Off Topic – Bus Stop Seat Heights

Welcome to my photoblog.  Today we’re off topic to reveal the variation observed in Chennai Bus Stop seat heights.  Why?  As a photographer one observers the light, determines the subject, creates the composition,… which is the art of “seeing”.

Over my 40 years in the Quality profession, I’ve been challenged to get the mean (average) on target to the standard while reducing the variation (standard deviation) whether it was a product dimension, a business process output or a production process parameter.  This has influenced me as to how I see the world.

As my eyes wander whether on a production shopfloor or the back seat of a vehicle, I’m always observing the environment and its variation.  One item that caught my attention is the varying heights of bus stop seats.

The variation seen for Chennai bus stop seat heights is only “a reflection of Chennai’s unique character” which by the act of variation, creates those wrinkles of interest forming our Chennai memories.  
Thanks for the memories.

I’ve observed the full spectrum of people waiting at Chennai bus stops with their “knees up in one’s torsal”, “legs in a full dangle” and “leaning against the seat” as it was too high to sit.

So I set out to understand the statistics of this issue with the premise that there must be assignable cause to the high variation and not just common cause variation one would observe in a normal distribution.

Extreme 1:  11″ Height in Action


Extreme 2:  36″ Height in Action


The “Stats”

1)Frequency Count, Sample n=40

Frequency Count With Outlier

2)Summary Statistics

Given a p-Value of 0.013, the data failed the test for normality – not a normal distribution as the value is not > 0.050.  There must be a special cause driving the

Mean 22.14′, StdDev 4.52, Min 11″, Median 22.5″ Max 36″

Statistical Summary With Outlier

Short Poll

Inferences and Conclusion

I Googled the ergonomic dimensions for bench set heights finding a 17″-19″ targeted range.  
If we accepted 18″ as our target then the 4″ shift on the sample average with a range of  25″ leads one to conclude we have special cause(s) at work as confirmed by a low p-value for normality.  This variation is not normal.
I leave the root cause of the special variation to the readers.

Post Script

After finishing this post, we found even a shorter bench seat height of 10″ for which I didn’t re-do the statistics.  Is there one taller than 36″?