Response quality in survey research with children and adolescents: The effect of labeled response options and vague quantifiers

Instead of using their parents or caretakers as informants, children are increasingly the principal informants about their own experiences, perspectives, attitudes, and behaviors (Scott, 1990. As a consequence, it is becoming an important question how well one can ask survey questions of young children and adolescents. Concerning adults, there is a substantial body of empirical evidence that shows effects of both respondent characteristics (especially cognitive abilities) and question characteristics (especially question difficulty) on response quality (e.g. Alwin Krosnick, 1991; Krosnick Fabrigar, 1997; Narayan Krosnick, 1996; Schwarz Hippler, 1995; Schwarz Knauper, 1999; Schwarz, Park, Knauper, & Sudman, 1998). Because children are still developing their cognitive and social skills, the quality of their responses to survey questions is of special interest. For example, earlier studies on surveying children have shown that vague and ambiguous words should be avoided (Borgers, 2002; Borgers Hox, 2001; De Leeuw & Otter, 1995), and that completely labeled response options helped children to produce more reliable responses (Borgers, 2002).