Modeling human behavior is one of those existential questions, that a social scientist dreads. After all Hard Scientists can do experiments is controlled circumstances in what ever conditions they require, but the poor Social Scientists doesn't has this luxury. So what's the solution? Should a Social Scientist just stop theorizing because the basic element of Social Science Theories i.e. "The Human Behavior" is just not model able.
Economics is sometimes referred to as "Dismal Science", and the Hard Scientists from Physics to Mathematics mock economists for their "Theories of Everything based on Nothing". But let's think objectively, is Human behavior really un model able?
Our best shot at modeling human behavior is to take its as completely random, which most of the social scientists do. But how can we prove that human behavior is just random? There are many models like the random walk model etcetera, that professes randomness of human behavior, but these models are not backed by hard data.
In this age of Data Deluge, theories can be backed by hard evidence, and the Social Scientist has an advantage in this regard. Since most of this data is produced by humans; that if we take out data produced by embedded systems in internet of things (IoT) devices, the social scientists can take all that data to model human behavior. Scientific experimentation on human behavior can be taken to the next level by formulating more realistic theories, and put them to test straight away using data. Then the cycle goes on, keep what words, discard whet doesn't.
The zenith of this approach is to look at the data first i.e. The Data Driven Approach, which is so abundant, and then recognize patterns. Once a true relationship is determined, only then theorize or what ever you may call it. Its not even theorizing, because theory is an endeavour to understand reality, but when reality is driving theories, then what would you call it, may I suggest "Realizations".
Social Scientists had a hard time explaining to newbies and novices that since elections don't have emotions, its easy to model them. Question is, how to model humans, who are a set of extremely complex emotions?
But that difficulty may be a thing of past, because now social scientists have so much data to work with, produced by humans on social networks, which is a minute subset of the mammoth amount of data available on internet, produced by humans. So my dear poor Social Scientist, you might not even know it, but you have truly arrived, you just have to open your eyes to see that.
Tough luck Hard scientists, you can't really delve that deeply into the emotions of electrons, as they don't have the ability to produce data themselves. Hard Scientists have to wait for the advent of particle embedded systems i.e. minute chips which can be put on electrons to give insights into their behavior, but as per the rate of technological progress, it's a long shot.
The source of this article was taken from: https://www.linkedin.com/pulse/modeling-human-behaviour-ammar-a-raja