Questionnaire format and item context affect level of belief in both scientifically unsubstantiated and substantiated phenomena.

Abstract
Belief in both scientifically unsubstantiated and substantiated phenomena was assessed with questionnaires that differed in format. The list format required respondents to check the phenomena they believed in. The graded format asked respondents to indicate the strength of their belief in each of the phenomena. In a variety of replications with university students, the list format always gave substantially lower estimates of the extent of belief. The discrepancy between Gallup-style survey results and previous university sample results can be attributed almost entirely to format differences. The context of the phenomena also had a significant influence on estimates of the extent of belief. It is argued that the Gallup-style surveys seriously underestimate the extent of belief in scientifically unsubstantiated phenomena. The high levels of belief in scientifically unsubstantiated phenomena are discussed as being symptomatic of students'' generally poor ability in assessing the quality of evidence.

This publication has 3 references indexed in Scilit: