Fluency in writing can be studied at the process level, the product
level or by combining these two. There is a lack of studies on L2 writing
fluency that use both product and process data to describe this complex
phenomenon. In this chapter, to study fluency in foreign language (L2) writing,
a multi-perspective approach comparing writing processes and products is used
(cf., Ellis & Yuan, 2004; Gunnarsson, 2012). The analysis of the processes
contained variables such as pause length and median transition times, while the
quality of the final product was assessed via the evaluation of the texts by
native speakers. To study the writers’ pausal behaviour and to create thus their
individual writer profiles (cf., Van Waes, 1992; Van Waes & Schellens,
2003), two different pause criteria were defined: a stipulated pause criterion
automatically provided by the tool (≥ 2 and ≥ 5 seconds) and individual pause
states.

This chapter presents the results of a small-scale experiment in
which the data were collected by using the keystroke recording tool ScriptLog (versions 1.07f and 1.07i), a computer
program which allows one to carry out research on the online process of writing.
The online data were produced by 11 Finnish university students of French. To
summarise, the objective of this research was to examine how writer profiles are
linked to L2 fluency during writing process and what their relationship is to
the quality of the end product. The following research questions were formulated
to shed light on this issue: How fluent is the participants’ writing process in
L2? What is the relationship between fluency and the final product in L2? How are
writer profiles related to fluency? These different aspects of fluency will
also be illustrated from the perspective of the SLA fluency framework proposed
by Segalowitz (2010, 2016), namely effortless processing (i.e., cognitive
fluency), performance (i.e., utterance fluency) and reader experience (i.e.,
perceived fluency).