## Academics - Experimental Design and Statistics

Steps Involved in the Scientific Method
compiled by Dr. Randy Alford, circa 2004

1.    State the Problem or Question

2.    Formulate the Hypothesis - what you believe will be the outcome of your
experiment or what you believe is the cause of a phenomenon you have observed.
This is your Research Hypothesis or Alternate Hypothesis. To conduct an
experiment, you will also need to formulate a Null Hypothesis.

3.    Design the Experiment

Factors to consider:
* Treatments to use to test the hypothesis
* Variables to measure/control – In experiments involving people the control often
requires more attention. (Pre-existing conditions and personalities need to be
accounted for, unlike subjects raised from embryos, these subjects are individuals.)
* Materials and Methods to use – Again, with human subjects this requires a little more
attention to carefully explain HOW you are accounting for individual history.
* Level of Precision of Measuring Tools (accuracy)
* Natural Variation in Population
* Choice of Sample (portion of population used in study)

4.    Make Observations / Collect Data - (i.e., write down your observations, the raw materials of research)

* common characteristic is variability

5.    Interpret the Data

* Look at the data; does it make sense at a glance?  Upon closer inspection?
* How can you explain the results?
* Statistically evaluate the Null Hypothesis

1. Draw Conclusions - Can/should we draw any conclusions about the data just yet, or make/take suggestions for improving upon the research design to be used in measuring the next experiments?
Statistics
We offer a Statistics class that allows students to learn about the type of statistical test they need to use for their particular experiment – why that test is appropriate, and how to interpret the mathematical results. We encourage the use of Vassar Stats  http://vassarstats.net/ for statistical calculations.