Leibovitz, D. P. (2011) Local Measure Reliability vs. Global Concept Validity. Has Cognitive Science Moved Beyond Behaviourism? (Insignificant Progress in Validating Cognitive Constructs p<.05). Poster presented at the Institute of Cognitive Science Spring Conference (ICSSC) of Carleton University, Ottawa, Canada. [doi: 10.13140/RG.2.1.2792.8801]
Abstract: Every cognitive experiment contributes to the factual accumulation of raw, stimulus-response behavioural data. The raw data are factual/indisputable in that 95+% scientists understand and can reproduce the operationalized procedure and measures despite validity and interpretation concerns. Nevertheless, there has been zero factual accumulation of cognitive constructs and interpretations as there is no 95+% agreement nor comprehension in the sea of hypotheticals. Indeed, the signal to noise ratio worsens (entropy increases) with every experiment as new micro-theories are created, rather than a scientific reduction (convergence) to unity.
Links: