The Washback Effects of High School Third Grade Exam on EFL Teachers’ Methodology, Evaluation and Attitude

Introduction

Washback or backwash refers to the effect of testing and assessment on teaching and learning
processes (Cheng, Watanabe & Curtis, 2004) and follows the idea that tests or examinations
can  and  should  drive  teaching  and  learning  processes.  Interchangeably  referred  to  as
measurement-driven  instruction,  the  concept  entails  a  match  between  the  content  and  the
format  of  the  test  and  the  format  and  content  of  the  instruction  (Cheng  et  al.,  2004).  The
consistency  or  match  has  also  been  termed  as  curriculum  alignment  (Shepard,  1990).  From
another point of view, the test effects and the scope of such effects persuaded Wall (1996) to
distinguish between test impact and washback. According to Wall, impact refers to the effects
of  a  test  on  individuals,  policies,  or  practices  in  different  contexts  including  the  classroom,
school, the educational system, and even society at large, while washback/backwash refers to
the  effects  of  tests  on  teaching  and  learning  processes.  Washback  is  inherently  believed  to
move  in  a particular direction to describe testing–teaching relations;  however,  Alderson and
Wall  (1993)  identified  the  bidirectional  nature  of  washback  as  either  positive  or  negative.
Negative  washback,  the  undesirable  effect  of  tests  on  teaching  and  learning,  happens  when
there is no match between the assumed goals of teaching and the focus of assessment. On the
other  hand,  positive  washback  is  described  as  the  positive  attitude  towards  the  test  and
cooperative  functioning  to  ensure  its  assigned  purposes.  According  to  Alderson  and  Wall
(1993, p.66) a test has a positive effect “if there is no difference between the curriculum and
teaching  to test.”  From  a  rather  different  perspective,  Watanabe  (2004)  described  washback
in  terms  of  its  dimensions,  aspects  of  learning  and  teaching  influenced  by  the  examination,
and the factors mediating the process of washback being generated.  
Washback  effect  has  attracted  great  attention  in  recent  years  in  different  educational
contexts and has been one of the main lines of research in both general education and foreign
or second language educational settings (e.g., Chapman & Snyder, 2000; Cheng, Sun & Ma,
2015; Davies, 1968; Green, 2007; Madaus, 1998; Shih, 2007; Spratt, 2005; Xie, 2015; Zhan
& Andrews, 2014).
As  an  important  washback  effect  of  high  stake  tests  and  contrary  to  the  perceived
common rule proposing that test comes  after teaching and  learning processes, the  priority  is
inverted in the case of many high stake tests (Cheng, 1997) so that in such testing situations
testing comes ahead of teaching and learning. This effect, in turn, influences different aspects
and stakeholders of the educational process. As Hughes (1993, p.2) asserts “in order to clarify
our  thinking  on  backwash,  it  is  helpful  to  distinguish  between  participants,  processes,  and
products in teaching and learning recognizing that all three may be affected by the nature of
 

the test”. Furthermore, that educational  systems  may  be  both directly and  indirectly affected
by high stake tests like school leaving examinations.  
Researchers  such  as  Swain  (1985)  underscored  the  positive  aspects  of  test  effects  on
language learning and language curriculum. Swain believed that teachers would “teach to the
test”. In other words, knowing the content and format of the test, the teachers would teach the
same  or  similar  content  more  effectively.  Similarly,  Wall  (2000)  believed  that  the  results  of
the tests’ ‘differentiating rituals’ are, sometimes, so effective in the testees’ future life that the
other stakeholders (e.g., teachers) do whatever necessary to help the learners pass the test and
the students’ parents ask them to do any possible activities to pass it.
The  effects  of  the  tests  on  teachers  and  learners  are  well  documented  and  various
studies  have  examined  this  effect.  It,  however,  seems  that  the  washback  effect  of  a
nationwide  high  stake  test  like  the  third  grade  school  leaving  final  exam  on
teachers’methodology, attitude, and assessment procedures in Iranian high school mainstream
educational context  is  still  understudied.  As a partial attempt to address the  need, this study
was conducted to examine the washback effects of  the third grade final exam as an annually
held  exam  in  the  Iranian  general  education  system  on  specifically  EFL  teachers’  teaching
methodology,  assessment  procedures,  and  attitude  towards  this  aspect  of  the  general
education.  
 
Review of the related literature
Testing  and  assessment  in  versatile  forms  are  integral  parts  of  every  system  of  education.
This is why assessment is primarily designed to service teaching and learning (Davies, 1990).
However, a role reversal  has recently occurred in  educational  settings  because of the  impact
high  stake  tests  exert  on  different  components  of  teaching  and  learning  process  which  has
altered teaching to be at the service of testing. This pernicious influence of tests on what goes
on  in  the  educational  environments  and  classrooms  in  particular  as  well  as  on  the  teachers’
teaching procedure has raised some concerns among EFL experts.  Additionally,  it has given
rise  to  a  plethora  of  studies  on  the  tests  and  their  possible  effects  on  the  stakeholders
including  participants,  test  developers,  and  administrators  (e.g.,  Alderson  &  Wall,  1993;
Bailey,  1996,  1999;  Chapelle  &  Douglas,  1993;  Damankesh  &  Babaii,  2015;  Hamp-Lyons,
1997; Shohamy, Donitsa-Schmidt & Ferman, 1996: Watanabe, 2004; Xie, 2015).
The aims and scope of washback studies have been quite versatile. Bailey (1996, 1999),
for  example,  proposed  that  washback  should  minimally  examine  both  washback  to  the
 

program (results of test-derived information provided to teachers, administrators, curriculum
developers,  counselors,  etc.)  and  washback  to  learners  (the  effects  of  test–derived
information  provided  to test takers)  from  teachers’  and  students’  perspectives.  According  to
Fulcher  and  Davison  (2007),  washback  studies  should  highlight  “those things that we do in
classroom  because  of  the  test,  but  ‘would  not  otherwise  do’  (p.221).  Furthermore,  in  their
washback  hypothesis,  Alderson  and  Wall  (1993,  p.117)  state  that “teachers and learners do
things they would not necessarily otherwise do because of the test”.  
Washback  researchers  attested  that  any  test  may  be  of  both  positive  and  negative
effects. According to Wall (2000), positive effect  is the drive that persuades testees to cover
all  subjects  completely,  complete  their  assigned  syllabuses,  and  get  familiar  with  other
teachers’ standards. On the other hand, quoting Wiseman (1961), Wall (2000) maintains that
the negative aspect of the test encourages teachers to watch the examiner’s  foibles  and  note
his  idiosyncrasies  to  prepare  students  for the  most  likely  test  items  that  might  appear  in  the
examination. This  negative washback effect restricts teachers’ teaching  styles and persuades
them  to  concentrate  on  the  ‘purely  examinable  side’  of  their  work  and  by  neglecting  other
areas.  Accordingly,  possible  positive  and  negative  washback  effects  of  such  tests  provide
ample opportunities and foci for the studies in this realm.  
Studies  on  the  washback  effects  of  high  stakes  tests  have  shown  that these  tests  make  
teachers focus on those points that are likely to appear in the tests and teachers usually do not
take pedagogical aspects of instructions into account as they usually teach to the test ( Hamp-Lyons, 1997). Furthermore, as Bachman (1990) believes, negative washback would result in
testing  determining  the  content  of  teaching.  However,  it  is  noteworthy  that  the  extent  and
nature of test consequences or washback effect depends on teachers’ educational background,
past learning experience, and beliefs about effective teaching and learning (Watanabe, 2004).  
In  an  effort  to  further  clarify  the  extent  and  nature  of  washback  effect,  Smith  (1991)
identified  five  components  of  change  as  a  result  of  washback  effects  of  tests  including  the
target  system,  the  management  system,  the  innovation  itself,  available  resources,  and  the
content  in  which  the  change  is  supposed  to  happen.  Also,  Hughes  (1993),  in  his  washback
model,  suggested  that  participants,  processes,  and  products  are  the  main  recipients  of  the
effect.  Participants,  in  Hughes’s  (1993),  included  “all  of  those  whose  perceptions  and
attitudes toward their work  may  be affected by  a  test”. The three elements of the  model are
described as:   

1.  Participants: Students, classroom teachers, administrators and material developers and
publishers  whose  perceptions  and  attitudes  towards  their  work  may  be  affected  by  a
test.
2.  Processes: Any actions taken by the participants which may contribute to the process
of learning.
3.  Products: What is learned and quality of the learning (Hughes, 1993, p.2)
As can be seen in the model, teachers constitute the most noticeable participants in washback
studies.  
          Concerning  washback  effect  type  and  degree,  Alderson  and  Hamp-Lyons  (1996)  in
their study on the washback effect of TOFEL test reported lots of variations among teachers’
perspectives. They  maintained that “our  study  shows  clearly  that  the  TOEFL  affects  both
what and how teachers teach, but the effect is not the same in degree or in kind from teacher
to  teacher”  (p.295).  Contrary  to  the  results  Alderson  and  Hamp-Lyons  (1996)  obtained,
Alderson  and  Wall  (1993)  examined  the  washback  effect  of  innovative  tests  on  Sri  Lankan
educational  system  and  found  that  tests  can  affect  content  of  teaching  but  less  likely  they
affect the teaching procedure. However, Cheng (1997) in a study on the revised Hong Kong
Certificate of Education Examination (HKCEE) found that “84%  of  the  teachers  believed
they  would  change their teaching methodology as a result of the introduction of the revised
HKCEE”(p.45). Similar to Cheng, Lam (1994) supported the washback effect of the tests on
the  teaching  methodology  but  he  further  noted  that  an  important  factor  affecting  the
methodology  change  as  a  result  of  tests’  washback  effect  is  the  teaching  experience  of  the
teachers.  He  stated  that  experienced  teachers  were  much  more  examination-oriented  than
their younger counterparts.
Some researchers have investigated the washback effect of different high stake tests on
teachers  and  students’  behavior  and  attitude  in  the  Iranian  educational  context.  Ghorbani
(2008),  for  instance,  conducted  a  survey  on  the  washback  effect  of  University  Entrance
Examination  (UEE)  on  the  teaching  practices  of  a  group of  pre-university  English  teachers.
Ghorbani  examined  the  six  dimensions  of  classroom  activities  and  time  management,
teaching  methods,  teaching  materials,  syllabus  design,  teaching  content,  and  classroom
assessment.  The  results  showed  that  all  of  the  participating  teachers,  regardless  of  their
demographics, were affected negatively by the nationwide high stake UEE.
Contrary  to  the  results  reported  by  Ghorbani  (2008),  Mousavi  and  Amiri  (2011)
investigated  the  washback  effect  of  Master  of  Arts  level  TEFL  University  Entrance

Examination on the academic behavior of students and professors. They used an observation
checklist  and  two  questionnaires  to  gather  the  required  data.  The  questionnaires  were
responded  by  32  university  teachers  and  210  students.  They  concluded  that  the  test  had  an
insignificant  effect  on  the  students  and  professors’  academic  behaviors.  Nikoopour  and
AminiFarsani (2012) evaluated the  washback effect of State and  Azad  UEE on Iranian EFL
candidates  and  high  school  teachers.  They  found  that  UEE  had  influence  on  teachers’
methodology,  content  of  educational  programs,  students’  learning strategies, and teachers’
method  of  evaluation,  students  and  teachers’  attitudes  and  students’  affective  domain.
Furthermore,  Razavi  Pour,  Riazi  and  Rashidi  (2011)  investigated  the  effects  of  teacher's
assessment literacy in moderating the washback effects of summative tests in the EFL context
of  Iran.  For  this  purpose  a  test  of  assessment  literacy  and  a  questionnaire  on  teaching
methodology  were  administered  to  53  EFL  secondary  school  teachers.  The  results  revealed
that  EFL  teachers  suffer  from  poor  knowledge  of  assessment  and  demands  of  external  tests
affect  their  teaching  and  assessment  procedures.  Moreover,  Nazari  and  Nikoopour  (2011)
investigated  the  washback  effects  of  high  school  examinations  on  120  female  Iranian  high
school  learners'  language  learning  beliefs  and  found that  first, the participants agreed on the
type of washback effect of the exams and second, there is a correspondence between different
factors of learners’ language learning beliefs and foreign language learning process.  
In another study, Mokhtari and Moradi Abbasabadi (2013) studied the washback effect
of  Iranian  school-leaving  tests  of  English  (ISLTE)  on  teachers’  perceptions  and
performances.  They  interviewed  and  observed  10  high  school  English  their  classes.  The
findings  verified  that  ISLTE  had  a  strong  negative  washback  effect  on  their  teaching
procedures.  The  negative  washback  effect  of  the  test  was  shown  in  the  form  of  materials
translation  by  teachers  and  the  absence  or  disappearance  of  communicative  activities  in  the
observed  classes.  They  suggested  that,  due  to  the  strong  impact  of  the  test  on  the teachers’
teaching  methodology,  the  format  of  ISLTE  was  in  need  of  serious  revision.  Finally,
Amengual  (2010)  examined  the  washback  effects  of  a  high-stakes  English  Test  (ET)  on
curriculum,  materials,  teaching  methods,  and  teachers’  feeling  and  attitudes  and  found  that
ET clearly affected curriculum and materials.
In  addition  to  University  Entrance  Examination  (UEE)  at  different  levels  of  BA,  MA,
and PhD, there are some other nationwide high stake tests held by the Ministry of Education
in the context of general education of Iran. One such a test is the third grade high school final
examination.  The  test,  as  a  gate  keeping  test,  plays  a  determining  role  in  the  candidates’
 

follow up studies. Moreover, the grade point average of the examinations  has a direct effect
on  the  high  school  students’  university  entrance  examination  results.  Hence,  due  to  the
significance of these examinations for both teachers and students, this study was designed to
probe into the potential effects of a less frequently studied high stake test on the high school
teachers’ teaching methodology, testing and assessment procedures, and attitudes towards the
educational system. Against this backdrop the following research questions were posed:
 
RQ1:  Does  the  nation-wide  third  grade  final  English  examination  of  high  school  have  any
washback effect on English teachers’ teaching procedures?
RQ2:  Does  the  final  English  examination  have  any  washback  effect  on  English  teachers’
classroom evaluation and assessment procedures?
RQ3:  Does  the  final  English  examination  have  any  washback  effect  on  teachers’  attitude
towards different aspects of the educational system?
 
Method
Participants
One hundred sixty EFL teachers who were teaching third grade courses in high schools were
chosen  based  on  convenience  sampling  procedure  to  participate  in  the  study.  The
participating  teachers  were  teaching  in  the  two  cities  of  Malayer  and  Boroujerd.  They  were
all  high  school  English  teachers  and  either  held  MA  in  TEFL  (15  %),  or  BA  in  English
literature,  translation  or  TEFL  (85  %).Thirty  percent  of  participants  were  female  and  70
percent  were  male  EFL  teachers.  Most of  the  participants  had  the  experience  of  teaching  at
different grades or levels of high schools and pre-university centers. Table 1 summarizes the
teaching experience and the number of the participants.

Instruments
The  main  instrument  in  this  study  was  a  researcher-made  five-point  Likert  scale
questionnaire. The first version of the questionnaire was developed based on a few previously
designed  questionnaires  (e.g.,  Cheng,  1997;  Mousavi  &  Amiri,  2011;  Nikoopour  &  Amini
Farsani, 2012) and the ideas the researchers received from some TEFL experts. The first draft
included  19  statements  to tap the  participants’  opinion  about the  three  intended  areas  of  the
washback  effects of the test. The early draft was  reviewed  by two TEFL experts  in order to
ensure  its  content  and  face  validity.  The  draft  was  reviewed  and  revised  based  on  the
suggestions and the comments of the TEFL experts.
Afterwards,  the  questionnaire  was  piloted  with  60  EFL  teachers.  Analyzing  the
obtained  data  through  principle  component  factor  analysis  (PCA),  6  items  (9,  11,  12,  13,
16and 14) were excluded from the final  version of the questionnaire due to poor correlations
and  factor  loadings  (less  than  0.3).  The  final  version  questionnaire  included  13  five  point
Likert  scale  items  ranging  from  strongly  disagree  (with  the  assumed  value  of  1) to  strongly
agree  (with  the  assumed  value  of  5).  Cronbach’s  alpha  internal  consistency  reliability  index
of the questionnaire was estimated to be 0.71 (α= 0.71),  and  hence,  deemed  acceptable.  In
addition,  Keiser-Mayer  Olkin  test of  adequacy  of  items  was  fairly  acceptable  (KMO=  0.68)
and  Bartellet’s  test  of  Spherecity  was  significant  (p=.000).  The  final  13  item  questionnaire
was  used  to  tap  the  participants’  ideas  on  three  factors  of  the  EFL  teachers’  teaching
methodology,  evaluation  procedures  and  attitudes  towards  the  education  system  (see
Appendix A).
 
Procedures
The  participating  teachers  in  both  pilot  and  main  study  were  met  in  their  schools.  Consents
were obtained prior to the administration of the questionnaire. At the pilot phase of the study,
60 high  school EFL teachers took the questionnaire. The  main aim of the piloting stage was
to  do  a  validation  study  on  the  instrument  and  estimate  the  reliability  of  the  questionnaire.
After the pilot study and the PCA  statistical procedure, a group of 100  high school teachers
were  asked  to  take  the  questionnaire  and  the  collected  data  were  descriptively  analyzed  to
answer the research questions.
 

Results
Pilot study results
As  mentioned  before, the  first version of the questionnaire was  first administered to 60 EFL
teachers who were teaching the third grade high school English course. As shown in Table 2,
the  questionnaire  had  an  appropriate  level  of  adequacy  since  the  observed  KMO  value
exceeded  the  minimum  acceptability  level  of  0.5  or  0.6,  (KMO=0.68>  0.5  or  0.6).  In
addition,  the  Bartlett’s  test  of  Spehercity  was  significant  showing  that  the  principal
component factor analysis was safe to be conducted.

As in Tables 3 and Table 4, the factor analyses confirmed the strong correlation of the
questionnaire items with three main factors.

In  addition,  the  initial  eigenvalues  of  only  the  first  three  components  exceeded  the
criterion  value  of  1(Pallant,  2013)  and  the  cumulative  percentage  of  the  three  components
explained a total of 60.01 percent of the variance.  
  
As  seen  in  Table  4,  the  variance  was  divided  among  13  items  and  6  items  out  of  the
total of 19 questionnaire items were discarded due to poor correlation and factor loading (less
than  0.3).  Finally,  the  rotated  factor  matrix  identified  the  more  strongly  correlated  items  for
each  factor.  As  such,  the  final  questionnaire  including  13  variables  which  tapped  altogether
the three main factors was achieved. The three factors were named as methodology (factor 1),
attitude (factor 2) and evaluation (factor 3). Items 8, 3, 7, 2, and 5 were loaded on factor one,
items 15, 17, 18, 19 tapped factor two and factor three was tapped by items 1, 10, 4, and 6 of
the  19  item  questionnaire.  Meanwhile,  it  should  be  noted  that  as  some  of  the  questionnaire
items  (items  2,  15,  and  14)  were  negatively  correlated  with  the  factors,  hence,  reversely
computed ad analyzed.

Main Study Results
In the second phase of the study 100 EFL teachers took the validated 13 item questionnaire.
The data collected were analyzed in terms of descriptive analyses and frequency counts. The
frequency for each level of the Likert scale of the items of each factor was computed and the
average frequency for each was obtained. In addition, in order to obtain the mean value of the
responses to each questionnaire item, considering the assumed value of the levels of the scale
(strongly  disagree  =1,  disagree=2,  undecided=3,  agree=4,  and  strongly  agree=5),  the  mean
value  for  all  items  was  calculated  and  finally  the  average  mean  score  for  the  factor  was
obtained.   
        As Table 5 presents, about 72 percent (45.8+26.2) of the teachers believed that the exam
affected their teaching methods in EFL classes and the average mean score for this factor was
fairly high (3.84).
      It is necessary to add that the questionnaire items in the following tables  were minimally
presented due to the limited space of the tables and the full report of the questionnaire items
is presented in the appendix.

Evidently, 43 percent of the respondents agreed that if they were to teach a third grade
final  exam  preparatory  course  (item  2)  ,  they  would  use  the  same  methods  and  techniques
 
they  were  using  in  their  regular  classes  of  high  schools  in  which  the  academic  skills  and
abilities are to be given the first priority. Thirty seven percent “strongly agreed” with item 2
which means that, added to the percentage of the teachers who  “agreed” with this statement,
80  percent  of  the  teachers  “change  their  teaching  methodology  so  that  they  could  prepare
their  learners  for  the  test  in  an  attempt  to  guarantee  their  learners’  success  at  the  intended
test”. Items 7 and 8 of the questionnaire referred to the teaching tips and tricks for successful
test  taking  of  the  learners  and  ultimate  success  in  passing  the  test.  The  percentage  of
responses to these questionnaire items at different points of Likert scale were quite revealing
(Table  5)  confirming  the  existence  of  teaching  to  the test  process  in  the  studied  educational
context. On the other hand, 42 percent of the respondents  ‘agreed’, and 33 percent ‘strongly
agreed’ with the statement in item 3 of the questionnaire stating that “I teach the material and
learning points according to their importance level in the exam”. This means that the content
of teaching was also strongly affected by the test content as well, as altogether 75 percent of
the respondents accepted the stated rationale for the choice of the content of their teaching.
Table 6 presents the descriptive statistic information for the ‘evaluation’ factor. A total
of  about  84  percent of  the  teachers  either  agreed  (49  %) or  strongly  agreed  (35  %)  that  the
test exerted a significant effect on their evaluation and assessment procedures. The total mean
score  for  this  factor  (4.12)  compared  with  the  methodology  factor  (3.84)  appeared  to  be
significantly  higher  which  was  indicative  of  even  stronger  influence  of  the  test  on  the
teachers’ evaluation and assessment procedures.

As is evident in Table 6, a total of about 92 percent of the teachers either agreed (43 %)
or  strongly  agreed  (49  %)  with  the  first  questionnaire  item  (item  1)  saying  that  I  consider
third  grade  final  exam  while  teaching  and  testing  in  my  classes.  A  mean  score  of  4.34  was
 

clearly  indicative  of  the  strength  of  the  effect  of  the  test  on  the  addressed  areas  of  the
teaching  and  assessment.  More  or  less  similar  effect  was  evident  above  for  the  other  items.
Roughly  speaking,  items  1  and  4  considered  how  of  testing  and  6  and10  focus  on  what  of
testing. In other words, the two categories of  items  addressed the content and the procedure
of testing the teachers use in their educational context. The percentages and the mean scores
presented in Table 6 are strongly indicative of the influence of the third grade final exam on
both what and how of the teachers’ assessment and testing.
Finally  concerning  the  test’s    washback  effect  on  the  teachers’  attitude,  the  obtained
results, presented  in Table 7, show that totally about 45 percent of participants believed that
the test affected their attitudes significantly (29% agreed, 16% strongly agreed); however, 19
percent  were  undecided,  and  36  percent  denied  the  tests’  impact  (28  %  disagree  +  8  %
strongly disagree) in this regard. The average mean score for this factor (3.16) was the lowest
compared to the other two factors i.e., teaching methodology and evaluation procedures.

The comparisons for the descriptive statistic information gained for the three factors are
presented  in  Table  8.  According  to  the  obtained  results,  the  teachers’  evaluation  and
assessment procedures were highly affected by the nation-wide third grade high school final
exam and  the magnitude of the test’s washback effect on the teachers teaching methodology
was  placed  in  the  second  place  of  importance.  However,  it  seems  that  the  participants’
attitude  towards the  Iranian  general  education  system  was  not  highly  affected  by  the  test  as
the total mean score obtained for this factor (3.16) was fairly close to the mid position of the
Likert  scale  that  was  neutral  in  value.  Consequently,  the  findings  roughly  indicated  that the
 

first and second null hypotheses of the study which denied any kind of washback effect of the
test  on  the  teachers’  teaching  methodology  and  evaluation  and  assessment  procedures  were
both rejected while the third hypothesis which rejected the effect of the test on the teachers’
attitude was confirmed.

Discussion
Testing  and  assessment  as  integral  parts  of  education  play  a  wide  range  of  prognostic  and
diagnostic  roles  in  education  process  and  help  the  pedagogical  or  educational  processes
which might precede or follow them. However, their positive contribution to education is not
free  of  some  negative  effects  on  the  same  processes.  Both  positive  and  negative  effects  of
testing  on  the  follow  up  teaching  and  learning  processes  have  been  termed  as  washback  or
backwash by the testing and assessment scholars (e.g., Alderson &Wall, 1993; Hughes, 1993;
Wall, 1996). However, the present concern  with  washback was  ignited  by Messick’s (1989)
introduction  of  the  notion  of  consequences  into  his  definition  of  validity  (Fulcher,  2010).
While the existence of washback effect is not in question, the how of this effect is not so clear
(Tsagari, 2009) and hence needs to be studied. The need for the study of the washback effects
of high stake tests is clearly more significant than the same need for low stake tests due to the
wider  scope  of  the  consequences  accompanying  such  tests.  High  stake  tests  are  considered
and used as agents of change (Luxia, 2005); however, as many empirical studies have shown
and the stated results of the present study confirmed, the use of high stakes tests is not usually
as effective as they are planned (Qi, 2004, as cited in Fulcher, 2010) and sometimes not in the
same way as their designers meant (Andrews, 1994). The current study partially attempted to
address the washback effects of a high stake nationwide achievement test that is administered
by the end of third year of high school in the Iranian general education system.
 

By the end of the third year of secondary high school all subject matters taught during
the educational  year are subject to this  nationwide evaluation through which students across
the  country  take  a  single  test  for  each  subject  at exactly  the  same  time.  The  test  results  are
influential in the candidates follow up academic studies in higher education centers, colleges,  
and  universities.  The  washback  effects  of  the  high  stake  test  of  English  as  a  Foreign
Language  (EFL)  on  the  English  teachers’  teaching  methodology,  assessment  and  evaluation
methods, and attitudes towards  educational processes were  studied  in this piece of research. 
The washback effect of the test on the teacher variables was  focused on here as teachers are
highly  decisive  and  hence  most  visible  participants  in  washback  studies  among  other
participants (Baily, 1999) owing to the direct effect of tests on their pedagogical behaviors.
In  this  study,  a  researcher-made  and  validated  questionnaire  was  administered  to  160
EFL  teachers  who  were  teaching  English  courses  of  the  third  grade  of  high  schools  in  the
pilot  and  main  study  phases.  The  results  indicated  that  the  third  grade  nationwide  final  test
significantly  affects  EFL  teachers’  teaching  methodology  and  increases  teaching  to  the  test
effect. As is described above, the participating teachers’ teaching methods were be under the
negative  impact  of  the  test  since  they  stated  that  they  change  their  teaching  method  so  that
they  could  guarantee  their  students’  success  at  the  test.  This  point  completely  confirms
Alderson  and  Wall  (1993,  p.117)  who  stated  that  "teachers  do  things  they  would  not
necessarily  otherwise  do  because  of  the  test".  The  participants  also  openly  agreed  with  the
focus  on  the  teaching  tips  and  tricks  of  taking  the  test  to  the  sacrifice  of  the  academic  and
pedagogical aspects and content of the course. Hamp-Lyons (1997) referred to a similar point
when he noted that these tests made teachers focus on points that were likely to appear in the
tests  and  they  usually  did  not  take  into  account  pedagogical  aspects  of  instruction  which
meant that they taught to the test. This  finding  is  also consistent with that of  many previous
studies  such  as  Alderson  and  Hamp-Lyons  (1996),  Watanabe  (1996),  Cheng  (1997),  Luxia
(2005),  Spratt  (2005),  Ghorbani  (2008),  Nikoopour  and  AminiFarsani  (2012),  SeyedErfani
(2012), Zhan and Andrews (2014) and Damankesh and Babaii (2015) all of which confirmed
the negative impact of high stake tests on the teaching methodology of teachers.
However,  the  finding  confirming  the  negative  impact  of  high  stake  tests  on  teaching
methodology  was  not  consistent  with  few  studies  such  as  Alderson  and  Wall  (1993)  who
concluded that the tests influenced the content of  teaching  but had  no  impact upon teaching
methodology. Similarly, Shin (2009)  suggested that teachers’  instruction was  not vulnerable
to the test impacts it exerted the micro-level contextual factors and teacher factors. 
 
In  addition  to  the  how  of  teaching  which  was  affected  by  the  third  grade  nationwide
test, the content or  what of teaching of the participating teachers was also highly affected, as
the  majority  of  the  teachers  (75  %)  chose  their  teaching  practice  based  on  the  test  content.
Conversely,  Wall  (2000)  maintained  that  one  of  the  negative  washback  effects  of  the  tests
happened  when  the  teacher  prepared  the  test takers  for  the  most  likely  test  items  that  might
appear  in  the  examination.  This  negative  washback  effect  would  persuade  them  to
concentrate  on  the  ‘purely  examinable  side’  of  their  work  and  the  other  areas  to  be
overlooked. The lack of attention to the other pedagogical aspects excludes the possibility of
measurement  driven  instruction  (Cheng  &  Watanabi,  2004)  which  favors  a  match  between
the  content  and  format  of  the  test  and  the  format  and  the  content  of  the  instruction.  In
measurement  driven  instruction  the  regular  course  of  instruction  is  to  be  reflected  in  the
format  and  content  of  the  test  while  teaching  to  the  test  entails  a  ‘role  reversal’  (Davies,
1990)  in that  it  is the content and  format of the test that  controls the process and content of
the preceding  instruction. In other words, teaching  is at the service of testing (Cheng, 1997;
Davies, 1990) while it is believed to be the other way round. The reported negative washback
effect  of  the  high  stake  tests on  the  content or  what  of  teaching  confirms  the  earlier  studies
results (e.g., Alderson & Hamp-Lyons,1996; Hamp-Lyons ,1997; Ghorbani, 2008; Nikoopour
& AminiFarsani , 2012;  and Cheng, Sun & Ma, 2015).
Furthermore,  the  results  of  the  present  study  confirm  an  even  stronger  significant
negative effect for the high stake test on the EFL teachers’ testing and assessment procedures
since  an  absolute  majority  of  the  respondents  (92%)  verified  that  they  consider  both  the
format and the content of the high stake tests in their own testing and evaluation practices. It
is concluded that both what of testing and  how of testing are affected by the  high stake test.
The effect on the teachers’ assessment and testing procedures seems to be even stronger than
the  effect  on  the  teachers’  methodology.  This  point  further  supports  the  finding  that  the
teachers  do  whatever  that  familiarizes  the  learners  with  the  content  and  format  of  the  high
stake  test  and  prepares  them  for  it  while  they  might  not  embark  on  the  same  course  of
teaching, testing and other pedagogical activities if it was not for the sake of the test or if the
test  did  not  exist  (Alderson  &  Wall,  1993).  In  other  words,  not  only  teaching  to  the  test  is
practiced  but  also  ‘testing  to the  test’  is  quite  evident.  Other  already  referred  to  researchers
like  Hamp-Lyons  (1997),  Ghorbani  (2008)  and  Nikoopour  and  AminiFarsani  (2012)  have
also  reported  the  negative  impact  of  high  stake  tests  on  the  testing,  assessment,  and
evaluation procedures of the teachers. 
 

Finally, the last finding of this study  verifies that, unlike teachers’ teaching and testing
methodology, their attitude towards different aspects of general education including teaching
and  learning  processes  are  as  strongly  affected  by  the  test  as  the  other  two  factors.  This,  in
turn, indicates that the teachers were applying quite strategic pedagogical practices to achieve
the most practical and institutionally valued objective that is to enable their learners pass the
test,  while  their  attitude towards  the  desirable  educational  processes  are  not  deeply  affected
and  altered.  A  probable  explanation  for  this  effect  might  be  the  fact  that  the  tests’
differentiating  rituals  (Wall,  2000)  sometimes  are  so  effective  in  the  testees’  future  life  that
the teachers ask the testees to do any possible activities to only pass the tests and quite clearly
they change their own pedagogical practices to serve this purpose. This last finding seems to
be  in complete accordance with the previous  findings as the existence of  negative  washback
effect projects the lack of positive attitude of the stakeholders towards the test. Alderson and
Wall  (1993)  believed  that  a  positive  washback  would  function  when  there  is  a  positive
attitude toward the test and there is a cooperative working to fulfill its assigned purposes. The
negative  washback  effect  of  the  third  grade  nationwide  English  exam  on  the  teaching  and
assessment  procedures  of  the  teachers  is  indicative  of  the  lack  of  a  positive  attitude  of  the
teachers towards the test. 
 
Conclusion
The  results  of  this  specific  study  provide  evidence  to  the  fact  that  content  and  format  of
teaching  are  to  a  great  extant  geared  towards  and  adapted  to  high  take  tests  content  and
format.  Both  what  of  teaching  and  how  of  teaching  of  the  EFL  teachers  were  negatively
affected by the content and format of the specific studied high stake test. It verifies the results
of  previous  studies  on  the  washback  effects  of  high  stakes  tests  on  teachers  attitude  and
methodology  in  EFL  classes.  However,  it  is  emphasized  that  EFL  teachers  spend  most  of
their  class  time  to  practice  the  material  which  are  likely  to  be  included  in  third  grade  final
exam  and the communicative  skills of the  language which were  not likely to be  included  in
the  studied  exam  were  all  neglected.  Evidently,  this  procedure  has  negative  and  detrimental
effects  on  the  overall  foreign  language  communicative  competence  development  of  Iranian
high  school  students  as  the  EFL  teachers  did  not  prioritize  this  main  aspect  of  foreign
language  learning over the  language related components which were deemed to be  included
in  the  third  grade  final  exam.  In  addition,  EFL  teachers’  classroom  assessment  procedures
and evaluation format were so designed that maximum similarity with the content and format
 
 
of  the  high  stake  tests  was  achieved,  maximally  preparing  them  for  the  test  in  advance.  All
such  pseudo-pedagogical  activities  were  carried  out  as  the  teachers  were  committed  to  do
everything to help their  learners pass the test  successfully, even though the true learning did
not  take  place  and  the  academic  and  educational  goals  were  not  achieved.  In  other  words,
teaching served testing and the teachers did “teach to the test” despite the fact that they were
aware  of  the  harmful  effects  of  this  behavior.  To  counteract  these  potential  negative
washback  effects,  as  Shohamy  (1993,  p.187)  argues,  a  continuous  and  cooperative  loop
between  external  test  developers  and  people  working  in  the  schools  seems  to  be  vital.  The
study results  necessitate a number of changes  in the program. The changes  include  not only
changes  in  the  test  content  and  format  and  the  testing  procedure  but  also,  as  Lam  (1994)
rightly attests to, changes in the teaching culture.