Modeling based on understanding

The models/tools that we use should adapt to our understanding of what we are modeling not the other way around: changing our understanding of something because it cannot be modeled!

Example

Say you want to use some objects in a Block Diagram (a modeling tool) to represent an individual (an object/system in the real world). Should I consider personality traits as "attributes" or "methods"?

Example: Some people angry and yell at others. Is it an attribute or a function (method) for the "Individual" block/class?

Let's say I don't know how to model this because I am not familiar with the tool quite well, yet.

• I can model this as an attribute ("yells=true/false"). But this means that all instances of this class (all people) constantly yell!
• or alternatively, as a method ("yells()"). It means people can yell; this is correct but how can I use it to distinguish various instances (people are different)

Or, I can stop there, thinking, what is my accurate understanding of an individual (one instance)? "They yell on average once a month" and now, how can I use attributes and methods to describe exactly that? Maybe have an attribute "yelling rate per month=1" and keep "yells()"

Later I you can keep working on the method's inputs: what are the inputs that the yells() function uses to decide if the person will actually yell?!