Banca de QUALIFICAÇÃO: DIÊGO CESAR LEANDRO

Uma banca de QUALIFICAÇÃO de DOUTORADO foi cadastrada pelo programa.
STUDENT : DIÊGO CESAR LEANDRO
DATE: 11/12/2019
TIME: 14:00
LOCAL: LAB 1 Instituto Ágora
TITLE:

Pre-task planning, Working Memory capacity and graph analysis: An exploratory study on L2 speech production

KEY WORDS:

Pre-task planning. Working memory. L2 speech production; English; Graphs.

PAGES: 100
BIG AREA: Linguística, Letras e Artes
AREA: Linguística
SUBÁREA: Linguística Aplicada
SUMMARY:

This research investigated the effects of the relationship between pre-task speech planning and working memory capacity (WMC) on L2 speech production (L2SP). Thirty-six individuals integrated the cohort of the study; they were 26 learners of English from Instituto Ágora and 10 teachers from Núcleo de Línguas (NucLi); both at Universidade Federal do Rio Grande do Norte (UFRN). Learners performed an oral task in which they were required to make a one-minute description of a picture, under three experimental speech planning conditions: (1) no-planning, (2) written planning and (3) oral planning. Immediately after performing the oral task, participants recorded a verbal protocol (ZACCARON, 2018) aimed at investigating the one-minute (MEHNERT, 1998), unguided planning processes (D’ELY, 2006; GUARÁ-TAVARES, 2008, 2016; ZACCARON, 2018). Finally, participants performed the Speaking Span Test (SSPAN) (DANEMAN; GREEN, 1986; DANEMAN, 1991; FORTKAMP, 2000; WEISSHEIMER, 2007; GUARÁ-TAVARES, 2008; PREBIANCA, 2009), aimed at assessing WMC and its relation to speech production. Participants’ oral performance was examined in terms of fluency (speech rate unpruned, speech rate pruned and silent pauses per minute), accuracy (errors per 100 words) and lexical density (proportion of words produced with grammatical and lexical properties). According to scores in the SSPAN, participants were placed in the high- or low-WMC group. Additionally, learners’ speech samples were compared to NucLi teachers’ speech samples through a graph analysis conducted with the software SpeechGraphs (MOTA et al., 2012, 2014, 2016, 2019), developed at the Brain Institute at UFRN. The study sought to investigate, in an exploratory manner, how speech graph attributes could sort out fluency patterns of learners who were still at an early stage of interlanguage development and the fluency patterns of near-native bilinguals. Speech measures were statistically compared according to planning conditions and WMC. The predictive power of graph analysis was also examined quantitatively, through correlational analyses. In turn, responses to the verbal protocol were examined in a mixed manner (DÖRNYEI, 2007), based on the interpretation of participants’ choice of words and the frequency with which the words occurred in the responses, represented graphically as wordclouds (SOUSA, 2014; LEANDRO, 2015; LEANDRO; WEISSHEIMER, 2019). Preliminary results of the comparison between planning conditions versus the no-planning condition indicate a statistically significant result on participants’ oral performance, especially regarding fluency (speech rate), independently of WMC. When memory is considered, no significant statistical results were found, differently from the expected, since low-WMC participants did not outscore high-WMC participants. This study may contribute to theoretical discussions regarding L2SP, the pedagogical treatment of speech planning and the use of speech graph analysis in the field of Second Language Learning.

BANKING MEMBERS:
Presidente - 1666189 - JANAINA WEISSHEIMER
Interna - 1153427 - MAHAYANA CRISTINA GODOY
Externa à Instituição - RAQUEL CAROLINA SOUZA FERRAZ D'ELY - UFSC
Externa à Instituição - NATALIA BEZERRA MOTA - UFPE