Abstract # 11:

Scheduled for Thursday, June 19, 2008 11:30 AM-11:40 AM: Session 1 (Meeting Room 2DEF) Symposium

Multimodal data input using touchscreen technology, voice recording and speech recognition in adverse field conditions

C. S. Ionica, K. L. Robbins, M. F.S.X. Novak and S. J. Suomi
Laboratory of Comparative Ethology, NICHD, NIH, NIH Animal Center, P.O. Box 529, Poolesville, MD 20837, USA
     Field conditions sometimes negatively impact computerized data recording, requiring researchers to change protocols, increase staff, or employ voice and video-recording to replace computers. We present a multimodal computer interface with touchscreen input and timed voice recording to retain the advantages of computerized recording, yet is adjustable to field conditions and single operator skills. Portable touchscreen packages available for behavioral data input (Noldus®) have small screens and do not use voice input, while specialized voice typing software (Nuance®) requires major adaptations for field deployment. Our alternative is a highly configurable software interface that integrates input from a pressure sensitive touchscreen and a wireless microphone in a tablet computer. The number and appearance of touchscreen elements are customized according to the project structure, screen dimensions, and observer preferences. The operator toggles between recording modes according to the rate of observed events and recording skill levels. Normally paced events are coded in full using the touchscreen elements, while fast paced events are time stamped in a voice file. Data editing allows instant access to each time stamp, voice playback, and manual code insertion, minimizing transcription time. Speech recognition may be used during data editing. The technique can be applied to most observation settings, but in extreme environmental conditions a rugged computer is required. This research is supported by the intramural research program of the NICHD at NIH.