What Bias Means in Psychology
A systematic error that consistently pushes results in a particular direction, rather than a random mistake that cancels out over many observations.
This distinction matters. Random errors are unpredictable and tend to balance out across a large enough sample. Bias is different: it operates in the same direction every time, so repeating a biased study just reproduces the same distorted result. More data does not fix a biased design.
Bias can enter a study at any stage: in how participants are selected, in how the researcher behaves during data collection, in how results are interpreted, and in which studies reach publication at all. Because psychology draws conclusions about human behaviour from samples of people, each of these entry points is a real vulnerability.
Students often use "bias" and "error" interchangeably. They are not the same thing. An error is random and unpredictable. Bias is systematic and directional. Confusing them leads to weak evaluation in exams.
Researcher Bias
Researcher bias occurs when a researcher's own expectations or beliefs shape how a study is conducted or how its results are interpreted, without the researcher necessarily being aware of it.
The clearest demonstration comes from Rosenthal and Fode (1963). Student experimenters were given identical rats but were told either that their rats were "maze-bright" (selectively bred for intelligence) or "maze-dull" (bred for poor performance). The rats were, in reality, no different. Despite this, the "maze-bright" rats outperformed the "maze-dull" rats. The experimenters' expectations had changed the way they handled and interacted with the animals, and that behavioural difference produced a measurable effect on the outcome.
This finding showed that researcher expectations are not merely a background condition: they can actively alter the data collected. The mechanism here is subtle. The experimenters were not deliberately falsifying results. Their beliefs leaked into their behaviour in ways they could not control or even detect.
Researcher bias is sometimes called experimenter bias or experimenter expectancy effect. All three terms point to the same phenomenon: the researcher's hypothesis influencing the result.