[News]

News: Welcome to The Oktave Forum.

[sri]

sri

sri

Show sri's last posts.

Show general stats for sri.

• 
• Administrator
• Sr. Member
• Karma: 7
• Posts: 264

As a teacher, one of the biggest challenges I face is to have this synchrony in abstractions. This even more so with theoretical subjects. A similar problem is also faced by practitioners of being in sync with the mental models of users.

Here is one example. We all have used elevators sometime or the other. The usual abstraction here is that, if we need to go up, we push the up button while waiting for the lift and if we need to go down, we push the down button and wait.

But I came across a completely consistent alternate mental model from an office boy in an office. Say if you are on the 6th floor and the lift is in the 1st floor (indicated by the panel reading), he pushes the up button to bring the lift up to the 6th floor. And if the lift were to be above, say in the 9th floor, he pushes the down button to bring the lift down to the 6th floor.

No matter how hard I try, I can't see any inconsistency in this mental model! It is perfectly implementable and works. I just found myself asking, "yeah, why not?"

I can only imagine the kinds of things that happen due to impedance mismatch of mental models, with purely intangible stuff like software..

So how do we know which mental model is the "right" one? Or the "better" one to use? How do we know that when interacting with someone, we are both understanding things in the same fashion?

[Abhijith]

Abhijith

Abhijith

Show Abhijith's last posts.

Show general stats for Abhijith.

• 
• Newbie
• Karma: 2
• Posts: 13

There is no right and wrong mental models.All you need to do is put a "what if" question and compare the student's answer with your expected answer.

[sri]

sri

sri

Show sri's last posts.

Show general stats for sri.

• 
• Administrator
• Sr. Member
• Karma: 7
• Posts: 264

Ok, let me put a "what if" question right away..

What if, I were a designer designing the elevator system? What model should I use? The first one which says, "Push the button in the direction of your intended travel" or the second one which says "Push the relevant button to fetch the elevator from above or below you"?

I'm sure the first one is somehow "more desirable" or "more intuitive" than the second. But it is not that obvious.

There are meta-heuristics like the "minimum description principle", "least-biased principle" etc. that designers often use. But yet, the entire question seems hopeless though.

Just imagine having to do a "what if" exercise for every one of the (say) 150 answer sheets that we need to evaluate!!

[Abhijith]

Abhijith

Abhijith

Show Abhijith's last posts.

Show general stats for Abhijith.

• 
• Newbie
• Karma: 2
• Posts: 13

Yes you have a point.
There is no way one can determine the mental model created in another persons mind without feedback.
So the communicator should be around to seek a feedback on the mental model created by my communication.
It all boils down to completeness of one time oneway communication.

[sri]

sri

sri

Show sri's last posts.

Show general stats for sri.

• 
• Administrator
• Sr. Member
• Karma: 7
• Posts: 264

Hmm.. no, I disagree again (sorry.. )

Here is an explanation why the former mental model for the elevator is better than the latter (office boy) one, even without any feedback from the users. This is based on the meta-heuristic called "minimum description principle". The idea here is that, if there are more than one ways to achieve the same thing, choose the one whose description is the smallest. Essentially the one that is easiest to explain. Again, the underlying idea behind MDL is rationality. As rational beings, we wish to optimize -- get the most benefit by the least effort.

The first design requires the user to know one piece of information (whether s/he is going up or down), while the second design requires the user to know two pieces of information (which floor s/he is in and which floor is the elevator on) and in addition, need to do a computation (compare my floor with the elevator floor) to decide which button to press.

Cheers
-Sri

PS: Just a gentle pointer to (http://forum.oktave.in/index.php/topic,7.0.html) to separate ideas from the person. My focus is only on the idea; no disagreement with the person behind the idea

[bharath]

bharath

bharath

Show bharath's last posts.

Show general stats for bharath.

• 
• Newbie
• Karma: 0
• Posts: 1

Sorry for bringing an old thread alive. The one thing I don't understand with elevators is the necessity for two push buttons. Why do we need two buttons when one will suffice? All the elevator will know until I get into the elevator is which floor the person who wants to use it is on? I do not know if elevators are more complex and if the up/down buttons actually influence the decision to move up/down when more than one person wants to use it.


Addendum : I find the one button elevator to be the simpler model and easier to use.

[sri]

sri

sri

Show sri's last posts.

Show general stats for sri.

• 
• Administrator
• Sr. Member
• Karma: 7
• Posts: 264

Usually, if you (say) press the up button and the elevator is already in a floor above you, it will not stop on its way down, if there is a lower floor calling it. So, the two buttons serve to prioritize the order in which the floors have to be visited by the elevator..

[sundar]

sundar

sundar

Show sundar's last posts.

Show general stats for sundar.

• 
• Global Moderator
• Newbie
• Karma: 0
• Posts: 2

Here is one example. We all have used elevators sometime or the other. The usual abstraction here is that, if we need to go up, we push the up button while waiting for the lift and if we need to go down, we push the down button and wait.

But I came across a completely consistent alternate mental model from an office boy in an office. Say if you are on the 6th floor and the lift is in the 1st floor (indicated by the panel reading), he pushes the up button to bring the lift up to the 6th floor. And if the lift were to be above, say in the 9th floor, he pushes the down button to bring the lift down to the 6th floor.
...
So how do we know which mental model is the "right" one? Or the "better" one to use? How do we know that when interacting with someone, we are both understanding things in the same fashion?

It's not just that one person whom you saw, but nearly 50% of employees in my previous office used to have the other mental model. From a systems perspective, this calls for a better interface that, at least, conveys what model the system designer had assumed, if no single "right" model could be found.

Some of these are "equivalent" protocols that get learned, taught, or enforced. Take driving protocols, transliteration schemes, writing systems, etc., for examples. Having said that, I don't have an answer to the "how do we know" question.