As part of the scientific method, a researcher may formulate a
hypothesis from which a prediction is derived. That prediction may
then be tested experimentally, and may be found to support or
contradict the hypothesis.
The methodology of the experiment may be flawed in such a way that it
will tend to bias the outcome, in which case there is a methodological
Consider a street survey where people are asked how much free time
they have. The survey may conclude that people have lots of free time.
However, because the survey is only based on people who had enough
free time to complete the survey, it may not be representative of the
population at large.
Consider a survey of the movement of animals. If the animals are
fitted with radio transmitters whose range is too limited, those
animals who move outside the range of radio reception will not be
counted in the survey, and the survey results will under-report the
average range of movement.
Consider a survey of educational level in a country where literacy
levels are low. If the survey involves written responses, it will show
a higher educational level than a similar survey conducted verbally -
because only literate people will complete the written survey and they
are likely to be more highly educated than the population in general.
These are examples of methodological bias - where the research is
flawed because of fundamental problem with the methodology of the
study (as opposed to, say, computational errors).
Google search strategy:
"methodological bias" "scientific method"