Citation

Howie PM, Roebers CM. Appl. Cogn. Psychol. 2007; 21(7): 871-893.

Copyright

(Copyright © 2007, John Wiley and Sons)

DOI

10.1002/acp.1302

PMID

unavailable

Abstract

The development of discrimination and realism was investigated in the event recall of 156 8-year-olds, 133 10-year-olds and 146 adults, using categorical confidence judgements. Target questions were either a mixture of misleading and unbiased ('non-bombardment'), or restricted to one question format ('bombardment'). The confidence judgements of all age groups discriminated between incorrect and correct responses to unbiased questions, but with misleading questions, this ability was undermined in the children, particularly when 'bombarded'. Calibration-style analyses of unbiased questions revealed a systematic confidence-accuracy association across age and question mix for unbiased questions. For misleading questions, however, the absence of a drop in performance from intermediate to low confidence at all ages suggested relative underconfidence at the lowest confidence level. At high confidence levels, there was evidence of realistic congruence between confidence and performance in adults, but this was not the case in the 10-year-olds when bombarded with misleading questions, or in the 8-year-olds, regardless of bombardment. Exploratory analyses of question difficulty revealed poor calibration across ages for difficult unbiased questions, and in the 8-year-olds, even for easy unbiased questions when intermixed with misleading questions. Bombardment with difficult misleading questions further undermined children's calibration. Implications for the role of social and cognitive factors in the development of metacognitive monitoring are discussed. Copyright © 2006 John Wiley & Sons, Ltd.

Language: en