I wonder what's in the newspaper

today...

One sunday morning...

hmm?

I know Dr. Wedge! He's really smart! but

alcohol and coffee glowing? That's preposterous!

This looks like a job

for...

Arf!

AMAZING DISCOVERY!Scientist Dr. Wedge has discovered that coffee and isopropyl alcohol GLOW when mixed together!

Dr. Luke Fusion,

and his trusty dog, Quark!

steph Hoechst

moments later... Hey, Dr. Fusion! What brings you to my lab?

Oh, skeptical, are you? Well, I'll tell you

exactly how I made this

groundbreaking discovery!

I was carrying my coffee into the lab the other day...

Well, Dr. Wedge, I read about

your discovery in the

newspaper, and I was rather

skeptical. How did you draw

the conclusion that coffee and alcohol

glow?

Isopropyl Alcohol

There was also a vat of alcohol in the room that my lab assistants had been using to sterilize equipment.

Isopropyl Alcohol

OOPS!

Then I dropped my coffee in the alcohol!

I turned around to get a paper towel so I could clean up, and the mixture was glowing when I looked back!

!Isopropyl

Alcohol

I then drew the

conclusion that coffee and alcohol

glow when mixed together!

Say, Dr. Wedge,

have you ever heard of

experimental design?

I think I see the

problem here...

No...

Now I'm famous for

my discovery!

Well, you

overlooked several

aspects in drawing your conclusion. If you had followed

good experimental design, your conclusion

might not be so absurd!

A chance observation isn't enough to draw a conclusion! You

need to design an experiment to

test your theory that alcohol and

coffee glow. that way, you'll

be able to draw a logical

conclusion.

First, you need a hypothesis. This is your

starting point for further investigation. It tells

what you're testing for. A hypothesis is an if/then

statement. In this case, your hypothesis would

have been, "if i mix coffee and alcohol together, then it will generate

light."

If you designed an experiment based on this

hypothesis, then you would also need a prediction. a prediction

connects your hypothesis to what might happen in the

experiment. In your case, your prediction would have been, "If

i pour a cup of coffee into a gallon of isopropyl alcohol,

then the mixture should generate some kind of visible

light."

Normally, in an experiment, you would need gradations. We

don't really need gradations, or different levels of measurement,

because our experiment is a yes/no question.

However, if we do find that our mixture glows, then we can make a new

experiment about its different levels of light

that it emits.

A control is an experiment that doesn't receive the independent

variable. YOu need one of these to make sure that the coffee and alcohol

are behaving normally and not messing up your

experiment. valid controls would be

Testing the coffee and alcohol by themselves.

Every experiment has an

Independent and a Dependent

variable. These are the factors that you

are changing and observing in the experiment. The

independent variable is what you are purposefully

changing; in this case, the coffee that's being added. the

dependent variable is what is expected to

change when you change the

independent variable. In this

experiment, it would be whether the mixture glows.

Sample size is pretty self-

explanatory. It's the number of different

samples you draw your data from. For example, we would need to use

several different kinds of coffee and alcohol

to determine that there's no one type of coffee or alcohol that causes the mixture to

emit light.

Replication is repeating the

experiment several times to

get the most accurate results.

That way, you won't run into any

freak coincidences or

outliers impacting your data.

Constants are really important as well. also known as

controlled variables, these are the parts of your experiment

that must not be out-of-the-ordinary! That makes sure that you

can effectively measure the effects of the independent and dependent variables. for us, it would be the heat of the room, the material of the bucket that the alcohol goes into, basically anything that could

affect the results of the experiment.

...

That's a lot of info to consume...

If you tested this again, the coffee and

Alcohol probably wouldn't glow. Then you

could draw a more logical conclusion based on a

thoughtful experiment.

Now that I think about it, alcohol

and coffee glowing seems pretty improbable. STill, I wonder what made that vat

glow in the first place...

Woof! Isopropyl Alcohol

Arf!

I think Quark has found an

explanation!

Isopropyl Alcohol

I think something's fishy about this vat...

Let's put on some gloves for

extra protection...

Isopropyl Alcohol

And check out what's in

here!

Did you find

anything?

Yeah, I feel something.

A Glowstick? My lab assistants must have put that in there as a

joke!

If I had used experimental design, then I would have drawn the conclusion that coffee and alcohol Don't glow, and I wouldn't have

embarrassed myself in front of the scientific community!

And that, Dr. Wedge, Is the

power of Experimental

Design!