The Inner Working Sympathy in the Eyes of a Data Scientist (Part 1)

Can you solve relational hardship using predictive modeling?

??As a high performer, you think you can change everything and you can make anything happen, right?

Until...you face 2 people?(not 2 things): your significant ones & yourself.

I know, you had tried your best:

- Common pattern: "If they changed this X, then I will change that y"

- Refined pattern: "If I can change my X (i.e. temperament), then the other will change their y."

You hit the wall again and again, but nothing seems to breakdown, except your self-confidence...being a data scientist.

I have been there & I am still there somehow sometime, but thanks God I am getting out the downward spiral quicker, with less efforts.

Here is a deep learning (mental) model that seems to work for me, as a data scientist (:

It's not to change your temperament, it's to practice & grow in an essential skill: Sympathy.

What is under the hood of Sympathy?

STEP1: Mirror their stories

Ask your significant ones what really happened to them. **Listen and paraphrase**:?

- "if I hear you right, this is what happened to you..." &?

- check back: "Did I get it right? Do I miss anything?".

Disregard whatever you know about what happened, mirroring is all about what the other party think it was the facts. Your quest is to know what their side of the story is, not yours.

STEP2: Validate their reasonings

Ask your significant ones what reasons they base & what purpose they want to achieve:?

- "Now it makes sense to me, you did y BECAUSE you want to do y', is it right ?".

- "I follow you from point to point & it now makes sense to me."

Disregard what you think is reasonable, here your quest is to know their reasons & their purpose. Validation is not for you to come to an **agreement** with the other party, but to **affirm** you hear their reasoning/logics.

STEP3: Sympathize their feelings

- "I imagine you must be feeling hard/upset/disappointed/..., Is that right?"

- "I understand that you feel..., it makes sense to me."

Sympathy is to putting yourself into their shoes and feel their pain. The risk for everybody is to stay on your comfortable shoes (: and criticize why the other react strangely with their seemly OK shoes. ):

Iterate through STEP1, 2, 3, 1, 2, 3, 1, 2, 3 ..., until you finish training this (mental) model:

Their stories = f(their purpose_reasoning, their feeling | context)

i.e. don't stop training your model until its train-validation gap increases i.e. your model overfits.

Hint: ensemble multiple mental models if you need to (:

What is your (mental) model(s) to solve relational challenges??How well it works for you? I am really curious.