Wednesday, April 6, 2016

E is for Experimenter Expectancy Effects

I've talked so far about how groups influence individuals and their behavior. But psychology has also contributed to our understanding of what controls (that is, what forces outside of what we are studying do we need to hold constant?) are necessary when we do research on people. And this contribution impacts many fields beyond psychology, such as medical research (like drug trials). One important thing to keep in mind is what effect the researcher might have on the participant, and how that researcher might inadvertently influence the participant's behavior. We call these "experimenter expectancy effects."

To take a step back and get (briefly) into the history of psychology: Psychology as a field grew out of two other fields - the physical sciences (like physics, chemistry, and so on) and philosophy. Early psychologists used methods from these respective fields for some of their first studies on psychological topics. But they learned that controls necessary in a physics study differ from those needed when studying people, so they had to develop new methods.

In the early 1900s, a man named Wilhelm von Osten gained media attention for his horse, Clever Hans.

Von Osten claimed that Hans could perform arithmetic, read and understand German, and keep track of time and the calendar. Von Osten would ask Hans a question, and Hans would respond by tapping his hoof. The German board of education sent psychologist Carl Stumpf to investigate the claims. They discovered that Hans was responding to the questioner's posture and facial expression to know when to stop tapping. (So Hans really was Clever, but at picking up social cues, not at arithmetic.)

Another famous example of experimenter expectancy effects came from research in the 1920s and 1930s at the Hawthorne Works factory in Cicero, Illinois. The researcher wanted to discover the best light conditions to maximize productivity in factory workers. They found that, regardless of lighting conditions used, worker productivity increased when changes were made, and decreased when the study ended. The conclusion is that the workers were more productive not because of the conditions, but because they knew they (and more importantly, their productivity) were being observed. This concept later became known as the Hawthorne effect.

We are social creatures, and we look for cues from others to make sure we're responding the way we're supposed to. This is great if you're in a new situation and want to fit in, but not so great when you're participating in a study. For this reason, new controls had to be added to studies of people to make sure the experimenter isn't unconsciously influencing the participant.

One way is through blinding. Single blind means the participant does not know what is being studied, or at least, what is being changed about his/her situation (the independent variable) to affect his/her response (the dependent variable); double blind means the experimenter also doesn't know what condition (level of the independent variable) the participant is in and/or what outcome is expected. Instead, a person who doesn't interact directly with the participant knows that information and uses it when analyzing data. In drug trials, this means that some people get the real drug and some get a placebo, and neither the participant nor the experimenter knows which the participant received.

But wait, there's more! There's another famous study of experimenter expectancy effects that I'll be blogging about later!

1 comment:

  1. It's really amazing that scientists have found tat even the tiniest particles of the universe behave differently if they are observed, and so many scientific papers contradict each other you have to wonder if they ever get to actually do research without influencing the outcome in some way or another! ~Liz