Content uploaded by Felipe Muñoz La Rivera
Author content
All content in this area was uploaded by Felipe Muñoz La Rivera on Sep 30, 2019
Content may be subject to copyright.
World Transactions on Engineering and Technology Education
2019 WIETE
Vol.17, No.3, 2019
260
INTRODUCTION
Building information modelling (BIM) methodology is a new industry standard for architecture, engineering and
construction (AEC), which improves the low productivity traditionally observed in infrastructure and building projects
[1]. The adoption of BIM has the result of promoting collaborative and integrated work among all agents in a project,
throughout the life cycle, optimising cost and duration through interconnected digital parametric models, along with
employing virtual construction processes [2]. In addition, BIM is able to incorporate environmental variables, facility
management, interference management and project co-ordination into the simulations, optimising design and
construction processes of architecture, structure and engineering, construction and maintenance [3].
Business information modelling counters fragmentation of the AEC industry through collaborative digital practices and
tools, in effect demanding that professionals work in a collaborative and interdisciplinary way throughout a project.
Faced with this new standard, the industry is now looking for professionals with training in BIM for various positions in
projects, with it becoming a basic requirement for recruitment [4].
This new professional should have not only the technical skills of the particular discipline and the use of BIM tools, but
also extensive interdisciplinary teamwork, communication and integration skills. There has been a shift from technical
assessment to collaborative practices [5]. Faced with these requirements, universities should promptly incorporate the
teaching of BIM into their curricula, to train professionals who can solve the challenges that the industry presents [6].
In recent years, many universities and training institutions have expanded training in BIM, focusing on teaching
computer modelling and simulation tools, through individual learning and without mixing disciplines [7]. Training in
BIM should not be considered only as learning a computer program, but as an integrated process of modelling
and management of the life cycle of a project, deepening the methodological aspects that underpin the collaborative
work [8].
A gradual systematic approach is required for the teaching of BIM, which starts with initial training in three-
dimensional parametric modelling, then moves on to planning and control, and then to aspects of collaboration and
integral management, always guided by a methodological vision over the strictly technological one [9].
Thus, the teaching of BIM requires methodologies that promote and form interdisciplinary collaborative work, applied
to real projects that allow the developing of a broad vision of the life cycle and the application of BIM [10].
Team-based learning (TBL) allows students to simulate real-work environments, assuming collaborative roles and
interactions. In this way, the student not only learns BIM, but also the learning process generates in parallel the
adoption of collaborative practices [11].
Impact on team performance of interaction with BIM
Rodrigo F. Herrera†‡*, Felipe Muñoz-La Rivera†, Juan C. Vielma† & Edison Atencio†
Pontificia Universidad Católica de Valparaíso, Valparaíso, Chile†
Pontificia Universidad Católica de Chile, Santiago, Chile‡
Universitat Politècnica de València, Valencia, Spain*
ABSTRACT: Building information modelling (BIM) is a methodology that is increasingly gaining strength in the
architecture, engineering and construction industry. Building information modelling is focused on the effective
collaboration and interaction of teams. It is for this reason that many classrooms around the world have decided to teach
BIM to engineering and architecture students, where team-based learning and project-based learning are the main teaching
methods. However, although students may work in teams, it is unknown how this interaction impacts on performance.
The objective of this work was to understand how interaction impacts the performance of teams adopting BIM.
A case study was carried out with 25 students in their final year of civil engineering and taking a Virtual Design and
Construction course. The students were organised into six groups, who then developed a project over 12 weeks. It was
concluded that there is a strong correlation between the interaction of the teams and their performance, which was shown
by the evaluation scores and the variability of the scores.
261
At present teamwork is considered a lever for achieving expected results in organisations. At the same time, teamwork
is the product of interactions among its members. However, finding synergies can be difficult in the face of the nature
of human relationships, individual differences and conflicts [12].
Interaction can be considered as the flows of information or instances of collaboration between people participating
in an organisation or project team. In the AEC industry, interactions have been evaluated through social network
analysis (SNA) [13]. For example, work interaction has been evaluated in architectural offices [14] or construction
teams [15].
A team of researchers simulated the interaction in a project team when BIM was used, and represented the interaction
with different SNA metrics [16]. Social network analysis is a method for studying interactions and relationships among
people; it provides both visual and quantitative analyses of human associations [17].
Generally, in university and engineering education, research activity has a higher priority than teaching, thus generating
an imbalance between these two [18]. This gap is widened by the prevalence of a traditional teacher model that prevails
in the classrooms of engineering schools, as opposed to a collaborative model. It is, therefore, interesting to demonstrate
the benefits of the application of this latter model, to promote it as a lever for better results.
Therefore, the objective of this research is to understand how interaction impacts the performance of different teams in
the context of BIM work. In this article, the following questions are addressed:
1) Does collaborative interaction in work teams have an impact on their performance?
2) Does the level of collaboration between students impact their performance?
RESEARCH METHOD
To achieve the objective, a case study was carried out with students who, in their final year of civil engineering,
participated in the Virtual Design and Construction course at the Pontificia Universidad Católica de Valparaíso. The 25
students were placed into six groups of four or five participants, to develop a project (divided into four-part deliverables)
over a 12-week period. The project consisted of BIM modelling, co-ordination of specialties, 4D modelling; and the
analysis and structural design of a building. Project-based and team-based learning were applied as teaching-learning
strategies. All this was in a classroom with the technological infrastructure to support collaborative learning [19].
Performance of the teams was evaluated by assessments of the different deliverables based on appropriate rubrics.
Some deliverables were group-based and others individual. Each student obtained a percentage achievement score
independently of the team to whom they belonged. The achievement percentage ranged from 0% to 100%, with the
minimum approval level being 60%. For the individual analysis, the average percentage of achievement of a student
was applied to all the tasks and deliverables. For team analysis, the average of all team members was applied, and the
standard deviation of the achievement percentage of each student in the team.
To assess the interaction, each course participant was asked with whom did they collaboratively interact during the 12
weeks. With this information, social network analysis was conducted: first, collaborative interactions were graphed
throughout the course; then, metrics were analysed by project team; and finally, metrics were analysed by student.
In the team analysis, the following SNA metrics were used [13]:
1) Density: ratio between number of real connections and number of possible connections in each team.
2) Average degree: average number of connections for each team member.
3) Diameter: minimum distance between the two farthest team members.
In the individual analysis, the following SNA metrics were used:
1) Indegree: the number of people who claim to have had a collaborative interaction with each node
(student/professor) [20];
2) Authority: an iterative algorithm to rank the nodes by number of in-links connections [21], i.e. the students other
students look for with whom to collaborate;
3) Page rank: an iterative algorithm that measures the importance of each node within the network; this metric is
normalised, therefore, the sum of the page rank is 1 [22];
4) Eigenvector centrality: a measure of the extent to which a node is connected to influential other nodes, therefore it
represents an informal power in the network [22].
Then, in the team analysis, the Spearman correlation between the indices of each team and the performance indices
(average and standard deviation) were calculated to define the strength and direction of the correlation. In addition,
the significance (p-value) was calculated to define whether, with the available data, the correlation was significant
defined as a p-value < 0.05. Finally, a similar analysis was made between the network indices and the percentage
achievement of each student.
262
DISCUSSION OF RESULTS
To visualise collaborative interactions, the free software, Gephi, was adopted to model the networks. Each node is
a person and each edge is a connection between people (nodes). In addition, each team was represented by a different
colour (Figure 1). In the graph, the team of professors is central and, in a position, to collaborate with all the teams of
students. The T1, T5, T3 and T6 teams seem compact, i.e. the collaboration focused on other team members during the
12 weeks of the course. Teams T2 and T4 do not seem compact; in the case of team 2, it is possible to imagine how
a student leaves the team as a result of the team’s poor academic performance. Team 4 never really managed to work as
a team and seemed to have had a traditional system of partitioning tasks among the members of the group.
Figure 1: Collaboration network: a case study.
For each team of n participants, calculated were the density, average degree and diameter. In a small team a density
(0 to 1) closer to one will imply a more connected network; the average degree (0 to n-1), if closer to n-1 would
demonstrate a more collaborative team; and the diameter (1 to n-1), if closer to 1 would imply a team with more
connections without intermediaries.
Table 1 shows Spearman’s correlation and p-value between each team’s network and achievement indices. Density and
average degree have a positive correlation with average student achievement and a negative correlation with standard
deviation; therefore, as there is greater collaboration, students improve their average performance and there are fewer
divergences among team members. Something similar can be concluded from the negative correlation between diameter
and average achievement, and positive correlation between diameter and standard deviation; that is, as fewer
intermediaries exist in a team, their performance will be higher, and their variability will be lower.
Table 1: Spearman correlation - team analysis.
Network index
Achievement index
(%)
Spearman
p
-
value
Decision
Density
Average
0.754
0.084
Strong positive correlation, but non
-
significant
Density
S
td
.
deviation
-
0.319
0.538
Weak negative
correla
tion a
nd
non
-
significant
Average degree
Average
0.812
0.0
48
Very strong positive correlation and significant
Average degree
Std
.
devi
ation
-
0.667
0.148
Strong negative correlation, but non
-
significant
Diameter
Average
-
0.617
0.192
Strong negativ
e correl
ation,
bu
t non
-
significant
Diameter
S
td
. deviation
0.648
0.164
Strong positive correlation, but non
-
significant
While all correlations calculated are strong or very strong, only the correlation between the network average degree and
the average achievement has a significant correlation (p-value < 0.050). Figure 2 shows the average degree of each team
versus the achievement percentage (left axis) and standard deviation (right axis). The circles with a blue border
represent the achievement percentage and the circles with a red border represent the deviation within each team, which
263
is represented by the same colour as in Figure 1. The positive and negative trends of the correlations presented
in Table 1 can be clearly seen in Figure 2.
It is important to highlight the differences with respect to the achievement between the T2 team (green) and the T6
(orange), given they have similar levels of connectivity. The T6 team is more aligned, while the T2 team has a lower
average performance and a high variability, which is due to the separation of one of the members of the team not
conforming to the team performance and this student had a much higher achievement than did peers.
Figure 2: Relation between network average degree and achievement percentage.
For each student, the following metrics were calculated: indegree, authority, page ranks and eigenvalue centrality.
All of these indicators are higher for students who are regarded as an informal authority among their peers. Therefore,
they are sought out for collaboration and consultation. Table 2 shows Spearman’s correlation and p-value between each
student index and the achievement percentage of those students. In all cases, strong positive correlations were found
with a high level of significance (p-value < 0.05), i.e. the students with the highest performance are the most consulted
and with whom the rest of the students most seek to collaborate.
Table 2: Spearman’s correlation: individual analysis.
Network index Performance index (%) Spearman p-value Decision
Ind
egr
ee
Achievement
0.800
0.000
Very
s
trong positive correlation and significant
Authority
Achievement
0.646
0.000
Strong positive correlation and significant
Page
-
ranks
Achievement
0.599
0.002
Strong positive correlation and significant
Ei
gen
value
centra
lit
y
Achievement
0.811
0.000
Very
s
trong positive correlation and significant
The correlations in Table 2 are shown graphically in Figure 3, where the horizontal axis represents the achievement
percentage of each student, the vertical left axis represents the degree of entry of each node, and the vertical right the
authority and the eigenvector centrality.
Figure 3: Relation between achievement and individual networks indices.
264
CONCLUSIONS
The results of the evaluation of the collaborative work of an advanced civil engineering course are presented.
The selected course was an elective course Virtual Design and Construction, which allowed the introduction of students
to technologies in business information modelling.
The results produced interesting findings, among which is the relationship between the level of collaboration and
the average performance of students. Likewise, it was possible to corroborate that, to the extent that the level of
collaboration between the members of a team is increased, the gaps in performance of the participants are reduced.
From the analysis of the indicators according to the p-value of Spearman’s correlation, only a significant correlation
was detected between the average degree of a network and the average achievement of each team, which reinforces the
positive evaluation of collaborative work.
Finally, the methodology has made it possible to detect the positive influence of the achievement percentage,
confirming that the participants with the greatest interaction (connection number) not only with the members of their
own team, but with the rest of the participants, including the professors of the course, had better achievements.
ACKNOWLEDGEMENTS
This research was conducted thanks to the contributions of the CORFO Project 14ENI2-26905 Ingeniería 2030-PUCV.
Also, the authors acknowledge the Collaborative Group of Engineering Education, the BIM Group from Pontificia
Universidad Católica de Valparaíso. Herrera acknowledges CONICYT - PCHA/National Doctorate/2018 - 21180884
for funding the graduate research.
REFERENCES
1. Boton, C. and Forgues, D., Practices and processes in bim projects: an exploratory case study. Adv. Civ. Engng.,
1-12 (2018).
2. Muñoz-La Rivera, F., Vielma, J., Herrera, R.F. and Carvallo, J., Methodology for building information modeling
(BIM) implementation in structural engineering companies (SEC). Adv. Civ. Engng., 8452461, 1-16 (2019).
3. Eastman, C., Jeong, Y., Sacks, R. and Kaner, I., Exchange model and exchange object concepts for
implementation of national BIM standards. J. of Comput. Civ. Engng., 24, 1, 25-34 (2010).
4. Uhm, M., Lee, G. and Jeon, B., An analysis of BIM jobs and competencies based on the use of terms in the
industry. Autom. Constr., 81, 67-98 (2017).
5. Succar, B., Sher, W. and Williams, A., An integrated approach to BIM competency assessment, acquisition and
application. Autom. Constr., 35, 174-189 (2013).
6. Markiewicz-Zahorski, P., Teaching building construction design using BIM: the benefits and difficulties. World
Trans. on Engng., and Technol. Educ., 17, 1, 54-59 (2019).
7. Zhang, J., Wu, W. and Li, H., Enhancing building information modeling competency among civil engineering and
management students with team-based learning. J. of Prof. Issues Engng. Educ. Pract., 144, 2 (2018).
8. Nushi, V. and Basha-Jakupi, A., The integration of BIM in education: a literature review and comparative context.
Global J. of Engng. Educ., 19, 3, 273-278 (2017).
9. Sacks, R., Radosavlijevic, M. and Barak, R., Requirements for building information modeling based lean
production management systems for construction. Autom. Constr., 19, 5, 641-655 (2010).
10. Murzi, H., Team-based learning theory appliedto engineering education: a systematic review of literature. Proc.
121st ASEE Annual Conf. & Expo. (2014).
11. Zhou, C., Kolmos, A. and Nielsen, J., A problem and project-based learning (PBL) approach to motivate group
creativity in engineering education. Inter. J. of Engng. Educ., 28, 1, 3-16 (2012).
12. Longenecker, C.O., Puertas y barreras a la cooperación y al trabajo en equipo. Harvard Deusto Bus. Rev., 104, 66-
75 (2011) (in Spanish).
13. Alarcón, D.M., Alarcón, I.M. and Alarcón, L.F., Social network analysis : a diagnostic tool for information flow in
the aec industry. Proc. 21st Annual Conf. of the Inter. Group for Lean Construct., 947-956 (2013).
14. Segarra, L., Herrera, R.F., Alarcón, L.F. and Pellicer, E., Knowledge management and information flow through
social networks analysis in chilean architecture firms. Proc. 25th Annual Conf. of the Inter. Group for Lean
Construct., 413-420 (2017).
15. Castillo, T., Alarcón, L.F. and Salvatierra, J.L., Effects of last planner system practices on social networks and the
performance of construction projects. J. of Constr. Engng. Manage., 144, 3, 1-10 (2018).
16. Al Hattab, M. and Hamzeh, F., Using social network theory and simulation to compare traditional versus BIM-
lean practice for design error management. Autom. Constr., 52, 59-69 (2015).
17. Chinowsky, P.S., Diekmann, J. and Brien, J.O., Project organizations as social networks. J. of Constr. Engng.
Manage., 136, 4, 452-458 (2010).
18. Sánchez, J.A., El ingeniero y la docencia universitaria. Rev. Educ., 330, 303-312 (2003) (in Spanish).
19. Herrera, R.F., Vielma, J.C. and Muñoz, F.C., BIM and teamwork of civil engineering students: a case study.
Global J. of Engng. Educ., 20, 3, 230-235 (2018).
265
20. Al Hattab, M. and Hamzeh, F., A process-social perspective for understanding design information flow. Lean
Constr. J., 11, 1-11 (2017).
21. Kleinberg, J.M., Authoritative sources in a hyperlinked environment. J. ACM, 46, 5, 604-632 (1999).
22. Lee, C., Chong, H., Liao, P. and Wang, X., Critical review of social network analysis applications in complex
project management. J. of Manage. Engng., 34, 2, 1-15 (2018).