Favourite Challenge Reducing cognitive biases

My favourite challenge is to understand the various biases that affect our decision-making process and try to reduce their influence on the same. This assignment deals with creating a low-cost experiment to find a solution for the challenge, and which could be used to verify its underlying hypothesis. However, every bias has different nature in terms of its triggers, effects and its entrenchment in our thinking process. If we assume that the solution to a problem is a function of the problem, then it would imply that every bias will have its unique solution. Therefore, honing in on a particular bias and designing a solution for the same will yield better results than an approach that sweeps all the biases under an umbrella and provides a common solution for them.

For this assignment, confirmation bias is chosen. Confirmation bias is the filter through which a person sees the world. Therefore, a solution for the same would be a falsification of our beliefs i.e, looking for pieces of evidence that disconfirm the initial assumptions, instead of looking for pieces of evidence that confirm them.

Prototype: The prototype of the solution is:

Part 1: Getting the person to know about his/her confirmation bias

Part 2: Steps for falsification

It is based on the underlying hypothesis that if a person gets to know about his confirmation bias and is presented with alternatives, he will be better equipped to undertake any new information in the future.

Low-Cost Experiment

It is a session of 3-5 candidates in a closed room and is led by a moderator. 

Part 1 of the prototype: Getting the person to know about his/her confirmation bias

Step 1  Moderator gives candidates a set of statements and asks them to write their inferences about the statements, on a piece of paper. For example, the statements might contain a description of an individual and the candidates would be required to guess the profession or nature of the said person. 

Step 2 After which, the moderator contacts each candidate individually and asks him to justify his inference. He then writes all the assumptions and justifications made by the candidate verbatim, trying to be as objective as possible. Please note that the candidates are not allowed to discuss anything among each other at until this step.

Step 3 Following table is created by the moderator:

Candidate

Inference

Assumptions/Justifications

This table is then made visible to every candidate (probably on a whiteboard or as a slide) and a discussion is started by the moderator. Now since it is not necessary that inference by every candidate is the same, therefore, every candidate gets to see what alternatives could be inferred from the given information. The moderator then tells them about their different assumptions and helps them to realize that whatever they inferred was based on their very subjective assumptions

Part 2 of the prototype: Steps for falsification of the underlying beliefs

Step 4  Now that the candidates have an idea of why they came to a particular inference, the moderator tells them about the falsification of their beliefs, i.e., he tells them that to counter their underlying assumptions, they should start finding pieces of evidence that are contrary to their beliefs. This is a group exercise where every candidate tries to give new inferences for the old set of assumptions and new assumptions for the original set of inferences. Therefore, the table looks like:

Candidate

Inference(old)

Assumptions(new)

Inferences (new)

Assumptions(old)

Venn Diagram 

This process aims to let the candidates realize the wide variety of possibilists that could arise from the given data set, i.e, there is more to the Venn diagram described above apart from the intersection area.

StoryBoard

Panel 2: Person tries to process the information, but he is not aware of his assumptions and the various dataset that the new information could come from

Panel 3: Hence, the uninformed solution

Panel 4 : Person is aware of his assumptions and the various dataset that the new information could come from

Panel 4: A more holistic solution

Resources required

1) Financial resource: low (costs of paper, pen etc.)

2) Human resource: Medium (more people would generate more inferences and increase the breadth of the solutions)

3) Time: 1-2 hours (of focussed attention)

Success Parameters

The success of the session could be defined as the number of data points in both the said tables. However, unless these non-intersecting Venn diagram areas become embedded in the candidate's psyche, we cannot say that it was a successful session. Only timely reviews from the candidates could tell us how effective the session was.

Points of caution

1)    The moderator should copy all the inferences made by the candidates verbatim. Any confirmation bias here could be resolved by having more than one moderator.

2)    The set of statements given to the candidates should be the same among the candidates, and it should be made sure that no candidate interacts with each other in the initial stages of the session.