O n f e r e n c e

Research questions 
Three main questions therefore guide this study.
1.
What impact do text-based chat-mediated tasks have on the quantity and quality 
(fluency, complexity, accuracy) of the language produced in follow-up face-to-face 
tasks? 
2.
What impact do text-based chat-mediated tasks have on perceptions of task-based 
self-efficacy in comparison with face-to-face tasks? 
3.
What impact do text-based chat-mediated tasks have on perceptions of task-based 
self-efficacy in follow-up face-to-face tasks? 
It is predicted that participation in text-based chat-mediated tasks will increase the 
complexity and accuracy of students’ oral production in follow
-up face-to-face tasks and 
have a positive impact on their levels of self-efficacy (see above). 
Method 
Participants: Chinese students, in particular, tend to have much poorer speaking skills 
than their other language skills (Guo, 2001). 20 pairs of Chinese (Mandarin) speakers of 
English as a second language were therefore recruited to participate in the study. They 
were all masters students studying English Language Teaching (ELT) and Applied 
Linguistics (AL) in the UK.
Design: The study had a within-participants design. The independent variable was task 
condition (text chat vs. face-to-face). The dependent variables were: (1) the quantity 
and quality of the language produced during follow-up face-to-
face tasks, (2) students’ 
task-based self-efficacy during the intervention, and (3) task-based self-efficacy for 
follow-up face-to-face tasks. Language p
roficiency, i.e. the students’ IELTS scores, was 
investigated as a covariate. 
Apparatus: All sessions were carried out in a quiet room equipped two laptop computers. 
Skype, a piece of software which provides facilities for text-based chat, was installed on 
both computers. The speaking tasks were provided to the students on paper. Students’ 
interactions during the text-chat sessions were recorded by copying their chat logs. 
Students’ interactions during the face
-to-face tasks were recorded using a digital voice 
recorder placed on the desk between the students.
Materials: The materials comprised: (1) a background questionnaire, (2) two jigsaw tasks 
(see below), each on a different topic, (3) a task-based self-efficacy self-report 
instrument. 

---

-171- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be 
Background ques
tionnaire: The background questionnaire asked for the participants’ age, 
gender, mother tongue(s), age of onset of English language learning, age of arrival in the 
UK and length of time spent in the UK. It also asked about previous visits to English 
speaking countries, specifically which countries they visited and how long for, and what 
other languages they speak. Further, it enquireed about participants’ experience using 
Skype and text chat applications more broadly.
Jigsaw tasks: Two jigsaw tasks (Pica et al., 1993) were used in this study. Jigsaw tasks 
were selected because they involve two-way information exchange and hence more 
balanced participation from the students in each pair. 
Task-based self-efficacy self-report instrument: The instrument was based on (National 
Capital Language Resource Center, 2000) and comprised four scales, one focusing on 
speaking, one on listening, one on task completion and one on future task performance 
in a face-to-face context. 
Procedure: The students participated in four 20 minute sessions each one week apart. 
Each session the participants completed a jigsaw task followed by the task-based self-
efficacy self-report instrument. In order to avoid order effects, the pairs of students were 
split into two groups and the order of presentation of the conditions and the tasks was 
counterbalanced (see below). In the final session, they completed the background 
questionnaire.