Given we can make causal claims from experiments and not from correlational studies, why don't we just experiment all the time? Give at least three different classes of reasons why we would do a correlational study instead of an experiment. This assignment should be about 1-2 single-spaced pages in length.

Jessica Bick Law
Abnormal Psychology
Module 3
9/26/18

A Time for Experimentation:
Causality vs Correlation

	Many of the most enlightening discoveries of the last several thousand years have been results of focused experimentation. Alchemists like Henning Brand in the 17th century sought via experimentation the “philosopher’s stone,” which would turn base metals into gold. What he accidentally discovered instead was phosphorous. Brand’s discovery was accidental, but his process was not. He began with a hypothesis: some stone as yet to be discovered would transmute metals into gold. Then painstakingly he went through a variety of combinations of materials, logging each process, hoping that some combination would fulfill his hypothesis. Of course, he never discovered the “philosopher’s stone,” but because he followed the scientific method and kept notes on each experiment, he was able to record the way in which he discovered phosphorous so that the process could be replicated. In psychology, it is not as easy to simply experiment until some result appears. Psychological experiments involve real people putting their lives at risk in the hands of researchers. Even though experiments produce causal data that correlational studies often cannot, there are many reasons to choose a correlational study over an experiment when we consider them in terms of neuroethics.
	The correlational approach is “designed to help us understand how specific factors are associated with one another” without any outside manipulation (Ray, 2015). Rather than relying on human analysis, we use a technique called the correlational coefficient that determines if there is an association within a relationship (Ray, 2015). A positive correlation on a scatterplot would look like a set of dots trending upwards along both the X and Y axes in tight concentrations. A negative correlation would go from a positive measure on the Y axis towards 0, while the X value increases accordingly. This technique, establishing that a relationship between two things exists, “may be the first step in dealing with a complex problem” (Ray, 2015).
	Correlational studies involve no manipulation by outside forces and thus cannot negatively influence a psychological subject. For instance, in our book it mentions that some scientists wanted to study the effect of dissociation during torture on the emotional outcome. It would be unethical to torture humans just to find out if there was a causal relationship between dissociation during torture and aftereffects like PTSD, so in this case correlational evidence is preferred to experimentation (Ray, 2015).
	Not only can it be unethical to experiment on people, often it is impossible to induce certain disorders in order to study them. We cannot induce autism, schizophrenia, or depression, for example, in healthy people in order to see what effects they have on the brain and body. With these conditions we must rely on correlational data. Likewise you could observe a pair of monozygotic twins as they aged to see what epigenetic quirks they displayed. For instance in some cases one twin would develop schizophrenia while the other wouldn’t. Experimentally it is impossible to develop two identical models, and researchers must rely on twins. There are studies that compare twins reared together and reared apart, but it would be unethical to purposefully separate a pair of twins at birth just to see what effects proximity and rearing had on each twin. Once again correlational studies surpass experimentation in this regard.