摘要：This paper discusses two things researchers should consider when selecting tasks for
cognitive noise studies and interpreting their findings: (a) The “process impurity” problem
and (b) the propensity of sound to capture attention. Theoretical and methodological
problems arise when the effects of noise on complex tasks (e.g., reading comprehension)
are interpreted as reflecting an impairment of a specific cognitive process/system/skill.
One reason for this is that complex tasks are, by definition, process impure (i.e., they
involve several, distinct cognitive processes/systems/skills). Another reason is that sound
can capture attention. When sound captures attention, the impairment to task scores
is caused by an interruption, not by malfunctioning cognitive processes/systems/skills.
Selecting more “process pure” tasks (e.g., the Stroop task) is not a solution to these
problems. On the contrary, it introduces further problems with generalizability and representativeness. It is argued that cognitive noise researchers should employ representative
noise, representative tasks (which are necessarily complex/process impure), and interpret
the results on a behavioral level of analysis rather than on a cognitive level of analysis.
关键词：noise; process impurity; cognition; behavioral level of analysis; theory; method
