The Effectiveness of Using a Deductive-Cum-Process Writing Process Approach to Teach Tone Formality

This study aims to investigate the usefulness of a deductive-cum-process writing approach to raise native Cantonese-speaking ESL students’ awareness of the inappropriate use of informal language in academic writing. Achieving tone formality by incorporating appropriate syntactic and semantic elements poses great difficulties to freshmen at the City University of Hong Kong. Furthermore, the effectiveness of using a deductive approach to teach word usage has been inconclusive despite numerous studies on this issue (e.g., Alzubi, 2015; Tammenga-Helmantel, 2014; Vogel et al., 2011). In the current study, seven types of informal language expressions and corresponding formal alternatives were first taught, followed by peer review and individual teacher-student consultations to reinforce the teaching. The experimental group consisted of 19 students taking an academic writing course in the summer semester 2017-2018. The end-of-course writing examination showed that the experimental group used significantly fewer informal expressions than did the control group in three measures—the total number of informal expressions and two component measures (p≤0.05). Questionnaire findings confirmed the usefulness of the current teaching approach, as perceived by the majority of the participants. Qualitative analysis revealed several types of informal expressions that were most difficult for the participants to identify and to make correction.

not been included in the syllabus of the subject "English Language" of the Hong Kong Diploma of Secondary Education. It is not unusual to find a large number of informal vocabulary and grammatical structures in the argumentative essays written by Year 1 students at the City University of Hong Kong before and even after the teaching of tone formality in academic writing lessons. As such, pedagogical research into the teaching of formal and informal words would benefit teaching practitioners as well as learners who learn English as a second language. The teaching of tone formality, similar to the teaching of other grammatical structures, involves deductive or/and inductive approaches. A deductive approach is characterized by the sequence of presenting general rules first, followed by illustrating the rules with examples and ended with the provision of practice; an inductive approach, on the other hand, is characterized by the provision of examples first, followed by practice and ended with the provision of general rules (Alzubi, 2015;Glaser, 2013). Alzubi (2015) succinctly pointed out several features of the two approaches. A deductive approach is a) rule-driven-that is, rules are taught directly and examples are provided afterwards; b) top-down, and c) teacher-front transmission style. An inductive approach is a) rule-discovery; b) examples are provided first, and rules will be induced by students; c) bottom-up; and d) student-centred (p. 188).
A deductive approach was found to be preferred by the majority of students in some research studies (e.g., Jean & Simard, 2013;Rahmatian & Zarekar, 2016;Vogel et al., 2011); similarly, a deductive approach was found to be preferred by teachers (e.g., Thornberry, 1999;Muthusamy & Farashaiyan, 2016). On the other hand, the students in Mohamed's (2004) study and in Valijärvi and Tarsoly's (2015) study indicated that both approaches had advantages and disadvantages. Overall, the effectiveness of using a deductive approach to teach word usage is not conclusive (Alzubi, 2015;Tammenga-Helmantel, 2014;Vogel et al., 2011). To add to the body of literature, this study aims to investigate the usefulness of a deductive approach supported by a process-writing approach to raise ESL students' awareness of the inappropriate use of informal language in academic writing. To be specific, this study is intended to investigate whether the number of informal expressions appearing in the problem-solution essays written by the experimental group will be significantly lower than that appearing in the essays written by the control group. It is hoped that the research findings will shed light on the pedagogical implications of the deductive-cum-process writing approach used in this present study.
The present study intends to address the following research questions: 1) Will the deductive-cum-process writing approach used in this present study help the participants who speak Cantonese as their first language learn tone formality more effectively than the inductive approach used for the control group?
2) What are the participants' views on the form of input enhancement used in class?
3) What are the participants' views on the effectiveness of the coding system used in this present study as the coding system as an error-correction tool? www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 2, No. 2, 2020 23 Published by SCHOLINK INC.

Literature Review
The features of deductive and inductive approaches, as described by researchers over the last twenty years, are highlighted as follows. DeKeyser (2003) viewed the two approaches from the dimension of explicitness and implicitness in terms of introducing general rules governing the use of a target language: explicit inductive involves the generating of rules explicitly whereas implicit inductive refers to unconscious acquisition of some general rules by exposing students to linguistic materials. The notion of Implicit inductive seems to have rooted in Krashen's (1982) notion of unconscious acquisition, which is triggered by exposing students to authentic language use. Takimoto (2008) further points out that implicitness and explicitness in introducing general rules are conceptualized as a process on a continuum rather than constituting a dichotomy; similarly, deduction and induction also reflect a continuum. Decoo (1996, pp. 2-3) described 5 modalities on the continuum with respect to deductive and inductive approaches: Modality A involves actual deduction.
Modality B refers to conscious induction as guided discovery. Students first encounter various examples, often sentences, sometimes embedded in a text. The conscious discovery of the grammar is then directed by the teacher to guide students to formulate the rules.
Modality C is characterized by induction leading to an explicit "summary of behavior" (Decoo, 1996, p. 2). The learner first practices a certain structure in an intense way. Through this practice the rule is somehow induced and internalized. At the end of the learning segment, the teacher summarizes the rule explicitly.
Modality D features subconscious induction using structured materials.
Modality E refers to subconscious induction using unstructured materials.
In addition to theoretical concepts concerning what constitutes deductive and inductive approaches, there have also been empirical research studies investigating the pedagogical effects of the two approaches. The explicit inductive approach was found to be more effective in raising post-test scores than did the deductive approach for both university and high school levels ((Alzubi, 2015;Jong-Won, 2007;Qi & Lai, 2017;Tammenga-Helmantel, 2014;Vogel et al., 2011). However, the implicit inductive approach was found to be less effective than did the deductive approach in raising post-test scores (Erlam, 2005;Robinson, 1996;Rosa & O'Neill, 1999).
Concerning students' preference over the two approaches, research results have varied. A number of studies have found that the majority of the participants preferred the deductive approach. Vogel, et al. (2011) found that 80% of the participants preferred the deductive approach; similarly, Haight (2008) found 73% of the participants favored the deductive approach. Reasons suggested by participants in support of the deductive approach included: Not requiring problem solving; more used to it (Jean & Simard, 2013); preferred authoritarian style (Rahmatian & Zarekar, 2016). Yet, some studies (e.g., Valijärvi & Tarsoly, 2015;Mohamed, 2004) found that students showed no particular inclination over the two approaches.
With respect to teachers' preference, Thornbury (1999) believed that a deductive approach is efficient and clear (p. 55). Muthusamy and Farashaiyan's (2016) study revealed that Iranian EFL instructors mostly used inductive approach (65%) to teach interlanguage pragmatics.
Effectiveness of input enhancement for a lesson constitutes another equally important considerationconducive to learning. Sharwood-Smith (1993) classified input into positive and negative input. The former refers to input that can "make salient certain correct form of the input" (p. 177), whereas the latter refers to flagging "given forms as incorrect, thus signalling to the learner that they have violated the targeted norms" (p. 177). Takahashi (2001) proposed four input enhancement conditions in terms of degree of input enhancement when teaching L2 pragmatics (making requests): a) explicit instructions; b) form comparison; c) form searched; and d) meaning-focused conditions.

Research Tools
Two research tools were used in the present study. a) End-of-course writing examination scripts (19 copies from the control group and also 19 copies randomly selected from the experimental group to match the size of the control group. Markers of the exam scripts were the language teachers in the English Language Centre of the City University of Hong Kong. A marker training session was held before markers started to mark the scripts. Class teachers were not assigned to mark their own students' scripts. b) A questionnaire completed by all the participants in the experimental group. A total of 44 copies were collected.

End-of-Course Writing Examination
The genre was a 700-wordproblem-solution essay. Students wrote on one of the two topics provided: Topic 1: Identify one of the problem(s) of the student canteen at City U, suggest at least two solutions and justify the solution you would recommend.
Topic 2: Identify one of the problem(s) of having too many brand-name chain stores in the shopping malls in HK, suggest at least two solutions and justify the solution you would recommend.

Questionnaire
Questions 1 and 2 required respondents to indicate their degree of agreement/disagreement with the following two statements on a 5-point Likert Scale

Participants
The participants were Associate Degree students, scoring Level 3 in the subject English Language in the Hong Kong Diploma of Secondary Education (HKDSE).
In the summer term 2017-18 (June, 2017), there were a total of five classes of English Enhancement Course for Associate Degree Students II (EECAD II), which was a reading and writing course. The course assignment included writing a problem-solution essay (incorporating external sources) using a process writing approach. There was an end-of-course writing examination, which required students to write a problem-solution essay.
Two classes taught by the present researcher constituted the experimental group and two classes taught by another teacher the control group. The experimental group consisted of 44 students, and the control group 19 students (Table 1): were chosen at random to match the size of the control group) Taught by the present researcher (2 classes totaling 44 students)

Treatment Applied to the Experimental Group
PHASE 1: The deductive approach (Modality A, Decoo, 1996) Stage 1: Teacher-front explicit explanations about seven types of informal expressions.

Seven types of informal expressions;
 Teaching material: structured input.
The teaching of tone formality to the experimental group included several phases. In Phase 1, Modality A (Decoo, 1996) was adopted, in which the teacher explicitly pointed out to students seven types of informal expressions and the corresponding neutral/formal expressions using a table contrasting the two versions of expressions(Appendix B). The seven types of informal expressions are underlined below for easy reference: 1) Informal words, for example, "kids", "a lot of"; 2) Sentence-initial connectives, for example, "And", "But" and "So"; 3) The omission of "that" before a noun clause, for example, "I agree (that) parents should take the major responsibility to monitor their children's use of the Internet"; 4) Contraction, for example, "What's more"; 5) Starting a sentence with a verb, for example, "Talk to your parents"; 6) Rhetorical questions, for example, "Should parents take responsibility to monitor their children's use of the Internet?" 7) The use of "I" in stating one's opinion in the body of an essay, for example, "I think that the housing prices in Hong Kong is too high." Stage 2: Teacher-front explicit explanations about the coding system used as a tool for error correction  The coding system included thirteen types of errors (e.g., organization, informal words, hedging, conjunctions (Appendix C);  Teaching material: structured input.
A handout was distributed to the students in the experimental group in the first hour of the lesson designated for peer review. Before peer review started, the thirteen types of errors, including the use of informal expressions in an academic essay, had been explained to students using the handout on the coding system.
Examples of errors included in the coding system are as follows: "The tone is considered too strong and thus not academic enough due to the lack of sufficient support. (Rule no. 7). Hedges (e.g., probably) should be used to weaken the tone. Refer to the handout on hedges for more examples." "Informal expressions should be avoided (Rule no. 8). For correction, please refer to the handout on informal and formal expressions." Because of the use of Arabic numerals to refer to errors, the coding system would allow students to see the types of their errors at one glance. Furthermore, as can be seen in the examples above, guidance for self-correction is provided (e.g., words underlined in Rules 7 and 8 above).

Stage 3: Student practice
Students reviewed their peers' essays, aiming to identity the thirteen types of errors appearing in the essays. In this exercise, the teaching materials were in fact their peers' essays, which can be considered as "unstructured input" -one of the features of Modality E (Decoo, 1996).

Phase 2: Process writing approach
Stage 3 also constituted the onset of the process-writing approach. After the peer review, students were supposed to make corrections focusing especially on the thirteen types of errors before the submission of the second draft of their essays to the class teacher. After the class teacher read the second draft of the essays, Individual teacher-student consultations, which lasted about 15 minutes, were held before the final submission, which would be marked and returned before the end-of-course writing examination.
The course materials used with the control group The control group was taught with the materials provided by the course. The teaching materials provided in the course book could be regarded as a manifestation of Modality Band Modality C (Decoo, 1996). The very first exercise in Unit 2 entitled "Distinguishing Text Genre" included two texts containing formal and informal expressions, and students were instructed to determine which expressions were formal/informal based on their prior knowledge. The second part of the question asked students to suggest features of formal/informal expressions. Generally, in Modalities B and C (Decoo, 1996), learners are provided with examples and then they are expected to induce rules underlying the examples with the guidance of the teacher (Modality B), or the teacher himself/herself summarizes the rules to the learners (Modality C). As such, the design of the first exercise in the course book can be regarded as a realization of these two modalities. The remaining exercises in Unit 2 of the course book involved: a) matching "Dos and Don'ts" with the corresponding formal/informal expressions; and b) determining the formality level of individual expressions in texts of different genres.
All of the seven types of informal expressions taught to the experimental group also appeared in the various sample texts in Unit 2. The process writing approach was also applied to the control groupthat is, Individual teacher-student consultations lasting about fifteen minutes each were also conducted among the control group.
A brief interview with the class of teacher of the control group confirmed that he used only the teaching materials in the course book, and the process writing approach, which was required by the course, was also adopted. This is to say, two drafts were also prepared by the students before the final submission.

Results
The data were analysed both quantitatively and qualitatively. Quantitative analysis involved T-test and frequency count using IBM SPSS version 24, whereas qualitative analysis was done manually.

Findings of the end-of-course writing exam
As can be seen in Table 2 b) The use of And/So/But at sentence-initial position  It is worth noting that the experimental group did not score significantly higher marks in the total score of the writing exam and in other such domains as content, organization, coherence and language at p≤0.05 than did the control group (Table 3). The insignificance in the various scores of the two groups in Table 3 seems to support the observation that the deductive-cum-process writing approach applied to the experimental group in the teaching of tone formality appears to have benefited the students in the group. The reason is that the experimental group, as indicated in the scores in Table 3, was not stronger than the control group in writing proficiency as measured by the four domains shown in Table 3, but the former was able to use significantly fewer informal words at p≤0.05.
The insignificant difference in "Language" might be explained as follows: "the use of formal expressions for an academic register" was not stipulated explicitly in the descriptors for "Language", which is copied from the course rubric verbatim below: "1) Accuracy and variety in grammatical structures and sentence types; 2) Accuracy (collocation, word form and spelling), appropriacy and variety in vocabulary use; and 3) Impact of errors on readability" (ELC Process Writing Essay & Final Writing Exam Rubric, 2017).
Although the assessment of "appropriacy and variety in vocabulary use" might be argued as inclusive of tone formality, the weighting of tone formality to be attributed to "language" did not appear to be clear to the markers; furthermore, it remains doubtful whether tone formality would indeed constitute a factor that markers would consider when assigning a mark to the domain "Language" given the lack of the actual words "tone formality" in the rubric.

Questions 1 and 2
Questions 1 and 2 required respondents to indicate their agreement/disagreement to the following two statements on a 5-point Likert Scale: (1=Totally disagree; 3=Neutral; 5=Totally agree) Findings:

Question 1
The informal expressions (in descending order of percentages) that he participants have become aware of in the process of learning to use a formal tone are shown in the following Table: As can be seen in Table 4, the three types of informal expressions the participants most aware of are: short form (77.3%), informal vocabulary (77.3%), and And/But/So at sentence-initial position (77.3%).

Question 2
The informal expressions (in descending order of percentages) that the participants believed that they had developed the ability to correct their own errors in the process of learning to use a formal tone are shown in Table 5: It is interesting to observe that the participants found it very easy to discern informal vocabulary (the highest-ranking), but they found it the third hardest to correct. This might be due to their lack of formal vocabulary. This difficulty was also reported by some participants in their open-ended comments.  Yes -97.7% The majority of the participants thought that it was useful to be provided with a table contrasting informal and formal expressions before they started to practice identifying to identify informal expressions in the peers' essays. The majority of students also believed that it would be more useful to correct their own errors of using informal words based on the prompts provided rather than having the class teacher provide the corrections directly.

Question 4
Do you think you will continue to have difficulty identifying informal expressions in your future writing?

Yes -50%
No -50% Question 5 Do you think you will continue to have difficulties correcting informal expressions in your future writing?
Yes -40.9% No -59.1% The participants seemed to need more teaching and practice concerning the use of formal words. About an equal number of the students thought that they would continue to have difficulties identifying and correcting informal expressions in the future writing.

Written comments provided by questionnaire respondents
The participants indicated the following comments: more examples of what constitutes informal words and formal words; more exercises to correct informal words; more model essays demonstrating the use of formal words; and more advice on how to handle their difficulty in memorizing the expressions that are generally considered informal (Table 6).

Discussion
There might be two possible reasons why deductive-cum-process writing approach used in this study seem to have benefited the participants in the experimental group. First, the participants in this study possessed a low level of English proficiency. As such, it might be easier for them to discern the differences between informal and formal expressions when they were directed to the inappropriate and appropriate expressions at the start of the lesson with the aid of a handout unambiguously contrasting informal expressions (i.e., "negative input") (Sharwood-Smith, 1993, p. 177) with the corresponding corrections (i.e., "positive input") (Sharwood Smith, 1993, p.177). unambiguously. On the other hand, an inductive approach would require learners to generate rules from random instances provided. Such a requirement might be difficult for learners possessing a low level of English proficiency and thus possibly not effective for learning. Second, the integration of the process writing approach with the deductive approach might have contributed to the less use of informal words by the experimental group.
In the individual teacher-student consultation sessions, the teaching of informal/formal words was reinforced by the teacher's referral to the table constructed under the principle of a deductive approach.
This view in support of the deductive approach is in line with Thornbury's (1999) view that a deductive approach is efficient and clear (p. 55). Also, the finding derived from this present study that the deductive approach helped students achieve higher scores than did the inductive approach was consistent with the findings that higher post-test scores were achieved for both students taught with the deductive approach at university and school levels (Erlam, 2005;Robinson, 1996;Rosa & O'Neill, 1999). Some studies, however, found students taught with the explicit inductive approach scored higher marks in the post-test (e.g., Alzubi, 2015;Jong-Won, 2007;Qi & Lai, 2017;Tammenga-Helmantel, www 2014; Vogel et al., 2011). The difference in findings might be due to English proficiency, the mental capabilities of learners and the subject matter being taught.
Pedagogically, a deductive approach in which negative input and positive input are contrasted in a table is recommended for learners who possess a low level of English proficiency to improve tone formality in writing problem-solution essays. To reinforce the learning effect, individual teacher-student consultation sessions-a practice commonly used in the process-writing approach-could be incorporated in the course schedule, in which teachers might guide students to correct their own mistakes concerning tone formality as a further learning opportunity. It is also recommended that teaching practitioners emphasize to students that the information with respective to formal/informal usage of word is available in dictionaries, thus possibly allaying students' fear for having to learn by rote a large number of informal expressions.
Recommendations for future researchers are threefold. First, such other language features of problem-solution essays as the use of complex noun phrases, subordination, and hedging might be investigated. Second, other genres such as argumentative essays might also constitute an interesting research area with respect to tone formality. Third, the involvement of learners possessing a higher level of English proficiency might reveal more pedagogical information about the effects of a deductive-cum-process writing approach.

Conclusion
This research study aimed to investigate whether a deductive-cum-process writing approach would benefit participants more than did an inductive-cum-process-writing approach. It was found that the former appeared to have reduced the experimental group's use of informal expressions effectively, as reflected in the significantly fewer informal expressions appearing in the end-of-course writing exam compared with those used by the control group (p≤0.05). Questionnaire findings reveal that the students in the experimental group were in favor of the deductive approach in which a handout contrasting inappropriate use of informal expressions with corresponding alternatives was first explained to students, followed by practice in the form of peer review using a grammatical coding system. Questionnaire findings also indicated that the three informal expressions the participants believed that they would be able to discern the most easily include: short form, informal vocabulary and And/But/So at sentence-initial position. On the other hand, the three informal expressions that the participants found most difficult to correct (in order of difficulty) include: rhetorical questions, omission of "that" to introduce a noun clause and informal vocabulary.
It is recommended that a deductive-cum-process writing approach be adopted when teaching tone It would be useful for parents to buy a filtering device. And it would also be effective for them to browse the Internet together with their sons It would be useful for parents to take the main responsibility, but parents are busy.
It would be useful for parents to take the main responsibility. However, parents are busy.
It would be useful for parents to take the main responsibility; however, parents are busy.
Despite the fact that it would be useful for parents to take the main responsibility, parents are busy.
It would be useful for parents to buy a filtering device, and it would also be effective for them to It would be useful for parents to buy a filtering device. In addition, it would also be effective for them to browse the Internet together with their sons and daughters.
It would be useful for parents to buy a filtering device; in addition, it would also be effective for them to browse the Internet together with their sons and daughters.
In addition to buying a filtering device, it would also be effective for parents to browse the Internet together with their sons and daughters Most parents have to work long hours in Hong Kong, so parents are too tired to look after their children after work.
Most parents have to work long hours in Hong Kong. As a result/ as a consequence/ consequently, parents are too tired to monitor their children's use of the Internet after work.

Most parents have to work long hours in Hong
Kong; as a result / as a consequence/ consequently, parents are too tired to monitor their children's use of the Internet after work.

Most parents have to work long hours in Hong
Kong, resulting in being too tired to monitor their children's use of the Internet after work.
As a result of having to work long hours in Hong Kong, Parents are too tired to monitor their children's use of the Internet after work. Wrong: x Most parents have to work long hours in Hong Kong; therefore, parents are too tired to monitor their children's use of the Internet after work.