Formulating Research Questions
-
Upload
ji-ying-leong -
Category
Documents
-
view
32 -
download
1
description
Transcript of Formulating Research Questions
-
5/19/2018 Formulating Research Questions
1/23
Formulating research questions in experimental doctoral
dissertations on Applied Linguistics
Jason Miin-Hwa Lim
Centre for the Promotion of Knowledge and Language Learning, Universiti Malaysia Sabah (Malaysian University of Sabah), Locked Bag 2073,
88400 Kota Kinabalu, Sabah, Malaysia
a r t i c l e i n f o
Article history:
Available online 17 April 2014
JEL classification:
2700 Communication Systems
2720 Linguistics & Language
3550 Academic Learning
Keywords:
Communicative functions
Genre analysis
Rhetorical shifts
Linguistic mechanisms
Dissertation writing
Teaching materials
a b s t r a c t
Research questions have often been regarded as an indispensable part of experimental
research dissertations, yet the ways in which the language varies in the formulation of
these questions have thus far remained an unexplored domain. This genre-based investiga-
tion analysed the language used for formulating research questions in 32 doctoral disser-
tations submitted to universities in the United States between 2001 and 2009. It examines
how candidates in experimental research actually use various communicative resources to
formulate research questions in the introductory chapters that determine the directions in
which their dissertations on Applied Linguistics will be developed. The aspects covered
include (i) the frequency and positioning of the questions, and (ii) categories of these
questions and their linguistic choices. The views of experienced supervisors in Applied
Linguistics were elicited to provide supportive explanations concerning the context in
which the research questions were formulated. Recommendations are given on how
teaching materials can be prepared to demonstrate the ways in which research questions
can be formulated using pertinent and authentic examples actually employed by doctoral
dissertation writers.
2014 Elsevier Ltd. All rights reserved.
1. Introduction
The writing of research introductions has attracted the interest of numerous scholars in the field of genre studies (e.g.,
Anthony, 1999; Feak & Swales, 2011; Samraj, 2005, 2008; Shehzad, 2008; Swales, 1981, 1990, 2004; Swales & Najjar,
1987) in recent decades. Scholars increasing fascination with the analyses of research introductions appears to be related
to both the important theoretical implications and multifarious practical applications of the findings obtained from such
genre-based investigations. Some of these studies have focused on the overall generic structures of research introductionswithin a single language and discipline (e.g.,Ahmad, 1997; Fakhri, 2004; Jogthong, 2001; Najjar, 1990; Ozturk, 2007), across
multiple disciplines (e.g.,Crookes, 1986; Samraj, 2005, 2008; Swales, 1990, 2004; Swales & Najjar, 1987), and across different
languages (e.g., Hirano, 2009; Loi & Evans, 2010; Sheldon, 2011; Soler-Monreal, Carbonell-Olivares, & Gil-Salom, 2011;
Taylor & Chen, 1991). Other studies (e.g., Lim, 2012; Shehzad, 2008, 2010, 2011), however, have opted to focus on only
selected communicative move(s) or step(s).
The aforementioned studies appear to have been largely grounded upon Swales (1990, 2004) seminal genre-based
analysis framework in which an introduction is considered as comprising three communicative moves. It is within this
http://dx.doi.org/10.1016/j.esp.2014.02.003
0889-4906/2014 Elsevier Ltd. All rights reserved.
Tel.: +60 88 320000x5026 (O), mobile: +60 16 8298305; fax: +60 88 435708.
E-mail addresses: [email protected],[email protected]
English for Specific Purposes 35 (2014) 6688
Contents lists available at ScienceDirect
English for Specific Purposes
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c at e / e s p
http://dx.doi.org/10.1016/j.esp.2014.02.003mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.esp.2014.02.003http://www.sciencedirect.com/science/journal/08894906http://www.elsevier.com/locate/esphttp://www.elsevier.com/locate/esphttp://www.sciencedirect.com/science/journal/08894906http://dx.doi.org/10.1016/j.esp.2014.02.003mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.esp.2014.02.003http://-/?-http://-/?-http://-/?-http://crossmark.crossref.org/dialog/?doi=10.1016/j.esp.2014.02.003&domain=pdf -
5/19/2018 Formulating Research Questions
2/23
framework that theoretical statements have been made to predict how writers schematically organise their research intro-
ductions. Swales (2004, p. 230)proposed that in research introductions, writers generally use Move 1 (i.e., establishing a
territory) to provide important background information about a topic by citing previous studies in an order of increasing
specificity (i.e., providing and/or citing general information before proceeding to more specific information). Having fur-
nished essential background information in the first move, writers may proceed to Move 2 (i.e., establishing a niche) by
(i) indicating a gap in past research; (ii) highlighting a need to extend the present knowledge (following a tradition or
research trend); or (iii) presenting positive justifications that foreground a need to solve an existing real-life problem via
research. Based on the niche established, writers rhetorically shift to Move 3 (i.e., presenting the present work) in whichthey announce their research purpose, present research questions or hypotheses, provide definitional clarifications, or
(briefly) summarise research methods.1 Some of these rhetorical steps, namely research purpose (or objectives), research
questions and research hypotheses, have been subsequently regarded as directional determinants ( Feak & Swales, 2011,
p. 112) that have a bearing on the way in which a research report or dissertation will proceed and develop. In particular, research
questions refer to interrogative sentences (and other closely related sentences) used by researchers to seek information about a
specific topic area which (i) has not, in fact, been addressed in past studies, and/or (ii) is worthy of investigation in the
discipline concerned (Sunderland, 2010, p. 11). According to Andrews (2003, p. 17), a dissertation needs to be driven by
research questions which perform the function of tying existing research literature with the rest of the dissertation. Syntactically,
these questions may appear in the form ofwh-questions, which begin with a wh-word (e.g., What, Why, Where, Who, To
what extent, etc.), as in What pedagogical models do teachers use in the multimedia classroom? (Harbon & Shen, 2010, p.
281). Alternatively, these questions may occur in the form of polar (yes/no) questions, which start with a primary or modal
auxiliary verb (e.g., is, does, will, etc.), as in Does focused written corrective feedback have an effect on intermediate ESL
learners acquisition of English articles? (Sheen, 2007, p. 260).Compared to hypotheses, research questions are at times considered to have more utility in cases where little is known
about a phenomenon. For instance, some phenomena in real life (such as those related to the use of technology) often enter
the mainstream research community faster, and as a result, writers have too few studies on which to base their research
hypotheses (Campbell, 2008). Under such circumstances, where there were fewer past studies upon which they can develop
hypotheses, writers are more likely to present research questions (Keyton, 2011).
It is interesting that research questions (RQs) have now been specified as one of the rhetorical steps in Move 3 (i.e., pre-
senting the present work) inSwales (2004)new theoretical model for predicting the organisation of a research introduction
(even though they were not previously viewed as part of Move 3 in his 1990 model). The latest inclusion of research ques-
tions in Move 3 has some noteworthy implications. Although research questions are now perceived by genre analysts as a
major rhetorical step in the introductory section that guides the development of a research report (Feak & Swales, 2011;
Swales, 2004), no previous studies, to my knowledge, have focused exclusively on how research questions are framed in rela-
tion to other rhetorical moves and how they are realised linguistically in doctoral dissertations. Several studies (e.g.,Ozturk,
2007; Shehzad, 2011; Sheldon, 2011; Soler-Monreal et al., 2011) have provided only limited clues with respect to researchquestions. In regard to frequencies of research questions, for instance,Sheldons (2011)study of Applied Linguistics research
article introductions (RAIs) showed that 33.3% (6/18) of the English first language (L1) writers incorporated research ques-
tions or hypotheses while only 5.6% (1/18) of Spanish L1 writers included them. Likewise,Soler-Monreal et al.s (2011)study
of computing doctoral thesis introductions revealed that 50% of the English thesis introductions incorporated research ques-
tions or hypotheses (while merely 10% of the Spanish thesis introductions included them). In another recent study, Loi and
Evans (2010, p. 2816)pointed out that writers used research questions to offer the readers a yardstick by which to measure
the success of the studies in educational psychology. They found that research questions appeared in more than a third
(35%) of the English RAs in educational psychology but only in 10% of the Chinese RAs in the same discipline, thus confirming
the relative importance of RQs in English research reports. In contrast, Shehzad (2011)found that although 32.14% of the
Computer Science RAIs contained research questions or hypotheses, merely 7% of the RAIs contained research questions
(not hypotheses). Her study, however, did not provide any figure relating to research questions in quantitative experimental
investigations.
In addition,Ozturks (2007)inquiry into Applied Linguistics RAIs focused on move sequences (e.g., M1M2M3M1M3)and acknowledged that research questions could be used by writers in Move 3, but did not study the linguistic realisations
and rhetorical shifts involving research questions. This means that studies have yet to analyse in greater detail how writers
rhetorically shift from other moves to research questions and how these questions are realised linguistically. In this regard,
our experience in teaching English for Academic Purposes (EAP) and supervising postgraduate students has also revealed
that although supervisors are often very much involved in shaping dissertation writers research questions, an additional
challenge lies in crafting the text that immediately precedes or follows the research questions. This explains the value of
conducting a thorough investigation into the rhetorical and linguistic mechanisms engaged in formulating RQs and their
surrounding context in a specific discipline.
To further understand the significance and nature of research questions, we need to review some related explanations
given by guidebooks on research methods and research writing in different disciplines. In education, for instance, Fraenkel
1 Other short rhetorical steps (related to the principal outcomes, value and structure of the research report) are only considered as probable in some fields(Swales, 2004, p. 232).
J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688 67
http://-/?-http://-/?-http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
3/23
and Wallen (2003)andParahoo (2006)are of the view that research projects are likely to lead to meaningful and insightful
outcomes when feasible and explicit research questions are formulated. In business management, the inclusion of research
questions in the statement of a business problem makes it easier to understand what is perplexing managers and indicates
the issues to be resolved (Zikmund, 2003, p. 98). Likewise, in communication, research questions largely control and influ-
ence subsequent parts of a research report; for instance, the method of a study needs to closely follow what the research
question has indicated (Hocking, Stacks, & McDermott, 2003). Overall, research questions need to be expressed precisely
to help researchers explore a topic in an appropriate manner and context so as to generate useful answers (Creswell, 2008).
Given that the focus of this study is specifically on quantitative experimental research, the following section will first re-view the scientific requirements and expectations of experimental research in general and examine how they are connected
with the formulation of research questions. Scholars (e.g., Fraenkel & Wallen, 2003; Gass, 2011) generally agree that an
experimental study has several major distinguishing features. First, an experimental study is deductive in nature in that
it adopts a theory-then-research approach, which differs from some non-experimental studies that use an inductive
(i.e., research-then-theory) approach (Gass, 2011, p. 8). An experimental study generally involves a randomly selected
group of participants or subjects who are randomly assigned to various treatment conditions and/or control groups, and
as such, the treatment described explicitly (in the subsequent method section/chapter) must also relate directly to the re-
search questions which guide the study (Gass, 2011; Hocking et al., 2003). In the context of deductive business studies,
for instance, researchers identify relevant concepts and existing theories and adjust them to the problem under scrutiny
(Ghauri & Grnhaug, 2002).
Although research questions play an important role in experimental research (Fraenkel & Wallen, 2003; Gass, 2011;
Hocking et al., 2003), we are relatively uncertain (i) how these questions are rhetorically linked with other key communi-
cative moves, and (ii) the ways in which they are realised syntactically to achieve their communicative functions. This meansthat the examples provided by guidebooks on research methods and research writing, though relevant and correct, may be
insufficient, and a comprehensive study needs to be conducted to ascertain how doctoral dissertation writers actually use
rhetorical shifts and language resources in the formulation of research questions. It is important to study such rhetorical
and linguistic mechanisms for two reasons. First, studying rhetorical shifts will reveal the extent to which different segments
are meaningfully linked with this communicative move to enhance the overall coherence of the introductory chapter. Sec-
ond, owing to the importance of framing research questions in the appropriate context of experimental research, we need to
use a broad range of examples to inform novice writers of the common linguistic mechanisms needed in formulating re-
search questions. At this juncture, it should be pointed out that the doctoral dissertations chosen for this analysis actually
passed and doctorates were awarded to the writers. Despite acknowledging that it is difficult to find the perfect text that
illustrates the desired linguistic exemplifications (Swales, 2009, p. 5), this paper argues that using a wide range of actual re-
search questions employed in dissertations would (i) demonstrate the extent to which syntactic structures suggested by
guidebooks (on research methods and research writing) are also commonly employed by dissertation writers in practice,
and (ii) reveal more explicitly the rhetorical strategies recurrently used by dissertation writers in real life. This means thatthe research questions identified in this study, in addition to the limited range exemplified in guidebooks and textbooks (e.g.,
Creswell, 2008; Fraenkel & Wallen, 2003; Gass, 2011), are likely to show us what needs to be emphasised in instructional
sessions aimed at helping novice writers formulate research questions.
It is interesting to note that research methods guidebooks and style manuals in Education and Applied Linguistics gen-
erally underscore the importance of formulating research questions appropriately in accordance with the type of study to be
conducted, and provide some related explanation on language choices. For instance, some recently published textbooks on
research methods (e.g.,Ary, Jacobs, & Razavieh, 2002; Drnyei, 2007; Gay, Mills, & Airasian, 2009; McBurney & White, 2004 )
have incorporated some enlightening discussion on the general differences between quantitative and qualitative research
questions.Gass (2011, p. 9)also provided an example of an acceptable experimental research question (i.e., Does focused
attention on nounadjective agreement in Italian promote learning to a greater extent than focused attention onwhat-move-
ment in Italian for beginning learners of Italian?). In relation to this, she specifically considered certain interesting questions
(e.g., Should language classes be introduced early in a schools curriculum?) as debatable but not researchable given the
vagueness as well as the wordshouldwhich implies some sort of right or wrong and which, as a result, cannot be empiricallyevaluated (Gass, 2011, p. 9). Other guidebooks (e.g.,Creswell, 2008; Fraenkel & Wallen, 2003) have provided suggestions
with respect to how questions can be written, but the examples furnished are limited and unvaried. In the context of quan-
titative educational research, for instance, Creswell (2008, p. 122)specified merely a polar research question (i.e., yes/no
questions) beginning with the operator do (e.g., Do parentteacher Internet communications affect student performance
in the classroom?) as a quantitative research question. More specifically, in a detailed discussion on experimental research,
Fraenkel and Wallen (2003, pp. 28, 291)specified exclusively polar questions beginning with the auxiliary do or copular
verb be (e.g., Does behaviour modification reduce aggression in autistic children?; Is there a difference in the social stud-
ies achievement of fourth-grade students according to the treatments of cooperative learning or traditional instruction and
according to gender across treatment groups?) as typical experimental research questions.
Given that textbooks (e.g.,Ary et al., 2002; Drnyei, 2007; Gay et al., 2009; McBurney & White, 2004) have generally men-
tioned the language used in research questions only in passing, and others (e.g., Creswell, 2008; Fraenkel & Wallen, 2003;
Keyton, 2011) have cited exclusively polar questions and evaluative wh-questions as instances of possible questions for
experimental studies without explicitly linking them to other communicative elements, it would be interesting (i) to deter-mine, in the first place, the extent to which research questions are used specifically in experimental doctoral dissertations on
68 J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688
http://-/?-http://-/?-http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
4/23
Applied Linguistics; (ii) to demonstrate how they are positioned and developed in the presentation of doctoral dissertations;
and (iii) to ascertain the extent to which experimental research questions tend to be presented in particular syntactic forms
in actual postgraduate dissertations. In view of the need to investigate the significance, frequency, positioning, range and
features of research questions in experimental dissertations, two research questions are formulated for this study as follows:
(1) How frequently are research questions incorporated in the introductory chapters of experimental doctoral disserta-
tions on Applied Linguistics and how are they positioned?
(2) What rhetorical shifts and linguistic mechanisms do doctoral candidates use to formulate research questions in thesedissertations?
The first research question seeks mainly quantitative data to ascertain the frequency with which the research questions
appear in the corpus of American doctoral dissertations.2 Analysis also focused on how research questions are positioned in
relation to other key elements to achieve their overall communicative purpose. The second question seeks largely qualitative
data to probe the possible range of rhetorical strategies and salient syntactic choices used by doctoral candidates to formulate
research questions as part of their endeavour to gain acceptance of the academic research community concerned. It also seeks
some quantitative data in regard to the frequencies of different categories of research questions and those of tenses employed
by doctoral candidates in formulating research questions. It should be pointed out here that research questions in experimental
studies were chosen because the English of an experimental research report is highly conventionalised, a fact that represents a
great advantage for non-native speakers as well as for language instructors ( Weissberg & Buker, 1990, p. iv). More importantly,
the conventions of experimental research reports are fairly consistent across a wide variety of scientific disciplines and if one
can master the conventions, one can replicate the genre in an acceptable form ( Weissberg & Buker, 1990, p. iv).
2. Methods
The research procedure consisted of three phases involving (i) selection of doctoral dissertations on experimental re-
search in Applied Linguistics; (ii) a thorough textual analysis of the doctoral dissertations; and (iii) face-to-face qualitative
interviews with American advisors supervising the writing of postgraduate dissertations. A total of 32 American doctoral dis-
sertations on experimental research were selected from a pool of 270 dissertations associated with the keywords language
experimental research using the ProQuest search engine. The sampling procedure was purposive in that it was based on two
criteria to ensure that (i) all the dissertations were written by doctoral candidates, and (ii) each of the dissertations selected
was genuinely based on quantitative experimental research. To obtain this sample, the abstracts and chapters characterising
the type of study reported in each dissertation were studied before each dissertation was categorised as one that was based
on experimental research. In selecting the dissertations, the criteria used in identifying quantitative experimental research
were grounded on the characteristics expounded byGay et al. (2009)andCreswell (2008). As explained above, all the studiesreported in the dissertations adopted a deductive approach, focused on objective reality to be discovered, and established
causeeffect relationships between a dependent variable and at least one independent variable. All the doctoral dissertations
were completed in partial fulfilment of the requirements for the degree of Doctor of Philosophy (Ph.D.) or Doctor of Educa-
tion (Ed.D.) and submitted to American universities during the period 20012009. The researcher selected only one disser-
tation from each of the 32 American universities that had produced doctoral dissertations on experimental research. The
dissertations were selected from universities in 21 states in the United States, and exactly half of the institutions were listed
as research 1 universities in the country. In this study, research 1 universities are those state institutions that offer a full
range of baccalaureate programs, are committed to graduate education through the doctorate, give high priority to research,
award 50 or more doctoral degrees each year, and receive annually $40 million or more in federal support (Weerts, 2002, p.
26). An asterisk () is used in the Appendix of this paper to indicate that a dissertation was submitted to a research 1 uni-
versity. The selection was done to ensure that the corpus did not exhibit merely the influence of individual supervisors or
the preferred requirements of particular universities. Such an evenly distributed purposive sample effectively precluded any
overreliance on the requirements and expectations of a small group of American universities as this large group constitutedmembers of the academic discourse community concerned.
The researcher included dissertations submitted by 32 writers in different U.S. universities for several reasons. First, a
broad range of research questions needed to be incorporated to form a larger and more representative sample that could
realistically reflect the general circumstances under which research questions were presented in experimental dissertations.
More specifically, rhetorical shifts used by different writers had to be considered to demonstrate how doctoral candidates
generally arranged information elements to meet different needs in introducing experimental studies on Applied Linguistics.
Second, in terms of linguistic mechanisms, more writers were involved in demonstrating the dissertation writers actual
usage of the language in fulfilling the pragmatic functions in presenting the questions. This was based on the notion that
writers in the same or related disciplines would share numerous norms and expectations concerning the use of linguistic
forms irrespective of their nationalities and first languages (Okamura, 2003; Shehzad, 2011). Nevertheless, it should be
2 In this study, the phrase American doctoral dissertations is defined as doctoral (Ph.D. or Ed.D.) dissertations prepared for and submitted to universities inthe United States of America regardless of the candidates countries of origin.
J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688 69
http://-/?-http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
5/23
pointed out here that so far as pedagogical implications are concerned, linguistic choices are recommended for the formu-
lation of research questions only if they are correctly, appropriately and recurrently used by different writers from several
institutions.
Swales (1990, 2004)movestep analytical framework was used to examine the texts with reference to communicative
purposes as reflected in the rhetorical segments. In this study, each introductory chapter was first analysed in terms of dis-
tinct units in a hierarchically organised framework whereby the entire chapter was divided into rhetorical moves which
were then subdivided into constituent steps. Attention was then devoted to all text segments directly connected with re-
search questions in the introductory chapters because a preliminary study had shown that research questions were generallypresented for the first time in the introductory chapters (although some doctoral candidates first presented their research
questions only after the end of their literature reviews in Chapter 2).
As the formulation of research questions constitutes part of Move 3 (i.e., presenting the present work), these segments
associated with research question presentations (RQPs) should fulfil two conditions. First, the text segments should be asso-
ciated with the research questions presented for the study being reported (i.e., presenting the present work in Swales Move
3) and not the research questions reported in past studies. Second, although the text segments might be written using any
language mechanisms, they should focus on (i) the postulation of questions that determine the direction in which the dis-
sertation would proceed; (ii) the focus, purpose or meaning of the research questions; and/or (iii) elaboration or additional
explanation that illustrates how and why the questions were formulated. The occurrence of each RQP was then marked in
every text so that its frequency could be identified. The frequency of RQPs refers to the number of rhetorical steps contain-
ing research questions (not the number of individual research questions or sentences). Every rhetorical step (i.e., each RQP)
may consist of (i) one or multiple research questions, and/or (ii) some preceding/ensuing explanations of the research ques-
tions. In determining the extent to which research questions are used in the doctoral dissertations, the words obligatory,quasi-obligatory and optional are defined here. In accordance with the common usage of the word obligatory in some
recent past studies (Lim, 2010,p. 231;Soler-Monreal et al., 2011, p. 8; Yang & Allison, 2003 , pp. 372374), a rhetorical move
is considered obligatory if it occurs in all (100%) of the texts, and quasi-obligatory or largely stable if it appears in 5199%
of the texts. However, a move is considered only optional if it occurs in only half or less than 50% of the texts.
The research questions presented in direct interrogatives were then classified usingHogues (2003)typology, in which a
sentence may be (i) simple (comprising merely one independent clause); (ii) compound (incorporating two or more indepen-
dent clauses); (iii) complex (comprising one independent clause and at least one dependent clause); and (iv) compound-
complex (including at least two independent clauses and at least one dependent clause). Several consecutive research
questions occurring in a row were considered as constituting one occurrence of RQP insofar as their occurrence was not
interrupted by any other rhetorical step. The frequencies of RQPs and the numbers of individual research questions in simple
and complex structures were also counted with reference to the numbers of times they appeared in the corpus. In the
subsequent portion of the analysis, prominent shifts from one segment (i.e., rhetorical move or constituent step) to an
RQP were analysed so far as they distinctly demonstrated linkages with other rhetorical segments.Lexico-grammatical choices employed to perform the functions were then studied with reference to sentence structures,
clause elements, categories of phrases, and parts of speech (if they appeared as prominent features of the RQPs). Instances of
the constituent steps related to research questions were recorded and analysed as authentic examples from the
dissertations. The analysis of prominent linguistic choices was conducted on the basis of (i) general linguistic descriptions
provided by Quirk, Greenbaum, Leech, and Svartvik (1985), Greenbaum and Quirk (1992), and Downing and Locke
(2006); (ii) Yadugiris (1986) descriptions of yes/no (polar) questions; and (iii) lexico-grammatical descriptions for
research-related genres as illustrated byThomas and Hawes (1994)andLim (2006, 2011a, 2011b). In regard to the tenses
used by the doctoral candidates in formulating research questions, all finite verbs (i.e., verbs showing tense distinction) used
by the writers (for presenting these questions) were highlighted before the frequencies of the verbs under each of the major
tenses were counted and compared.
In the third phase, four specialist informants were engaged to provide supportive qualitative data related to the formu-
lation of research questions. The informants were selected based on Bhatias (1993, p. 34) criteria for the selection of spe-
cialist informants in that every specialist informant included in this study should be (i) a practising member of thedisciplinary culture in which the genre is routinely used, and (ii) an experienced individual who was able to confirm
the analysts findings, bring validity to his insights, and add psychological reality to his analysis. In this context, all
the informants had successfully supervised experimental research in Applied Linguistics at doctoral degree level in the
United States and were familiar with the requirements involved in writing doctoral dissertations in the discipline. Initially,
detailed information was elicited from the first two specialist informants via semi-structured interviews conducted to en-
sure that problems arising from the two earlier phases could be posed to the informants if necessary. The semi-structured
interviews employed open-ended questions (Berg, 2004; Lynch, 1996) aimed at eliciting views and responses from the
supervisors in regard to (i) the significance of research questions in doctoral dissertations on experimental studies, and
(ii) the basic requirements and expectations of the research community with respect to the formulation of research ques-
tions. Verbatim transcriptions of the interviews were prepared immediately after each interview so that the spoken data
could be used to facilitate further analysis (although the information elicited from informants only played a supportive role
in this study that is largely based on the rhetorical analysis). Subsequently, another two specialist informants who had
successfully supervised the doctoral dissertations included in this corpus were invited via e-mails to provide responses tounresolved questions concerning the absence of research questions in some dissertations.
70 J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
6/23
Steps were taken to ensure validity and reliability of this study, although they were comparatively more relevant in the
quantitative component (rather than the qualitative component) of this study. Validity in this study refers to the defensi-
bility of the inferences researchers make from the data collected (Fraenkel & Wallen, 2003, p. 119). Three aspects relating to
content validity were in focus, namely the (i) significance of the research questions in doctoral dissertations; (ii) tense usage
in formulating RQs; and (iii) identification of rhetorical shifts and language mechanisms. First, inferences made in regard to
the significance of RQs and their related tense usage were ensured by using multiple specialist informants (rather than a sin-
gle specialist informant), and their views were compared point by point to ensure that the related conclusive statements
could be defended. Second, in cases where the informants views differed, their spoken data were compared with those infer-ences based on the textual analysis to minimise discrepancy in the inferences made. Third, inferences made regarding rhe-
torical shifts and language mechanisms were also verified by considering whether salient features appear in multiple
doctoral dissertations submitted by writers in different universities. With respect to reliability, defined as the degree to
which a method consistently measures whatever it is measuring (Gay et al., 2009, p. 158), consistency was ensured in
the categorisation (i.e., coding) of text segments as being RQ-related. Given that such categorisation could affect the calcu-
lation of the frequencies of RQPs in the corpus, intra-rater reliability was ensured by using Petrics (2007, p. 242)technique
for repeated categorising and re-categorising of the text segments in the experimental introductory chapters at different
points in time, whereby each categorisation was conducted six months after the previous one. In this study, four rounds of
coding were conducted. The intra-rater agreement attained after the second round of coding was 94.8%, and subsequently
the agreement reached after the third round of coding (another six months later) was 98.7%. To further minimise errors
in coding, the coded categories indicated in the fourth round were compared with those assigned in the third round. As
the categories indicated in the fourth round matched completely with those assigned in the third round, an intra-rater agree-
ment of 100% was finally attained. To deal with the discrepancy emerging in the re-categorisation of the segments, the basiccriteria used were that (i) the segments categorised as RQPs should raise questions regarding the possibility of the expected
research outcomes, and (ii) the segments had to be directly associated with the main focus of the study.
3. Results and discussion
In this section, some qualitative findings based on an analysis of the specialist informants statements are first reported to
explain the overall significance of research questions in experimental settings and to illustrate the context in which research
questions are generally formulated. Subsequently, results on the frequency and positioning of research questions are
reported. Finally, results on the analysis of the language used by the doctoral candidates are presented with reference to
categories of research questions and their salient linguistic choices.
3.1. Significance of RQPs based on specialist informants spoken data
The specialist informants views were largely consistent with respect to the role of research questions in directing the
entire study. Specialist Informant A (SIA) pointed out that doctoral candidates often struggle with figuring out what the
appropriate literature is and they can spend a lot of time on literature which is pretty peripheral to their experiments.
SIA considered the trickiest part in the introductory chapters to be what she called the right breadth that demonstrated
the scope of appropriate past research papers which had been reviewed. She was of the view that research questions served
as a guide that could help candidates make the right decisions concerning what to focus on using past researchers findings
and experience in the early chapters of a dissertation.
According to Specialist Informant B (SIB), however, the full form of an introductory chapter is generally not written until
the findings have been obtained even though students often want to write that thing up completely before they get to writ-
ing up the research itself. The introductory chapter is difficult for doctoral candidates to write because on the one hand
they (doctoral candidates) feel they want to discuss every single issue that can possibly impinge on the research, which
means that many of them want to write a hundred pages of literature review. As such, to avoid getting sidetracked and
deleting numerous unrelated portions at a later stage, supervisors expect doctoral candidates to formulate research ques-tions in order to advance issues that guide the choice of the experiment and write a little about each of these and be
prepared for the possibility of flushing it out substantially only when they subsequently see how the data come out.
SIB also viewed research questions as essential not solely in the writing process, but also in the reading process given that
the questions would (i) remind the student researcher himself/herself of what he or she is looking for in the doctoral study,
and (ii) inform the readers about what to be looking for (i.e., the results that readers are expected to find subsequently) in
the dissertation.
3.2. Frequency and positioning of research questions
Apart from the overall significance of research questions, another aspect that deserves some attention is the frequency of
research questions. A small portion (i.e., 18.8% or 6/32) of the introductory chapters do not contain research questions. In the
six doctoral dissertations that include no research questions in their introductory chapters, four dissertations (i.e., 12.5%)
have research questions presented for the first time after the end of their literature reviews in Chapter 2. One dissertation
J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688 71
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
7/23
writer presents his research question for the first time at the beginning of his Chapter 3 (on research methods) after his lit-
erature review, and another writer only states her research purpose without presenting any research question at all in her
doctoral dissertation. In regard to these rare cases, Specialist Informant C, who had successfully supervised the writing of a
doctoral dissertation included in this corpus, was of the view that she was not particularly concerned about the means by
which the purpose of the study is communicated a question, a hypothesis, a prediction or whether it is implicit or explicit
so long as its purpose is logical or motivated by theory and previous research. Specialist Informant D, however, acknowl-
edged that he would prefer the phrases we believe or we expect that indirectly indicates a hypothesised phenomenon
even though researchers occasionally are guided by a general question or purpose. Although these specialist informantsresponses differ to a certain extent, at this juncture it appears that research questions are prevalent, and if RQs are used, they
are generally linked with purpose statements. In some rare cases where RQs are not used, purpose statements (considered
absolutely essential in every dissertation) are employed to perform the function of determining the direction in which the
doctoral dissertation should proceed. Such a general view points to the need to consider some quantitative data about RQPs
based on a textual analysis. In this regard,Table 1indicates that in terms of frequency, RQPs appear in the vast majority (i.e.,
81.3%) of the dissertations with a mean frequency of 1.19, thus showing that presenting research questions constitutes a
quasi-obligatory (largely stable) rhetorical step in the experimental dissertations.
In regard to word count, however, the number of words needed to form a research question ranges from 13 to 50 while
the average number of words used for a research question is 24.2 (i.e., 2880 words in 119 RQs; see Table 1). In the corpus,
more than half (i.e., 61/119 or 51.3%) of the research questions (in direct interrogatives) are presented in complex sentences.
Even the shorter simple sentences in the corpus also exhibit a considerable degree of complexity that engages the use of
postmodifying structures (seeTable 2and the explanations in Section3). Interestingly, none of the research questions (in
direct interrogatives) are formulated in compound and compoundcomplex structures (although some of the ensuing
Table 1
Frequencies of RQs and numbers of words used in RQs in the experimental introductory chapters (EICs).
Dissertation
introduction
Freq. of
RQPs (steps
with RQs)
Freq. of RQs
in simple
sentences
Freq. of RQs in
complex
sentences
No. of words
used in RQPs
No. of RQs
(in direct
interrogatives)
No. of words used
in RQs (in direct
interrogatives)
EIC1 1 3 0 75 3 53
EIC2 3 12 0 248 12 202
EIC3 1 0 2 69 2 63
EIC4 1 1 0 31 1 23
EIC5 2 2 3 200 5 108
EIC6 1 0 5 170 5 137
EIC7 4 6 4 261 10 229EIC8 1 3 1 64 4 52
EIC9 1 1 2 229 3 53
EIC10 1 4 0 73 4 67
EIC11 0 0 0 0 0 0
EIC12 1 0 4 217 4 200
EIC13 2 3 0 84 3 84
EIC14 1 4 3 202 7 193
EIC15 1 0 2 86 2 73
EIC16 1 0 3 129 3 87
EIC17 2 2 8 322 10 263
EIC18 1 0 0 17 0 0
EIC19 1 0 10 328 10 264
EIC20 0 0 0 0 0 0
EIC21 1 0 3 104 3 89
EIC22 0 0 0 0 0 0
EIC23 1 0 2 100 2 85EIC24 0 0 0 0 0 0
EIC25 2 4 1 81 5 112
EIC26 2 0 2 77 2 75
EIC27 0 0 0 0 0 0
EIC28 1 10 0 210 10 135
EIC29 1 1 2 95 3 80
EIC30 3 2 1 97 3 84
EIC31 0 0 0 0 0 0
EIC32 1 0 3 74 3 69
Total 38 58 61 3643 119 2880
Mean 1.19 1.81 1.91 113.80 3.72 90.00
No. of EICs (n) 26 15 19 26 25 25
Percentage of EICs (n/26 100%) 100.0 57.7 73.1 100.0 96.2 96.2
Note:Frequencies of RQs in compound and compound-complex sentences are not indicated here because they are completely absent in the corpus.
72 J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688
http://-/?- -
5/19/2018 Formulating Research Questions
8/23
sentences used to explain research questions may be presented in compound and compoundcomplex structures). To
further illustrate the complexity of the RQs, more instances are discussed in Section 3.3.
While the aforementioned quantitative results have given us an important overview of the frequency of RQPs, the ways in
which they are realised rhetorically in doctoral dissertations on experimental research will be more comprehensible if we
analyse (i) some related inter-move, inter-step, and/or intra-step shifts, and (ii) their salient linguistic features. In the ensu-
ing analysis, selected expressions in each instance are underlined if they constitute salient linguistic choices characterising
Table 2
Salient syntactic choices used in formulating three categories of polar research questions.
Syntactic choice Freq.of
RQs
No. of
EICs (x)
% of EICs (x/
26 100%)
Instance of research questions
Primary auxiliary verb do in the simple present
used as an operator in an interrogative VaSVmstructure to ask about a regular event
25 10 38.5 Do learner factors, such as gender, time spent abroad,
and proficiency in other languages have an effect on
learner understanding of T/V? (EIC5: 9)
The research questions that guided the study were: (a)Does systematic explicit phonics instruction using the
Get Reading Website (Get Reading, 2006) contribute to
an increase in phonological awareness among adult
female Arabic speaking Emirati nationals aged 18-25 in
a government tertiary foundation level ESL program?
(EIC15: 15)
The research questions that will guide this study are as
follows:
1. Do basic editing and proofreading skills improve
among Business Information students enrolled in
BI118, a business English course, as compared to first-
semester Law Enforcement and Paralegal Studies
students who did not enrol in the business English
course?... (EIC23: 7)
Primary auxiliary verb will in the simple future
(used as an operator) in an interrogative V aSVmstructure to ask about a possible future action
9 2 7.7 Will students administered learning-style assessments
and strategies on the basis of Building Excellence (BE)
(Rundle & Dunn, 2000) results score significantly higher
on the CUNY?ACT Compass Reading Test (CUNY Office
of Assessment, 2006), compared with students taught
with traditional study-skills materials and methods?
(EIC14: 6)
Will BE elements predict the performance of the
experimental group on the CUNY/ACT Compass Reading
Test? (EIC14: 7)
Will students administered learning-style assessments
and strategies on the basis of BE results have
significantly lower anxiety levels, as measured by the
State-Trait Anxiety Inventory for Adults (STAI)
(Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, (1983),compared with students taught with traditional study-
skills materials and methods? (EIC14: 7)
Primary copular verb be in the simple present used
as a main verb to ask about the existence of a
relationship or difference
31 10 38.5 In summary, this study will attempt to address these
issues with the following general research questions:
1. Is there a relationship between writing self-efficacy
and written language performance in (of) middle school
students?... EIC1: 7)
Are adults with aphasia as accurate and fast as healthy
controls when switching between rules?... (EIC9: 22)
This study attempts to answer some specific questions
about the reading comprehension in ESL classes in
secondary stage in Saudi Arabia and examines the
students attitudes toward cooperative learning and the
level of students motivation toward reading:
1. Is there a significant difference in ESL reading
comprehension (RC) performance of secondary stage in
Saudi public girls schools when comparing students
taught using cooperative learning and those taught
using traditional methods?. . .(EIC16: 6)
J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688 73
http://-/?- -
5/19/2018 Formulating Research Questions
9/23
the rhetorical move/step concerned. In addition, rhetorical shifts are highlighted in figures to demonstrate how writers stra-
tegically direct their studies in dissertation introductions by placing their research questions in proximity to selected text
segments. This means that the qualitative analysis of the rhetorical positioning of research questions is considered with ref-
erence to the connections between the questions and their preceding and/or ensuing rhetorical category (i.e., move or step).
3.2.1. Inter-move shifts from territorial establishments, via niche establishments, to research questions and other related steps
Writers often employ inter-move shifts using research questions in a way that engages territorial establishment and gap
indications, both of which involve reviews of past research (seeFigure 1). According to SIB, it is crucial that the questions beconnected to gaps in the literature as there is a tendency for doctoral candidates to summarise past literature rather than
to identify the problems with past literature, and as a result, that inert lump of talk about past research may not highlight
any signpost or important issue that merits investigation. To SIB, the review of literature must identify what has not
been understood, and the questions must pick up on them because by identifying gaps in the literature, you motivate
the research questions in such a way that readers should have no surprises when the research questions are presented.
Rhetorical connections between real-world problems, research gaps and research questions have been identified in this
corpus.Figure 1illustrates that writers accomplish a complete 1-2-3 cycle that ends with the presentation of a research
question. In establishing a territory, an emphasis on the importance of researching a topic is combined with a description
of a notable real-world problem that has been discovered and observed by the writer (e.g., serious sociolinguistic conse-
quences caused by non-native speakers who have made a wrong pragmatic choice as a result of their lack of appropriate
understanding of the second language). Alternatively, a language-learning problem reported in past research (e.g., auditory
training which did not transfer to reading and spelling scores despite an improved transfer of temporal processing at the
sound and phoneme level) may form the theoretical basis for a writers argument that an issue is central and noteworthy. Onthe basis of such theoretical arguments (i.e., statements used to predict, assume or suggest that a possible phenomenon ex-
ists in the process of language learning and teaching), writers proceed to foreground previous research gaps using noun
phrases (NPs) indicating gaps in past investigations (e.g., few studies, limited research, etc.) or NPs acknowledging meth-
odological shortcomings in previous studies (e.g., research limitations, flawed design and methodology, small sample size,
imprecise and inconsistent inclusion criteria, etc.). These gap indications are deployed to (i) explain why past researchers
failed to arrive at a reliable conclusion, and (ii) justify the need to formulate new research questions which precisely address
the specific previous research gaps (e.g., questions that ask (i) whether foreign language learners of Russian are able to accu-
rately judge the pragmatic appropriateness of the use of Russian address term in different situations, or (ii) about the effect
of temporal auditory training on reading and on the expressive and receptive language skills of children with diagnosed
temporal auditory processing deficits) as exemplified inFigure 1.
An alternative inter-move shift that merits attention in the qualitative analysis has to do with how writers use not merely
descriptions of real-world problems and past research gaps but also alternative steps in Move 3 as lead-ins to lay the ground
Fig. 1. Inter-move shifts from a territorial establishment to a niche establishment and a research question presentation.
74 J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688
http://-/?-http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
10/23
for a full presentation of their new study.Figure 2shows that writers first establish a research territory in Move 1 by (i) using
past researcher(s) theoretical statements regarding a possible existing phenomenon in language learning (e.g., According to
Bromley (2000) when children begin formal writing, their oral-language is the basis for much of what they write . . ., For
VanPatten, moreover, explicit focus on form instruction is required to reachdistinguishedgrammatical accuracy. . ., Ishihara
and Chi (2008), proposed that the use of entire video films is better than the short video clips. . ., etc.), (ii) underscoring topic
centrality (through stressing the potential positive impact of an intervention on learning performance), or (iii) foregrounding
Fig. 2. Shifts from a territorial establishment, through a niche establishment, to a research question presentation and alternative probable steps in Move 3.
J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688 75
-
5/19/2018 Formulating Research Questions
11/23
real-world problem/s associated with language teaching and learning. Having demonstrated the prominence of a research
domain or noteworthiness of a real-world problem, writers continue to direct readers attention to an existing research
dearth using quasi-negative expressions highlighting the scarcity of past studies (e.g., seldom investigated, do not receive
as much attention, minimal number of studies, etc.) in Move 2. Based on such research lacunas functioning as a preliminary
lead-in, writers proceed to present selected steps in Move 3 (i.e., presenting the present work), which may not always begin
with a research question presentation (as exemplified inFigure 1). It is at this juncture that writers generally deploy descrip-
tions or statements of the research purpose, methods and/or value of the study in Move 3, which involves the use of deictic
expressions signalling their new studies (e.g., this quasi-experimental study, the present study, etc.) to set the stage for afull formulation of research questions. Having furnished such preparatory information pertaining to the new study in Move
3, writers could (i) further align these research questions with either a set of closely matched hypotheses, and/or (ii) subse-
quently support their questions with a brief reiteration of their original research purpose, methods (or methodological
limitations) and positive contributions (or values of their studies).
3.2.2. Inter-step shifts from purpose announcements to research questions
The following analysis demonstrates that when research questions are included in a doctoral dissertation, they are closely
connected with purposive and descriptive announcements in the same dissertation. The contents of research questions are
therefore contingent upon the specific ways in which doctoral candidates associate their purpose statements with research
questions in this highly standardised subgenre of experimental dissertations.Figure 3illustrates how purposive and descrip-
tive announcements (functioning as rhetorical lead-ins) are linked with research questions in terms of semantic content. For
instance, the initial purpose statement specifies that the study aims to find out whether an experimental intervention (i.e.,
oral-language rehearsal) will have any effect on the students writing. This possible effect is then broken down into dif-ferent specific facets in six subsequent research questions, which ask (i) whether such an oral-language rehearsal will have
an effect on four possible dependent variables of the students writing, comprising the students independent writing, the
number of sentences they write, the number of content words they use, and their sentence complexity (as in RQs 1
through 4), (ii) whether these dependent variables will change overtime (as in RQ 5), and (iii) whether there is a relationship
between the baseline measures and the major dependent variables mentioned above (as in RQ 6).
Doctoral candidates generally incorporate a purpose statement indicating the type of instructional intervention or treat-
ment introduced in the research and the dependent variables to be measured, and each of these variables is then specified
separately in every subsequent research question. In more than a third (i.e., 42.3% or 11/26) of the doctoral dissertations con-
taining RQPs, research questions immediately follow purpose statements. RQPs rarely immediately follow topic generalisa-
tions in Move 1 (i.e., in only five of the 26 doctoral dissertations). By contrast, in a vast majority (i.e., 84.6% or 22/26) of these
dissertations, RQPs immediately follow a step in presenting the present work in Move 3 (such as a purpose statement, sum-
mary/preview of research methods, or value/structural indication). More specifically, although purpose statements may not
Announcing the present research
purposively and/or descriptively
Presenting research questions
This study examined students writing
after using a teaching technique called
oral-language rehearsal. The effect of
this experimental intervention (oral-
language rehearsal) on student writing
was compared with students writing
created without the structured oral
language practice... (EIC2: 2)
This study investigated the following questions:
1. To what extent does oral-language rehearsal affect first-grade
students independent writing?
2. To what extent does the treatment group differ from the comparison
group in the number of sentences they write?
3. To what extent does the treatment group differ from the comparison
group in the number of content words they use?
4. To what extent does the treatment group differ from the comparison
group in their sentence complexity due to the practice of orally
rehearsing sentences before actually writing?
5. Is there a changein any of the dependent variables over time?
6. Is there a relationship between the baseline measures and the
dependent variables? (EIC2: 2-3)The current study investigates the effect
of awareness of derivational morphology
on lexical acquisition in Arabic as a
foreign language. More specifically, the
study is designed to explore the
effectiveness of training in the main
word formation process of root and
patterns in Arabic on the acquisition of
Arabic vocabulary by native speakers of
English at the low proficiency level.
(EIC17: 6)
The central research question of this thesis is: Does morphological
awareness of the main word formation process of root and patterns in L2
Arabic facilitate learners' ability to infer meanings of unknown words,
coin new words, and retain words? This main research question is broken
down into seven questions formulated as follows:
1. Are students who receive morphological training on roots and patterns
better at inferring meaning of unfamiliar words when root information is
provided as compared to those who do not receive morphological
training?
2. Are students who receive morphological training on roots and patterns
better at coining unfamiliar words when root information is provided as
compared to those who do not receive morphological training?
3. Are students who receive morphological training on roots and patterns
able to benefit from their morphological awareness equally in their
receptive inferring and productive coining of unfamiliar words when root
information is provided?... (EIC17: 6)
Fig. 3. Inter-step shifts from purpose statements to research questions.
76 J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688
http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
12/23
always occur immediately before these research questions, their semantic linkages with research questions are well estab-
lished. Overall, in nearly all (i.e., 96.2% or 25/26) of the dissertations containing research questions, RQPs are consistently
positioned after purpose statements (although they only immediately follow purpose statements in some dissertations as
explained above), thus showing that specific research questions are, in general, meaningfully formulated only when the fun-
damental purpose of a study has been lucidly presented in a doctoral dissertation introduction.
3.2.3. Intra-step shifts involving main and specific research questions
Building on the aforementioned purpose statements, dissertation writers may further develop specific questions based ona central research question, which is semantically more closely connected with a general purpose statement.Figure 4shows
that the semantic features of each specific research question are covered in advance by parts of the central research question.
The relationship between central (general) and specific research questions in this context is only partly similar to that (i)
between major research questions and subsidiary research questions (Blaikie, 2000, p. 68), or (ii) between an overarch-
ing research question and its subordinate research questions (which are grouped hierarchically under an overarching re-
search questions) as a rule (Sunderland, 2010, p. 15). Rather, the one and only central research question is closely linked
with (and almost equivalent to) the research purpose (or general objective) which is often expressed in a single sentence. For
instance, as shown inFigure 4, the central research question incorporates three aspects (i.e., inferring meanings, coining new
words and retaining words). These three aspects covered by the main research question are subsequently dealt with sepa-
rately in different specific research questions. More precisely, when the writer raises a question regarding whether morpho-
logical awareness of the main word formation process of root and patterns in L2 Arabic will facilitate learners ability to infer
meanings of unknown words (which is the first aspect of the central research question), the aspect concerned (i.e., inferring
meanings) is further subdivided into different components in the subsequent specific questions, which are (i) inferringmeaning of unfamiliar words when root information is provided as in specific RQ 1, and (ii) inferring meaning of unfamiliar
words when root information is not provided as in specific RQ 4. This means that one aspect may be broken down into mul-
tiple components (or situations) to be examined separately via different research questions (as in RQs 16 in Figure 4).
In contrast, different selected aspects in the main research questions may also coalesce and be examined comparatively
via one subsequent specific research question. As exemplified inFigure 4, in the central research question, the writer asks
whether morphological awareness of the main word formation process of root and patterns in L2 Arabic will facilitate learn-
ers ability to coin new words and retain words. These two aspects in the overarching research question are then covered in
one subsequent subordinate research question (i.e., RQ 7) when the writer asks whether students who receive morphological
Fig. 4. Intra-step shifts from a main research question to specific research questions.
J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688 77
http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
13/23
training on roots and patterns are able to benefit from their morphological awareness both in their (i) receptive retention
(i.e., ability to retain words), and (ii) productive vocabulary retention (i.e., ability to coin new words).
3.2.4. Intra-step shifts from asking research questions to explaining research questions
An alternative type of intra-step shifts will appear lucid and comprehensible if we divide presenting research questions
into two components consisting of (i) asking research questions, and (ii) explaining research questions.Figure 5illustrates
that immediately after research questions are asked, reasons are often given to (i) explicate the difference between the re-
search questions, and (ii) justify the necessity of having different questions (to seek answers or data via different method-ological means). More specifically, explanations of the original questions may appear in the form of paraphrases beginning
withreformulative conjuncts(e.g., that is, in other words, etc.). The writers often switch from direct questions in the present
tense to indirect statements in the form of anticipatory it-clauses in the past tense (e.g., it was expected that . . .would not
take, etc.) which function as paraphrases that further explain the preceding research questions in an attempt to gain the
readers acceptance. The following subsections present findings on the language used in research questions at a deeper level.
3.3. The language used in formulating research questions
Given the contextual descriptions and semantic relationships between the moves and/or steps, the following sections will
focus on the linguistic mechanisms that the doctoral candidates opt to use in presenting their experimental research
questions.
The range of linguistic choices that the dissertation writers made in presenting these research questions appear far broad-
er than those reported for experimental research in textbooks (e.g.,Creswell, 2008; Gay et al., 2009) which consist mainly ofyes/no questions beginning with primarily auxiliaries/operators will and do. Research questions used in the doctoral
dissertations comprise a conspicuously wider range of both polar questions (i.e., yes/no questions) and non-polar wh-
questions although the experimental dissertations are all quantitative experimental studies focusing on the effects of
treatment variable(s). Overall, the corpus includes six major categories of research questions with different syntactic
structures, comprising three categories of polar questions and another three types ofwh-questions.
Before considering the functional aspect, we may first examine the syntactic choices of the first three categories of polar
research questions (as illustrated inTable 2). The first two categories, which involve the use of the operators do and will in
auxiliary verbsubjectmain verb (VaSVm) structures, appear 25 and 9 times respectively in the corpus, and are used in
Fig. 5. Intra-step shifts from asking a research question to explaining a research question.
78 J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688
http://-/?-http://-/?-http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
14/23
38.5% (10/26) and 7.7% (2/26) of the experimental introductory chapters (EICs) containing RQPs respectively. The third
category, which engages the primary copular verb in the simple present (i.e., is, are, etc.) in an existential interrogative
structure, occurs 31 times in the corpus, and is used in 38.5% (10/26) of the EICs. Overall, polar research questions (inclusive
of all the three categories) are used in 69.2% (18/26) of the doctoral dissertations. Three points merit some attention. First, in
most cases, the simple present (involving the operators do, does, is, are, etc.) is used in polar research questions, and the
dissertation writers often prefer to present them as enquiries seeking data to confirm the existence of a regular event or gen-
eral phenomenon. Second, the simple future is used in some (30.8%) of the doctoral dissertations containing RQPs, thus
showing that in some cases, doctoral dissertations might foreground the predictive nature of a study (in contrast to theaforementioned descriptive nature of an existing/current phenomenon involving the simple present). It is used as a mech-
anism to draw readers attention to the predictability of an expected occurrence as part of the researchers attempt to plan
the direction in which their studies will proceed. Third, the simple past is also used in indirect statements that explain or
justify research questions. Such segments, as exemplified inFigure 5, are found in 96.2% (25/26) of the dissertations contain-
ing RQPs. The high incidence of the simple present in presenting research questions is ascribable to the fact that both direct
interrogatives and indirect statements related to research questions are presented in the tense concerned. While direct
interrogatives in the simple present play a central role in suggesting possible acceptance of general phenomena, indirect
statements in the same tense (which introduce and justify research questions) also function as a mechanism to indicate
current relevance.
As these polar questions (shown inTable 2) are presented in the introductory chapter of an experimental dissertation on
Applied Linguistics, the propositions are negated by referring to a quantitative cut-off point or line of demarcation distin-
guishing whether the answer should be affirmative or negative. To be precise, the final response can be considered affirma-
tive or negative only after inferential statistical data have been provided and/or explained. For instance, if the asymptoticvalue is lower than the cut-off point of 0.05 (or 0.01 in other cases), a candidate is expected not merely to state that there
is a significant difference between a control group and an experimental group (i.e., to give an affirmative answer yes to an
ordinary polar question), but also to use the related inferential statistical data to demonstrate, explain or justify how the
affirmative answer is reached. In 55.6% (10/18) of the EICs containing polar questions, answers to the polar research ques-
tions (presented in the introductory chapters) are supported by findings (in the last two chapters) that cover far more than
merely a yes/no answer. In this regard, Figure 6provides additional information which shows how doctoral dissertation
(DD) writers actually report their results (in the last two chapters when results are presented or summarised) in response
to their research questions (initially formulated in the first chapter of each dissertation). It has been found that the polar
Fig. 6. Rhetorical connections between presenting a research question and reporting findings on the research question.
J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688 79
http://-/?- -
5/19/2018 Formulating Research Questions
15/23
research questions are generally not addressed with only a yes/no answer. For instance, when a polar research question asks
whether there is a relationship between teacher ratings and student beliefs (or whether Explicit Experienced Grammar
Instruction has a positive effect on learners grammatical accuracy), writers would use different mechanisms that indicate
a need to move beyond providing merely a yes/no answer. First, additional boosters or downtoners (e.g., highly signifi-
cantly correlated with, not significantly related with, etc.) may be deployed to describe the extent to which an experimen-
tal intervention (independent variable) has on the subjects performance (dependent variable). Second, in answering polar
research questions, the relationship may be subdivided into different parts (e.g., relationship between teacher ratings
and student appraisals of ability beliefs for Conventions and relationship between teacher ratings and student efficacyresponses for Language and Story Construction subscales) that require specific results beyond a yes/no answer. Third, in
some cases, it is not possible to give a straightforward yes/no answer given that the findings are mixed, and an initial
prediction (that an interventional training will lead to better performance) is only partially supported by the data (as shown
in the instance from EIC17 inFigure 6).
In addition, three types of non-polar questions have been found in 42.3% (11/26) of the EICs containing RQPs. The first
type of non-polar question is a special type of wh-question (e.g., What is the impact of SIOP on teachers and English
Language Learners at the elementary level?) that appears like a factual question but is functionally a convergent question
which usually falls within a finite range of acceptable accuracy. Syntactically, this type of convergentwh-question involves a
wh-pronoun and a copular verb in an SVC interrogative structure, which is Subject (What) + Verb (is/are) + Complement (NP
associated with experimental effect) (seeTable 3). Appearing 14 times in the corpus, it occurs in 23.1% (6/26) of the EICs
concerned. On the surface, this type of factual wh-question seems to solicit reasonably simple and straightforward answers
based on obvious facts, yet it pragmatically entails an in-depth discussion given that the sentence complement requires the
use of a noun phrase indicating experimental or instructional effect (which can be reported only with reference to a numer-ical cut-off point when inferential statistical data are presented). As such, this category of research question in doctoral dis-
sertations differs considerably from other wh-questions in daily conversational contexts (e.g., What is your name?, Who
did you talk to?) in which direct answers in noun phrases may be considered adequate. Nonetheless, despite the higher level
of complexity involved, this special category (involving effect measurement) still seeks answers that clearly differentiate
strong effects from moderate and weak ones. Alternatively, these questions may require data that ascertain the difference
between control and experimental groups as significant or insignificant.
The second type ofwh-question appears more evaluative but actually also seeks data to ascertain the effects of treatment
variables or the degrees of difference between control and experimental groups. This category of research question begins
with a prepositional phrase (i.e., to what extent) and is followed by a V aSVmstructure (which comprises the primary oper-
ator e.g., do/does, the subject, and the main verb respectively). Appearing 17 times in the corpus, it occurs in only 11.5%
(3/26) of the EICs containing RQPs, thus showing that evaluative wh-questions beginning with to what extent constitute a
less prevalent choice for experimental studies. These evaluative questions apparently necessitate sophisticated levels of
cognitive and/or emotional judgement as doctoral candidates are expected to combine a logical thinking process and acomparative framework to analyse the experimental data collected from different perspectives before providing an extended
response using their newly synthesised information. In this case, the dissertation writers need to use numerical data to
indicate the degree to which the treatment effect is significant or the group differences are prominent. The third type of
wh-question seeks circumstantial descriptions involving the primary verb do in a VaSVm structure. Appearing seven times
in the corpus, it occurs in 7.7% (2/26) of the EICs. Even though these questions that begin with how appear to be typical
questions in qualitative research, they also occur in quantitative experimental dissertations in this corpus. Nevertheless, they
are relatively rare compared to the polar questions and convergent wh-questions mentioned above.
Overall, the three types ofwh-questions explained above perform the same function of identifying the degree to which an
experimental treatment has an effect or the extent to which various groups differ from each other. As mentioned above, the
answers to a polar question can cover far more than merely yes or no. These direct interrogatives may have different syn-
tactic structures, but they actually need to be interpreted using the readers knowledge of inferential statistics explained
above. This means that the reader needs to consider the context in which the research questions are presented by referring
to the cut-off point that writers normally use to decide whether a treatment has a significant or insignificant effect on thebehaviours of the subjects.
A careful analysis of the verbs used across the 32 dissertations showed that the simple present, simple past and simple
future are more commonly employed than other tenses (including the present perfect and present perfect continuous) in
formulating research questions (seeTable 4).3 In this regard, SIA was of the view that research questions might not necessarily
be expressed in the form of direct interrogatives, and this explains why indirect statements or declarative sentences in the sim-
ple past (apart from the common direct interrogatives) pertaining to the prediction of experimental research outcomes have
also been found in this study of the writers formulation of research questions.
Figure 5shows that declarative sentences in the simple past are used when writers look at their work retrospectively (as
opposed to focusing on the present text) while explaining their preceding research questions (expressed in direct interrog-
atives). For instance, EIC 9 has a high frequency of verbs in the simple past, which constitute a third of the total number of
3 The present perfect and present continuous are used in RQPs in only two dissertations (EICs 23 and 24) for describing and distinguishing groups of subjects(e.g., English-speaking college students who are learning a romance language. . ., etc.).
80 J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688
http://-/?-http://-/?-http://-/?-http://-/?-http://-/?- -
5/19/2018 Formulating Research Questions
16/23
past tense verbs used (seeTable 4) because explaining research questions (which may require the simple past) is consis-
tently used after the preceding sub-step (i.e., asking research question) as illustrated inFigure 5. SIB expected doctoral can-
didates to use the simple present for research questions although there should be some transitional phrases or sentences in
Table 3
Salient syntactic choices used in formulating three types ofwh-questions that guide the research in EICs.
Syntactic choice Freq.of
RQs
No. of
EICs (y)
% of EICs (y/
26 100%)
Instances of research questions
Factual wh-questions (involving a wh-pronoun and a
copular verb) in an SVC interrogative structure
14 6 23.1 Specifically, the question that this study asks is:
What are the individual and collective effects of
voiced-pronunciation and stroke sequence animation
in flashcards on beginning CFL learners production ofChinese characters? (EIC4: 5)
The questions answered in this study included: What
is the effect of temporal auditory training on reading,
and on the expressive and receptive language skills of
children with diagnosed temporal auditory
processing deficits? (EIC13: 14)
The overall research question posed for this study
was: What is the impact of SIOP on teachers and
English Language Learners at the elementary
level?...(EIC25: 11)
The research question was broken down into more
specific questions. The specific research questions
were:
1. What is impact of SIOP training on the instructional
practices of teachers with regard to presentation ofcontent for ELI students?...EIC25:12)
What are teacher perceptions of the SIOP model
aligned with the results regarding the change in
instructional practices? (EIC25: 12)
Evaluative wh-questions seeking data that ascertain
the effects of treatment variables or the degrees of
difference between control and experimental
groups (in VaSVmstructures involving the operator
do)
17 3 11.5 5. To what extent does oral-language rehearsal affect
first-grade students independent writing?
6. To what extent does the treatment group differ
from the comparison group in the number of
sentences they write?...EIC2: 25)
The research questions are:
1. To what extent, if any, do students who are taught
reading using cooperative learning have higher
reading performance in vocabulary than those taught
in (using) traditional methods?...(EIC6: 4)In order to achieve this purpose the following
research questions will be answered:
1. To what extent do differences in treatment
contribute to the variance in the overall achievement
as measured by the summed score Phonological
Awareness Literacy Screening (PALS) of the second
grade students who participated in the non-
traditional instructional literacy program compared
to the achievement of second grade students who
participated in the comparison group in the
traditional model?. . .(EIC12: 8)
Circumstan-tial wh-questions using the primary verb
do in an VaSVm structure
7 2 7.7 1. Through the use of vestibular stimulation, how
does participation in Rowes (2005; 2001a, 2001b,
1000a, 1999a, 1999b, 1999c) NN activities influencethe attainment of skills associated with at-risk
educational status of Kindergarten students (i.e.,
phonological awareness and rapid automatic naming
skills)?...(EIC19: 11-13)
Research Question 2: How does teaching with the use
of cooperative learning strategies affect the academic
achievement of sixth-grade students with disabilities
on the Georgia Performance Standards test in
language arts? (EIC31: 15)
J.M.-H. Lim / English for Specific Purposes 35 (2014) 6688 81
-
5/19/2018 Formulating Research Questions
17/23
Table 4
Frequencies and categories of verbs in the three major tenses used for formulating research questions in experimental introductory chapters.
Dissertation introduction Simple past Simple present Simple
Total no. of verbs 36 168 20
Mean no. of verbs 1.13 5.25 0.63
No. of EICs (z) 13 25 8
% of EICs (with RQPs) (z/26 100%) 50.0 96.2 30.8
Instances of verbs illustrated in
simplified exemplary structures
of RQPs
Verbs denoting likelihood in it-clauses
that explain RQs
It was expected that Group A would
not behave as effectively as Group B.
It was expected that V2 would be
related with V3.
Copular and stative verbs in existential interrogative
structures, wh-questions, and polar questions
Is there a relationship/difference between V2 and
V3?
Is there a significant difference in V2 (performance)
when comparing Groups A and B?
What is the effect/impact of V1 on V2?
Is Group A as accurate and fast as Group B in
situation S?
Does V1 have an effect on V2?
Stative
Will
level of
Procedu
Will
Test X a
Test
question
Verbs signifying cause-effect relationships
How/To what extent does V1 affect/differentially
influence /facilitate V2?
Does V1 predict/contribute to an increase in V2?
Verbs denoting variations
To what extent does Group A differ from Group B on
Test X?
Does V2 (skill) improve among students in Group A
as compared to those in Group B?
Does V1 decrease V2 (performance) of Group A in
situation S?
Verbs s
Will
Note:An independent variable is represented by V1. Dependent variables are represented by V2 and V3, while A and B refer to different groups of subject
a test or an assessment.
-
5/19/2018 Formulating Research Questions
18/23
the simple past (as exemplified in the explanations of research questions inFigure 5). The analysis of overall tense usage in
RQPs also shows that research questions are far more commonly expressed in the simple present (96.2%) while merely half of
them (50% of the verbs in RQPs) use the simple past in indirect statements referring to research questions (seeFigure 5).
Overall, the comparative prevalence of the simple present may be ascribed to the doctoral candidates greater propensity
to foreground the current relevance of an expected existing phenomenon.
It is necessary to consider tense usage in relation to the respective categories of verbs in order to provide a comprehensive
account of the language resources needed to formulate research questions. As illustrated inTable 4(and indirectly inTables
2 and 3andFigures 35), the formulation of research questions is characterised distinctly by (i) copular or stative verbs inthe simple present or simple future (as in the simplified exemplary structure Is there a relationship/difference between var-
iable 2 and variable 3?, Does/will variable 1 have an effect on variable 2?); (ii) verb phrases signalling causeeffect rela-
tionships in the simple present or simple future (as in the pattern How/To what extent does/will variable 1 affect/
influence/facilitate/predict variable 2 (performance) of the learners?); (iii) verbs denoting variations in the simple present
(as in the simplified structure To what extent does Group A differ from Group B?, Does variable 1 decrease variable 2 (per-
formance) of Group A in situation S?, etc.); (iv) past tense verbs denoting likelihood used for explaining research questions
(as in the pattern it was expected that Group A would not behave as effectively as Group B); and (v) procedural verbs in the
simple future (as in the exemplary structure Will students in Group A score significantly higher on Test X as compared with
students in Group B?).
Another common linguistic device employed by doctoral candidates in RQPs has to do with the use of noun phrases
indicating connections between factors and dependent variables (e.g., effect, impact, relationship, predictor, etc.). More
distinctly, as illustrated inTable 2, writers often use head nouns postmodified by (i) adjectival relative clauses (e.g., students