Ask versus tell: Potential confusion when child witnesses are questioned about conversations

Stacia Roosevelt, Kelly McWilliams, Thomas D. Lyon

Research output: Contribution to journalArticlepeer-review

8 Scopus citations


Children's potential confusion between "ask" and "tell" can lead to misunderstandings when child witnesses are asked to report prior conversations. The verbs distinguish both between interrogating and informing, and between requesting and commanding. Children's understanding was examined using both field (Study 1) and laboratory methods (Studies 2-4). Study 1 examined 100 5- to 12-year-olds' trial testimony in child sexual abuse cases, and found that potentially ambiguous use of ask and tell was common, typically found in yes-no questions that elicited unelaborated answers, and virtually never clarified by attorneys or child witnesses. Studies 2 to 4 examined 345 maltreated 6- to 11-year-olds' understanding of ask and tell. The results suggest that children initially comprehend telling as saying, and thus believed that asking is a form of telling. As such, they often endorsed asking as telling when asked yes-no questions, but distinguished between asking and telling when explicitly asked to choose. Their performance was impaired by movement between different use of the words. Child witnesses' characterization of their conversations can easily be misconstrued by the way in which they are questioned, leading questioners to misinterpret whether they were coached by disclosure recipients or coerced by abuse suspects.

Original languageEnglish (US)
Pages (from-to)447-459
Number of pages13
JournalJournal of Experimental Psychology: Applied
Issue number4
StatePublished - Dec 2017


  • Ask
  • Child sexual abuse
  • Child witnesses
  • Linguistic development
  • Tell

ASJC Scopus subject areas

  • Experimental and Cognitive Psychology


Dive into the research topics of 'Ask versus tell: Potential confusion when child witnesses are questioned about conversations'. Together they form a unique fingerprint.

Cite this