OBJECTIVE: The objective was to study the involuntary capture of attention by spoken words varying in intonation and valence.

BACKGROUND: In studies of verbal alarms, the propensity of alarms to capture attention has been primarily assessed with the use of subjective ratings of their perceived urgency. Past studies suggest that such ratings vary with the alarms' spoken urgency and content.

METHOD: We measured attention capture by spoken words varying in valence (negative vs. neutral) and intonation (urgently vs. nonurgently spoken) through subjective ratings and behavioral measures. The key behavioral measure was the response latency to visual stimuli in the presence of spoken words breaking away from the periodical repetition of a tone.

RESULTS: The results showed that all words captured attention relative to a baseline standard tone but that this effect was partly counteracted by a relative speeding of responses for urgently compared with nonurgently spoken words. Word valence did not affect behavioral performance. Rating data showed that both intonation and valence increased significantly perceived urgency and attention grabbing without any interaction.

CONCLUSION: The data suggest a congruency between subjective ratings and behavioral performance with respect to spoken intonation but not valence.

APPLICATION: This study demonstrates the usefulness and feasibility of objective measures of attention capture to help design efficient alarm systems.