Designing an Experiment

1
Designing an Experiment
2
The Scientific Method

• A problem solving technique based on a logical
  progression of steps

State the Problem
Gather info
Propose a hypothesis
Test a hypothesis
Analyze data
Conclusion
3
I have a great idea!

• As a scientist working for Exxon, I want to prove
  Exxon gets better mileage than Lukoil!

Lets design an experiment!
4
Gather info
5
Lukoil experiment
Fill a Hummer with Lukoil gas.
Drive as fast as you can
Drive the Hummer up Mt Pocono
6
Exxon experiment
Fill a SmartCar with Exxon gas.
Drive as slow as you can
Drive the SmartCar DOWN Mt Pocono
7
Analyze Data

• Lukoil 1 mi/gal
• Exxon 100 mi/gal

Lets make some commercials and sell some Exxon!!!
8
Fair Experiment?

• bias

Favoritism, or prejudice towards a certain outcome
Where does bias come from?
Past experience
Personal gain
Peer pressure
9
Variables in an experiment

• Variable factor which changes

How many variables should a fair experiment
have?
10
Variables in an experiment

• You need to change ONE thing at a time

Then do the experiment and see what happens.
11
Variables in an experiment
IndependentVariable

• You need to change ONE thing at a time

Then do the experiment and see what happens.
DependentVariable
12
Variables in an experiment
13
Variables in an experiment
These are sometimes called controls.
14
Independent variables answer the question "What
do I change?". Dependent variables answer the
question "What do I observe?". Controlled
variables answer the question "What do I keep the
same?"