The phenomenon of expectation-related observation : an exploration of nature, associations and causes
In the context of the current study, an expectation-related observation (ERO) was defined as the biased collection or interpretation of data, influenced by a desire to reach a predetermined conclusion. EROs are common in science classrooms, and consequences are thought to be largely problematic, encouraging an adherence to unscientific methods and the continued existence of misconceptions. In order to investigate EROs in a school science setting, three largely quantitative research phases were devised, the general aims of which were: a) Exploration into the nature of EROs. b) Determination of whether students who were committing EROs differ from other, more scientifically behaving individuals. c) Finding out why students ERO. d) Testing if encouragement of EROing could actually improve learning. To provide further focus for the study and help differentiate the many distinct issues in play, seventeen operational research questions (RQs) were formulated from these general aims, and the three phases were planned and executed using the RQs as a framework. The general lesson phase involved delivering a set of fifty lessons, enabling EROing to be explored and recorded. The convergent phase allowed for variable reduction, supplying ‘cleaner’, less cluttered data, and examinations of associations between EROing and other variables were possible by means of questionnaire and interview methods. The final, experiment phase was a series of educational experiments to determine the effectiveness of a particular type of practical lesson in correcting a misconception. Findings from the three phases demonstrated how student observations of experiments were frequently biased by personal theory. A variety of behaviours are described where learners purposely manipulated apparatus, invented results or carried out other improper operations to either collect data which they believed were scientifically correct, or achieve social conformity. Unconscious behaviours guided by these same motives were also thought to have occurred. Rationalisations were offered, though were frequently irrational; however, honest admissions of inappropriate evidential considerations were forthcoming at interview. Generally, EROing students differed inherently from more scientifically behaving peers, tending to be male, less able, high-risk takers and competitive. Inferences from triangulated data proposed that a trio of initiators constituted motivation for EROs: a desire to determine the scientific answer, a need to co-ordinate data with theory, and a wish to confirm the observations of peers. Affectual arousal provided a conduit so allowing these internal influences to drive ERO-related behaviours. Notwithstanding the often undesirable effects of EROing, an intervention is described where EROs were encouraged in attempts to boost affectual arousal and increase engagement with the science, and the proposition that these processes facilitate memory and favourable conceptual change was supported by experimental data.