Variables

• In the simplest cases, cause and effect are important in hypothesis generation.
• Increasing the amount of oxygen causes the candle to burn faster.
• Again, in the simplest case there is an independent variable, the amount of oxygen, and
• a dependent variable, how fast the candle burns.
• One key is the measurement of these variables. There has to be some mechanism to measure.
• This of course has brushed over the problem of controlled variables. Even in a simple case like oxygen and candles, there are a lot of potential problems.
• Candles, temperature, wind are ones that I expect really matter, but what about other things like, time of day, light, presence of a person. It's really hard to say what might be a problem.



• This one independent variable one dependent variable everything else controlled type of set up is the best.
• If you can do this, you have a pretty strong claim to truth.
• One way you can slip is to use mutliple variables.
• Of course statistics can help here.