Due to the fact that CF, provided in (L2) writing instruction, “typically consists
of negative feedback teachers provide in response to learners’ actual or perceived
errors” (Park, 2004), and the fact that it forges enhancement of saliency for target
features in text (Han et al, 2008), CF can be perceived as one type of reactive FonF
pedagogical intervention. In the past two decades, myriads of studies have been
conducted on the efficacy of CF on students’ linguistic accuracy in writing, with
positive results. However, in this line of inquiry, the essence of CF as a pedagogical
intervention in SLA (reactive technique in focus on form) and inextricable theoretical
base concerning the role of noticing and processing in learning have received
relatively little exploitation, which will shortly be reviewed below.
Many studies have examined the efficacy of CF (Bitchener, 2008; Bitchener &
Knoch, 2010; Bitchener, Young, & Cameron, 2005; Chandler, 2003; DeKeyser, 1993;
Ferris & Robert, 2001; Sheen, 2007). Most showed that CF is beneficial to students’
writing accuracy, after overcoming a few flaws.
In earlier CF studies, a common flaw concerned the research design (Sheen,
2007). The lack of a control group, for example, limited the generalizability of the
efficacy and application of CF. Another insufficiency was that the effect of CF was
mostly confined to the revision of students’ original work. The effect of CF to new
pieces of writing was yet to be investigated. In light of the aforementioned
insufficiency, later studies on CF enlisted control groups as well as extended the
experiment into examining whether the effect of CF would be sustained in the
composition of new writing pieces. Along with these improvements of earlier flaws,
the number of target structures has also been narrowed down, from multiple into
single one, so as to facilitate learner noticing. Other factors related to efficacy of CF,
such as analytic ability, were also examined. Sheen (2007) found that individual
learners with higher analytic ability tend to benefit more from CF.
The efficacy of CF has been examined along the criterion of directness.
Bitchener (2005) questioned “whether certain types of CF (more direct) are more
likely than others (less direct) to help L2 students improve the accuracy of their
writing” (cited in Bitchener, Young, Cameron, P193). In an attempt to address the
above inquiry, issues regarding directness of CF were then explored in many studies.
Bitchener et al.’s (2008) study is a case in point. According to Bitchener, “direct CF
may be defined as the provision of the correct linguistic form or structure above or
near the linguistic error…written meta-linguistic explanation … and/or oral
meta-linguistic explanation” (Bitchener et al., 2008, p. 105). He sought to find out
whether different corrective feedback (indirect vs. direct) would have a different
effect on accuracy and whether this accuracy, if any, would be capable of being
sustained into new pieces of writing. Seventy five students took part in this study.
They were divided into four groups. Group one (17 students) received direct error
correction for each targeted error category, as well as written and oral meta-linguistic
explanations. Group two (18 students) received direct error correction for each
targeted error category and written meta-linguistic explanation. Group three (20
students) only received direct error correction for each targeted error category. Group
four was the control group (20 students). The procedure followed
pretest-treatment-posttest-delayed posttest. It was found that written CF had a
significant effect on improving accuracy in the use of the English article system and
that this level of accuracy was retained two months later without additional feedback
or instruction (Bitchener et al., 2008).
The efficacy of CF has also been examined in terms of the nature of the CF
provided to learners. Bitchener and other researchers investigated whether CF of a
different nature (written vs. oral; explicit vs. implicit; individual five-minute
conference) given to 53 adult migrant students on three types of error (prepositions,
the past simple tense, and the definite article) resulted in improved accuracy in new
pieces of writing over a 12-week period (Bitchener, Young, & Cameron, 2005). They
found that the CF with both written and individual oral meta-linguistic explanation is
significantly more effective than that with only written meta-linguistic explanation,
which is yet more effective than mere CF.In light of his finding, Bitchener (2007)
thus suggests that direct feedback reduces confusion when students deal with errors.
This is especially true with lower proficiency learners. Direct feedback is preferred in
the pedagogical setting (Ferris & Helt, 2000).
What also influences the efficacy of CF is whether errors are treatable or not
(Ferris, 1999). Ferris suggests that treatable errors are those whose correct usage can
be sought and consulted in reference materials such as grammar books. Treatable
errors are the errors made on verb tense and form, subject-verb agreement, article
usage, plural and possessive noun endings, to name just a few. What they have in
common is that they are all rule-governed. Untreatable errors, on the contrary, are
those whose usage cannot be easily guided or manipulated in reference materials.
It would be ideal if treatable errors can be treated with direct CF. This view has
been examined in some CF studies (Ashwell, 2000; Butler, 2002; Chandler, 2003;
Sheen, 2007). So far, the result is positive. For instance, Bitchener et al. (2005) found
that the combination of full, explicit written feedback and one-to-one conference
feedback enabled the learners to use the past simple tense and the definite article with
significantly greater accuracy in new pieces of writing than was the case with their
use of prepositions. It can be inferred that the use of prepositions is relatively
untreatable, compared with definite articles and verb tense, which are rule-governed.
The efficacy of CF can also be influenced by the number of forms that are
targeted in a given pedagogical session (one vs. two or more). In reviewing existing
CF literature, Sheen (2007) found that some CF studies did not yield positive result
and that this disparate finding might be attributed to the fact that “the linguistic
feedback was not sufficiently focused and intensive” (Sheen, 2007). Multiple targeted
forms would distract learners’ attention than single targeted form. With insufficient
attention to form, there will be little subsequent processing, which in turn affects
learning.
Such processing issues have not been extensively studied in the CF literature. It
is assumed that upon receiving CF, learners’ attention to the content would be directed
to the linguistic form (the input), and that if attention is successfully directed, intake
of the target form will follow. However, input cannot be transformed into intake
without noticing and processing (Schmidt, 1990). Although techniques aimed at
directing learners’ attention to the target form are a prerequisite for noticing and
processing of the form, they do not necessarily guarantee that learners would
linguistically process the target form. Therefore, while literature suggests that CF is
facilitative of the perceptual saliency of the target form, it cannot be assumed that it
also speaks to learner processing. Furthermore, given the fact that there is not always
a match between external saliency (consciousness raising feedback provided by
teachers, e.g., CF) and internal saliency (learners’ attention determined by learning
agenda, readiness of linguistic development), CF does not always ensure that learners
are able to proactively and actively analyze the enhanced input, turning it into intake.
In other words, CF is a necessary but insufficient condition for learning the target
form.
Some researchers, particularly psycholinguists (VanPatten, 1990), argue that the
enhanced input needs to be carefully structured to cater to learners’ universal
processing strategies, so as to create an optimal encoding/decoding environment for
the learners. Such a view will be reviewed in section 2.4, processing instruction in
focus on form.