Sharpening the Tool of Language: Examining Anchors and AMBIGUITIES

Language, though a powerful human tool, can often be imprecise. Such underspecifications often lead to misconceptions and misrepresentations in communication. These issues arise not only in everyday speech, but can also be present in forms of psychological evaluation. Many researchers, psychologists in particular, regularly rely on qualitative measures in the form of subjective response. Methods employing Likert scales use lexical choices to denote cognitive meaning, yet much variation in what is meant by specific words remain. Here we investigated the ambiguity that is inherently involved in such communication. We documented the responses of 94 participants on the quantitative value they placed on scale anchors such as “agree”, “disagree”, strongly agree”, etc. Participants rated each of these terms on a scale from -100 to +100. Also, participants rated terms related to reasonability and doubt on a 0 to 100 point scale. Results indicated that positive valence anchors fell significantly further from the midpoint of the scale, as compared to peer, negative valence anchors. Intervals between each anchor were not consistent across the spectrum. We concluded that the imprecision of language can be a significant source of confusion and discord leading to human-system failure regardless of the reliability of the technology to hand.