Experimenter Bias

[Creepy-]

Experimenter Bias

As scientists, psychologists try to conduct their studies in an objective, unbiased way so that their own views will not influence results. However, objectivity is a goal that scientists strive for, not an accomplished fact that can be taken for granted (MacCoun, 1998). In reality, most researchers have an emotional investment in the outcome of their research. Often they are testing hypotheses that they have developed themselves and that they would like to see supported by the data. It is understandable, then, that experimenter bias is a possible source of error in research.

Experimenter bias occurs when a researcher's expectations or preferences about the outcome of a study influence the results obtained. Experimenter bias can slip through to influence studies in many subtle ways. One problem is that researchers, like others, sometimes see what they want to see. For instance, when experimenters make apparently honest mistakes in recording subjects' responses, the mistakes tend to be heavily slanted in favor of supporting the hypothesis (O'Leary, Kent, & Kanowitz, 1975).

Research by Robert Rosenthal (1976) suggests that experimenter bias may lead researchers to unintentionally influence the behaviour of their subjects. In one study, Rosenthal and Fode (1963) recruited undergraduate psychology students to serve as the "experimenters". The students were told that they would be collecting data for a study of how participants rated the success of people portrayed in photographs. In a pilot study, photos were selected that generated (on the average) neutral ratings on a scale extending from -10 (extreme failure) to +10 (extreme success). Rosenthal and Fode then manipulated the expectations of their experimenters. Half of them were told that they would probably obtain average ratings of -5. The other half were led to expect average ratings of +5. The experimenters were forbidden from conversing with their subjects except for reading some standardized instructions. Even though the photographs were exactly the same for both groups, the experimenters who expected positive ratings obtained significantly higher ratings than those who expected negative ones.

How could the experimenters have swayed the participants' ratings? According to Rosenthal, the experimenters unintentionally influences their subjects by sending subtle non-verbal signals as the experiment progressed. Without realizing it, they sometimes smiled, nodded, or sent other positive cues when participants made ratings that were in line with their expectations. Thus, experimenter bias may influence both researchers' observations and their subjects' behaviour (Rosenthal 1994).

In the realm of Socionics we almost invariably act as both experimenter (or rather, observer) and subject. Due to bias we may ignore information that contradicts with our own ideas, or unintentionally misinterpret the information to comply with our ideas.

The bias also applies to our analysis of others and their perceived type.

I don't know how much we can do about the issue (double-blind studies are quite impractical for us), but I think we should at least be aware of it.