from perception to practice: native and non-native english...
TRANSCRIPT
25
English Teaching, Vol. 71, No. 1, Spring 2016
DOI: 10.15858/engtea.71.1.201603.25
From Perception to Practice: Native and Non-Native English Teachers’ Use of Error
Correction in Written Feedback*
Cheol Baek**
(Hankuk University of Foreign Studies)
Sookyung Cho
(Hankuk University of Foreign Studies)
Baek, Cheol, & Cho, Sookyung. (2016). From perception to practice: Native and
non-native English teachers’ use of error correction in written feedback. English
Teaching, 71(1), 25-45.
This study compares native English teachers (NETs) and non-native English teachers
(NNETs) in their perceptions of errors as well as their actual feedback. Studies
comparing NETs and NNETs have focused on actual feedback practice (Green &
Hecht, 1985; T. Kobayashi, 1992), with very few studies relating this feedback to their
actual perceptions of error correction (Hyland & Anan, 2006; Kim, 2007). In order to
better understand this phenomenon, 26 NETs and 24 NNETs completed a
questionnaire and provided feedback on a sample academic essay. The results reveal
that while both groups showed differing degrees of perceptions, they did not
significantly differ from each other in actual feedback, except that NETs preferred
coded feedback than NNETs by explaining errors. This study implies that NNETs are
as reliable as NETs in correcting errors, but that they differ in how they give feedback.
Key words: error correction, written feedback, native English teachers, non-native
teachers
1. INTRODUCTION
As the importance of feedback in teaching writing has increased, much research has
focused on whether native English teachers (NETs) and non-native English teachers
* This work was supported by Hankuk University of Foreign Studies Research Fund of 2015.** Cheol Baek: First author; Sookyung Cho: Corresponding author
© 2016 The Korea Association of Teachers of English (KATE)This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits anyone to copy, redistribute, remix, transmit and adapt the work provided the original work and source is appropriately cited.
26 Cheol Baek ∙ Sookyung Cho
(NNETs) differ from each other in their approach to written feedback. Regarding error
correction more particularly, while some studies found no statistical difference between
NETs and NNETs (Hinkel, 1994; H. Kobayashi & Rinnert, 1996; Zhang, 1999), others
have shown that one main difference does exist (Green & Hecht, 1985; Hughes &
Lascaratou, 1982; T. Kobayashi, 1992; Schmitt, 1993): NNETs are stricter towards student
errors than NETs. These researchers speculate that this strictness toward student errors
originates from a difference in their learning experiences and teaching experiences. That is,
in comparison with NETs, NNETs have been exposed to limited authentic English input,
and years of teaching EFL students who are learning English as a foreign language, not
their native language, has made them more conscious of student errors. Given that these
fundamental differences cannot but exist between NETs and NNETs, it is necessary to
explore how NETs and NNETs differ in their perceptions and how these perspectives are
reflected in their actual feedback practice.
Several studies have examined NETs’ and NNETs’ thoughts about their own error
correcting behavior (Hyland & Anan, 2006; Kim, 2007; Montgomery & Baker, 2007; Park,
2014). These scholars have found that NETs and NNETs evaluate errors based on different
grounds; while NETs place a priority on intelligibility, meaning the extent to which they
can understand student texts, NNETs are more interested in rule infringement, or whether
or not the student violates a certain grammatical rule. However, the scope of these studies
on perceptions of NETs and NNETs has been restricted to error gravity, that is, how
serious they consider student errors to be and why, and thus these studies do not seem to
take into account teacher perceptions of other aspects, including how and why they correct
student errors. In order to broaden our understanding of NET and NNET perceptions of
error correction in general, this study examines how NETs and NNETs perceive error
correction in three areas—principles for error correction, error format, and effectiveness of
error correction—and analyzes how these perceptions are reflected in their actual error
feedback practice. The guiding research questions are as follows:
1. How do NETs and NNETs perceive of error correction?
2. How do NETs and NNETs correct errors?
3. How do NET and NNET perception relate to their error correcting behaviors?
2. LITERATURE REVIEW
2.1. Error Correction
Whether writing teachers should give feedback on student grammar has been heavily
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 27
debated (Ferris, 1999, 2004; Goldstein, 2004; Truscott, 1996, 1999). Some scholars argue
that providing error feedback does not help students make fewer errors (Campbell, 1998;
Truscott, 1996, 1999). Truscott (1996, 1999) indicated that error correction is not only
ineffective, but also harmful to students because they are deprived of the chance to
concentrate on developing other areas such as content or organization. Campbell (1998)
also suggested that error correction can be stressful for students, limiting their ability to
develop awareness of global issues in their composition.
On the other hand, several researchers have argued that error correction produces several
positive effects and is necessary to improving student writing (Ferris, 1997, 2006; Lalande,
1982; Polio, Fleck, & Leder, 1998). These researchers theorized that error feedback helps
students notice the differences between adequate and inadequate usages in the target
language and provides them with better chances to recognize error types. Ferris (1997),
who gave feedback including error correction to 47 English as a second language students
and investigated how they responded to it, found that much of the error correction led
students to produce better revisions. Furthermore, Ferris (2006) examined various issues
related to how error feedback is given to students and how they use it, finding that the
number of student errors was significantly reduced in five major categories (verb tense,
verb form, lexical errors, articles, and sentence structure) in their final essays.
As the research has shown that error correction can be helpful, then we should consider
how teachers correct errors. Several studies have indicated that indirect feedback—
indicating the presence of errors without correcting them—is more beneficial than direct
feedback—correcting errors (Bates, Lane, & Lange, 1993; Ferris, 2002; Frantzen, 1995;
Lalande, 1982). Ferris (2002) indicated that indirect feedback is more helpful for learners
because in giving direct feedback, teachers might misinterpret student meanings. Lalande
(1982), who compared students who received indirect feedback with those who received
direct feedback, found that although the results showed no statistically significant
differences in linguistic accuracy, the indirect format resulted in fewer student errors. He
concluded that the indirect feedback format offers stronger guidance than direct feedback.
Semke (1984) claimed that writing progress is enhanced by written commentaries and
questions alone without error correction. Accordingly, he suggested that direct error
feedback did not increase writing accuracy, writing fluency, or general language
proficiency, and that direct feedback could have a negative influence on students’ attitudes,
as students did not learn how to make corrections independently.
In giving feedback, teachers should also decide whether to identify errors, that is, whether
to code or explain errors (Ferris, 2002; Lee, 1997; Raimes, 1991). Identifying errors is often
regarded as effective for low-proficiency-level students, because it provides specific
responses to errors (Aljaafreh & Lantolf, 1994; Chandler, 2003). According to Raimes (1991),
error identification is helpful as a first step to make students think of their own errors.
28 Cheol Baek ∙ Sookyung Cho
Similarly, comments in the margin are also considered as an effective means for giving
feedback about specific aspects of a composition, because the proximity of the comments to
the portion of the text referred to makes it easier for students to understand them (Snow,
1996). While indicating the error location in the margins, they can indicate the error location
directly in order to identify more errors (Robb, Ross, & Shortreed, 1986) or indicate the error
location indirectly to train their students to find further errors (Lee, 1997).
2.2. Native vs. Non-Native English Teacher
While the practice of NETs and NNETs regarding error correction has been debated, the
findings are inconclusive. Some researchers have argued that NETs and NNETs do not
differ from each other in error correction (Fujita & Sakamoto, 1998; Jolivet, 1997). Fujita
and Sakamoto (1998), who examined error correction in two compositions assessed by 17
Japanese EFL teachers and 17 native EFL teachers, found no statistical difference between
the general tendencies of the two groups. Jolivet (1997) found a similar result in his
quantitative and qualitative analyses of errors corrected by NETs and NNETs, in the
number of error corrections as well as in error categories after comparing four native and
four non-native French teaching assistants.
On the other hand, several scholars have argued that NETs and NNETs do differ in error
correction (Davies, 1983; Hughes & Lascaratou, 1982; Hyland& Anan, 2006; T.
Kobayashi, 1992; Schmitt, 1993; Takashima, 1987). After comparing assessments of 18
NNETs and 20 NETs, Schmitt (1993) found that NETs tended to focus on writing
comprehensibility, while NNETs marked significantly more grammatical errors. Hughes
and Lascaratou (1982) compared three groups of graders—ten native English teachers, ten
native English non-teachers, and ten Greek teachers of English—and found that, in terms
of error gravity, English non-teachers were the most lenient toward grammatical errors,
basing their judgment on the intelligibility of content, while Greek teachers resorted more
to infringement of grammatical rules. Similarly, Hyland and Anan (2006) compared three
groups—native English teachers, Japanese teachers of English, and native English non-
teachers—in their practice of error correction and found that Japanese teachers were the
strictest toward student grammatical errors, being more likely to consider stylistic
variations as errors. Hyland and Anan imply that this difference might originate from
differences in their prior experiences, that is, while native English teachers were exposed to
various alternative expressions to a target expression, Japanese teachers has limited
exposure as well as limited English proficiency.
In the Korean context, Kim (2007) replicated Hyland and Anan’s (2006) study and
found similar results, in that NETs corrected more errors than NNETS, which was based
on different variables: NETs corrected errors focusing on intelligibility, while NNETs
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 29
emphasized rule violation criteria. After comparing an NET’s and an NNET’s assessment
behaviors for 110 Korean high school students before and after instruction, Cheong (2012)
found that the NET evaluated student essays with lower scores than the NNET in general,
and that his evaluation of students’ language usage did not change positively even after
writing instruction, while the NNET evaluated students’ language use as improved after
instruction. Similarly, Park (2014) compared seven NETs’ and seven NNETs’ scoring
behaviors for 39 student essays and found that NETs were more stringent than NNETs and
had higher intra-rater consistency. The analysis of their comments justifying why they gave
the particular scores to a given essay reveals that their decisions were based on different
perspectives of important components of writing: it seems that NETs gave more weight to
organization than NNETs, who valued other categories such as contents, grammar,
vocabulary, and mechanics.
It has been suggested that the differences between NETs and NNETs may originate from
differences in their identities as English teachers, including their language awareness and
experience (Hyland & Anan, 2006). It is often assumed that NETs have a number of
advantages over NNETs, including higher familiarity with idiomatic error correction and
distinguishing acceptable and unacceptable versions of the language. NNETs often believe
that their awareness of the language they teach is inadequate and that their lack of insight
into language can make it difficult for them to carry out their work. There is a significant
gap between the language awareness of non-native and native teachers, as the latter have
intuition and experience with grammar knowledge and use (Borg, 2003; Hughes &
Lascaratou, 1982), leading some NNETs to feel reluctant to mark some student errors.
According to Hyland and Anan (2006), NNETs often lack confidence in their English and
are uncertain about appropriate usage, and this can lead to a prescriptive attitude to
grammar correction and reluctance to mark non-standard errors, such as informalities.
However, although NNETs have difficulties related to English knowledge and intuition,
they have the distinct ability to relate to their non-native students because they share the
students’ first language and have gone through the same process of learning the second
language. This signals a connection between NNETs and EFL students, as teachers can
share the same perspective with their students (Kramsch, 1998). Davies (1983), examining
error feedback in terms of the viewpoints from which NETs and NNETs approach errors,
found that NNETs were perceptively familiar with the deviant structures of their students’
works and that they understood the constructions that originate from the transfer of first
language structures. H. Kobayashi and Rinnert (1996) found similar results. They gave 16
versions of student compositions containing Japanese English and American English
patterns to 465 native and non-native teachers. They discovered that the Japanese teachers
performed more strongly in terms of understanding Japanese English patterns. These
results suggest that non-native teachers’ deeper understanding of their students’ first
30 Cheol Baek ∙ Sookyung Cho
language makes it possible for them to have an advantage not only in the correction of
constructions that originate from their first language.
Drawing on these studies that emphasize the close relationship between teacher
perceptions and teaching practice, this study tries to explore both the perceptions and actual
error correction practice of NETs and NNETs in order to see how their views on error
correction are reflected in their actual error correction practice.
3. METHOD
3.1. Participants
The participants were 26 NETs and 24 Korean NNETs from various universities. All the
NETs were from English-speaking countries, 23 of whom majored in English in their
native countries, and three of whom had not studied English-related majors but had taken
English composition courses during their undergraduate or graduate periods. Although six
participants of the NETs were not currently teaching, all the participants had taught or were
teaching English writing. All the NNETs were native Koreans who learned English in
Korea and in English-speaking countries; 21 of the NNETs majored in English, while the
other three did not, but had taken English-related courses. Although six were not currently
teach English writing, they had taught English writing previously.
3.2. Data Collection
A survey was designed by modifying Icy Lee’s (2003) framework for investigating
teacher perceptions of error correction. The survey, consisting of multiple choice and short-
answer questions (See Appendix A), is divided into two portions: Section 1 investigates
teacher perceptions of writing education, and Section 2 focuses on (1) perspectives on error
feedback and (2) feedback principles. Along with the survey, each participant was asked to
provide error feedback on a sample essay (See Appendix B) written by one of the authors
of this study in 2010 as they usually did. It was intentionally chosen due to the variety of
errors participants could comment on (e.g., grammar errors, Korean English, organization,
and informality).
3.3. Data Analysis
The answers in the survey were fell under three themes: principles for error correction,
error format, and effectiveness of error correction. Error correction from the essay was
analyzed by error type and format. Error types and formats were coded by the two authors
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 31
and a third rater who has worked as an EFL teacher for more than eight years. The coding
of the three raters was in agreement for more than 80% of the data.
TABLE 1
Error Categories Error Type Description Example
Word choice 1. Wrong word choice2. Synonyms
Preposition “in”/“to”
Verb tense Verb tense “am”/“was”Verb form Verb form “way to enjoying”/“way to enjoy” Word form Word form “easy”/“easily”Articles Articles “a”/“an”Singular–plural Singular–plural “tool”/“tools”Pronouns Pronouns “The man who”/“He”Run-on Probability between sentences “As a result”/“To sum up” Punctuation Comma splices and fragments Local issues, such as a comma instead
of an en dashSpelling Includes capital and small letters “Tehcer”/“Teacher”Sentence
structure Includes missing and unnecessary words, phrases and word order problems; teachers changed sentences to read more naturally
“My father bought a computer to me”/“My father bought me a computer”
Informal Refers to register choices considered inappropriate for academic writing
“Have fun together”(Comment) too informal, not academic
Idiom Refers to errors in use of idiomatic expressions
“Agree (to)”/“(with)”
SV agreement Subject–verb agreement “It (make)”/“(makes)”Miscellaneous Errors that cannot be otherwise classified
TABLE 2
Error Format Label Description Example
Direct Indicate (underline/circle) errors and correct them
has went(gone)
Direct-coded Indicate (underline/circle) errors, correct them, and categorize them
has went(gone)(verb form)
Indirect Indicate (underline/circle) errors, but don’t correct them
has went
Indirect-coded Indicate (underline/circle) errors and categorize them, but don’t correct hem
has went(verb form)
Direct error location Explain error location(marginal note)
(In margin) Grammar! Third paragraph, line number 10
Direct error location-coded
Explain the location of errors and categorize them (marginal note)
(In margin) Grammar! Third paragraph, line number 10: You should change the word.
Indirect error location No explanation of error location(marginal note)
(In margin) Grammar!
Indirect error location-coded
No explanation of error location, but categorization or explanation of the errors (marginal note)
(In margin) Grammar! You should focus on the word choice.
32 Cheol Baek ∙ Sookyung Cho
Table 1 shows the error categories used in this study. Error types were coded by adopting
and modifying the error coding method used by Ferris (2006). The original coding scheme
contained 15 error categories, but the ‘fragment’ category, which was rarely found in the
sample student essay, was not used.
Error format was coded using Lee’s (2003) method with slight modifications. Table 2
outlines the revised error format coding scheme. While Lee classified error format into six
categories (i.e., direct, direct-coded, indirect, indirect-coded, indirect error location, and
indirect error location-coded), we added ‘direct error location’ and ‘direct error location-
coded’ to examine teacher feedback in the margin more closely. 4. RESULTS
4.1. Teacher Perception
4.1.1. Principles for error correction
The analysis of the survey results shows that NETs and NNETs showed similar results in
the number of errors they corrected. Seeing that no NETs or NNETs chose the no-marking
option, it seems that both groups mark errors; furthermore, both groups prefer to
selectively mark errors (76.92% for NETs and 79.16% for NNETs) instead of marking all
errors (23.07% for NETs and 20.83% for NNETs). However, they differed in terms of the
number of errors they selected for marking.
TABLE 3
Amount of Errors Selected for Marking Amount of Errors NET NNET
About 1/3 42.30% 8.69% About 2/3 15.38% 60.86% More than 2/3 23.07% 21.73% All errors 19.23% 8.69%
According to Table 3, NNETs believed that they are likely to mark relatively more errors
than NETs did in the essay. The majority of NETs (42.30%) answered that they mark about
one-third of the total errors, but 60.86% of NNETs reported that they correct about two-
thirds of errors.
Furthermore, NETs and NNETs seem to base their selection of errors to mark on
different grounds, as displayed in Table 4.
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 33
TABLE 4 Principles for Error Selection
Principles for Error Selection NETs NNETs Grammar instruction in class 25.72% 38.46% Students’ specific needs 42.85% 0.00% Suggestion of English panel 2.85% 0.00% Ad hoc basis 8.57% 57.69% Others 20.00% 3.84%
While student’s specific needs had the strongest influence on the error selection of NETs
(42.85%), with the second most frequently-used criterion being class instruction (25.72%),
a total of 57.69% of NNETs responded that the errors were selected on an ad hoc basis,
with classroom instruction being the second most-commonly used criterion (38.46%).
Other answers not included in the survey options were “based on the purpose of the
assignment,” “to give feedback/hints,” “related to style and organization,” “related to
discourse organization and mechanics,” “to match students error frequency,” and “related
to repeated errors.”
4.1.2. Format of error correction
As seen in Table 5, direct error format is the most preferred error feedback format for
both groups, although a slightly higher percentage of NNETs (62.5%) chose this as their
preferred feedback format compared to NETs (50%). Interestingly, while some native
teachers do not prefer the direct format (30.7%), indicating that they never or rarely use
direct format, no NNETs responded that they never or rarely used the direct format.
TABLE 5 Error Correction Format
Label Never or Rarely Sometimes Always or Often NET NNET NET NNET NET NNET
Direct 30.7% 0.0% 19.2% 37.5% 50.0% 62.5% Direct-coded 53.8% 12.5% 30.7% 70.8% 15.3% 16.6% Indirect 42.3% 45.8% 46.1% 54.1% 11.5% 0.0% Indirect-coded 46.1% 33.3% 50.0% 50.0% 3.8% 16.6% Indirect error location 61.5% 62.5% 30.7% 25.0% 3.8% 12.5% Indirect error location-coded 61.5% 62.5% 34.6% 20.8% 3.8% 16.6%
On the other hand, more NETs (11.5%) than NNETs (0%) always or often use the
indirect format. However, Table 5 shows that more NNETs (16.6%) than NETs (3.8%)
always or often prefer the coded format when they used the indirect format. Interestingly,
the coded format was also preferred by NNETs (16.6%) when they gave feedback in the
margin (indirect error location-coded format), with only 3.8% of NETs having this
preference.
34 Cheol Baek ∙ Sookyung Cho
4.1.3. Effectiveness of error correction
When asked about the need to provide feedback on student errors in writing, each group
had a slightly different perspective, as shown in Table 6. The majority of NNETs (70.8%)
strongly disagreed with the claim that there is no need for teachers to provide feedback on
student errors in writing, with 20.8% disagreeing; only 53.8% of NETs strongly disagreed,
with 42.3% disagreeing. However, both groups considered selective marking as positive—
about 83% of NETs and about 75% NNETs thought that teachers should selectively
provide feedback on student errors.
When asked about whether it is the teacher’s job to locate errors and provide corrections
for students, both groups gave similar answers: 76.8% of NETs and 79.1% of NNETs
agreed. Although both groups thought that teachers should locate errors and provide
corrections, they had different perceptions regarding varying the error format in accordance
with the type of error: More NNETs (41.6%) than NETs (26.9%) strongly agreed. While
more NNETs (57.6%) strongly agreed that teachers should vary their error feedback
techniques according to the type of error, NETs (37.5%) were slightly more lenient.
TABLE 6
Effectiveness of Error Correction (%)
Item description Strongly Disagree Disagree Agree Strongly Agree NET NNET NET NNET NET NNET NET NNET
There is no need for error correction
53.8 70.8 42.3 20.8 3.8 4.1 0.0 4.1
Error correction is teacher’s job
7.6 4.1 15.3 16.6 53.8 62.5 23.0 16.6
Various techniques according to error types
0.0 4.2 15.3 16.6 57.6 37.5 26.9 41.6
Coding errors with the help of a marking code
15.3 8.3 11.5 29.1 53.8 45.8 19.2 16.6
Marking codes should be easy
3.8 0.0 3.8 8.3 30.7 16.6 65.3 75.0
Equal attention for all errors
46.1 29.1 26.9 41.6 19.2 12.5 7.6 16.6
Students should learn to locate their own errors
3.8 4.1 3.8 0.0 42.3 54.1 50.0 41.6
Students should learn to locate and correct their own errors
3.8 4.1 3.8 8.3 42.3 41.6 50.0 45.8
Students should learn to analyze their own errors
0.0 4.1 0.0 0.0 42.3 54.1 57.6 41.6
When asked about the effectiveness of their marking codes, the groups responded
similarly. A slightly higher percentage of NETs (73%) felt their marking codes were
effective in helping students self-correct errors compared to NNETs (62.4%). Regarding
the comprehensibility of error codes, the majority of both groups strongly agreed that
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 35
marking codes were easy for students to follow and understand (65.3% of NETs and 75%
of NNETs). However, in the order of priority among error types, NETs and NNETs
showed different perceptions. More NETs (46.1%) strongly disagreed that all student errors
deserve equal attention, compared to 29.1% of NNETs. In addition, 26.9% of NETs
disagreed that all student errors deserve equal attention, compared to 41.6% of NNETs,
who had a relatively lenient perception of the order of priority of error types.
When asked if students should learn to locate errors, both groups gave almost identical
answers; both NETs (92.3%) and NNETs (95.7%) agreed that students should learn to locate
errors. In addition, both groups had the same perceptions when asked if students should learn
to locate errors and correct their own errors (92.3% of NETs; 87.4% of NNETs). Almost all
participants responded that students should learn how to analyze theirown errors (99.9% of
NETs and 95.7% of NNETs). However, they differed in their degree of agreement. More
NETs (57.6%) strongly agreed that students should learn to analyze their own errors than
NNETs (41.6%), while NNETs were more lenient (42.3% of NETs and 54.1% of NNETs
agreed). This suggests that NETs are more likely to believe that their students should take the
responsibility of taking care of their own errors than their NNET counterparts.
4.2. Teacher Feedback
4.2.1. Error types The analysis of the error types marked by teachers reveals that the average number of
error types differs for NETs and NNETs. Table 7 displays the average number of error
types in both groups.
TABLE 7 Descriptive Statistics for Error Types
Error type NETs (Mean) NNETs (Mean) Word choice 6.5 4.1 Verb tense 2.1 2.4 Verb form 1.2 0.9 Word form 3.5 3.8 Articles 1.9 2.1 Singular–plural 1.4 1.1 Pronouns 0.3 0.08 Run-on 1.1 0.3 Punctuation 0.3 0.08 Spelling 0.8 0.5 Sentence structure 8.8 5.1 Informal 3.0 0.8 Idiom 0.6 0.3 SV agreement 0.0 0.0 Miscellaneous 1.6 2.2
36 Cheol Baek ∙ Sookyung Cho
NETs marked more errors across almost all categories, except for verb tense, word form,
articles, and miscellaneous. This indicates that while NETs focus on all types of errors,
NNETs focus on a certain number of key error types like verb tense, word form, and
articles.
In order to see whether these differences between NETs and NNETs are statistically
significant, a MANOVA test was conducted for these 15 error types, as shown in Table 8. A
multi-way MANOVA revealed a significant multivariate main effect only in the informal
category (F = 8.175, p = .006), out of these 15 error types.
TABLE 8
MANOVA Results of NETs’ Error Type Dependent Variable SS df MS F Sig.
Word choice 72.501 1 72.501 2.184 .146 Verb tense 1.133 1 1.133 .240 .626 Verb form 2.302 1 2.302 1.154 .288 Word form .821 1 .821 .078 .781 Articles .333 1 .333 .070 .792 Singular–plural 1.109 1 1.109 .294 .590 Pronouns 1.133 1 1.133 .850 .361 Run-on 8.337 1 8.337 2.415 .127 Punctuation .628 1 .628 1.411 .241 Spelling .862 1 .862 .454 .504 Sentence structure 172.513 1 172.513 3.358 .073 Informal 60.685 1 60.685 8.175 .006 Idiom .970 1 .970 2.660 .109 Miscellaneous 5.026 1 5.026 .370 .546 * p < .05
The results indicate that NETs marked informal errors that were inappropriate for
academic writing more often than NNETs did. However, no significant differences in the
other 14 categories show that NETs and NNETs are not statistically different, although
NETs showed higher average numbers in many categories. Based on these results, it can be
concluded that the two groups differ from each other only in terms of their approach to
informal errors.
4.2.2. Error correction format
The actual error corrections were analyzed quantitatively for the eight error formats.
Table 9 shows averages for these different types of formats in the actual essay assessment.
The direct format was preferred by both groups, although NETs made more frequent use of
this format, using it 21.5 times on average, than NNETs, using it 12.6 times on average. On
the other hand, the average number of uses for the direct-coded format was relatively
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 37
smaller in both groups than the direct format (2.5 times for NETs and 3 times for NNETs).
TABLE 9
Means of Error Formats Mean NETs NNETs
Direct (D) 21.5 12.6 Direct-coded (D-C) 2.5 3.0 Indirect (I) 2.1 2.0 Indirect-coded (I-C) 5.5 2.0 Indirect error location (IL) 0.0 0.0 Indirect error location-coded (IEL-C) 1.3 0.3 Direct error location (DL) 0.5 0.0 Direct error location-coded (DL-C) 3.6 0.9
It seems that both groups tended to give direct answers without explaining the errors.
For indirect feedback, no striking difference was found between the groups, though,
interestingly, a difference was found between indirect-coded format uses. NETs used the
indirect-coded format 5.5 times on average, while NNETs used this format two times on
average. That is, NETs tended to categorize errors more often than NNETs when giving
indirect feedback.
Likewise, NETs tended to categorize errors more when they gave feedback in the margin
(error location formats). The average number of indirect error location format uses was
higher for NETs (1.3 times) than for NNETs (0.3 times). The native teachers also had a
higher average number of direct error location-coded format uses (3.6 times for NETs, and
0.9 times for NNETs).
TABLE 10
MANOVA Results of NETs’ Error Format Dependent Variable SS df MS F Sig.
Direct (D) 973.787 1 973.787 3.242 .078 Direct-coded (D-C) 3.120 1 3.120 .065 .800 Indirect (I) .283 1 .283 .017 .896 Indirect-coded (I-C) 149.262 1 149.262 4.740 .034* Indirect error location-coded (IEL-C) 10.857 1 10.857 6.753 .012* Direct error location (DL) 3.120 1 3.120 3.698 .060 Direct error location-coded (DL-C) 96.148 1 96.148 5.832 .020*
In order to test for significance between the groups, a MANOVA test was conducted for
these eight error formats. The MANOVA results, as displayed in Table 10, revealed that the
groups significantly differ from each other in terms of indirect-coded format, indirect error
location-coded format, and direct error location-coded format, but not in the other five
categories. These results suggest that NETs tended to give more explanations of errors in
the three coded formats: indirect-coded format (F = 4.740, p = .034), indirect error
38 Cheol Baek ∙ Sookyung Cho
location-coded format (F = 6.753, p = .012), and direct error location-coded format (F =
5.832, p = .020).
4.3. Teacher Perceptions and Feedback
The comparison between teacher perceptions and feedback reveals both consistency and
discrepancy in error correction format. As the majority of both NETs (50%) and NNETs
(62.5%) indicated their preference for direct error correction in perceptual results, their
actual feedback reveals that both groups gave direct error correction more frequently than
other types of feedback (21.5 times for NETs and 12.6 for NNETs).
On the other hand, some discrepancies were found between teacher perceptions and
feedback. For instance, although NETs as well as NNETs answered that they never or
rarely used indirect-coded (46.1%) or indirect-error-location-coded (61.5%), their actual
feedback results revealed that NETs used more of these coded formats. NETs’ significantly
strong preference for these coded types of error correction, however, corroborate the
survey results, in that more NETs than NNETs strongly agreed that student should learn to
both locate their own errors (50% vs. 41.6%) and correct them (50% vs. 45.8%). By using
these types of coded formats or categorizing errors, NETs support student independence in
analyzing their own errors.
5. DISCUSSION
The comparison between teacher perceptions and their actual feedback practice reveals a
discrepancy in terms of principles for error correction and error format but consistency in
terms of effectiveness of error correction. According to the perception results, it seems that
more NETs than NNETs tend to correct fewer errors (over 40% of NETs correct about one-
third of student errors) for different reasons from those of NNETs—while NETs consider
students’ specific needs in selecting errors, NNETs select on ad hoc basis. However, the
comparison of their actual feedback practices does not demonstrate any statistically
significant difference in error types except for informal use of language. As a number of
previous studies have shown (Davies, 1983; Green & Hecht, 1985; Heilenman, 1991;
Hyland& Anan, 2006; T. Kobayashi, 1992; Schmitt, 1993; Takashima, 1987), the native
teachers tended to focus more on informality than the non-native teachers.
There is also a discrepancy in error format between the perception results and actual
practice. The format results for error corrections in the survey showed that the coded
formats are more preferred by NNETs when they give feedback in the margins. However,
the actual feedback reveals that NETs use more coded formats, even in correcting errors in
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 39
the margin. On the other hand, this preference of NETs for coded error format in actual
practice can be explained by their encouraging students to develop independence as
learners, which was evidenced in their perceptions. In the survey, more NETs than NNETs
strongly agreed that students should learn to locate, correct, and analyze their own errors.
These findings imply that both NETs and NNETs are reliable sources for teaching
writing, particularly when they correct errors. Although some concerns exist regarding the
professionalism of non-native writing teachers in the EFL context, this study found that
NETs were not significantly different from NNETs in providing feedback: NETs simply
focused more on the “informal” errors in students’ compositions. Moreover, using the
coded format (identification and categorization of errors) would be an advantageous
method for NNETs, which might enable students to become more independent learners.
Further studies are necessary to analyze the differences between NETs and NNETs more
deeply. The same type of study could be conducted in a real-world situation. That is, the
questionnaire could be given to teachers and then teacher comments on actual student
papers could be analyzed in order to see how NETs and NNETs perceive and respond to
their target students in the classroom setting. Also, it would be insightful if teachers’
teaching experiences—that is, how long they have taught writing and the types of writing
courses they have taught, whether they are NETs or NNETs—are incorporated into the
analysis to ascertain the role that teaching experiences play in their feedback. In spite of
these limitations, the results of this study imply that NETs and NNETs do differ from each
other not in whether they correct errors or not, but in their manner of correcting errors.
REFERENCES
Aljaafreh, A., & Lantolf, J. P. (1994). Negative feedback as regulation and second language
learning in the zone of proximal development. The Modern Language Journal,
78(4), 465-483.
Bates, L., Lane, J., & Lange, E. (1993). Writing clearly: Responding to ESL compositions.
Boston, MA: Heinle & Heinle.
Borg, S. (2003). Teacher cognition in grammar teaching: A literature review. Language
Awareness, 12(2), 96-108.
Campbell, C. (1998). Teaching second language writing: Interacting with text. Boston,
MA: Heinle & Heinle.
Chandler, J. (2003). The efficacy of various kinds of error feedback for improvement in the
accuracy and fluency of L2 student writing. Journal of Second Language Writing,
12(3), 267-296.
Cheong, S. H. (2012). Native- and nonnative-English-speaking raters’ assessment behavior
40 Cheol Baek ∙ Sookyung Cho
in the evaluation of NEAT essay writing samples. English Language Teaching,
24(2), 49-73.
Davies, E. (1983). Error evaluation: The importance of viewpoint. ELT Journal, 37(4),
303-311.
Ferris, D. R. (1997). The influence of teacher commentary on student revision. TESOL
Quarterly, 31(2), 315-339.
Ferris, D. R. (1999). The case for grammar correction in L2 writing classes: A response to
Truscott (1996). Journal of Second Language Writing, 8(1), 1-11.
Ferris, D. R. (2002). Treatment of error in second language student writing. Ann Arbor,
MI: The University of Michigan Press.
Ferris, D. R. (2004). The “grammar correction” debate in L2 writing: Where are we, and
where do we go from here? (and what do we do in the meantime . . . ?). Journal of
Second Language Writing, 13(1), 49-62.
Ferris, D. R. (2006). Does error feedback help student writer? New evidence on the short-
and long-term effects of written error correction. In K. Hyland & F. Hyland (Eds.),
Feedback in second language writing: Contexts and issues (pp. 81-104).
Cambridge: Cambridge University Press.
Frantzen, D. (1995). The effects of grammar supplementation on written accuracy in an
intermediate Spanish content course. The Modern Language Journal, 79(3), 329-
344.
Fujita, M., & Sakamoto, M. (1998). EFL composition evaluations by native EFL and
Japanese EFL instructors. JACET Bulletin, 29, 137-154.
Goldstein, L. M (2004). Questions and answers about teacher written commentary and
student revision: Teachers and students working together. Journal of Second
Language Writing, 13(1), 63-80.
Green, P., & Hecht, K. (1985). Native and nonnative evaluation of learners’ errors in
written discourse. System, 13(2), 77-97.
Heilenman, L. K. (1991). Writing in foreign language classrooms: Process and reality. In J.
Alatis (Ed.), Linguistics and language pedagogy: The state of the art (pp. 273-288).
Washington, D.C.: Georgetown University Press.
Hinkel, E. (1994). Native and nonnative speakers’ pragmatic interpretations of English
texts. TESOL Quarterly, 28(2), 353-376.
Hughes, G., & Lascaratou, C. (1982). Competing criteria for error gravity. English
Language Teaching Journal, 36(3), 175-182.
Hyland, K., & Anan, E. (2006). Teachers’ perceptions of error: The effects of first language
and experience. System, 34(4), 509-519.
Jolivet, C. A. (1997). Comparing native and nonnative speakers’ error correction in foreign
language writing. Texas Papers in Foreign Language Education, 3(1), 1-14.
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 41
Kim, K. J. (2007). Error perceptions of native and nonnative English-speaking writing
teachers. English Teaching, 62(4), 221-239.
Kobayashi, H., & Rinnert, C. (1996). Factors affecting composition evaluation in an EFL
context: Cultural rhetorical pattern and readers’ background. Language Learning,
46(3), 397-437.
Kobayashi, T. (1992). Native and nonnative reactions to ESL compositions. TESOL
Quarterly, 26(1), 81-112.
Kramsch, C. (1998). Language and culture. Oxford: Oxford University Press.
Lalande, J. F., Jr. (1982). Reducing composition errors: An experiment. The Modern
Language Journal, 66(2), 140-149.
Lee, I. (1997). ESL learners’ performance in error correction in writing. System, 25(4),
465-477.
Lee, I. (2003). L2 writing teachers’ perspectives, practices and problems regarding error
feedback. Assessing Writing, 8(3), 216-237.
Montgomery, J. L., & Baker, W. (2007). Teacher-written feedback: Student perceptions,
teacher self-assessment, and actual teacher performance. Journal of Second
Language Writing, 16(2), 82-99.
Park, M. (2014). A comparison of assessment behavior between native and nonnative
English speaker raters in English writing. Unpublished doctoral dissertation,
Chonbuk National University, Jeonju, Korea.
Polio, C., Fleck, C., & Leder, N. (1998). “If only I had more time”: ESL learners’ changes
in linguistic accuracy on essay revisions. Journal of Second Language Writing, 7(1),
43-68.
Raimes, A. (1991). Errors: Windows into the mind. College ESL, 1(2), 55-64.
Robb, T., Ross, S., & Shortreed, I. (1986). Salience of feedback on error and its effect on
EFL writing quality. TESOL Quarterly, 20(1), 83-96.
Schmitt, N. (1993). Comparing native and nonnative teachers’ evaluations of error
seriousness. Japanese Association of Language Teachers Journal, 15(2), 181-191.
Semke, H. D. (1984). Effects of the red pen. Foreign Language Annals, 17(3), 195-202.
Snow, D. (1996). More than a native speaker: An introduction for volunteers teaching
abroad. Alexandria, VA: Teachers of English to Speakers of Other Languages.
Takashima, H. (1987). To what extent are nonnative speakers qualified to correct free
composition? A case study. British Journal of Language Teaching, 25(1), 43-48.
Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language
Learning, 46(2), 327-369.
Truscott, J. (1999). The case for “The case against grammar correction in L2 writing
classes”: A response to Ferris. Journal of Second Language Writing, 8(2), 111-122.
Zhang, W. X. (1999). The rhetorical patterns found in Chinese EFL student writers’
42 Cheol Baek ∙ Sookyung Cho
examination essays in English and the influence of these patterns on rater response.
Unpublished doctoral dissertation, Hong Kong Polytechnic University, Hong Kong.
APPENDIX A Teacher Questionnaire
This questionnaire aims to find out how you mark grammar errors in students’ writing, your beliefs
about error feedback, and the concerns you may have regarding the subject. All your answers will be
treated confidentially.
-The last page of this questionnaire is a real student’s essay. Please, give the essay your feedback as
if you give feedback to your student.
Section 1
1. What is your nationality?
2. What country did you acquire/learn English in?
3. Do you have an English-related degree (TESL/TEFL, linguistics, translation, literature)?
4. Have you taken an English writing course? What kind of course was it?
5. How many students do you teach? (English writing only)
6. Does your school/academy offer you a particular teaching method related to error correction?
Section 2
1. In your opinion, what is the main purpose of providing feedback on students’ errors in writing?
2. Which of the statements below best describes your existing error feedback practice? Please tick the
most appropriate box.
A. I don’t mark students’ errors in writing.
B. I mark ALL students’ errors.
C. I mark students’ errors selectively.
*If your answer to Question 2 is “A”, please specify the reason.
3. Tick the amount of errors selected for marking.
A. About 1/3
B. About 2/3
C. More than 2/3
D. All errors
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 43
4. Which of the following best describes the major principle for error selection? A. The selected errors are directly linked to grammar instruction in class. B. The selected errors are related to students’ specific needs. C. The selected errors are suggested by the English panel. D. The errors are selected on an ad hoc basis. E. Others (Please specify in the relevant box) 5. Rate the frequency with which you use each of the following error feedback techniques according to the scale below. Please circle the appropriate number.
Item Description 1
Never or rarely2
Sometimes 3
Always A. I indicate (underline/circle) errors and
correct them. -e.g. has went(gone) B. I indicate (underline/circle) errors, correct
them and categorize them.-e.g. has went (gone) (verb form)
C. I indicate (underline/circle) errors, but I don’t correct them. -e.g. has went
D. I indicate (underline/circle) errors and categorize them, but I don’t correct them -e.g. has went (verb form)
E. I hint at the location of errors F. I hint at the location of errors and categorize
them. G. Others (Please specify) 6. Indicate the extent to which you agree with the following statements according to the scale below. Please circle the most appropriate box for each statement.
Item Description 1
Strongly disagree
2 Disagree
3 Agree
4 Strongly
agree A. There is no need for teachers to provide
feedback on student errors in writing. 1 2 3 4
B. It is the teacher’s job to locate errors and provide corrections for students.
1 2 3 4
C. Teachers should vary their error feedback techniques according to the type of error.
1 2 3 4
D. Coding errors with the help of a marking code is a useful means of helping students correct errors for themselves.
1 2 3 4
E. Marking codes should be easy for students to follow and understand.
1 2 3 4
F. All student errors deserve equal attention.
1 2 3 4
G. Students should learn to locate their own errors.
1 2 3 4
44 Cheol Baek ∙ Sookyung Cho
H. Students should learn to locate and correct their own errors.
1 2 3 4
I. Students should learn to analyze their own errors.
1 2 3 4
7. Do you have any concerns and problems regarding providing error feedback on student writing?
Please elaborate.
APPENDIX B Sample Essay
Various use ways of laptop or computer
Nobody doubts that computers are affecting humans’ lives in the world. There are various ways to
use computers. These kinds of uses have broadened people’s views. I agree this fact because
computers have impacted in my life. There are four different ways to use a computer: gaming,
business uses, searching for information, and social networking. This essay is consisted like these
orders.
First of all, gaming is common use of computers. I have an enjoyment with computer games since
my father bought a computer to me when I am young. It was fantastic to me to enjoy my time
because I could experience very interesting games. Actually, gaming helped me reduce my stress in
school and relationships between others. As I grow up, I spent little time playing computer games, but
sometimes I still enjoy them. When my friends and I don’t have anything to do, we can meet in on-
line game and have fun together.
Another way for using a computer is for working. I have an experience from a cloth company
when I took a semester off from my university. I arranged the warehousing of goods on a chart with
my computer. If I tried to arrange this work without the computer, I would have written the numbers
one by one through hands. Using a computer was also helpful to design something. When my boss
ordered me and my colleague to designing sleeves, the design program and photo shop helped us
make it easy. It was easy to change the design when we didn’t satisfy it after a discussion. Using a
computer was absolutely necessarily to do this job.
The last one thing is social networking with others. Because human relationships are complex, it
helps manage them easier. I can notice a friend’s birthday through receiving message or talk with my
friends studying overseas. Also, I can make new friends in online chatting. Many people use social
networking to do this every day.
As a result, the computer is effectiveness through various ways. It enriches people’s lives with
many contents to accomplish their own purpose. The computer frequently helps with my tasks
through various ways as if it was a teacher, a friend, and family member.
From Perception to Practice: Native and Non-native English Teachers’ Use of Error Correction in … 45
Applicable levels: Elementary, secondary, tertiary
Cheol Baek
Department of English Linguistics
College of English, Hankuk University of Foreign Studies
107 Imun-ro, Dongdaemun-gu
Seoul 02450, Korea
Phone: 02-2173-2266
Email: [email protected]
Sookyung Cho
Department of English Linguistics
College of English, Hankuk University of Foreign Studies
107 Imun-ro, Dongdaemun-gu
Seoul 02450, Korea
Phone: 02-2173-3194
Email: [email protected]
Received on December 1, 2015
Reviewed on January 15, 2016
Revised version received on February 15, 2016