept of

“participatory”. We present a children’s health

education project based on the framework. Then, we

discuss how a health problem structure can be

changed from the viewpoint of a sustainable school

system. Finally, we introduce a risk recognition

system that we developed as a future health education

tool.

2 “CHANGE THE

CHANGEABLE” FRAMEWORK

The three-step approach presented in this section

describes how to establish a strong community health

program in one’s community. The conceptual

framework of change the changeable is shown in

Figure 2.

The three steps are as follows:

1. Share the value of the change the changeable

framework with community members.

2. Systematize the process of a problem structure

change.

3. Build a sustainable system.

2.1 Step 1: Share the Value of Change

the Changeable with Community

Members

To change the A-C structure, it is necessary to

identify variable A as a target and possible B

variables through co-designing with community

members and other collaborators from various fields.

In public health, the following are critical for program

implementation: engaging all stakeholders,

developing a trusting relationship, and more

importantly, realizing that all stakeholders have their

own B’s.

In this step, the content-driven approach is key.

Most researchers are very likely to start with data

collection because the data indicate the direction of

their research. However, it is not easy for community

members to experience the value of data, and data

collection is usually a huge burden for a community.

CSEDU 2018 - 10th International Conference on Computer Supported Education

362

Figure 2: The change the changeable framework using the 1) content-driver approach, 2) kludge approach, and 3) technology-

use approaches.

Therefore, a relatively easy and useful activity,

such as a one-hour workshop or a health education

class, should first be presented so that community

members can learn visible positive outcomes in return

for accepting a small change. Then, they realize

research participation is valuable.

2.2 Step 2: Systematize the Process of

Problem Structure Change

When community members see positive changes for

a particular issue and are ready to make other changes,

it is important to keep and sustain their ability to

change. For example, if community members decide

to implement a health education program at school,

many questions arise: e.g., who provides the classes,

when the program is offered, and who is in charge.

Therefore, step 2 is finding a way to systematize the

process of making changes while developing a

trusting relationship.

Creating a new system from no system or from an

existing system as an independent system is usually

difficult for a community. To overcome such a barrier,

the authors suggest the kludge approach. According

to the dictionary (Dictionary.com, 2017), kludge is

defined as “a software or hardware configuration that,

while inelegant, inefficient, clumsy, or patched

together, succeeds in solving a specific problem or

performing a particular task”. This tells us that it is

important to realize that addressing an issue is a

priority even if the performance is awkward.

Researchers need to find a system that is already

working in a community and expand that system.

Expanding a system could be easier than embedding

a new system in a community.

2.3 Step 3: Build a Sustainable System

When researchers successfully systematize the new

system by expanding the originally existing system, it

is still necessary to strengthen it. To do so,

technologies play a huge role because they can reduce

the task load and maximize the efficiency for work

(the technology-use approach). In reality, however,

many people tend to stay away from new

technologies rather than accept them. If people

understood that technologies can accelerate their

efforts to reach their goal, then they would be more

likely to accept them.

“Change the Changeable” Framework for Implementation Research in Health

363

3 CASE STUDY

In this section, we discuss a case study of injury

prevention education for children and its outcomes,

and then reflect it based on the three steps of the

framework.

3.1 Project Overview: Injury

Prevention Education Curriculum

A comprehensive safety education curriculum was

developed, implemented, and evaluated for 5

graders of Fujimidai Elementary School in Tokyo.

This school was certified as the International Safe

School (ISS) in 2016. ISS is a school that carries out

activities to promote school safety and is awarded by

the World Health Organization Collaborating Centre

on Community safety Promotion at the Karolinska

Institutet, Stockholm, Sweden (International Safe

Schools Certifying Centers, 2017). The school’s

teachers, students, and community members

collaborate each other to improve school environment

by identifying B variables and changing the

changeable. The safety education curriculum we

developed consisted of injury prevention

explanations, playground safety, indoor safety, a

photovoice project, and preventive action. Photovoice

is a qualitative community-based participatory

research method using photographs to reflect a

community’s strengths and concerns (Wang, 1997).

From the perspective of problem structure change

theory, B variables can be classified into three

categories, known as the 3E’s: Enforcement,

Environmental modification, and Education. To make

these traditional three E’s of injury prevention more

suitable for school children, we modified them to the

children’s three C’s: Creating rules, Changing the

environment, and Communicating to friends and

family. In the photovoice project, students were asked

to take pictures that showed risks as well as factors

promoting school safety. Then, they wrote how to

reduce the risk and to increase safety based on the

3C’s.

3.2 Health Education Outcomes

Thirty-eight students took the curriculum. Students

presented 27 pictures from the viewpoints of the 3C’s.

As one example of promoting factors, a student took

a picture of a counseling room and said, “The state of

our mind affects injury risk. I feel safe when I consult

a school counselor.” For a risk factor, students

mentioned the risk of a fall by going up the stairs by

skipping every other step and the risk of a crossing

collision in front of a restroom when students run.

Examples of photovoice pictures are shown in Figure

Students also developed a photo R-map (Figure

4). An R-map is a 4×4 matrix that shows risk levels

divided into two levels of frequency of occurrence on

the vertical axis and two levels of injury severity on

the horizontal axis. The R-map is widely used in

engineering fields for risk assessment. Students

placed their pictures based on their opinions about

injury frequency and severity. Sometimes they

relocated the pictures based on class discussion.

Figure 3: Examples of photovoice.

Figure 4: Photo R-map.

During the preventive action time, students

divided into three groups, one for each of the three

C’s, and discussed what they could do to improve

school safety. The first C group created new rules.

The second C group cleaned their classroom in places

where students might trip over an obstacle. Moreover,

for a place that was difficult for the students

CSEDU 2018 - 10th International Conference on Computer Supported Education

364

themselves to modify, they wrote a request letter that

stated where to modify and the reasons behind the

requests. This letter was given to the school principal.

The third C group created educational posters, and

some of them wrote a manuscript for a school

broadcast to promote safety behaviours schoolwide.

To evaluate the effectiveness of the project, we

conducted a survey asking students if they gained new

knowledge of injury prevention and if the curriculum

raised their willingness to learn more about injury

prevention. The willingness was asked on a scale of

one (low) to ten (high). The Wilcoxon signed-rank

test was used to determine its effectiveness.

The evaluation survey results showed that 91% of

the students said that they gained new knowledge of

injury prevention. As examples of new knowledge,

students stated, “One’s state of mind and

environmental factors cause injury, much more than

what I expected” and “To prevent injuries, it’s critical

to consider one’s behaviours and environmental

conditions around us.” In addition, we found that the

willingness to learn was significantly increased from

6.24 on average to 7.56 (p = .001) (Figure 5).

Figure 5: Changes in students’ willingness to learn.

As a result of sending a request letter to the

principle, a tile floor at the entrance that caused slips

during rainy days was changed to non-slip floor tiles.

3.3 Change the Changeable

Perspectives of the Project

3.3.1 Sharing the Value of Change the

Changeable by using the

Content-Driven Approach

When this project had just started, we recommended

the school to start injury data collection to identify the

school’s injury risk. However, as discussed in

subsection 2.1, it just did not work out. Then we

learned that school teachers were very interested in

science-based tools for injury prevention education

because their expertise is not an injury prevention,

and it is hard for them to gather evidenced-based

injury prevention messages for students. Thus, we

developed educational materials by using our

expertise gave the materials to the teachers. A

snapshot of the materials is shown in Figure 6. By

collaborating with teachers to implement an effective

injury prevention class, we successfully built a

trusting relationship with the school teachers.

Figure 6: Snapshot of the class materials.

3.3.2 Systematization of a Problem

Structure Change by using the Kludge

Approach

When we discussed how to start a health education

curriculum, we found out that a unit for injury

prevention was required as a part of the health and

physical education (PE) curriculum for 5

graders.

Thus, we integrated the curriculum into the existing

PE class and taught it over a five-week period. The

PE teacher was in charge of playground safety, indoor

safety, and a part of the photovoice project, and we

took responsibility for an explanation of injury

prevention, the photovoice presentation, and the

“Change the Changeable” Framework for Implementation Research in Health

365

prevention action time. By integrating our education

system into the existing school curriculum system,

our developed injury prevention course has become a

required unit for 5

graders at Fujimidai Elementary

School.

3.3.3 Building the Sustainable System using

the Technology-Use Approach

One of the huge successes during collaborations for

the injury prevention class was that school teachers

started to understand the meaning of data and to want

to collect injury data at school. As a means of

developing effective and convincing class material

for students, the values of the data outweighed the

time and effort of data collection. This is where

technologies came in. We developed an injury

surveillance system for schools. The system is

capable of collecting, aggregating, analysing, and

searching data of injuries. The system also has a

function called “Body-graphic Information System”

(“BIS”) that enables us to express, collect, retrieve,

and analyse external injury geometric data (Figure 7).

Figure 7: Injury surveillance software.

The injury education project has been conducted

since 2013, and Fujimidai Elementary School decided

to use the surveillance system at the end of their

school fiscal year. Between April 2014 and March

2017, 704 instances of injury were collected from the

school. As of April 3

, 2017, the total number of

students enrolled in Fujimidai Elementary School was

302 children. As shown in Figure 8, the number one

cause of injury was falls (245 cases) followed by

collisions (202 cases), cuts (60 cases), and pinching

(34 cases).

Figure 9 shows the body parts that were injured

due to falls. A red area indicates a high frequency of

injury. The analysed data were integrated into the

health education class and helped students recognize

their own school’s problems.

Figure 8: Types of injury.

Figure 9: Fall injury analysis using BIS.

Moreover, we developed an injury recognition

system by using pictures of the photovoice project.

The system is based on the convolutional neural

network (CNN)-based model “RotationNet”

(Kanezaki, 2016; Savva, 2017). The system takes

multi-view images of an object as input and estimates

both the object’s pose and category. The method

treats the pose labels as latent variables, which are

optimized to self-align in an unsupervised manner

during training with an unaligned dataset.

Figure 10 shows the result of stair recognition.

The system automatically extracts view features (top

photo), and then chooses a specific viewpoint based

on the scores of the predicted category. Once the

system recognizes an object, it indicates possible

injuries that commonly occur in and around the

recognized object (bottom photo). Students used this

recognition system to discuss what preventive actions

can be taken to avoid injuries.

CSEDU 2018 - 10th International Conference on Computer Supported Education

366

Figure 10: Output of the risk recognition system.

4 DISCUSSION

Change the changeable framework suggested a way

of bridging the gap between community members and

researchers. Using three steps with three approaches

is critical to establishing a sustainable system. A

sustainable system combines the power that

community members have and the knowledge and

skills from external resources, and does not proceed

with the project based only on the researchers’

intentions. As in the definition of kludge, even though

researchers think that the way of proceeding on the

project may seem inefficient, ineffective, or clumsy,

whatever action acceptable to community members is

most important because this is how community

members keep participating in a project.

Moreover, carrying out the project according to

the community members’ desires helps them to enjoy

changing the changeable, and this is an invaluable

outcome for community-based programs. As

emphasized in subsection 2.1, starting with data

collection to develop a community-based

participatory system usually does not work.

Researchers need to convince community members

that starting a project is actually valuable to them even

though they need to undertake some tasks. So far, the

authors have successfully implemented community-

based health education programs at five schools,

including a middle school, by using the change the

changeable framework. In addition, community-wide

injury surveillances were started in two cities.

Our developed injury surveillance system and risk

recognition system helps people learn the injury risks

around them. Injury data collected from the

surveillance system (BIS) are integrated into class

material development to promote student learning

after the system was installed. Understanding injuries

occurred in their school using BIS and predicting

plausible injuries using the AI-based recognition

system brings home the fact that injuries can actually

occur to them. Injury data localization based on the

technologies is powerful to heighten the perception of

injury susceptibility. We developed these systems to

promote interactions between human intelligence and

artificial intelligence. Utilization of these systems is

one example of future injury education tools that will

connect people and people, communities and

communities, people and knowledge, and problems

and solutions.

5 CONCLUSIONS

We suggested the change the changeable framework

and discussed a health education program based on

the framework. Our previous experience of

implementing a community-based health program

showed three useful approaches for program

implementation: 1) content-driven approach, 2)

kludge approach, and 3) technology-use approach.

We conducted a case study of an injury prevention

education project that consisted of classroom lectures

and a photovoice project. During the photovoice

presentation class, students presented 27 pictures

from the viewpoints of the children’s 3C’s (Creating

rules, Changing the environment, and

Communicating to friends and family). They also

created photo R-maps using their photovoice pictures.

The project’s evaluation survey indicated that the

students’ willingness to learn about injury prevention

was significantly increased.

We hope that the change the changeable

framework will be applied to various community-

based programs and expand our knowledge of

implementation science.

ACKNOWLEDGEMENTS

This research is supported by the Strategic Basic

Research Program (CREST) from Japan Science and

Technology Agency, JST.

“Change the Changeable” Framework for Implementation Research in Health

367

REFERENCES

Dictionary.com., 2017. Definitions: kledge (Accessed 23

October 2017). Available from:

http://www.dictionary.com/browse/kludge.

International Safe Schools Certifying Centers., 2017.

History (Accessed 25 December 2017). Available from:

http://internationalsafeschool.com/history.html.

Kanezaki A., Matsushita Y., Nishida Y., 2016. Rotationnet:

Joint learning of object classification and viewpoint

estimation using un-aligned 3D object dataset. arXiv

preprint arXiv:1603.06208.

Minkler, M., 2004. Ethical challenges for the “outside”

research in community-based participatory research.

Health education & behavior. Vol.31, No.6, pp.684-

697.

Nishida, Y., Kitamura, K., Oono, M., Tamanaka, T., 2017.

Smart transfer of social problem into industry by linking

living data center with social function library: Case

study of toothbrush injury prevention. In Proc. Of 3

IEEE annual international smart cities conference

(ISC2 2017).

Rodríguez, MA., Barrios, JDC, Schultz, ES., 2009. The use

of an innovation classroom: a perspective in the

introduction of ICT in elementary schools, In

Proceedings of the First International Conference on

Computer Supported Education, Vol.1, pp.173-180.

Sanders E.B.N., Stappers, P.J., 2008. Co-creation and new

landscapes of design. CoDesign, Vol.4, No.1, pp.8-19.

Savva, M., Yu, F., Su, H., Kanezaki, A., Furuya, T.,

Ohbuchi, R., et al., 2017. SHREC’17 Track Large-scale

3D share retrieval from ShapeNet Core55. (Accessed 23

October 2017). Available from:

https://shapenet.cs.stanford.edu/shrec17/shrec17shapen

et.pdf.

Schuurman, D., De Marez, L., Ballon, P., 2016. The impact

of living labs methodology on open innovation

contributions and outcomes. Technology Innovation

Management Review. Vol.6, No.1, pp.7-16.

Schuurman D. De Moor, K., De Marez, L., Evens T., 2011.

A living lab research approach for mobile TV.

Telematics and Informatics Vol.28, No.4, pp.271-282.

Terton, U., White, I., 2014. A computer-based educational

adventure challenging children to interact with the

natural environment through physical exploration and

experimentation. CSEDU 2014 In Proceedings of the 6

International Conference on Computer Supported

Education, Vol.3, pp.93-98.

Wallerstein, N., Duran, B., 2010. Community-based

participatory research contributions to intervention

research: the intersection of science and practice to

improve health equity. American Journal of public

health Vol.100 Suppl.1, pp.S40-46.

Wang, C. ， Burris, M.A., 1997. Photovoice: concept,

methodology, use for participatory needs assessment.

Health education & behaviors, Vol.24, No.3, pp.369-

387.

CSEDU 2018 - 10th International Conference on Computer Supported Education

368