Bias, Misinformation and the Paradox of Neutrality

Abstract

What is normally described as bias? A possible definition comprises attempts to distort or mislead to achieve a certain perspective, i.e. subjective descriptions intended to mislead. If designers were able to exclude bias from informing systems, then this would maximize their effectiveness. This implicit conjecture appears to underpin much of the research in our field. However, in our efforts to support the evolution and design of informing systems, the way we think, communicate and conceptualize our efforts clearly influences our comprehension and consequently our agenda for design. Objectivity (an attempt to be neutral or transparent) is usually regarded as non-biased. However, claims for objectivity do not, by definition, include efforts to inquire into and reflect over subjective values. Attempts to externalize the mindset of the subject do not arise as part of the description. When claims to objectivity are made, this rarely includes any effort to make subjective bias transparent. Instead, objectivity claims may be regarded as a denial of bias. We suggest that bias can be introduced into overt attempts to admit subjectivity. For example, where people are asked to give subjective opinion according to an artificially enforced scale of truth-falsity (bi-valued logic), they may find themselves coerced into statements of opinion which do not truly reflect the views they might have wished to express. People do not naturally respond to their environment with opinions limited to restricted scales; rather, they tend to use multivalued logic. This paper examines the impact of bias within attempts to establish communicative practice in human activity systems (informing systems).