Supporting Decision-making Activities

in Multi-Surface Learning Environments

Lili Tong

, Audrey Serna

, Sébastien George

and Aurélien Tabard

Univ Lyon, CNRS, INSA-Lyon, LIRIS, UMR5205, F-69621, Villeurbanne, France

UBL, Université du Maine, EA 4023, LIUM, 72085 Le Mans, France

Univ Lyon, CNRS, Université Lyon 1, LIRIS, UMR5205, F-69622, Villeurbanne, France

Keywords:

Decision-making, collaboration, Computer Supported Collaborative Learning (CSCL), Computer Supported

Cooperative Work (CSCW), Multi-Surface Environments (MSE).

Abstract:

Collaborative decision-making is one of the emphasized student skills required by educators from different

domains including science, technology and society. Multi-surface environments (MSE) appear particularly

well suited for such learning activities, with large shared surfaces dedicated to the overview of information

and context awareness, while personal surfaces serve browsing and analytical purposes. In this study, we

present the design of Pickit, a MSE tool supporting collaborative decision-making activities. We show how

Pickit is used by four groups of high school students as part of a learning activity. We analyze students’

interactions with digital devices that are related to given phases of the decision making process. Our results

show that MSE are particularly interesting for such learning activities as they enable to balance personal and

group work. The introduction of personal devices (tablets) makes free riding more difﬁcult, while enabling

development of personal judgment. By using Pickit, students successfully made their decisions and meanwhile

better knew about the decision-making process.

1 INTRODUCTION

Collaborative decision-making broadly consists of ex-

ploring alternative options, deﬁning criteria, weight-

ing options, and choosing one among them. Educa-

tors have highlighted the importance of collaborative

decision-making skills in scientiﬁc education (Rat-

cliffe, 1997; Grace, 2009; Evagorou et al., 2012), as

it can help students to appropriating knowledge. Col-

laborative decision-making becomes even more rele-

vant in multi-disciplinary education, such as when ed-

ucators tackle socio-scientiﬁc issues (Evagorou et al.,

2012), e.g., sustainable development. These issues

are often ill-structured, and always involve consider-

ation for technical, economical, and ethical aspects.

However, the literature (Ratcliffe, 1997; Grace,

2009; Evagorou et al., 2012) emphasize that students

struggle with such complex issues, especially with

multi-dimensional analyses. The underlying prob-

lems lie in difﬁculties to: become familiar with the

material, evaluate choices and reach a decision to-

gether, follow a process, know what others are doing

and adjust behavior accordingly.

In the meantime, tabletops have demonstrated

their ability to support collaborative activities (Shen

et al., 2003). They can lead to greater awareness of

group activities (Rogers and Lindley, 2004). They

afford the introduction of external regulation mecha-

nisms (DiMicco et al., 2004) through the design of ap-

plications or visual indicators of participation. Multi-

surfaces environments (MSE) seem suited to support

decision-making and problem-solving activities (Dil-

lenbourg and Evans, 2011). Large shared displays

are excellent at providing an overview of informa-

tion, while personal devices support individual activi-

ties and enable participants to conduct analytical tasks

in parallel.

Our focus in this article is to better understand

how digital devices and MSE can support learning

decision-making process. We co-designed Pickit with

teachers, a learning environment enabling students to

explore various locations on a map on a shared dis-

play and decide which is the most appropriate regard-

ing several criteria. Students can use personal devices

to analyze the characteristic of each location. We con-

ducted an in-situ study of Pickit with 12 students from

a vocational school.

Our study shows that MSE is well suited for

Tong, L., Serna, A., George, S. and Tabard, A.

Supporting Decision-making Activities in Multi-Surface Learning Environments.

DOI: 10.5220/0006313200700081

In Proceedings of the 9th International Conference on Computer Supported Education (CSEDU 2017) - Volume 1, pages 70-81

ISBN: 978-989-758-239-4

tightly-coupled collaboration and loosely-coupled

parallel work in the decision-making process. Groups

reached motivated decisions for their choices, while

broadly following the decision-making process we

implemented in our application. We outline the de-

vice use in MSE and link them to the behaviors in the

decision-making process. We further derive a set of

considerations for designing decision-making activi-

ties using MSE in classrooms.

2 RELATED WORK

2.1 Decision-making Process in

Education

Decision-making is a process of gathering informa-

tion, identifying and weighing alternatives, and se-

lecting among various alternatives based on value

judgments (Uskola et al., 2010). Several models of

decision-making processes have been introduced in

the learning literature over the past years (Janis and

Mann, 1977; Aikenhead, 1985; Kraemer and King,

1988; Ratcliffe, 1997; Roberto, 2009). These models

differ mostly in how they scope the decision-making

process. They deﬁne however similar stages and all

underline the non-linear nature of the process. The

most generic model, proposed by (Ratcliffe, 1997),

draws upon common elements in normative and de-

scriptive decision-making models (Janis and Mann,

1977). This decision-making process consists of 6

stages that can be intertwined: 1) Listing options;

2) Identifying criteria; 3) Clarifying information; 4)

Evaluating the advantages and disadvantages of each

option according to the criteria previously identiﬁed;

5) Choosing an option based on the analysis under-

taken; 6) Evaluating the decision-making process, and

identifying any possible improvements.

This process of critical analysis of information is

widely used in basic scientiﬁc education and is well

suited to the co-construction of knowledge (Hong

and Chang, 2004; Grace, 2009; Evagorou et al.,

2012). Making complex decisions in science courses

involves many aspects of critical thinking, such as

understanding procedures for rational analysis of a

problem, gathering and using of available informa-

tion, clarifying the concerns and values raised by the

issues, or coordinating hypotheses and evidence (Rat-

cliffe, 1997; Siegel, 2006). Students evaluate the rel-

evance and relativity of evidence and make choices

by considering and respecting different viewpoints.

Whenever a group makes a collective decision, each

member of the group should reach the same decision

individually (Deneubourg and Goss, 1989). Students

should develop their own judgment and understand-

ing on the problem, and exchange opinions with the

group.

To improve students’ ability to make decisions, it

is necessary not only to focus on the result of their

discussions but also on how the students carry out the

decision-making process, such as how they are able

to evaluate and take into account the available infor-

mation individually (Uskola et al., 2010). Learning

environments supporting decision-making processes

should enable both collective and individual activi-

ties, including exploring and analyzing data, model-

ing, voting, or analyzing decisions.

2.2 Collaborative Learning with

Multi-surface Environments

Tabletops have been widely recognized as well suited

to co-located collaborative activities (Shen et al.,

2003). They provide shared spaces for organizing and

controlling simultaneously information (Dillenbourg

and Evans, 2011), and support for cognitive ofﬂoad-

ing and shared awareness (Judge et al., 2008).

Former studies have shown that using tabletops

in collaborative learning activities promotes higher-

level of thinking and more effective work (Kharrufa

et al., 2010; Higgins et al., 2012). Collaborative

decision-making applications can thus leverage table-

top’s properties: promoting equal access to all group

members, and direct interaction with digital informa-

tion displayed on an interactive surface (Rogers et al.,

2004; McCrindle et al., 2011). However, support-

ing the development of personal judgments in these

shared environments can be a challenge (Gutwin and

Greenberg, 1998; Häkkinen and Hämäläinen, 2012).

Introducing personal devices, such as tablets can be

helpful to address that problem.

Multi-surface environments (MSE) typically ex-

pand shared tabletop or vertical surfaces with extra

devices for personal data exploration. For instance,

in Caretta (Sugimoto et al., 2004), groups of students

work together on urban planning tasks with a shared

surface combined with several PDAs. The shared sur-

face supports the city construction, on which students

can manipulate physical objects and make decisions

on the city evolution. PDAs are used as personal

spaces where students can experiment ideas running

simulations on different parts of the city.

Other research on MSE applications in analytics

or decision-making range from basin (oil/gas) explo-

ration (Seyed et al., 2013), to emergency response

planning (Chokshi et al., 2014). In most cases, large

shared displays support groups in prioritizing infor-

Supporting Decision-making Activities in Multi-Surface Learning Environments

mation, making comparisons, and structuring data

that embodies the working hypotheses (Wallace et al.,

2013). While auxiliary displays are used to simul-

taneously control and exploit additional information

(Seifert et al., 2012).

Former work in decision-making support with

tabletops and MSE focused mostly on the system de-

sign and collaboration results. This article focuses

more precisely on how students pursue decision-

making learning activities by using the combination

of devices. To this end, we deﬁne below the be-

haviors relevant to decision-making activities derived

from previous literature and teachers interviews.

3 APPLICATION DESIGN

3.1 Decision-making Behaviors

Supported

To understand teachers’ requirements and expecta-

tions regarding the design of an environment for

learning decision-making processes, we organized

several workshops with four teachers from a voca-

tional high school. Based on the discussions and

workshops, we found that the decision-making pro-

cess was often too complex for our target students.

Teachers emphasized the needs for developing

students’ analytical skills. Considering Ratcliffe’s

model, three stages (stage 3, 4 and 5) were par-

ticularly relevant to teachers’ pedagogical concerns.

These concerns are strengthening students’ abilities to

analyze and evaluate different options, weighing the

beneﬁts and drawbacks, expressing their own reason-

ing, considering others’ opinions, and reaching group

decisions. Other stages of the decision-making pro-

cess such as deﬁning criteria or searching for options

(Siegel, 2006) are not the main focus for the targeted

learners. The knowledge about the activity, including

the options, criteria and context would be developed

in previous classes. Teachers would prepare these

supporting material beforehand.

Our proposition thus focuses on supporting the an-

alytical process in decision-making activities. In or-

der to structure the design and evaluate the activity,

we identiﬁed four broad categories of behaviors rele-

vant to decision-making activities based on the litera-

ture and the discussions with the teachers:

1. Exploring content;

2. Discussing options;

3. Maintaining group and activity awareness;

4. Regulating the activity.

Exploring is mostly an individual behavior, with

which students can develop their own opinions. Ex-

ploring consists of browsing content, running simu-

lations or conducting data analyses. It corresponds

to behaviors linked to clarifying information and sur-

veying described in several models (Janis and Mann,

1977; Ratcliffe, 1997).

Discussing is the main collaborative behavior in

the decision-making process. It happens when stu-

dents are talking about and exchanging their ideas.

According to the different models (Aikenhead, 1985;

Kraemer and King, 1988; Ratcliffe, 1997), this be-

havior occurs when participants are building common

ground (Clark and Brennan, 1991) on the options and

collectively evaluating them. Ratcliffe identiﬁed sev-

eral subcategories when analyzing discussing includ-

ing discussing options, discussing criteria, discussing

information, comparing options, and choosing with

reasoning.

Awareness is necessary for collaborative activities

to enable participants to adjust to the group progres-

sion. It involves being aware of the activity as it un-

folds: its progress, what other people are doing and

how the group is behaving as a whole (Dourish and

Bellotti, 1992; Hornecker et al., 2008).

Regulation is built upon awareness and relates to

people’s ability to plan, monitor, and evaluate the

joint activity (Vauras et al., 2003; Volet et al., 2009;

Rogat and Linnenbrink-Garcia, 2011). It is important

in group activities, even more so in a learning context,

in which students are still in the process of develop-

ing collaboration skills (Hadwin et al., 2011). In a

classroom environment, regulation can come from the

teachers or the students themselves.

3.2 Scenario

The decision-making activity is part of a larger paper-

based learning game focusing on a non-determinist

and pluri-disciplinary pedagogical situation. The goal

of the game is to set up a sustainable company breed-

ing and selling insects. Students are split into groups

of three people and must choose one kind of renew-

able energy and the family of insect to breed and sell.

Our application focuses on the selection of the

best location to establish the insect farm. Students

must analyze the geographical and abiotic data of four

optional locations and decide which location is best.

To do so, they should consider the breeding and living

conditions of their insects, logistics requirements, and

sustainable development principles. Teachers deﬁned

six criteria before the session for students to help them

analyze: 1) moisture for breeding the insect; 2) tem-

perature for breeding the insect; 3) feasibility of using

CSEDU 2017 - 9th International Conference on Computer Supported Education

wind energy; 4) feasibility of using solar energy; 5)

accessibility (transport, communication routes); and

6) neighborhood.

To support students’ progress throughout the ac-

tivity, but also to monitor and evaluate the analytical

process, we decided to structure the activity around

three steps:

1. Survey: students explore, analyze the data and rate

the four locations before they can move to the

next step. We merge clarifying information (stage

3) and evaluating option (stage 4) of Ratcliffe’s

model into this step because of the close relation

between these two stages and the non-linear na-

ture of the process.

2. Choice: students decide which location to choose.

Once a location is picked, they can move to the ﬁ-

nal step. This step is based on choosing an option

(stage 5) of Ratcliffe’s model.

3. Justiﬁcation: students produce an explanation of

their choice, explaining how the location meets

their requirements based on the criteria. This step

is essentially required by the teachers for the eval-

uation of students analytical skills.

3.3 Application

Pickit, the application developed, uses the combina-

tion of a tabletop and tablets to support analytical

decision-making process (Figure 1). The tabletop is

mostly dedicated to viewing the decision-making con-

text and the tablets are dedicated for browsing the data

about each location, i.e. the options in the decision-

making model.

Figure 1: Three students using Pickit.

The tabletop displays a large map with four mark-

ers corresponding to the four locations to consider

(Figure 2). To get the information about a location

on their tablets, students need to tap a marker on the

map, then tap their avatars in the pop-up box. An in-

formation button in the menu bar in the top left corner

lets students get information about the energy and the

insect they chose.

On their tablets, students can then see data about

light intensity, wind strength, soil temperature, and

humidity (Figure 3). The students can analyze each

location based on six criteria deﬁned by the teachers.

Students must rate each location based on these crite-

ria on a scale from one star to ﬁve stars. While explor-

ing, they can also submit comments about the loca-

tions to help them build an argument and support later

discussions. Such arguments can also be recorded to

build a justiﬁcation of the ﬁnal decision. Each student

has an individual color for their comments.

These evaluations appear as four cards represent-

ing the four locations to support group discussion

(brown cards in Figure 2). On these cards, students

can see each other’s comments. When all the group

members ﬁnish their rating, they can get the aver-

age of the group rating results on the cards. These

cards can be dragged and scaled, which allows stu-

dents to organize and compare different options when

discussing. It also allows students to orient cards eas-

ily when sitting on the side of the tabletop.

We introduced several features to foster awareness

and facilitate group regulation. For example, when a

student picks a location to explore, the background

color of his/her avatar on that option card will change

on tablets in order to indicate who is exploring which

location. Once a student has ﬁnished rating a location,

a green checkmark appears on that option card next to

his/her avatar as a sign of completion. On the menu

bar, there are rating progress bars for each student and

Table 1: Functionalities of the application to support

decision-making behavior.

Behaviors Functionalities

Exploring

(tabletop)

- Information button to show the decision

making context (energy and insect)

- Map to show the geographical

condition of locations

Exploring

(tablet)

- Showing data of locations

- Rating tool based on criteria provided

- Writing comments and arguments

Discussing

(tabletop)

- Sharing everyone’s comments

- Providing average of the group’s

rating result

- Showing location cards for comparison

Awareness

Regulation

(tabletop)

- Showing who is exploring what

- Progress bars for the rating processes

- Check marker for the location that

has been rated

- List of steps in the menu bar

Supporting Decision-making Activities in Multi-Surface Learning Environments

Figure 2: Pickit on tabletop. A screenshot of the survey stage.

Figure 3: Pickit on one student’s tablet, from the top to bot-

tom showing the location title, location information, rating,

comment for the location, and comment in general.

step-lists with the current step highlighted, to help stu-

dents understand their progress and the ongoing task.

Table 1 shows the functionalities of the application

that support these decision-making behaviors.

We used AngularJS to develop a Web applica-

tion running on tabletops and tablets. All the de-

vices connect wirelessly to an external server hosted

on Heroku, and they communicate via Web sockets.

4 STUDY

We conducted the activity described above in the high

school to study whether and how our application sup-

ported the four categories of behaviors that were de-

ﬁned before, and how the devices were used when stu-

dents were performing those behaviors.

4.1 Participants

Four groups of three students from the same high

school class took part in the activity (12 in total). The

students were between 16 and 19 years old (mean:

17.5, SD: 0.78), including seven females and ﬁve

males. They knew each other well and had already

worked together as groups in their former classes.

Two teachers joined us to follow groups and gave in-

structions on task. Teachers underlined that it was

challenging for students from this class to collaborate.

CSEDU 2017 - 9th International Conference on Computer Supported Education

4.2 Apparatus

Each group had a capacitive 27-inch horizontal touch

screen with a resolution of 1920x1080 pixels, and a

tablet per student (Galaxy Note 8 with protective cov-

ers). The touch-screens were positioned on round ta-

bles. Students all sat on the same side of the table and

could rest the tablets on the table in front of the touch-

screen (see Figure 1). All the devices were wirelessly

connected to the network.

4.3 Task Organization

Each session lasted about 30 minutes. The study

took place in the high school library. Two groups

performed the task at the same time. A teacher fol-

lowed each group to give instructions. Before the

task, the teacher explained the application to the stu-

dents, and let them play with the application to fa-

miliarize themselves with all the functionalities. Sev-

eral bookshelves isolated the two groups, to avoid any

external inﬂuence during the task. Competition or

collaboration between groups was neither encouraged

nor discouraged.

4.4 Data Collection / Recording

We used one camera set in front of the table to record

the tabletop screen, students’ interactions, gestures,

and postures. We also collected logs to quantify in-

teractions on the tabletop and tablets and comple-

ment the videos, e.g. know what was displayed on

the tablets of the participants. After the experiment,

students were asked to ﬁll in a questionnaire con-

taining 5-point Likert-scale and free-form questions

about their perception of learning and their feelings

on using the combination of devices to carry out the

activity.

4.5 Analysis Method

We focused our analysis on the four behaviors iden-

tiﬁed earlier. We used two type of data sources, the

recorded videos, and logs. The goal of the analysis

was not to provide quantitative data of the activity but

to give qualitative insights on how students behaved

and collaborated in MSE. We tried to understand how

the design of our application impacted activity and

device usage, and contributed to the decision-making

process. To do so, we deﬁned a coding scheme for the

analysis of videos based on the literature presented

previously (Ratcliffe, 1997; Hornecker et al., 2008;

Rogat and Linnenbrink-Garcia, 2011).

Two coders analyzed the videos independently on

different samples. They did an inter-rater reliability

test with the result of 85% agreement. Then one re-

searcher analyzed three groups and the other analyzed

one group. Both of them went through the videos

twice. The ﬁrst time, they used the coding scheme

and the second time, they noted down devices usage

(for instance who was interacting on the shared dis-

play or how students used their tablets and the table-

top to discuss a speciﬁc option).

4.5.1 Video Coding Scheme

For exploring behaviors, we noted the sequence of ex-

ploring options of each student on a timeline. This

timeline is used to analyse how students chose the op-

tions to explore, how long they spent on each option,

and whether their exploration strategies were inﬂu-

enced by others.

To analyse discussing behaviors, we used the cat-

egories deﬁned by Ratcliffe’s (Ratcliffe, 1997):

D-O. discussing options;

D-C. discussing criteria;

D-I. discussing context (background knowledge or

information);

D-CMP. discussing beneﬁts and drawback of op-

tions (comparing);

D-R. choosing with reasoning.

We coded awareness behaviors when they were di-

rectly linked to the MSE conﬁguration and to the dis-

tribution of the interface among the different devices,

for instance when someone was looking or taking a

glance at one another to check what others are doing

(Rogers and Lindley, 2004):

AW. looking what others are doing.

We split regulation between group regulation

(RG) and teachers initiating group regulation (RT).

We coded regulation when observing monitoring and

planning of the task, as in (Rogat and Linnenbrink-

Garcia, 2011).

RG-1. one group member reminds others of the

time or the task progression;

RG-2. one group member offers help to others;

RT. teacher gives instructions on how to use the

application or provides instruction/information to

help students understand the decision they should

make.

When analyzing these behaviors, we also ob-

served how students were using the different devices.

We observe device usage from two dimensions: the

number of students involved in the activity, and the

devices students focused on.

Supporting Decision-making Activities in Multi-Surface Learning Environments

4.5.2 Log analysis

We focused our analysis on the number of actions on

the large display including zooming and dragging the

map or the location card, pressing buttons, selecting

avatars, etc. We computed students’ ratings to see

whether their ﬁnal choice was the location for which

they gave the highest score. We also counted the num-

ber of comments and justiﬁcations of each student.

5 RESULTS

We now describe how the different groups performed

the decision-making activity, focusing on the behav-

iors deﬁned previously.

5.1 Overview of the Activity

All the groups succeeded in analyzing locations, mak-

ing decisions and providing reasonable justiﬁcations.

Groups spent different amount of time on their tasks

(G1 = 38m14s, G2 = 25m37s, G3 = 31m33s, G4

= 23m02s). We were interested in observing how

groups structured the analytical process and which be-

haviors were involved in the different steps. We ob-

served that three groups (G1, G2 and G3) explored

options and began to discuss and compare them at the

same time, during the survey step. Discussions al-

ways happened when one student ﬁnished browsing

an option and wanted to exchange ideas with others.

Such discussions only happened when students were

checking or had already checked the location. But if

other students did not check the location yet, the dis-

cussion would not be initiated.

Group 4 acted differently. The students explored

options individually and did not discuss during this

step, which made it much shorter comparing to other

three groups (time percentage of the survey step for

each group: G1 = 65.4%, G2 = 77.7%, G3 = 77.8%,

G4 = 55.1%). Group 4 also had fewer interactions on

the shared display (G1 = 102, G2 = 156, G3 = 50, G4

= 35) but a higher amount of individual comments for

each location (G1 = 14, G2 = 15, G3 = 16, G4 = 26).

G1, G2 and G3 spent a little time on the choice

step as they already had discussions and changed their

ideas in the former step. On the contrary, G4 be-

gan to discuss various options in the choice step, us-

ing the individual comments written on the previous

step. Consequently, this step was longer than for the

other groups (time percentage of the choice step: G1

= 1.6%, G2 = 6.5%, G3 = 3.2%, G4 = 14.3%).

5.2 Exploration

Exploration happened mostly during the survey step.

Students checked the information about breeding re-

quirements of their insects, their chosen energy and

geography of locations on the shared surface, and

browsed each option and criteria (such as the tem-

perature for breeding the insect or the feasibility of

using wind energy) on their tablet. In G1, G2 and G3,

students did not double check options in the last two

steps, which indicated they knew well about these op-

tions after the survey step. Only G4, who did not dis-

cuss during the survey step, re-visited some locations

in the choice step.

Figure 4: Location analysis sequence of each student based

on the timeline. Four colors represent four locations.

During the exploration, students acted individu-

ally when choosing a location to analyze on the map.

We observed nevertheless implicit coordination in the

choice of locations. From the ﬁgure 4 we can see that

three groups (G2, G3 and G4) started the exploration

with the same location. When a participant selected

another location, s/he was later followed by the two

others. According to the speed of exploration of each

student, we can observe adjustments and switches in

the exploration sequences. For instance, in G3, af-

ter the second location visited, students 2 and 3 broke

the order chosen by student 1, who is quicker than

the other two. They chose the location that student

1 was exploring at this moment so they could dis-

cuss together. We can ﬁnd several periods in these

three groups when students were checking the same

location together. One reason for this could be they

CSEDU 2017 - 9th International Conference on Computer Supported Education

Table 2: Discussion in survey: the number of discussions on

locations, criteria, context and comparing locations by each

group (expect group 4 who did not discuss in this step).

Discussion Group 1 Group 2 Group 3

Locations 12 8 14

Criteria 5 3 10

Context 6 1 2

Comparing locations 1 3 1

were inﬂuenced by the interface, choosing the marker

that showed up on the display. The other reason

could be that students wanted to evaluate together the

same locations so they could discuss and share opin-

ions. Only students of G1 adopted a different strategy.

They chose different locations to explore on purpose

during the whole survey.

5.3 Discussion among group

5.3.1 Discussion in survey

We counted the number of discursive acts in the sur-

vey step and we observed how devices were used dur-

ing the discussions. Options and criteria were mostly

the main focus of discussions whereas discussing the

context and comparing options occur few times in this

step (Table 2). We observed mostly two participants

focusing on one tablet during discussions followed by

two or three students focusing on the shared display

and hybrid chat using both shared display and tablet.

Students had most discussions dyadically and used

tablets to discuss, especially when the discussions

were related to options and criteria. They used tablets

to check a speciﬁc option, analyzing the data accord-

ing to the criteria and exchanging ideas. The shared

display was often used when students discussed about

the context and their preferences (example of a dis-

cussion in Table 3).

Regarding group coordination, groups that tended

to explore the same options in parallel on their tablets

(G2 and G3) preferred to discuss using their tablets.

Group 1, where students explored the location inde-

pendently on their tablets, used mostly the tabletop to

discuss. Group 4 did not have discussion in this step.

5.3.2 Discussion in choice and justiﬁcation

Discussions in the choice and justiﬁcation steps were

more concentrated in comparison to the survey step

where they were fragmented. Therefore, we consid-

ered discussion durations instead of counting occur-

rences in our analysis (Table 4). Group 4 spent more

time in these steps compared with other groups since

they did not discuss during the survey. All the groups

Table 3: An extract of G3 discussion on a location.

Student A and B have the same location on their

tablets:

- A: “What do you think about the humidity?"

(A looks at B.)

- B: “It’s too humid." (B looks at his own tablet.)

- A:“Too humid? What about the wind speed?"

(A takes a look at her tablet, then looks at B.)

- B: “We don’t care about wind, we use solar

energy." (B looks at A.)

- A: “Yes, but the light intensity is also not good

enough." (A looks again at her tablet, then looks

at B.)

- B: “No, it only has 1017 lux." (B looks at his

tablet.)

Table 4: Duration of discussions in choice and justiﬁcation

on: comparing locations and choosing with reasoning.

Discussion G1 G2 G3 G4

Comparing locations 142s 20s 37s 184s

Choosing with reasoning 12s 16s 11s 53s

discussed around tabletop. This tends to indicate that

the shared display supported groups in synchronizing

opinions and reaching a decision. In group 4, we also

observed one student looking at the tabletop while

two others looked at their tablets, when they had difﬁ-

culty deciding which option was better and needed to

check the data on their tablets again. Possibly, tablets

served as supportive tools for debating during discus-

sion when more information was needed.

Location cards played an important role during the

discussion. The ratings on the cards helped students

to quickly ﬁnd the options with the highest scores.

Students could zoom out and put aside the options

with lower scores and only focus on the best ones.

We observed several times students were organizing

the cards on the shared display together to compare

options. The comments they added to the cards also

reminded them of their reasoning and supported them

in building justiﬁcations. In the end, each student

submitted his/her justiﬁcations through his/her tablet.

We observed two ways students used for submitting

their justiﬁcations. In two groups, students submitted

the group justiﬁcation that they discussed and agreed

upon. In two other groups, students distributed their

justiﬁcations, each one submitting the justiﬁcations

that concerned some speciﬁc aspect. From the logs,

we saw that members of a group did not always have

the same preferences, but all groups chose the option

that had the highest average score.

Supporting Decision-making Activities in Multi-Surface Learning Environments

5.4 Awareness and Regulation

Awareness and regulation behaviors happened mainly

in the survey step, when students were individually

exploring options. The maintenance of awareness

was subtle. For example, when one student chose

an option on the tabletop, other students working on

their tablets would take a glance at the tabletop. This

slight head movement indicated that students could

be aware of others actions on the tabletop. In group 2

and 3, although we did not observe students explicitly

looking at each other, they still knew which option

their partners were exploring and talked with them

about that option. Awareness was mainly maintained

by the shared surface, such as one student looking at

another’s interaction on the tabletop.

We did not observe many regulation behaviors in

which students reminded others of their progress or

time (RG-1) (mean = 1.25, SD = 0.96). In con-

trast, regulation in the form of supporting one another

(RG-2) happened more frequently (mean = 17, SD =

10.7). For example, in group 1, student A was look-

ing at the location card on the tabletop and saw stu-

dent B dragging the map looking for the next location

to explore. A pointed at a location card and said to

B: “This one you haven’t evaluated." Such behaviors

happened mostly over the tabletop when they were

tapping avatars for others, showing others the picture

of a location on the map, or passing location cards.

At a more minor level we also observed regulation

behavior while students shared a tablet.

Teachers intervened mostly at the beginning of

each step to regulate the activity. It mainly involved

explaining the task and the application functionalities.

They also answered students’ questions about the cri-

teria and the decision-making context, ensuring rele-

vance of students’ analysis.

5.5 Learning Experience

After the activity, we gathered students’ feedback

(Figure 5). They were all positive about their learn-

ing experience and the skills developed during the ac-

tivity. In particular, they felt more competent in col-

laborating with others, analyzing problems and taking

reasonable decisions at the end of the activity. Be-

sides that, they also enjoyed the activity (see Figure

6). Most of them thought using a personal tablet with

a shared display helped them collaborating.

We also interviewed teachers after the activity.

They were all satisﬁed with the activity progress.

Teachers were positively surprised to see that students

collaborated well and could listen to others opinions

as they used to have problems on collaborating.

Figure 5: 5-point Likert questionnaires on learning results:

do you feel more competent in collaborating with others

(top), analyzing problems (middle), and taking reasonable

decisions (bottom).

Figure 6: 5-point Likert questionnaires on learning experi-

ence: enjoy the activity (top) and using a personal tablet

with a shared display helps collaboration (bottom).

6 DISCUSSION

We compare Pickit with former systems supporting

decision-making that mentioned in the related work

(Table 5). Some of these systems do not focus on

decision-making processes, but still support some

decision-making stages, such as Caretta (Sugimoto

et al., 2004) and ePlan (Chokshi et al., 2014).

6.1 Main Findings

Our study showed that the analytical process con-

ducted during decision-making activities in classroom

can beneﬁt from MSE properties. The combination

of devices enabled tightly-coupled collaboration and

loosely-coupled parallel work in the decision-making

process, avoiding free-riding situations. The shared

display seemed to increase students’ awareness of the

ongoing activity, and led students to explore the same

option synchronously, but with a high level of free-

dom and without interfering with each other. In this

sense, the tablets improved independent exploration

of options within groups. Students were able to de-

velop their own judgments on the various options us-

ing criteria. And hence increased their understand-

ing. During discussions, the shared display supported

students in synchronizing opinions and reaching a de-

cision whereas tablets served as supportive tools for

debating when more information was needed.

From an educational point of view, according to

teachers, students all succeeded in making reasonable

decisions and providing justiﬁcations to support their

decisions. Students learned to argue and to struc-

ture better a decision-making activity. Teachers also

underlined the positive effects of the application on

CSEDU 2017 - 9th International Conference on Computer Supported Education

Table 5: Comparison of Pickit with former systems supporting decision-making. For the systems that use MSE, we also

clarify the role of devices.

Systems

Apparatus

Application

context

Stages of decision-making supported and associated functionalities

Summary

1) Listing

options

2) Identifying

criteria

3) Clarifying

information

4) Evaluating

options

5) Choosing

an option

6) Evaluating

decision

Convince me

(Siegel, 2006)

Science and

sustainability

issues

Listing

options and

evidences

Linking

hypotheses

and evidence

Evaluating

evidence and

belief

Obtaining

feedback of

evaluation

Making individual

decisions, no

collaboration

Argue-WISE

(Evagorou et

al. 2012)

Socioscientic

issues

Browsing

information

Writing

arguments

Writing

arguments

Mainly focusing on

building arguments

T-vote

(Mccrindle et

al. 2011)

Tabletop

Finding a shared

topic of interest in

a Museum

Voting for topics

Not supporting the

analytical process

Finger talk

(Rogers et al.

2004)

Tabletop

Calendar design

Browsing

images

Choosing

images

Not supporting the

analytical process

Caretta

(Sugimoto, et

al. 2004)

MSE

Urban planning

Browsing

information

(individual)

Simulating

plans

(individual)

Agreement

button

(shared)

Backtracking

(shared)

No clear decision-

making ow

ePlan

(Chokshi, et

al. 2014)

MSE

Emergency

response planning

Browsing / sharing

information (shared);

annotating (individual)

Focusing on analysis

rather than taking

decisions

Pickit

(our study)

MSE

Pluri-disciplinary

sustainability

issues

Browsing

information

(shared &

individual);

writing notes

(individual)

Rating

options based

on criteria

(individual)

Rating

averages,

validation

(shared)

Writing

arguments

(individual)

Focusing on the

analytical process

and collaboration in

a decision-making

activity with a

dened ow

students’ cross-disciplinary skills. They were more

inclined to collaborate with others even though they

usually have great difﬁculties to listen to others and

consider other opinions. Former studies have demon-

strated that prior friendship had a signiﬁcant, large

negative impact on group performance (Maldonado

et al., 2009). Our ﬁndings suggest that MSE may ad-

dress this issue as using Pickit allows students to col-

laborate better than in a classic classroom situation.

6.2 Implication for Design

6.2.1 Structuring Decision-making Applications

and Taking Advantages of MSE

The design of decision-making activities should take

into account the properties of MSE to balance indi-

vidual and collaborative work and to help students in

structuring their analytical process.

As underlined by previous literature (Roberto,

2009), we observed that according to the groups, the

stages of the analytical process could be strongly in-

tertwined. Many times students were discussing dur-

ing the individual exploration building gradually their

choice. The structure promoted more independent an-

alytical work nonetheless. Students rated their op-

tions independently and were able to construct their

personal opinions. We did not observe free-riding be-

haviors in class that was prone to it according to the

teachers. We attribute this partly to the design of the

application: the analysis being connected to each par-

ticipant through their tablets, which made everyone

accountable in a way.

In addition, distributing content on tablets enabled

to avoid conﬂicted viewing and analyzing, especially

in our case where the information was abundant and

challenging to display at once.

6.2.2 Supporting Discussion

Overall, tablets became supportive tools for debat-

ing, and supported discussions. Tablets also en-

abled students to freely analyze options at their own

pace. However, analyzing different options in parallel

tended to impede discussion. We observed how con-

fusion could arise from students discussion a speciﬁc

criteria but not referring to the same option/location.

We also hypothesize that group 4 did not discuss

at all in the survey stage, because one student did not

let others look at her tablet. We assume that letting

students going through options together may promote

more discussions. Moreover, giving an overview of

analysis elements on the shared surface can support

students in comparing options. We observed several

times that students were organizing the cards accord-

ing to the scores of options. The comments added

Supporting Decision-making Activities in Multi-Surface Learning Environments

during the survey and displayed on the cards also sup-

ported students in building justiﬁcations.

6.2.3 Better Supporting Awareness

In our application, we distributed the controls be-

tween the shared display and tablets. For instance,

in order to control what was displayed on their tablet,

students needed to use the tabletop. An alternative

could have been to offer tabs on the tablets so that

all the locations would be directly available. How-

ever, we noticed that our strategy led to a good aware-

ness of students’ actions, status and progress. Stu-

dents could notice when their partners ﬁnished one

location and switched to the next. In addition, the

features designed in the application, such as avatars

showing who is exploring which option, progress bars

and step-list, contributed to maintaining a good level

of awareness.

6.3 Limitations

The design of Pickit was focused on the analytical

process in decision-making activities. In agreement

with the teachers, we decided to not support several

aspects of the process, such as searching for options,

and identifying criteria. We assume that more features

would be required in the application when support-

ing the whole process of decision-making. Also, we

should investigate deeper whether the design of the

application could inﬂuence or introduce some bias in

the behaviors we analyzed. In particular, it remains

difﬁcult to establish if the sequences of location ex-

plored during the survey followed a speciﬁc strategy

or if it was a consequence of the interface, displaying

the last location chosen by someone.

7 CONCLUSION

We presented Pickit, an application supporting

decision-making process for multi-surface environ-

ments. Our application design is grounded in work-

shops with teachers and an in-depth analysis of the

literature on decision-making process. Based on these

dual requirements from teachers and the literature,

we focused on the analytical process of the decision-

making activity, which involves four broad categories

of decision-making behaviors: exploring, discussing,

awareness and regulation.

We conducted a study with 12 high school stu-

dents to understand how our application would sup-

port in a complex decision-making scenario, involv-

ing the consideration of multiple options, criteria and

background information. A second purpose of the

study was to understand how students interact with

personal and shared devices in such an environment.

Our results show that the proposed MSE application

supported the analytical decision-making process. It

helps students develop their own ideas and makes

free-riding more difﬁcult. Students were all succeed

in making reasonable decisions and providing justiﬁ-

cations to support their decisions.

ACKNOWLEDGEMENTS

This research would not have been possible with-

out the teachers Cerise, Pascaline, Xavier, Cécile,

Sébastien and Caroline who participated to the design

and made the activity possible, as well as the amaz-

ing students from LPA la Martellière in Voiron. This

work was partially funded by the China Scholarship

Council PhD program and the ANR project JENlab

(ANR-13-APPR-0001).

REFERENCES

Aikenhead, G. S. (1985). Collective decision making in

the social context of science. Science Education,

69(4):453–475.

Chokshi, A., Seyed, T., Marinho Rodrigues, F., and Maurer,

F. (2014). eplan multi-surface: A multi-surface envi-

ronment for emergency response planning exercises.

In Proceedings of the Ninth ACM International Con-

ference on Interactive Tabletops and Surfaces, ITS

’14, pages 219–228, New York, NY, USA. ACM.

Clark, H. H. and Brennan, S. E. (1991). Grounding in com-

munication. Perspectives on socially shared cogni-

tion, 13(1991):127–149.

Deneubourg, J.-L. and Goss, S. (1989). Collective patterns

and decision-making. Ethology Ecology & Evolution,

1(4):295–311.

Dillenbourg, P. and Evans, M. (2011). Interactive table-

tops in education. International Journal of Computer-

Supported Collaborative Learning, 6(4):491–514.

DiMicco, J. M., Pandolfo, A., and Bender, W. (2004). In-

ﬂuencing group participation with a shared display. In

Proceedings of the 2004 ACM conference on Com-

puter supported cooperative work, pages 614–623.

ACM.

Dourish, P. and Bellotti, V. (1992). Awareness and coor-

dination in shared workspaces. In Proceedings of the

1992 ACM conference on Computer-supported coop-

erative work, pages 107–114. ACM.

Evagorou, M., Jimenez-Aleixandre, M. P., and Osborne,

J. (2012). “Should we kill the grey squirrels?” A

study exploring students’ justiﬁcations and decision-

making. International Journal of Science Education,

34(3):401–428.

CSEDU 2017 - 9th International Conference on Computer Supported Education

Grace, M. (2009). Developing high quality decision-

making discussions about biological conservation in

a normal classroom setting. International Journal of

Science Education, 31(4):551–570.

Gutwin, C. and Greenberg, S. (1998). Design for indi-

viduals, design for groups: tradeoffs between power

and workspace awareness. In Proceedings of the 1998

ACM conference on Computer supported cooperative

work, pages 207–216. ACM.

Hadwin, A. F., Järvelä, S., and Miller, M. (2011). Self-

regulated, co-regulated, and socially shared regulation

of learning. Handbook of self-regulation of learning

and performance, 30:65–84.

Häkkinen, P. and Hämäläinen, R. (2012). Shared and per-

sonal learning spaces: Challenges for pedagogical de-

sign. The Internet and Higher Education, 15(4):231–

236.

Higgins, S., Mercier, E., Burd, L., and Joyce-Gibbons, A.

(2012). Multi-touch tables and collaborative learn-

ing. British Journal of Educational Technology,

43(6):1041–1054.

Hong, J.-L. and Chang, N.-K. (2004). Analysis of ko-

rean high school students’ decision-making processes

in solving a problem involving biological knowledge.

Research in science education, 34(1):97–111.

Hornecker, E., Marshall, P., Dalton, N. S., and Rogers, Y.

(2008). Collaboration and interference: awareness

with mice or touch input. In Proceedings of the 2008

ACM conference on Computer supported cooperative

work, pages 167–176. ACM.

Janis, I. L. and Mann, L. (1977). Decision making: A psy-

chological analysis of conﬂict, choice, and commit-

ment. Free Press.

Judge, T. K., Pyla, P. S., Mccrickard, D. S., Harrison, S.,

and Hartson, H. R. (2008). Studying group decision

making in afﬁnity diagramming.

Kharrufa, A., Leat, D., and Olivier, P. (2010). Digital mys-

teries: designing for learning at the tabletop. In ACM

International Conference on Interactive Tabletops and

Surfaces, pages 197–206. ACM.

Kraemer, K. L. and King, J. L. (1988). Computer-based sys-

tems for cooperative work and group decision making.

ACM Computing Surveys (CSUR), 20(2):115–146.

Maldonado, H., Klemmer, S. R., and Pea, R. D. (2009).

When is collaborating with friends a good idea? in-

sights from design education. In Proceedings of the

9th international conference on Computer supported

collaborative learning-Volume 1, pages 227–231. In-

ternational Society of the Learning Sciences.

McCrindle, C., Hornecker, E., Lingnau, A., and Rick, J.

(2011). The design of t-vote: a tangible tabletop appli-

cation supporting children’s decision making. In Pro-

ceedings of the 10th International Conference on In-

teraction Design and Children, pages 181–184. ACM.

Ratcliffe, M. (1997). Pupil decision-making about socio-

scientiﬁc issues within the science curriculum. In-

ternational Journal of Science Education, 19(2):167–

182.

Roberto, M. A. (2009). The art of critical decision making.

Teaching Company.

Rogat, T. K. and Linnenbrink-Garcia, L. (2011). Socially

shared regulation in collaborative groups: An analy-

sis of the interplay between quality of social regula-

tion and group processes. Cognition and Instruction,

29(4):375–415.

Rogers, Y., Hazlewood, W., Blevis, E., and Lim, Y.-K.

(2004). Finger talk: collaborative decision-making us-

ing talk and ﬁngertip interaction around a tabletop dis-

play. In CHI’04 extended abstracts on Human factors

in computing systems, pages 1271–1274. ACM.

Rogers, Y. and Lindley, S. (2004). Collaborating around

vertical and horizontal large interactive displays:

which way is best? Interacting with Computers,

16(6):1133–1152.

Seifert, J., Simeone, A., Schmidt, D., Holleis, P., Reinartz,

C., Wagner, M., Gellersen, H., and Rukzio, E.

(2012). Mobisurf: improving co-located collabora-

tion through integrating mobile devices and interac-

tive surfaces. In Proceedings of the 2012 ACM inter-

national conference on Interactive tabletops and sur-

faces, pages 51–60. ACM.

Seyed, T., Costa Sousa, M., Maurer, F., and Tang, A. (2013).

Skyhunter: a multi-surface environment for support-

ing oil and gas exploration. In Proceedings of the 2013

ACM international conference on Interactive table-

tops and surfaces, pages 15–22. ACM.

Shen, C., Everitt, K., and Ryall, K. (2003). Ubitable: Im-

promptu face-to-face collaboration on horizontal in-

teractive surfaces. In International Conference on

Ubiquitous Computing, pages 281–288. Springer.

Siegel, M. A. (2006). High school students’ decision mak-

ing about sustainability. Environmental Education Re-

search, 12(2):201–215.

Sugimoto, M., Hosoi, K., and Hashizume, H. (2004).

Caretta: A system for supporting face-to-face col-

laboration by integrating personal and shared spaces.

In Proceedings of the SIGCHI Conference on Human

Factors in Computing Systems, CHI ’04, pages 41–48,

New York, NY, USA. ACM.

Uskola, A., Maguregi, G., and Jiménez-Aleixandre, M.-P.

(2010). The use of criteria in argumentation and the

construction of environmental concepts: A university

case study. International Journal of Science Educa-

tion, 32(17):2311–2333.

Vauras, M., Iiskala, T., Kajamies, A., Kinnunen, R., and

Lehtinen, E. (2003). Shared-regulation and motivation

of collaborating peers: A case analysis. Psychologia,

46(1):19–37.

Volet, S., Vauras, M., and Salonen, P. (2009). Self-and

social regulation in learning contexts: An integrative

perspective. Educational psychologist, 44(4):215–

226.

Wallace, J. R., Scott, S. D., and MacGregor, C. G. (2013).

Collaborative sensemaking on a digital tabletop and

personal tablets: prioritization, comparisons, and

tableaux. In Proceedings of the SIGCHI Confer-

ence on Human Factors in Computing Systems, pages

3345–3354. ACM.

Supporting Decision-making Activities in Multi-Surface Learning Environments