arrow_drop_up arrow_drop_down


Bias are faults in withdrawing information out of the empery. The term is original from radio signal technology and has to do with the clarity of a signal.

The origin of bias

A hundred years ago – nowadays maybe even more – the radio was invented and became a widely used medium from broadcasting messages and music. Due to distance, trees, house blocks, atmospheric interferences and so on, the broadcasted signal becomes unclear and the radio starts producing noise. The amount of noise was called bias. A clear received signal is unbiased.

What is meant with bias in measuring

In science the term bias means about the same. When a scientist want to gather information from the empery he needs clear input. Like in radiotechnology the input isn’t always clear. The amount of unclarity of the information is called bias.

Bias can be due to mistakes of the respondent (he is not understanding what is being asked, or he gives not the right answer, the answer he wants to give is not possible and so on) or due to mistakes by the scientist (not understanding the answer, misinterpreting the answer, wrong questions and so on). Bias can also be due to the environment. A survey that is answered with some friends telling the respond to hurry-up, isn’t a god setting for getting reliable answers. And a research that’s is being done in a laboratorial setting, might give a different information compared to a research that is accomplished in the natural settings.

Some type of bias occur only in special types of research. They are given separate names. If an observer makes a mistake it is called observer bias. An interviewer who misinterprets the information, it is called interviewer bias. If a respondent isn’t secure in his answer, it is called enquiry bias.

One way to reduce bias is – like in radio technology - collecting a lot of signals and mix these signals to a more clear one by computing a mean. If that is allowed should be tested by implying statistical techniques like factor analysis and computing a reliability score.

Related topics to bias:

  • Observing
  • Interviewing
  • Instrumentation
  • Transferring data
  • Generating data

  • Validity
  • Reliability
Place comment