Interactive Landslide Simulator: Role of Social Norms in
Learning against Landslide Risks
Pratik Chaturvedi 1,* and Varun Dutt1
1Applied Cognitive Science Laboratory, Indian Institute of Technology Mandi, Kamand, In-
* Defence Terrain Research Laboratory, Defence Research and Development Organization,
Delhi, India - 110054
Abstract. Landslide disasters, i.e., movement of hill mass, cause significant
damages to life and property. People may be educated about landslides via sim-
ulation tools, which provide simulated experiences of cause-and-effect relation-
ships. The primary objective of this research was to test the influence of social
norms on people’s decisions against landslides in an interactive landslide simu-
lator (ILS) tool. In a lab-based experiment involving ILS, social norms were
varied across two between-subject conditions: social (N = 25 participants) and
non-social (N = 25 participants). In social condition, participants were provided
feedback about investments made by a friend against landslides in addition to
their investments. In non-social condition, participants were not provided feed-
back about friend’s investments, and they were only provided feedback about
their investments. People’s investments were significantly higher in the social
condition compared to the non-social condition. We discuss the benefits of us-
ing the ILS tool for educating people about landslide risks.
Keywords: Human Factors · Human-systems Integration · Systems engineering
· Simulation games · Social Norms · Experience · Landslides
Worldwide, landslides cause huge losses in terms of human fatalities, injuries, and
infrastructure damages . In fact, landslides are a major concern for disaster-
prevention groups in regions with steep terrain, especially in the Himalayan moun-
tains . Due to the catastrophic effects of landslides, it is essential to make people at
risk understand the causes-and-consequences of landslide disasters [7-8]. This under-
standing is likely to help vulnerable communities make informed decisions against
landslide disasters. However, prior research suggests that people in developing coun-
tries (like India), who reside in landslide-prone areas, possess misconceptions about
landslide risks [3-5]. One way of improving people’s understanding about landslides
may be via exposure to landslide simulation tools and games [7-8].
Prior research shows that interactive simulation tools and games have been effec-
tive in providing experience and visibility of underlying system dynamics to people
across a wide variety of problems [10, 11, 31]. For example, references [10-11] de-
veloped a Dynamic Climate Change Simulator (DCCS) tool, which provided feed-
back to people about their decisions against climate change. This feedback enabled
people to reduce their misconceptions about climate change. In a similar way, refer-
ences [8-9] developed an Interactive Landslide Simulator (ILS) tool based on the
hypothesis that experience and recency of events through a simulation exercise, may
be influential in improving the public’s awareness and perceptions about landslide
disasters [10, 11]. Results revealed that repeated feedback in ILS improved the peo-
ple’s decision-making against landslide risks [8-9].
Although the use of feedback in improving learning in simulation tools has been
explored across a variety of domains [6-11], there is less literature on how social
norms, i.e., an acceptable behavior in a particular group, influences learning in simu-
lation tools. Prior literature [15, 20, 22] suggests that social norms may play a signifi-
cant role in shaping people’s decision-making in the real-world. Also, social norms
may be effective in multi-agent simulations [23, 24]. However, the applications of
social norm in simulation tools for educating people about landslide disasters has not
been explored in literature.
The current research addresses this literature gap by investigating the impact of so-
cial norms on people’s decision-making against landslides in the ILS tool. The central
hypothesis under test is that monetary contributions against landslide risks in the ILS
tool (which is an indicator of improved understanding) will be larger under the influ-
ence of social norms compared to when these norms are not present. Based on results
of a lab-based experiment involving human participants, we propose some benefits of
the use of social norms in the ILS tool for educating people about landslide risks.
In what follows, first, we discuss background literature about use of simulation
tools and social norms. Next, we report an experiment in which human participants
performed in the ILS tool in the presence or absence of social norms. Finally, we
report results from the experiment and highlight the potential of using social norms in
simulation tools for imparting landslide awareness and education.
Prior research has clearly established that social norms not only spur but also guide
action in direct and meaningful ways [12-15]. A number of researchers have applied
the concept of social norms in computer simulation of real-world phenomena [16, 32-
35]. For example, reference  applied and examined the use of psychological mod-
el of social identity and social norms to computer models of pedestrian motion during
evacuation in an emergency situation. Similarly, reference  compared empirical
data on human behavior to simulated data on agents with values and norms in a psy-
chological experiment involving the ultimatum game. Given this asserted power of
social norms, there has been a surge of programs that have delivered normative in-
formation as a primary tool for changing socially significant behaviors, such as alco-
hol consumption, drug use, disordered eating, gambling, littering, and recycling [18,
19]. Such social-norms marketing campaigns have emerged as an alternative to more
traditional approaches (e.g., information campaigns, moral exhortation, fear inducing
messages) designed to reduce undesirable conduct . In a study, reference 
conducted a field experiment in which normative messages were used to promote
household energy conservation and found that a descriptive normative message de-
tailing average neighborhood usage produced either desirable energy savings or the
undesirable boomerang effect, depending on whether households were already con-
suming at a low or high rate. They also found that adding an injunctive message
(conveying social approval or disapproval) eliminated the boomerang effect .
Social norms have also been used in public goods games (PGGs) for studying how
people exhibit cooperative behavior [20, 21, 36]. In PGG, participants may either
contribute an amount to the public good or defect by contributing nothing. Reference
 has shown that punishments in PGG as a societal norm can foster cooperative
Although there is extensive literature on social norms in simulation tools [16, 32, 33,
34, 35], applications of social norms to decision-making against landslide risks in
simulation tools has not been explored. References [8-9] have used the ILS tool to
understand the role of amount and availability of feedback in decision making against
landslides. In this research, we use social norms in the ILS tool to improve people’s
decision-making and investments against landslides. We create norms in the ILS tool
by exposing participants to the investments of other participants against landslides.
We believe that an increase in investments by others in the ILS tool will make people
contribute more against landslides and these contributions will enable people to im-
prove their understanding about landslides. Thus, we expect participants’ investments
against landslides to be higher when norms are present in the ILS tool compared to
when these norms are absent.
3 INTERACTIVE LANDSLIDE SIMULATOR (ILS) TOOL
The ILS tool is an interactive dynamic system for studying people’s decisions
against landslide risks [8, 9]. Details about the ILS tool were already discussed by [8,
9], and here we briefly cover the tool’s working. Figure 1 shows the investment
screen of ILS tool, where people need to invest some part of their daily income
against landslide risks. Investment against landslides would be used to mitigate land-
slide risks via interventions like building reinforcements and planting trees.
Feedback is shown to participants in three ways (Fig. 2A, 2B): monetary
information about total wealth, text messages about different losses, and imagery
corresponding to losses. There is a decrease in the daily income due to an injury or
fatality due to a landslide. Also, damages to property due to landslides cause a loss of
property wealth. In addition, in situations involving the use of social norms, the
feedback screen shows the contribution from a friend against landslides (Fig 2A). The
friend function is linear: It starts at half of the total investment (=292/2) and increases
towards 292 by the last trial. A normally distributed noise is added to the linear
function (normal distribution’s mean = 0 and standard deviation = 5). The noise
component makes it difficult to find patterns in the friend function. In situations not
involving the use of social norms, the friend function is missing from the feedback
screen (Fig 2B). Positive feedback is presented to participants on the feedback screen
if a landslide does not occur in a certain trial (Fig 2C). The friend’s investments
against landslide across trials is shown in Fig 2D.
Human induced landslide risks were generated using the probability equation from
reference . The likelihood of landslides due to physical factors is calculated con-
sidering the combination of the effects of rainfall, slope, and soil properties. It is
based on the method proposed by reference . The probability of a landslide due to
rainfall P(R) is a random event, based on the computed value z, where
Fig 1. The different components on the investment screen of ILS tool. (A): The text box
giving choice for daily investments to reduce landslide risk. (B): Game parameters window
showing values of parameters used in ILS model. (C): Dynamic plots of changing outcomes
with every decision.
With the daily (DR), 3-day cumulative (3DCR) and 30-day antecedent rainfall
(30DAR) as significant predictors influencing slope failure. These rainfall values
were calculated using daily rain data from Indian Metrological Department (IMD).
Five years of daily rain data (2010-14) was averaged to find the average probability
of landslide due to rainfall P(R) over an entire year.
The slope and soil characteristics are expressed as P(S), which represent the local
probability of landslides, given the geological features of the location. The determina-
tion of spatial probability of landslides, P(S) is done from Landslide Susceptibility
Zonation (LSZ) map of the area [8-9].
Fig 2. Feedback screen in the ILS tool. (A) Feedback screen with information on the contri-
bution of a friend as well as negative feedback due to a landslide’s occurrence. (B) Feedback
screen with no information on the contribution of a friend as well as negative feedback due to a
landslide’s occurrence. (C) Feedback screen with positive feedback when a landslide does not
occur. (D) Friend’s investments against landslide risk across trials
A landslide occurs on a certain day when a independent random number (~ U(0,
1)) become less than or equal to the corresponding net probability of occurrence of
landslide which is a weighted sum of landslide probability due to environment (spa-
tial and triggering factors) and human factors. The natural environmental probability
of a landslide event is the product of the two probabilities, P(S) and P(R). Data from
reference  helped set the monetary levels in the ILS tool.
4 EXPERIMENT INVOLVING SOCIAL NORMS IN THE
The ILS tool considers both environmental factors (spatial geology and rainfall)
and human factors (people’s investments against landslides) for calculating landslide
risks. The simulation, its validation, and full functions are explained in different pub-
lications [8-9]. In this paper, we present results of an experiment in which human
participants interacted with the ILS tool in the presence or absence of social norms.
We expected participants’ investments against landslides to be higher when norms
were present compared to when norms were absent.
The study was approved by the ethics committee at the Indian Institute of
Technology Mandi, India. Fifty participants from diverse fields of study participated
across two feedback conditions: social (N = 25) and non-social (N = 25). In the social
condition, participants were provided feedback about contributions made by a friend
against landslide risks on the feedback screen. In non-social condition, feedback
about contributions made by a friend against landslide risk was absent. Data were
analysed for all participants in terms of their investment ratio in both the feedback
conditions. The investment ratio was defined as the ratio of total investments made by
participants up to a trial divided by the total investments that could have been made
up to the trial. Given the effectiveness of feedback and social norms in simulation
tools [17, 18, 28-30], we expected participant investments to be greater in the social
condition compared to the non-social condition.
All participants were from Mandi district and adjoining areas and their ages ranged
in between 18 and 26 years (Mean = 21.4 years; Standard Deviation = 2.81 years).
Around 30% participants were from science, technology, engineering, and
mathematics (STEM) backgrounds and the remaining were from non-STEM
backgrounds. Out of a total of 50 participants, 30 cited basic understanding about
landslides, 14 cited little understanding about landslides, 5 responded as
knowledgeable about landslides, and 1 possessed no idea about landslides. All partic-
ipants received a base payment of INR 50 upon completing their performance in the
ILS tool. In addition, a performance incentive was also provided to participants in the
ILS tool. To calculate the performance incentive, participants were ranked based
upon the total wealth remaining at the end of their ILS play and top-10 participants
were put in a lucky draw. After the entire study was over, one participant was ran-
domly selected and awarded a cash prize of INR 500. Participants were told about the
lucky draw via written instructions before they started performing in the ILS tool.
Participants were invited to a landslide awareness study via a flyer advertisement
in Mandi, India. Participants signed a consent form and participation was entirely
voluntary. Participants were first provided with instructions on the task in the study.
Participants started their study once they were ready to begin and did not have any
questions on the study. Participants started their study by providing demographic
information and knowledge about landslides. Next, participants interacted in the ILS
tool for a 30-day simulated time period. Experimental sessions were about 30-minutes
long per participant. Participants were not given any information concerning the na-
ture of the environment or conditions in the ILS. They were told that their goal was to
maximize their total wealth across repeated decisions in the ILS tool.
We performed a repeated-measures ANOVA with condition as a between-subjects
factor and investment-ratio over the 30-day period as a within-subjects factor. As per
our expectation, the average investment ratio was significantly higher in the social
condition (0.70) compared to that in the non-social condition (0.46) (F (1, 48) =
21.53, p < 0.05, η2 = 0.31) (see Fig 3).
Fig 3. Average investment ratio in social and non-social conditions. The error bars show
95% CI around the point estimate.
Furthermore, the trend of investment ratio across trials in the social condition was
different from the trend of investment ratio across trials in the non-social condition
(F (29, 1392) = 2.38, p < 0.05, η2 = 0.05; see Figure 4). It can be clearly seen in Fig-
ure 4 that there is a higher increase in investment ratio in the social condition com-
pared to non-social condition. Overall, the results suggest that social norms feedback
helped participants to increase their investments for landslide mitigation.
Fig 4. Average investment ratio in social and non-social conditions across 30 trials.
6 Discussion and Conclusions
In this research, we investigated the use of social norms in a simulation tool, Inter-
active Landslide Simulator (ILS), against landslide risks. On average, people’s con-
tributions against landslides were significantly higher when social norms were present
compared to when these norms were absent. Also, there was a consistent increase in
people’s contributions against landslides over trials in the presence of social norms
compared to in the absence of social norms.
Our results are in agreement with those reported in prior literature [15, 21-24, 28-
30]. In fact, reference  found that telling people about what other people did in
the neighborhood enabled people to agree with the societal norm. In the current
research involving ILS, the presence of friend’s contributions created a similar effect.
In the current research, participants found the friend to be contributing against
landslides and did not want to deviate from this norm. However, when the friend’s
contribution was missing, there was an absence of the norm. In the absence of the
norm, people were not affected by it and they kept their contributions low.
This research has a number of implications involving the use of simulation tools.
First, this research showed that social norms are impactful even in situations where
there is a prevailing uncertainty about when landslides occur. Second, via a simula-
tion, this research showed that social norms could be used to change people’s deci-
sions against landslide disasters. Overall, the presence of social norms in simulation
tools like ILS is likely to be an effective method for educating people about landslide
disasters. Researchers may also use tools like ILS to perform “what-if” analyses in
the presence of social norms. These analyses may stretch over certain time periods
and cover certain geographical locations. However, the assumptions made in the ILS
tool may have to be first evaluated for the study area before it is used for policy re-
In the current research, we made certain assumptions in the ILS tool, and these as-
sumptions may be different in the real world. For example, the model behind the ILS
tool assumed that people’s contributions against landslides influenced the probability
of landslides linearly. Second, the damages due to landslides were fixed at the same
constant value across all trials. We plan to overcome some of these assumptions as
part of our future research.
First, as part of our future research, we plan to assume different models of how
people’s investments in the ILS tool may influence the probability of landslides. Sec-
ond, we plan to try a number of models concerning damages due to landslides, where
the landslide damages are different over trials. Also, as part of our future research, we
plan to compare the proportion of investments against different insurance models
available in the real world. Some of these ideas form the immediate next steps in our
research program on landslide education and awareness.
This research was partially supported by the following grants to Varun Dutt:
IITM/NDMA/VD/184 and IITM/DRDO-DTRL/VD/179. We thank Akshit Arora for
developing the website for ILS. We also thank students of IIT Mandi who helped in
data collection in this project.
1. Margottini, C.; Canuti, P.; Sassa K. Landslide Science and Practice. Proc. of the Second
World Landslide Forum, Rome, Italy, 2011, Vol. 2.
2. Pratik Chaturvedi et al.; Remote Sensing Based Regional Landslide Risk Assessment.
International Journal of Emerging Trends in Electrical and Electronics (IJETEE –
ISSN: 2320-9569) Vol. 10, Issue. 10, Oct. 2014
3. Oven, K. (2009). Landscape, livelihoods and risk: community vulnerability to landslides in
Nepal (Doctoral dissertation, Durham University).
4. Wanasolo, I. (2012). Assessing and mapping people's perceptions of vulnerability to
landslides in Bududa, Uganda (Doctoral dissertation).
5. Chaturvedi, P. & Dutt V. (2015). Evaluating the Public Perceptions of Landslide Risks in
the Himalayan Mandi Town. Accepted for presentation in the 2015 Human Factor &
Ergonomics Society (HFES) Annual Meeting, L.A.
6. Knutti, R., Joos, F., Müller, S. A., Plattner, G. K., & Stocker, T. F. (2005). Probabilistic
climate change projections for CO2 stabilization profiles. Geophysical Research
Deleted: Also, r
7. Wagner, K. (2007). Mental models of flash floods and landslides. Risk Analysis, 27(3), 671-
8. Chaturvedi, P., Arora, A., & Dutt, V. (2018). Learning in an interactive simulation tool
against landslide risks: the role of strength and availability of experiential feedback.
Natural Hazards and Earth System Sciences, 18(6), 1599-1616.
9. Chaturvedi, P., Arora, A., & Dutt, V. (2017). Interactive Landslide Simulator: A Tool for
Landslide Risk Assessment and Communication. In Advances in Applied Digital
Human Modeling and Simulation (pp. 231-243). Springer, Cham.
10. Dutt, V., & Gonzalez, C. (2012). Human control of climate change. Climatic
change, 111(3-4), 497-518.
11. Dutt, V., & Gonzalez, C. (2012). Why do we want to delay actions on climate change?
Effects of probability and timing of climate consequences. Journal of Behavioral
Decision Making, 25(2), 154-164.
12. Aarts, H., & Dijksterhuis, A. (2003). The silence of the library: Environment, situational
norm, and social behavior. Journal of Personality and Social Psychology, 84(1),
13. Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative
conduct: A theoretical refinement and reevaluation of the role of norms in human
behavior. In Advances in experimental social psychology (Vol. 24, pp. 201-234).
14. Griskevicius, V., Goldstein, N. J., Mortensen, C. R., Cialdini, R. B., & Kenrick, D. T.
(2006). Going along versus going alone: when fundamental motives facilitate
strategic (non) conformity. Journal of personality and social psychology, 91(2), 281.
15. Schultz, P. W., Nolan, J. M., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V. (2007).
The constructive, destructive, and reconstructive power of social norms.
Psychological Science, 18(5), 429-434.
16. von Sivers, I., Templeton, A., Künzner, F., Köster, G., Drury, J., Philippides, A., ... &
Bungartz, H. J. (2016). Modelling social identification and helping in evacuation
simulation. Safety science, 89, 288-300.
17. von Sivers, I., Templeton, A., Köster, G., Drury, J., & Philippides, A. (2014). Humans do
not always act selfishly: Social identity and helping in emergency evacuation
simulation. Transportation Research Procedia, 2, 585-593.
18. Donaldson, S. I., Graham, J. W., & Hansen, W. B. (1994). Testing the generalizability of
intervening mechanism theories: Understanding the effects of adolescent drug use
prevention interventions. Journal of behavioral medicine, 17(2), 195-216.
19. Neighbors, C., Larimer, M. E., & Lewis, M. A. (2004). Targeting misperceptions of
descriptive drinking norms: efficacy of a computer-delivered personalized normative
feedback intervention. Journal of consulting and clinical psychology, 72(3), 434.
20. Roos, P., Gelfand, M., Nau, D., & Lun, J. (2015). Societal threat and cultural variation in
the strength of social norms: An evolutionary basis. Organizational Behavior and
Human Decision Processes, 129, 14-23.
21. Hauert, C. (2010). Replicator dynamics of reward & reputation in public goods games.
Journal of Theoretical Biology, 267(1), 22-28.
22. Fehr, E., & Fischbacher, U. (2004). Social norms and human cooperation.
Trends in cognitive sciences, 8(4), 185-190.
23. Pan, X., Han, C. S., Dauber, K., & Law, K. H. (2007). A multi-agent based
framework for the simulation of human and social behaviors during
emergency evacuations. Ai & Society, 22(2), 113-132.
24. Chu, M. L., Parigi, P., Law, K., & Latombe, J. C. (2014, April). SAFEgress:
flexible platform to study the effect of human and social behaviors on egress
performance. In Proceedings of the Symposium on Simulation for Architecture & Ur-
ban Design (p. 4). Society for Computer Simulation International.
25. Hasson, R., Löfgren, Å., &Visser, M. (2010). Climate change in a public goods game:
investment decision in mitigation versus adaptation. Ecological Economics, 70(2),
26. Mathew, J., Babu, D. G., Kundu, S., Kumar, K. V., & Pant, C. C. (2014). Integrating
intensity–duration-based rainfall threshold and antecedent rainfall-based probability
estimate towards generating early warning for rainfall-induced landslides in parts of
the Garhwal Himalaya, India. Landslides,11(4), 575-588.
27. Parkash, S. (2011). Historical records of socio-economically significant landslides in India.
J South Asia Disaster Studies, 4(2), 177-204.
28. Sterman, J. D. (2008). Risk communication on climate: Mental models and mass balance.
Science, 377, 532-533.
29. Cialdini, R.B. (2003). Crafting normative messages to protect the environment. Current
Directions in Psychological Science, 12, 105–109.
30. Dutt, V. (2012). Social influence encourages conservation behaviours. Retrieved from
Financial Chronicle website: http://www.mydigitalfc.com/knowledge/social-
31. Sterman, J. D. (1994). Learning in and about complex systems. System Dynamics Review,
32. Mercuur, R., Dignum, V., & Jonker, C. (2019). The Value of Values and Norms in Social
Simulation. Journal of Artificial Societies and Social Simulation, 22(1), 1-9.
33. Savarimuthu, B. T. R., Purvis, M., Purvis, M., & Cranefield, S. (2008, May). Social norm
emergence in virtual agent societies. In International Workshop on Declarative
Agent Languages and Technologies (pp. 18-28). Springer, Berlin, Heidelberg.
34. Axelrod, R. (1986). An evolutionary approach to norms. American political science review,
35. Gilbert, N. (2004). Agent-based social simulation: dealing with complexity. The Complex
Systems Network of Excellence, 9(25), 1-14.
36. Kumar, M. Aggarwal, K., & Dutt, V. (in press). Modeling Decisions in Collective Risk
Social Dilemma Games for Climate Change using Reinforcement Learning.
Accepted in IEEE Conference on Cognitive and Computational Aspects of Situation
Management (CogSIMA) 2019. Las Vegas, Nevada.