Académique Documents
Professionnel Documents
Culture Documents
entitled
by
Submitted to the Graduate Faculty as partial fulfillment of the requirements for the
_________________________________________
Dr. Douglas W. Coleman, Committee Chair
_________________________________________
Dr. Sharon Barnes, Committee Member
_________________________________________
Dr. Ruslan Slutsky, Committee Member
_________________________________________
Dr. Patricia R. Komuniecki, Dean
College of Graduate Studies
December 2014
Copyright 2014, Rosemary Kathryn Sorg
This document is copyrighted material. Under copyright law, no parts of this document
may be reproduced without the expressed permission of the author.
An Abstract of
by
Submitted to the Graduate Faculty as partial fulfillment of the requirements for the
Master of Arts Degree in
English
December 2014
(ESL) student writers have been assigned to work with non ESL-focused tutors rather
than ESL- focused tutors. Because this has adversely affected the ESL students writing
in previous semesters, this study was aimed at finding out if studenst in the masters
tutors gave more accurate error identification and correction on ESL texts. To test this, a
stimulus text was created containing ten British dialect variants and ten ESL errors to test
the participants accuracy on error identification and correction. The results show that
Writing Center tutors were more accurate at error identification and correction than the
iii
Table of Contents
1 Literature Review.....................................................................................................1
1.5 Training in the ESL Masters Program Versus the Writing Center ..................6
2 Methodology ..........................................................................................................9
References ..........................................................................................................................22
iv
List of Figures
2.2 Table of DVs and ESL Errors in the Stimulus Text ..............................................11
v
List of Abbreviations
L1 ...............................First Language
L2 ...............................Second Language
UF ..............................Ultimate Frisbee
WC .............................Writing Center
WCTs .........................Writing Center Tutors
vi
Chapter 1
Literature Review
1.1 Introduction
writer better at writing. For English as a second language (ESL) writers, error
identification and correction from instructors and tutors needs to be accurate and tailored
to their needs and focused on helping them improve their writing because they are not
writing in their first language. The errors that typically appear in ESL writing differ from
the ones that appear in native English speakers (NES) writing (Harris and Silva, 1993).
At the University of Toledo, two groups of student tutors predominately aid in the
identification and correction of errors in the writing of ESL writers enrolled in the ESL
freshman composition courses. The first group consists of the graduate students in the
ESL masters degree program, studying to become future ESL instructors. The second
group consists of students who are undergraduates employed at the Writing Center as
tutors. This study strives to see who gives ESL writers more accurate responses to local
errors in a text containing global and local errors, students in the ESL masters program
Two main error types appear in all writing, including ESL writing: global and
local errors. Global errors are the errors that interfere with the intended readers
understanding of the text (Harris and Silva, 1993, p. 526). Local errors, while still
errors, do not affect the meaning of the text (Harris and Silva, 1993). In an ESL writing,
global errors should take precedence for correction over a local error as global errors
However, by no means should ESL writers local errors be ignored. Local errors
are either as important as, if not more important than, meaning-related concerns [global
errors] (Zamel, 1985, p. 82). Local errors demonstrate the complex problems ESL
writers have when writing (Stanley, 2004). Local errors can be caused by the ESL
writers non-native ability to assess and fix an error, past instructional errors, bizarre
English as a second language writers commit both global and local errors in their
writing. The reason for the errors stem from their non-native English background (Ferris
and Hedgcock, 2005). ESL writers are broadly defined as anyone whose first language is
university. While they share the same non-native English language background, writers
are at different levels, have different language skills, have different writing styles, write
in different accents, and may not be familiar with the cultural expectations of the English
2
speaking country in which they are studying but still share common writing errors despite
their diverse L1 backgrounds and English proficiency levels (Harris and Silva, 1993).
The main issues all ESL writers have to improve on are understanding writing
assignments, overcoming inherent weaknesses in their own writing, avoiding content and
organization errors, and avoiding grammatical errors. When it comes to ESL writing, a
basic cause of global error they have is in understanding what kind of writing is being
asked for in a given writing assignment. ESL writers who cannot explain the assignment
or what they wrote in response to the assignment struggle to do well in writing courses
(Al-Zahabe, 2013). If they cannot explain what the assignment asks them to do, then the
error they will produce is a global error. Some reasons writers struggle to understand an
assignment are their understanding of the precise language in the assignment and/or the
For content and organization, which are global errors, ESL writers struggle to stay
focused on what they want to say in their writing. Writing that is riddled with errors can
block the writers messages (Bates, Lane, and Lange, 1993). Contextual issues can stem
understand ESL writing. Depending on the effect of the error, grammar issues could
result in global errors or local errors (Harris and Silva, 1993). Causes of these
3
conventions, the use of L1 grammar in L2 writing, and little to no experience with
English rhetoric (Ferris and Hedgcock, 2005). These grammar issues vary across
spelling, nouns, verbs, articles, prepositions, and word choice. For verbs, ESL writers
may make mistakes with inflectional morphology, verb formation, verb deviation, verb
completion (Harris and Silva, 1993), verb tense (Woodward, 2013), passive construction,
modal construction, and subject-verb agreement (Ferris and Hedgcock, 2005). Noun
errors ESL writers make can be with inflection, derivation (Harris and Silva, 1993),
collective nouns, plural endings, and progressive endings (Ferris and Hedgcock, 2005).
Articles may be used in the wrong context, used in the wrong place, used when they are
not needed, and missing when they are needed (Harris and Silva, 1993). With
prepositions, ESL writers struggle to know which one goes with which particular nouns,
verbs, adjectives, and adverbs (Harris and Silva, 1993). Word choice errors may be
issues ESL writers have with writing in English, they need help with error identification
and correction on their writing tailored to these errors, especially global errors, as they
have different struggles in writing than native English writers (Harris and Silva, 1993).
Native English speaker (NES) writers do not have the same issues as ESL writers
when it comes to writing. The most obvious difference is one group grew up with
English as their first language and the other group did not. ESL writers definitely write
differently than NES writers (Harris and Silva, 1993). Because of issues in second
4
language acquisition, even high level ESL writers may not produce as advanced levels of
While ESL writers could be strong writers in their native language, they could
struggle to write well in English because their writing is inadequate (Zamel, 1995), and
not proficient (Harris and Silva, 1993). Zamel says,My experience with teaching ESL
students is that they have often not received adequate English instruction to complete the
required essay texts and papers in my class; she goes on to say these students may
have adequate intelligence to do well in the courses, but their language skills result in low
grades (Zamel, 1995, pp. 102-103). Perhaps if ESL writers had been better prepared in
lower level English courses in their native country before they stepped into an English
speaking, college level writing atmosphere, they would be better English writers.
ESL writers need more time, help, and slower pacing than NES writers (Harris
and Silva, 1993). L2 student writers have gaps in morphological, syntactic, and lexical
knowledge that are more pronounced than those of L1 writers (Ferris and Hedgcock,
2005, p. 262). ESL writers have no native speaker like language development, and less
English language exposure (Ferris and Hedgcock, 2005). Adult ESL writers plan less,
write with more difficulty primarily due to a lack of lexical resources, reread what they
have written less, and exhibit less facility in revising by ear, that is, in an intuitive manner
- on the basis of what sounds right than their NES peers (Harris and Silva, 1993, p.
529). Since ESL writers and NES writers write differently, it is natural that ESL writers
will produce different writing errors, both global and local errors than the NES writers.
5
1.5 Training in the ESL Masters Program Versus the Writing Center
At the University of Toledo, the training that the graduate students in the MA
ESL program (ESLs) receive differs from the training Writing Center tutors (WCTs) have
received. The ESLs learn to identify and give feedback on both global and local errors in
ESL writing through the course Approaches to ESL Writing. This course assists students
in how to instruct ESL students through the entire writing process (Ferris and Hedgcock,
2005). In this course, students are equipped with theoretical and practical knowledge of
the L2 writing process, as well as an awareness of the unique linguistic and cultural
expertise of L2 writers (Ferris and Hedgcock, 2005, p. 24). In this course, students are
snapshots of other students writing. WCTs typically do the following in their sessions
with student writers: greet them, converse with them about their writing and assignment,
read aloud through the paper, highlight major/prevalent errors and or what the students
want to focus on, praise the student for what they do well, and summarize the session
(Al-Zahabe, 2013). While this process is an excellent one to follow for tutoring writers,
if the tutor has little experience with ESL writers, the session can quickly devolve into
The issues that can pop up when tutors with little to no ESL tutoring experience
are that tutors will be overwhelmed, will not know where to begin, will get frustrated
while trying to help the ESL writers, will have lower tolerance levels, will not put forth
6
their best effort, will not commit to the session, will have a lack of patience (Al-Zahabe,
2013, Woodward, 2013), and will discount their [ESL writers] intelligence with the
way they write (Woodward, 2013). Most likely with these issues, the ESL writers will
not receive accurate tutoring and their writing can severely suffer. Tutors need to
understand that ESL students are culturally and linguistically unique in order to tutor
them (Blau et al., 2001) because of the ESL writers non-native grammar, different
writing accent, and nonlinear writing (Harris and Silva, 1993). Ergo, tutors need to be
trained to understand how to accurately help ESL writers with error identification and
correction (Zamel, 1995). Tutors who are inexperienced with ESL writers will be caught
off guard and will view their writings as basic, unorganized, in need of a lot of help, and
could correct local errors over global errors (Harris and Silva, 1993).
After content, those who give feedback on ESL texts should focus on error
identification and correction of global errors over local errors. British dialect variants are
used in this study to represent a differing style of writing, and represents local errors in
the stimulus text. Global errors are represented by general ESL errors in the stimulus
text. At the University of Toledos Writing Center in past semesters, ESL writers were
consistently being tutored by NES tutors who have limited to no training in working with
them (Woodward, 2013). This study is focused to see if graduate students enrolled in the
ESL masters degree program, who are future/current ESL instructors (ESLs) give better
error identification and correction than undergraduate NES Writing Center tutors (WCTs)
on a text that contains not only ESL errors that represent global errors, but also British
7
dialect variants, which represent apparent t local errors. Based on the research, will the
ESLs provide more accurate error identification and correction than the WCTs?
8
Chapter 2
Methodology
2.1 Participants
The participants of this study are five graduate students from the English as a Second
Language masters degree program and five undergraduate students who are employed as
tutors at the University of Toledos Writing Center. All participants are native speakers
of English. The age range of the participants is twenty to forty-two years old. A small
number of participants were recruited because the university has a small number of
graduate students in the ESL program and a small number of Writing Center tutors. Only
five for each group were used to keep the numbers of participants even for both groups.
The data from the graduate students (ESLs) are labeled as Participants 1-5. The data
from the tutors (WCTs) are labeled as Participants 6-10. All participants have experience
working with student writers. Of the five graduate students, all have taken a course in
providing feedback to ESL writing, all but one had taught or were currently teaching an
ESL freshman composition course at the time of data collection, and one had experience
as an ESL writing tutor in the Writing Center. Of the five WCTs, two were currently
taking the course in providing feedback to ESL writing during data collection, two of
them had no training in tutoring ESL writers, and one had limited experience, but no
9
Instructions: In the following text, please identify any errors you find by circling
them. After you have identified an error, please write a correction above the error,
below the error, or in the margins.
Ultimate Frisbee Text
Ultimate Frisbee (UF) is my favorite sport because it is a very active and
fulfiling sport. I learned how to play UF when I was in high school. On my very
first UF game, nobody betted I would catch the frisbee, and scored my very first
goal. Now I play UF at least every week; usually at the weekend with my friends.
UF is a bit like playing American football or rugby, but with a frisbee. Anyone from
any skill range can play UF. UF can be played in almost any open area with almost
any number of players for any length of time or to however high a score.
Before playing Ultimate Frisbee, it is important to warm up by stretching and
throwing the frisbee around with friends. If a person does not stretch, he could pull
a muscle; which would be very painful. Did you ever pull a muscle before? I had
got one once and it spoilt all my fun because it hurt too much to run; so I leant
against a tree to stretch and watch that game. After warming up, players can be
splited up into teams in any manner they see fit; numbering off, guys versus girls,
flip disk, etc.
Once teams have got picked, the game begins. Each team starts off on
opposite ends of the field behind their goal. One team member throw the frisbee to
the other team. The goal of each frisbee team are to score a goal by having one
team member catch the frisbee behind the other teams goal. Teams can deside how
high they will play to, score wise. Keeping score mentally is a great way to practise
math. In a team, usually only one person per team keeps score. I have kept score
on more often than not.
Ultimate Frisbee is not only an American sport, but is also an international
sport. There are leagues, some of which are professional, at both Europe and Asia.
This sport is so serious, that in the professional leagues, players have gotten to take
drug tests regularly so to make sure they is not taking performance enhancing drugs.
The European Union are especially adamant about this. The EU has got very
strict rules and regulations about drug testing because of the past issues with athletes
who used performance enhancing drugs that.
One day I will hopefuly be a good enough player that I can join a professional
Ultimate Frisbee league.
10
training, in tutoring ESL writers. The two tutors who had taken the course in providing
feedback on ESL writing could affect the outcome of the study, but there were a limited
number of subjects to select from, so they were chosen as participants. Each participant
was approached separately and given the stimulus text, Figure 2.1, to read over and
provide error identification and correction. Figure 2.1 below lists the DVs in bold and
the ESL errors in italics. The actual stimulus text had the DVs and the ESL errors in
italics. Figure 2.2 below is a table showing what DVs and ESL errors are contained in
Figure 2.2: Table of DVs and ESL Errors in the Stimulus Text
The stimulus text for this study was created with advice from Dr. Douglas W.
Coleman because an authentic text containing exactly ten similar British dialect variants
(DVs) and ten corresponding ESL errors could not be found. The stimulus text was
created and not an authentic text so it would have an equal number of DVs and ESL
11
errors. The stimulus text contains two types of the following categories for the DVs and
ESL errors: spelling, verb deviations/errors, prepositions, have/got errors, and collective
nouns/plural confusion errors. There are two errors per matched pairs for both the DVs
and the ESL errors. The matched pairs were specifically created to be similar between
the DVs and the ESL errors so that the differences between the two are almost
unidentifiable by the participants. The text describes the sport ultimate frisbee, and the
DVs and ESL errors are distributed throughout the text so that similar errors/variants are
2.3 Procedure
grammatical level. At the grammatical level, the results of this study are quantitatively
calculable and can reveal the differences between the ESLs and WCTs ability to identify
and correct errors in an ESL text. The participants are instructed at the top of the text to
please identify any errors you find by circling them. After you have identified an error,
please write a correction above the error, below the error, or in the margins (from
stimulus text). Participants were assigned the same text in which to give feedback. The
independent variable in this study is the WCTs and their prior training and the dependent
Participants were given the stimulus text to read and provide error identification
and correction. All participants read through the stimulus text once before providing
error identification and correction. After that, they provided error correction they saw fit.
Participants 1 (ESL), 2 (ESL), 5 (ESL), 6 (WCT), and 7(WCT) provided error correction
12
as exactly as instructed. Participant 3 (ESL) did not provide error correction on a
grammatical level instructed in the stimulus text, but instead provided feedback on
content and organization; this participants results were not added into the studys results.
Participants 4 (ESL) and 5 (ESL) both provided feedback on content and organization as
well as the instructed error correction in the stimulus text. Participant 9 provided error
comma usage. Participant 10 provided error correction as instructed and comma usage.
Once each participant had completed providing their error correction on the
stimulus text, the data was compiled in R Portables R Commander. For the DVs, if the
participants marked them as an error, which is inappropriate, they received the value 0,
and if they did not mark them as an error, which is appropriate, the value 1. For the
ESL errors, if the participants marked them as an error, which is appropriate, they
received the value 1, and if they did not mark them as an error, which is inappropriate,
the value 0. The scores could range from 0-20. If a participant ignored all ten DVs, but
marked all ten ESL errors as errors, they would receive a score of 20. For example,
Participant 1 (ESL) appropriately identified only two DVs as not being errors (2 points)
and appropriately identified seven ESL errors as errors (7 points), yielding a total score of
only one DV as not being an error (1 point) and appropriately identified six ESL errors as
13
Error Type Label in Data Table
This study is focused on whether graduate students who are enrolled in the ESL
masters degree program and who are future/current ESL instructors (ESLs) perform
superior on error identification and correction than NES Writing Center tutors (WCTs) on
a text containing DVs and ESL errors. Figure 2.3 explains which error types were used
and their corresponding data label in Figure 2.4, the data set. The data set (Fig. 2.4)
shows the results for each participant on each DV/error they marked.
14
Particip 1 2 3 4 5 6 7 8 9 10
ant
Type ESL ESL ESL ESL ESL WCT WCT WCT WCT WCT
BDsp1 0 1 1 1 0 0 1 0 0 0
BDsp2 1 0 1 0 0 1 0 0 0 1
BDvb1 0 0 1 0 0 0 1 0 0 0
BDvb2 0 0 1 0 0 0 1 0 0 0
BDpr1 0 0 1 0 0 0 1 0 0 0
BDpr2 0 0 1 1 1 1 1 1 1 0
BDhg1 1 0 1 1 0 0 0 0 0 0
BDhg2 0 0 1 1 0 1 0 0 1 1
BDpn1 0 0 1 1 0 0 0 0 0 1
BDpn2 0 0 1 1 0 0 1 0 0 1
ESLsp1 0 0 0 1 0 1 1 0 1 1
ESLsp2 0 1 0 0 1 0 0 1 1 0
ESLvb1 1 1 0 0 0 1 0 1 1 1
ESLvb2 1 1 0 1 1 1 1 1 1 1
ESLpr1 0 1 0 0 0 0 0 1 1 0
ESLpr2 1 1 0 1 1 1 1 1 1 1
ESLhg1 1 1 0 1 1 1 1 1 1 1
ESLhg2 1 1 0 1 0 1 1 1 1 1
ESLp1 1 0 0 0 1 1 1 1 0 1
ESLp2 1 1 0 1 1 1 1 1 1 1
Score 9 9 10 12 7 11 13 10 11 12
15
Chapter 3
Each participant, four ESLs (participant 3 was removed from the study) and five
WCTs, were given a score based on their accuracy of identifying and correcting ESL
errors over British Dialect Variants (DVs). Figure 3.1 is the breakdown of each
participants score and Figure 3.2 is a table that shows the average score of the two
groups. Figure 3.1 also shows what group the participants were in (ESL or WCT), if they
took the course on providing feedback on ESL writing (yes or no), and what score they
Figure 3.1: Participants Scores . *(Did not give feedback as instructed. Not calculated in
group results).
16
Figure 3.2 shows the t test results for both groups. The results show that on
average, the WCTs had better accuracy of error identification than the ESLs:
The result of this study is unanticipated. Based on the research and experience in
the ESL masters degree program, the ESLs, especially with their training and experience
were expected to have a higher average score than the WCTs who had less to no
17
For those who were currently taking the course Approaches to ESL Writing, it
was expected that they would have had a higher average score than those who had not
taken the course. Figure 3.3 shows the t-test results for these two groups. The results
show that on average, those who did not take the course had better accuracy of error
The results of this study show that not only did the WCTs give more accurate
error identification and correction on the stimulus text containing DVs and ESL errors,
but also, the participants who did not take the course Approaches to ESL Writing also
gave more accurate error identification and correction than those who did take the course.
While the WCTs did do better than the ESLs, both groups actually did rather
poorly on their accuracy in error identification and correction on the stimulus text. The
ESLs group average is 9.6/20, which is 48%, an abysmal percentage. The WCTs did
only slightly better with their average as 11.4/20, or 57%, a higher but still terribly low
percentage.
If the DVs were to be taken out of the scoring, the results get better for both
groups. Without the DVs affecting their scores, the ESLs average 6.75/10, or 67.5%,
which is still a poor percentage of accuracy for the group that is or will be ESL
instructors. Without the DVs affecting their scores, The WCTs average 8.2/10, or 82%,
a modestly acceptable percentage. Perhaps another study could be done that looks only
at both groups error identification and correction with a text of only ESL errors and does
not include any DVs to see if the WCTs once again do better than the ESLs.
18
In this study, the expectation was that the ESLs would be more accurate at error
identification and correction than the WCTs. The experience the ESLs had in working
with ESL writers varied from zero experience and just training in the masters degree
Participant 1 had no experience with ESL writers and scored 9, Participant 2 had one
semester of ESL teaching experience and scored 9, Participant 4 had five semesters worth
of ESL teaching experience with ESL writers and scored 12, and Participant 5 had four
semesters of ESL teaching experience and scored 7. With just this limited data from
these four ESL participants, it appears that the scores neither improved nor worsened
with experience as Participants 1 and 2 scored better than Participant 5 who had more
The ESLs training could potentially have caused these unexpected results. The
course they took taught them about working with ESL writers and how to properly give
feedback. Learning from a given program will be promoted in direct proportion to its
implementation (Merrill, 2002, p. 44). When learning a new skill, once the skill is
acquired, it is integrated into the learners world (Merrill, 2002). A possible reason the
ESLs were not better at error identification and correction compared to the WCTs is that
because they are more used to looking at ESL writings they became blind to seeing
anything but ESL errors in the text. Since they are trained with ESL writing, they were
expecting ESL writing and proceeded as they were trained; anything that seemed foreign
to them, like the DVs, looked like errors. The ESLs learned to give error identification
on ESL writing, so when they saw the stimulus text, they transferred the knowledge they
19
learned from the course (Driscoll and Driscoll, 2005), which is giving feedback to ESL
writing, and they implemented the procedure they would use on a regular ESL text to the
stimulus text.
Another possible cause for this unexpected result, and a critical flaw in the study,
is the limited number of participants because participants were only recruited from one
university. Ten participants, and having calculable data from only nine participants, is
too low a number to make a definitive result in this study. Another study could be done
using the stimulus text as it is or edited to remove all the DVs, and then tested on a larger
group of participants who are ESLs and WCTs from multiple universities in order to find
out if ESLs give better error identification and correction than WCTs.
The hypothesis of this study was that the students enrolled in the ESL masters
degree program (ESLs) would be better at error identification and correction than the
undergraduate writing center tutors (WCTs) on the stimulus text that contains both DVs
and ESL errors. However, in this study, that was not the case. Over all, the WCTs were
better at error identification and correction on the stimulus text than the ESLs. While
there were issues in the past with ESL writers being tutored by WCTs, it appears based
on this limited data that this may no longer the case at least with regard to local errors.
20
References
Al-Zahabe, L. (Writing center ESL tutor) in discussion with the author, November 2013.
Bate, L., Lane, J., & Lange, E. (1993). Writing clearly: Responding to ESL
Blau, S.R., Hall, J., Davis, J., & Gravitz, L. (2001). Tutoring ESL students: a different
Harris, M. & Silva, T. (1993). Tutoring ESL students: issues and options. College
Woodward, D. (Writing center ESL tutor) in discussion with the author, November
2013.
Zamel, V. (1995). Strangers in academia: The experiences of faculty and ESL students
across the curriculum. In Corbett, E.P.J., Myers, N., & Tate, G. (Eds.), The
Writing Teachers Sourcebook (pp. 100-112). New York, NY. Oxford University
Press.
21