There's a Madness in the Method

Abstract

Recent meta-analysis of research on citizen-initiated contacting of government officials revealed that the findings are basically an artifact of the method used. Macroanalytic studies using aggregate data find a negative or curvilinear relationship between contacting and socioeconomic status (SES), and microanalytic studies using survey data find a positive or null relationship between contacting and SES. This author argues that almost all survey researchers have used an invalid operational measure of contacting. Analysis of results from a Birmingham, Alabama, survey, which defined contacting in the conventional way and with specific, bounded-and aided-recall questions, strongly suggests that much of what researchers know or think they know about contacting is suspect.