Organizational Biases in International Humanitarian … faculteit...decision making process....

Organizational Biases in International Humanitarian Disaster ResponseProf. Dr. Tina Comes

Centre for Integrated Emergency ManagementUniversity of [email protected]

The Impact of Technology and Information

The development of a more technology-oriented approach to humanitarian action is essential – and inescapable – to take advantage of the opportunities to improve, for example, information gathering, analysis, coordination, action or fund-raising.

Bekele Geleta, Secretary-General of the IFRC, World Disasters Report 2013

Tina Comes: Organizational Biases in International Humanitarian Disaster Response


Local Decisions;Delegation of

Authority;Democratisation

Community Engagement

Tremendous Data Collection Efforts;

Control and Monitoring at

(remote) HQ and Donour levels

Findings From Field Research

�Our most urgent challenge is: what are we collecting and why? Who is it for?”

(Excerpt from interview with NN (UNICEF) in Manila, December 2013.)



“We are providing information to donors and Headquarters. We are only feeding the beast…”

Interview in Manila

�In Syria, it is impossible to separate operational from strategic or political information. More than anywhere else, information is power here�

(Excerpt from interview with NN (NN) in Amman, May 2014.)


As there is a total lack of a protocol on how we should deal with information, the prevailing mode of official operation is to stick to secrecy – unless explicitly told otherwise by their superiors.

(Excerpt from interview with NN (NN) in Amman, May 2014.)


This atmosphere of mistrust is ubiquitous, and hampers exchange even of such information that is common knowledge between all actors.


IM in Complex and Natural Disasters: Challenges

Natural disaster response:• Fast, streamlined, automated production of standard products

• Reduced the flexibility in tailoring to actual needs, or even lack of clarity about the and lack of clarity about needs

• Access as driver of quality

Complex disaster response:• Information sensitivity and a lack of adequate procedures to handle this

• Lowest common denominator in information product quality as only that information of which one is absolutely certain is being published

• Sensitivity and barriers to sharing as driver of quality


B. Van de Walle and T. Comes, “On the Nature of Information Management in Complex and Natural Disasters,” Humanit. Technol. Sci. Syst. Glob. Impact, vol. 107, pp. 403–411, 2015.

• Data a source of status, power, and control – there is no such thing as neutral or objective data.

• Information is only a small part of the organizational decision making process. Decision makers tend to simplify a problem, discard or re-interpret information to the point at which the problem is manageable in the given time and fits their perceptions.

• Negotiations, habit, rules of thumb, and "muddling through" have far more influence on decision making than rational application of information.

Power, Politics, and IM


MODELING/MUDDELINGTHROUGH?


Recognition Primed Decisions

• Experts choose plausible actions without analyzing many options• Experts withstand time stress because they use pattern matching• Experts can decide without analyzing courses of action


Experience the situation

in a changing context

Re-asses situationSeek more information

Activation of Information from MemoryPlausible Goals / Relevant CuesExpectanciesActions 1,…,n

Will it work?

Complex

RPD

Strategy

Evaluation via mental simulation

IMPLEMENT

Modify

Are expectanciesviolated?

Is situation familiar?

noYes, but

no

revise


Recognition-

primed decision

(RPD) model

(Klein 1985)

Features

• First option is usually workable and NOT random generation and selective retention

• Serial generation/evaluation of options; NOT concurrent evaluation • Satisficing, NOT optimizing• Evaluation through mental simulation, NOT MAUA, Decision Analysis or

Bayesian stats• Focus on elaborating and improving options, NOT choosing between

options• Focus on situation assessment, NOT decision events

• DM primed to act, NOT waiting for complete analyses


Concerns

• One cannot derive prescriptive statements from descriptive premises (let alone theories!)

• Focus on experts, not on rationality principles… but what is an expert?

• Computational model(s)?


The role of Sensemaking


• provide a framework to structure chaotic streams of information into meaningful patterns that are the basis for decision-making

• continuously on-going process, adding to the narrative of expected events or patha

• social activity

“In decision-making you have to keep moving, you cannot sit still and reflect too long, by taking a decision you are moving and then you can adjust and learn.”(Excerpt from an interview in the response to Typhoon Haiyan, Tacloban, December 2013).


THE ROLE OF BIASESField-based decision-makers vs. HQ and Donours?



T. Comes, “Cognitive and Motivational Biases in Humanitarian Sensemaking and Decision-Making,” in IEEE International Multi-Disciplinary Conference on Cognitive Methods in Situation Awareness and Decision Support (CogSIMA), 2016.

Biases Across Disaster Phases


Bias Operational DM Donors

Anchoring and insufficient adjustment

High: decisions made at the onset ofan operation are based on uncertainor lacking information. Given the timepressure, this is an initially effectivestrategy. As initial decisions provide thebasis for the response, they have apowerful impact. While decisions shouldlater on be adjusted in the light of newinformation, the anchoring bias canprevent adequate adaptation, and isparticularly difficult to overcome ifcombined with confirmation bias.

Moderate: Possibility for fact checking online and trends analysis in near real-time enables remote donors to use counter-anchors.

Wishful thinking and illusion of control

High in complex crises: the influence ofhumanitarian aid in conflicts is limited.Nevertheless, humanitarian narratives(internally and externally) rely on theability to alleviate suffering inducing amotivational bias. This may lead toneglect of risks and limitedcommunication with local actors orNGOs with controversial views.

High: donors have only anindirect impact on theresponse operations. Hence,there are motivationalincentives to exaggerate theimpact of such work. Thisleads to irrelevant orredundant information,and many informationrequests adding to the loadof field responders.


“Headquarters and donors want to control the game. And they want to choose their own reality.” (Excerpt from an interview of the author on the Syria Crisis, May 2014).


Bias Responders DonorsLogical

reconstruction and hindsight bias

High: driven by the need tojustify the efforts towardsdonors, the public, and policy-makers there is a tendency toprovide explanations that makethe order of events morepredictable and logical.Hindsight biases in this case willcombine cognitive andmotivational factors.

High: since donors aredetached from thecontext, it is difficult forthem to fully understandthe complexities anduncertainties that field-based decision-makersfaced. Outcomeknowledge is thusparticularly easilyassimilated with what isknown about thesituation. Such ease ofassimilation is a driverfor the hindsight bias .

DESIGN OF DECISION SUPPORT SYSTEMS IN THE BIG DATA ERA


• Obtain data, status, views• Monitor conditions• Obtain expertise, liaison, action takers, reporters• Draft contingencies• Validate options• Obtain approvals, delegate authority• Coordinate actions, take actions, evaluate actions• Evaluate outcomes

> Modify scenarios and plans> Modify community and operations

Requirements


• Timely tacking of what is happening is critical

• Delegation of authority a must and• Providing related data and information up, down, and

laterally is critical• Plans are in constant modification• Supporting confidence in a decision by the best possible

timely information

© Bartel Van de Walle 2004

System features


Research Topics

• Decision models • Heuristics vs Opimisation• Protracted vs. sudden onset

• Identify information relevance and triggers to reduce information overload and its negative effects when it occurs

• Irreversibility and temporality of decisions made• Design of (training, gaming) scenarios• Biases in decision-making and the role of virtual control


[email protected]

tinacomes


QUESTIONS?GET IN TOUCH !

• Chan, J. & Comes, T (2014): Innovative Research Design – A Journey into the Information Typhoon. Humanitarian Technology: Science, Systems and Global Impact 2014, HumTech2014. ProcediaEngineering. Available online

• Comes, T. & Van de Walle, B. (2014): A Journey Into the Information Typhoon. DRL Field Research Report. Available online: http://www.disasterresiliencelab.org/haiyan-report.html

• Laguna, L., Lauras, M. & Comes, T. (2015): Humanitarian Value Stream Mapping: Application to the EBOLA Outbreak. ISCRAM2015, Kristiansand, Norway

• Laguna, L., Lauras, M., Van de Walle, B. & Comes, T. (2015): Towards More Relevant Research on Disaster Management Coordination. ISCRAM2015, Kristiansand, Norway.

• Van de Walle, B. & Comes, T. (2014) Risk Accelerators in Disasters. In Advanced Information Systems Engineering, 8484, 12-23. Springer International Publishing. DOI: 10.1007/978-3-319-07881-6_2

• Van de Walle, B. & Comes, T. (2015): On the Nature of Information Management in Complex Disasters. HumTech2015, Boston, USA.

• Van De Walle, B., Gralla, E. & Goentzel, J. (2013). Field-Based Decision Makers’ Information Needs. Geneva, Switzerland. Avalaible online: http://digitalhumanitarians.com/content/decision-makers-needs

• Turoff, M., Chumer, M., Van de Walle, B. A., & Yao, X. (2004). The design of a dynamic emergency response management information system. Journal of Information Technology Theory and Applications, 5(4), 1–36.


Selected Readings

Organizational Biases in International Humanitarian … faculteit...decision making process....

Documents

Transcript of Organizational Biases in International Humanitarian … faculteit...decision making process....