Define some environmental problem.

Step 1: Tabulate the set of emotional and passionate responses and viewpoints that exist on the problem. These viewpoints can be irrational.

In the real world, progress beyond this step is rarely achieved and different groups just end up arguing over various irrational viewpoints, none of which can be supported.

However, to be effective in this arena, one has to be aware that such irrational viewpoints exist and to develop methods and tactics to diffuse them.

Step 2: Design experiments of data acquisition techniques that can actually address the individual issues raised above. Most of the group exercises will be devoted to this critical step.

Step 3: Perform __ unbiased __ data analysis and go where the data leads,
instead of using the data to support a cherished notion or personal
prejudice.

Step 4: Develop policies and solution spaces based on premises that can be defended and are consistent with the available data.

## Beware of three main problems with Environmental Data:
- Its usually very noisy
It is often unintentionally biased because the wrong variables are being measured to address the problem in question. A control sample is usually not available. |

Elementary Data Analysis:

The first steps are to produce a distribution (histogram) of the variables and define the statistical components of that distribution (usually just the mean and the standard deviation.

After doing that one wants to attempt a simple linear regression to search for correlation in the data.

Benefits of Linear Regression:

- Defines the dependency of one variable on another in a relatively
simple manner Linear: Y = ax + b; (most all
relations in nature, however, are non-linear!)
- Allows for predictions to be made for values of X larger than the
data set. Good for trend extrapolation (provided that the relation really
is intrinsically linear).
- Using the dispersion around the relation, you can quantify your predictive power.