Measuring (online) word segmentation in adults and children

We conducted three experiments in which we tested three groups of adults and one group of children on online and offline learning on a word segmentation task (Saffran, Aslin, & Newport, 1996). In the first experiment, we tested whether the click detection paradigm could be used to infer online word segmentation in adults (Gómez, Bion, & Mehler, 2011; Franco, Gaillard, Cleeremans, & Destrebecqz, 2015)). Participants listened to 8 minutes of uninterrupted syllables. The language consisted of four disyllabic words. Click sounds to which participants had to respond as fast as possible were added to the speech stream, either within words or between words. Sensitivity to word boundaries was hypothesized to result in faster response times to clicks between words compared to clicks within words. Offline learning was tested using a test phase during which participants had to choose between words (TP=1) and part-words (TP=0.333). We did not find evidence for online nor offline learning in the first experiment. In the second experiment, we tested whether adults and typically developing school-aged children would show learning of word boundaries in the same word segmentattion task without the click detection task. We did not find evidence for learning in either adults or children. Finally, in the third experiment we changed the test phase such that participants (adults) had to choose between words (TP=1) and non-words (TP=0), instead of part-words (TP=0.333). Again, there was no evidence of sensitivity to word boundaries.