Experimental Design

Conducting an experiment

Can be set up in an office or a tent

You can do it anywhere

Even in your underwear

But if you need a group

Like a dancing troupe

Alone and together each plays

There are two standard ways

The between-group and within-group

Choose which method before each tastes the soup

Each member of the between-group

Only gets to taste one of each soup

However, each member of the within-group

Is permitted to try each different one

It depends always on what it is you want

Choose within-group if you want to know

what each thinks of each kind

Or between-group should each be blind

To all the other options they could have had

And in this case, NO learning is good not bad

Though there is one more choice

But do not yet rejoice!

It takes quite some organisational skill

To craft this hybrid experimental drill [1]

And blend between and within

For this new test type to begin

Allow each player to taste one soup

Then move them around as a troupe

Shifting their places along by one

To try a different soup on each run

Keep moving them around the table

Until each taster has been able

To offer an opinion or a thought

On whether the score is one or naught

Giving a label

To each soup on the table

Stating when each soup was tasted

First, second or third position noted and pasted

To indicate the objectivity of their opinion

On whether the soup tasted like carrot, pea or onion

And if their taste buds got confused

By the third tasting, even bemused

At least each soup was tried on the first run

Clear of thought, on the mind only one

This omits learning and fatigue

Instead curiosity and intrigue

But on the first run also consider

The individuality of the figure

What each person brings to the table

Affects the opinion placed on the label

Subjectivity is key to an independent opinion

Without it we would all be minion under imposing dominion

But this is a topic for another time

For now this is already quite a long rhyme

Just to round off with a conclusion

And offset any delusion

Weigh and consider all the labels on the table

The testers have placed willing and able

Giving thought to cons and pros

When choosing the best tongue and nose

For the direction of your results

Will stir up the soup and raise exults!

[1] Jonathan Lazar, Jinjuan Heidi Feng and Harry Hochheiser (2010) Research Methods in Human-Computer Interactions, p49, sect. 3.3.2.

Advertisements