Effects of an explicit rime-based phonics method on the phonological awareness outcomes of Ethiopian EFL children

Abstract

The present study aimed to assess whether a training-based intervention using an explicit rime-based phonics method improves the phonological awareness outcomes of EFL children in Ethiopia. The participants were two sections of 3rd grade children (N = 70) from two public primary schools in Hawassa, Ethiopia. A quasi-experimental research design was used in which a section of the children (N = 35) were randomly assigned to the experimental group and the other section of the children (N = 35) to the control group. The experimental group was explicitly exposed to a training-based beginning reading intervention using an explicit rime-based phonics method. The control group stayed in their regular classroom with their regular beginning reading lessons with the conventional way. The phonological awareness test (PAT) was used to measure phonological awareness outcomes in the two groups before and after the intervention. The findings showed that the experimental group achieved significant improvement on the phonological awareness at post- test and that the magnitude of the intervention’s effect was very large for all sub-subtests and the total phonological awareness. The findings also indicated that brief and direct beginning reading instruction with an explicit rime-based phonics method helps children in a non-English language context with a first language (L1) significantly different from English to rapidly improve their phonological awareness performances. The findings further suggested that just knowledge of letters alone is not a sufficient condition for phonological awareness to develop.

Keywords:

• rime-based phonics
• phonological awareness
• beginning reading
• young Ethiopian EFL learners

1. Introduction

Learning to read, write and spell is a complex process that comprises the development and integration of a number of linguistic and cognitive skills such as phonological awareness, alphabetic principle and adequate strategies for decoding words (Ehri, 2005; Share, 2004). Phonological awareness is a set of knowledge and skills about the sound structure and characteristics of spoken words in a given language. Phonological awareness allows one to identify and manipulate the sound units that make up words. Research has demonstrated that phonological awareness is one of the basic skills needed in early acquisition and development of reading skills (Bing et al., 2013; Yeung et al., 2013).

A meta-analysis report from the National Reading Panel (Ehri et al., 2001) suggests that in order to read words accurately and fluently, children should first develop phonological awareness or an understanding of the structure of sound units (i.e., syllables, onset-rime, and phonemes) in words in a specific language. Phonological awareness enables children to recognize sound units in spoken language, and this knowledge provides a basis for matching letters to sounds as they learn to read words (Munro, 1998).

Language teaching research used to link failures in early development of reading skills to children’s inability to identify, recognize and process the visual or orthographic components (i.e., letters and letter clusters) in words. However, recent evidence has shown that the problems are largely due to deficiencies in recognizing, identifying and processing auditory or phonological components (i.e., syllables, onset-rime and phonemes) rather than visual or orthographical components (Milankov et al., 2021; Otaiba et al., 2012). Children with low phonological awareness cannot decode words accurately and quickly, and inability to identify and process the sound units of spoken words is the most popular cause of children’s reading difficulties (Choi & Zhang, 2021; Torgesen et al., 1997; Varghese, 2015).

Phonological awareness cannot be acquired naturally or automatically, but is developed gradually through various developmental stages (Cunningham et al., 2004). According to a progressive theory of phonological development described by Treimal (Goswami, 1991), “as children develop phonological awareness, they can progress from the ability to divide spoken words into syllables, to the ability to divide syllables into onsets and rimes, and finally to the ability to segment onset and rime in to phonemes” (p. 525). Therefore, Blachman et al. (2000) consider phonemic awareness which is the ability to identify and work with the smallest sound units—phonemes, to be the highest level of phonological awareness, and teaching phonological awareness provides a strong foundation for developing phonemic awareness.

Phonemic awareness is vital to developing reading skills because of the following reasons. First, it enables children to become aware of the alphabetic principle. Second, it allows children to develop the knowledge of grapheme-phoneme correspondence (GPCs), which is crucial for children to foster their sight word recognition necessary for fluent and efficient word reading. Third, it helps children to identify and manipulate (delete or replace) the initial, middle, or final sounds in words, and this enables them to guess unfamiliar words more accurately based on one or more of these sound units (Catts & Kamhi, 2005; Torgesen et al., 2005). Insufficient or poor development of these skills can lead to impaired reading development (Seymour et al., 2003; Sucena et al., 2015; Vellutino et al., 2004).

There are two theoretically driven approaches to teach children to read in English. The first approach is often referred to as the whole-to-part or whole-word approach. The whole-word approach teaches children to copy and store the orthographic images of words and to memorize the pronunciation of words without considering the letter-sound correspondences (i.e., the alphabetic principle) (Duwenhoegger, 2021; Synder & Golightly,). Proponents of such an approach believe that reading is a natural process that develops in a similar way to listening and speaking, and that children can learn the phonological components of words and letter-to-sound correspondence implicitly from the words they speak and listen without being exposed to explicit instructions (Christensen & Bowey, 2005; Connelly et al., 2001). The whole word is challenged because there is nothing natural about the alphabetic codes invented to represent language (McGuinness, 2004). Since the whole word approach does not encourage an explicit teaching of phonological awareness and alphabetic principle, empirical evidence (e.g. Foorman, 1995; Jeynes & Littell, 2000) has shown that the approach can help children read only words they have been exposed to, but not new or unknown ones.

On the other hand, there is growing evidence that a phonologically based approach such as phonics is more effective in developing children’s reading skills than the whole word approach (Jeynes & Littell, 2000). The phonics approach is also called a systematic and explicit phonologically- based approach that values the phonological and orthographic components of words as its focus teaching word recognition (Honig et al., 2000). Phonics has been described as both “…a system for encoding speech sounds into written symbols … [and a way of] teaching learners the relationships between letters and sounds and how to use this system to recognize words” (Mesmer & Griffith, 2005, p. 366–367).

In this instructional approach children are explicitly exposed to letter-sound and syllable or onset-rime segmentation, blending, manipulating and ways of employing these knowledge and skills to decode words (Maddox & Feng, 2013). The advocates of phonics approach argue that explicit phonics instruction is particularly beneficial in helping children to develop phonological and/or phonemic awareness, which allows children to read both familiar and new or unfamiliar words and increase their sight vocabulary (Ehri et al., 2001; De Graaff et al., 2009; Tunmer & Nicholson, 2011). Once children have developed phonological and/or phonemic awareness, they can easily master the alphabetic principle and are able to begin making orthographic representations of words on their own and make reading development more learners’ dependent over time (Anthony & Francis, 2005; Christensen & Bowey, 2005; Connelly et al., 2001; Share, 2004).

Teaching methods in the phonics approach vary according to the level and type of phonological awareness they intend to develop or the size of the functional unit they use while teaching phonological awareness. In this regard, the National Reading Panel has identified two methods of phonics, namely synthetic phonics, which uses the smallest units in words—the phonemes to develop phonological awareness at phoneme level (i.e., phonemic awareness) and analogy or rime based phonics that considers larger units in words -onset-rime or syllable to develop phonological awareness at word, syllable, onset-rime and phoneme levels (Ehri et al., 2001; Ehri & Robbins, 1992; Goswami, 1986; 1988, Goswami, 1991; Goswami & Goswami & Bryant, 1992; Greaney et al., 1997; Wylie & Durrell, 1970).

In recent years there has been an ongoing debate on the effectiveness and choices of the explicit phonics methods -synthetic or analogy? When 2004), explained this issue, they noted that “Indeed it is arguable that the important questions about phonologically- based interventions are now no longer “do they work?” but rather, “which are the optimally efficacious interventions?” and “how do they operate?” “(p.212). The debate over the choice of phonics methods has been further accentuated following the work of Ziegler and Goswami (2005) on psycholinguistics grain size theory. According to this theory, phonological awareness is the foundation upon which phonics is built and the level of phonological awareness may vary with the depth of orthography of the language being learned. In transparent orthographies, where a letter corresponds in principle only to a specific sound unit (phoneme), phonological awareness at phoneme level (i.e., phonemic awareness) may be a more important predictor of reading success, and synthetic phonics method that uses small grain sizes (i.e., graphemes and phonemes) can be an effective method of teaching. Once children have memorized and stored the grapheme to phoneme connections, they will be able to decode or pronounce any words in that language, as well as invented, meaningless words (i.e., pseudo or non-words) (Ziegler & Goswami,2005). However, in less transparent orthographies like English, it is an onset- rime unit that is highly consistent in representing sounds than a letter does. Due to this reason, English reading acquisition and development is predicted by phonological awareness both at rime and phoneme levels (Ziegler & Goswami, 2005). Exposing children to instructions that teach phonological awareness at rime and phoneme levels therefore facilitate the acquisition and development of children’s English reading skills.

Rime-based phonics instruction teaches children to identify, analyze and manipulate the phonological structures of rime units and the rime families of words (Johnston, 1999). Thus the method helps children to become aware of rhyme phonological units and rime spelling patterns in words, and this orthographic insight in turn helps to develop phonemic knowledge of all the constituent phonemes in a given word (Booth & Perfetti, 2002; Ehri & Robbins, 1992; Gaskins, 2004; Goswami, 1998; Lovett et al., 2000; Wood, 2000).

Empirical evidence from numerous intervention studies in English- speaking children (e.g. Conrad & Levy, 2011; Evans, 2013; Hines, 2009; Ouellette & Sénéchal, 2008; Sénéchal et al., 2012; Walton et al., 2001; White, 2005) have already demonstrated the functional importance of the rime-based phonics intervention for improving children’s reading performances. In spite of this fact, very little has been known so far about the effects of such an instruction on the reading growth of ESL/EFL children since there is a lack of research in the area. Few studies have been attempted in ESL/EFL settings and indicated the importance of the method. Zuriyatiaslina, et al., (2018) for example investigated whether or not the onset- rime instruction improves phoneme blending and word reading outcomes in Malaysian EFL learners and concluded that the use of the onset -rime instruction significantly improved students’ outcomes in phoneme blending and in new word reading. Similarly, Ng and Yiakoumetti (2010) found in their experiments that a systematic and explicit rime-based phonics intervention program resulted in significant improvements in word recognition outcomes of Hong Kong Chinese-speaking children. Can the results for monolingual English speakers be generalized to EFL children who have very little social and environmental support and exposure to spoken and written English and whose first language (L1) is significantly different from English? It is worth finding the answers to this fundamental question and to minimize the existing knowledge gap in this area. It is hoped that the specific Ethiopian context of the study will help answer the question and provide insights into the role of such an instruction in the acquisition and development of phonological awareness of children in a non-English language speaking context with transparent first language (L1) background.

Therefore, the aim of this study was to examine whether an explicit rime-based phonics training based intervention helps to improve the phonological awareness outcomes of Ethiopian primary EFL learners.

2. Statement of the problem

Since the introduction of modern education in Ethiopia during the early 20th century/in 1908, English has been serving as a language of instruction at different levels of the country’s education system and being taught as a school subject (Pankrest, 1968). The rationale and importance/justification for using English within the Ethiopian education system has been indicated in various national policy and strategy documents, including, among others, the FDRE Education and Training Policy (Ministry of Education MoE, 1994), The New Education Roadmap (MoE, 2020), Curriculum Framework for Ethiopian Education (MoE, 2009), the Education Sector Development Program I-IV (ESDP, 2003–2015), and the General Education Quality Improvement Packages (GEQIP I-III). While the 1994 policy states that “cognizant of the pedagogical advantage of the child in learning in mother tongue (MT) and rights of nationalities to promote the use of their languages, primary education will be given in nationality languages … ” (MOE, 1994, p. 23), the policy made clear that “English will be taught as a subject starting from Grade 1” and “will be the medium of instruction for secondary and higher education” (MoE, 1994, p. 24). Indeed in the new education roadmap, the government defines one of the main learning goals of the country as “multilingual proficiency: every learner will be proficient in the mother tongue, the national communication official language(s), interregional language(s), and English as an international language of communication” (Ministry of Education MoE, 2020, p. 34). As can be gleaned from these and similar other documents, English is and will continue to own an influential role in Ethiopian education system.

Because of the growing role of English in Ethiopia, many parents today prefer to enroll their children in English-medium private schools, particularly in relatively middle- and high-income urban areas (Negede, 2012; Teklemariam, 2006; Tonkyn, 2017). However, the only way for low income parents to give their children an education is through government-sponsored public schools. Most teachers in these schools are unqualified or semi-qualified and use poor teaching methods to teach children to read in English. Children are learning to read in a language that is orthographically different from their first language (L1), may not have been previously familiar with and lack social and environmental support systems (see: Kim et al., 2021; Almaz, 2015; Belilew, 2016; Smith et al., 2012; (2017); Hugh et al., 2007; Fekede & Hailu, 2018; NEAEA, 2000, 2004,2008, 2012b).

Because of this restricted linguistic, social and educational environment, various assessment results have shown that Ethiopian primary school children’s reading performance in English falls below the standards set by the Ministry of Education for each grade level and they are illprepared for the next level of education (secondary and tertiary) in which English is used as a tool to learning other subjects. For example, the first, second, third, fourth, and fifth National Learning Assessments (NLA) conducted by the National Education Assessment and Examinations Agency (NEAEA) in 2000, 2004, 2008, 2012, and 2016, respectively, demonstrated that the reading achievement of primary school children is below average (50%) and steadily declining in Ethiopia. Besides, the Early Grade Reading Assessment (EGRA) revealed that Ethiopian children’s reading achievement and primary schools’ EFL reading instruction in the country is poor (Smith et al., 2012). These results appear to be similar to those found in studies by Almaz (2015); Belilew (2016); Enyew et al. (2015a, b); Fekede and Hailu (2018); Solomon and Chanyalew (2019) and Jemberu, (2012). These studies indicated that primary school children in Ethiopia have clear reading difficulties that require further study and intervention activities ranging from curriculum revision to classroom instruction.

In fact, there are many possible reasons why children have trouble learning to read, write, and spell in their early school years. There are two major categories of causes in this regard. Extrinsic, or experiential, causes include lack of instruction, poor quality instruction, and/or a home environment lacking in literacy cultural capital- LCC (Catts & Kamhi, 2005; Tunmer & Nicholson, 2011; Vellutino et al., 2004). The other main causes are intrinsic or internal causes. These may include genetic/neurological causes (e.g. dyslexia), or language based causes such as problems with the type and nature of the writing system (orthography) the children are acquiring (Lyytinen & Richardson, 2014; Seymour et al., 2003; White, 2005).

Range of research findings (see: Cunningham et al., 2004; White, 2005; Ehri et al., 2001; De Graaff et al., 2009; Vellutino et al., 2004) underscored that adequate and effective reading instruction in the early grade influences the acquisition and development of reading skills more than any other single factor. Studies by Ojanen et al. (2013); Lyytinen and Richardson (2014) revealed that ineffective early-grade reading instruction affects millions of EFL/ESL children in Africa and elsewhere. Similarly, most studies in the Ethiopian context (see: Almaz, 2015; Belilew, 2016; Negede, 2012; Smith et al., 2012) demonstrated that children in Ethiopia are not learning to read well before completing primary school, mainly because of the poor or ineffective teaching methods of the teachers. The teachers were not taught the basic elements of reading and how to teach these skills. Most English teachers in public schools in Ethiopia tend to teach to read in English through memorization or using a method that shares characteristics of the whole word approach. Teachers often begin by teaching the names of the letters of the alphabet rather than teaching the sounds of the letters. Teachers typically point to words written on the blackboard or in student texts and model their pronunciations, which are then recited by the rest of the class. Teachers encourage children to copy the orthographic images of the words and memorize their pronunciation by reading the words repeatedly. Teachers do not explicitly teach phonological or/and phonemic awareness and alphabetic principle- letter/s-sound or/and pattern-sound correspondences (Almaz, 2015; Belilew, 2016; Gove & Cvelich, 2010; Negede, 2012; Smith et al., 2012).

On the bases of the findings of the studies discussed so far and the actual classroom observations, the researchers of the present study believed that one of the causes of the poor English reading performance of primary school children in Ethiopia might be their poor phonological awareness, which are a result of poor instructional practices. According to Gunderson (2014), EFL children who have had limited or no exposure to spoken and written English should be exposed to phonologically based instruction that explicitly teaches phonological and/and phonemic awareness and alphabetic principles in the early grades. Empirical evidence from some intervention studies has shown that an explicit rime-based phonics method promotes phonological awareness and reading growth in young children from diverse language backgrounds and with English as a second or foreign language (Ng & Yiakoumetti, 2010; Zuriyatiaslina, et al., 2018). However, to the best of the researchers’ knowledge, no such intervention study has been conducted in Ethiopia. Given the current research findings and theoretical assumptions, the researchers designed and implemented an explicit rime- based phonics training based intervention and examined its effects on the phonological awareness outcomes of 3rd grade EFL children in public schools in Ethiopia. The following specific research question was addressed.

1. Will children with an explicit rime based phonics method perform better at the posttest than the control group children on phonological awareness tasks?

3. Methods and materials

The main purpose of this study was to determine if there was a significant difference in the phonological awareness outcomes among Grade Three EFL students at the end of twelve weeks who were taught:

1. beginning reading lessons with an explicit rime based phonics method.

2. beginning reading lessons not in explicit rime based method.

3.1. The research design

This study followed a quasi-experimental design with pre- and post-testing. It compared the phonological awareness outcomes of children who were randomly assigned either into experimental or control group.

3.2. The research setting

The intervention study was carried out at two public primary schools, Nigist Fura and Tabor Primary Schools in Hawassa, Ethiopia. The study site is within the region where the highest number of non-word readers (68%) was found in the 2010 National EGRA (Piper, 2010). The participating schools are the typical Ethiopian public primary schools, where the children’s mother tongue is used as the language of instruction and English is taught as a school subject from the first grade. Five 40-minutes English lessons were given by non-native English teachers.

3.3. Participants and sampling technique

The target groups of the study were 3rd grade children studying in two public primary schools in the second semester of the 2021 school year, and their English teachers. Grade 3 was deliberately chosen as the official word study lesson starting with this grade in Ethiopia. All of the children in this study spoke their native language at home. The parents of these children were mostly lower or lower-middle class citizens who would not communicate with their children in English at home. In general, children were only exposed to English at school in English classes. There were 70 students, 44 boys and 26 girls, who took the pretest and started the intervention, and only 42 boys and 25 girls completed the intervention, and their mean age was 10 years. There were two participating teachers, one for each group. The teachers came from different schools. This was done to avoid communication regarding the intervention lesson between the teachers of the experimental and control groups and to minimize the possibility of contamination of the control group. Both teachers had English diplomas and were studying for their B.Ed. degrees in the part time summer program. The mean teaching experience of the teachers was 12 and 13 years, respectively. Due to COVID-19 protocols, each school in the city administration was allowed a maximum of 35 students per class. The researchers used this number as a benchmark to determine the sample size and set it at 70.

A mixed sampling design was used to select participating schools, students and teachers. The participating schools were carefully selected to ensure that they were comparable in terms of students’ socio-economic background, school resources, teaching methods, and English teachers’ qualifications and experience. In order to ensure this, data on the training-based intervention schools were first collected from the city administration’s education department and later confirmed by observations and interviews made with school heads. Finally, based on the data, two schools were selected, Nigist Fura and Tabor primary schools. These schools were comparable in terms of resources and teachers’ profiles, and consisted of large numbers of children from families with low socio-economic backgrounds.

Sample children were selected from each participating school using probability sampling techniques. Sections were used as the unit of randomization and two 3rd grade classes and their English teachers were randomly selected. A pre-test was conducted to examine the initial phonological awareness performances of the children in the two sections. The results indicated that the children’s phonological awareness performances were comparable as there were no statistically significant differences in children’s outcomes across all component sub-tests and the total phonological awareness. Accordingly, the section at Nigist Fura primary school was randomly assigned to the experimental group and the section at Tabor primary school to the control group.

3.4. The training based intervention lessons

The training based intervention was intended to teach phonological knowledge and processing skills at word, onset-rime and phoneme levels based on patterns of English letters such as rimes and word families containing these rime units. The control group was taught the actual word study and vocabulary lessons from the existing Grade Three students’ textbook, while the lessons for the experimental group was designed based on 200 monosyllabic word families with 37 common rime units compiled by Wylie and Durrell (1970). Words with CVC, CCVC, CVCC, and CCVCC structures (e.g. cat, chop, pink, and shock) were selected from existing English textbooks.

The model proposed by White (2005), for designing effective rime-based phonics instruction was adapted and used for instructional design. Accordingly, a total of 36 lessons of 40 minutes each were prepared and used for teaching the children in the experimental group. The goal of these lessons was to help the children develop their phonological awareness at word, rhyme, and phoneme levels by identifying and working with each of the sound units that make up the target rime and keywords in each lesson. The activities in each lesson were aimed at practicing sentence segmentation, identifying and producing rhymes and alliteration of words, segmenting and blending onset and rime in words, segmenting words into phonemes, blending individual phonemes into words, and deleting and substituting phonemes to form new words. The lessons were compiled into a manual and made available to the teacher in the experimental group to use it as a teacher’s guide to the effective delivery of each lesson. The handbook included detailed lesson plans prepared for each lesson, rimes and keywords, lesson objectives, a list of teaching resources, activity descriptions with time breakdowns, and language and phrases used by the teacher.

3.5. Measurements

In this research project, the main data gathering tool was a test. Tests were administered twice, before and after the intervention. A pre-test was given before the intervention began for both the control and the experimental groups to examine the initial phonological awareness performances of the groups. After the children in the experimental group were taught for 12 weeks a post-test was conducted in order to see if there was any significant difference in the phonological awareness outcomes of the control and experimental groups.

A test was developed based on the Phonological Awareness Test (PAT) by Robertson and Salter (Robertson & Salter, 1997) and used to collect data for the study both. The test score was able to provide information about the phonological awareness performance of a given student, as well as the entire group. The test consisted of 11 component subtests. Each component subtest contains seven items and the total number of items was 77. The reliability of the tests was checked using a test-retest reliability check before they were used to collect the data. The types and nature of each component subtest is described below.

(a). Sentence Segmentation

The purpose of this subtest was to evaluate children’s ability to identify the words that make up a sentence. To do this, seven simple sentences with three to five common words were developed and used. The children had to listen to each sentence and say the words that compose the given sentence.

(b). Rhyme Recognition

This subtest aimed to measure children’s ability to recognize rhymes and alliteration of words. Seven rhyming and non-rhyming word pairs were selected from students’ English text books and used for this purpose. The children had to hear and identify pairs of words that rhymed and didn’t rhyme.

(c). Rhyme Production

This subtest aimed to measure children’s ability to generate or produce rhyming words. Seven high-frequency words were selected from children’s English textbooks and used in this test. The children were given a word and asked to say their own word that rhyme with the one given.

(d). Onset-Rime Segmentation

The main purpose of this subtest was to examine the children’s ability to recognize the onset and rime that comprise a particular word. Seven words were selected from children’s English textbooks and read aloud to the children to divide them into onset and rime.

(e). Onset-rime Blending

This subtest aimed to assess the children’s ability to blend the given onset and rime to form words. The children were given seven pairs of common onset and rime and were asked to blend each pair and form a word.

(f). Phoneme Segmentation

In this subtest, the children were assessed on their ability to enunciate the sounds that comprise a simple word. Seven common words were selected from the children’s English text books and used in this test.

(g). Phoneme Blending

This subtest required the children to blend together verbally presented phonemes to form a word.

(h). Phoneme Deletion

This subtest contained two parts. The first one was for deletion of the initial phoneme in the presented words and one for deletion of the final phoneme in the presented words. For example, children were asked “can you say ‘top’ without the/t/?”, and “can you say ‘same’ without the/m/?”

(i.) Phoneme Substitution

This subtest also had two sections, children were asked to swap first the initial phoneme in the target words for a specified alternative, and then to manipulate the final phoneme in the target words for specified alternatives (e.g. “can you say “ball” not with a/b/at the beginning but with a/k/and “can you say ‘park’ with a/t/at the end instead of a/k/?”).

3.6. Procedures of the experiment

The experimental group was explicitly taught beginning reading lessons designed based on 37 dependable rime units and keywords contain these rimes. The intervention lasted 12 weeks, 40 minutes long, three days a week (Monday to Wednesday), with a total teaching time of 24 hours. The intervention was conducted by a 3rd grade English teacher who was trained by the researcher prior to the intervention. The teacher was encouraged to conduct the training mainly in English and to use the children’s native languages when the children could not understand the instructions. The researcher observed the intervention class twice a week and talked to the teacher to evaluate the implementation of the training, provide professional and technical assistance to the teacher, and collect the teacher’s feedback on each of the intervention lessons.

Instruction was conducted through an oral activity-based approach. The teacher first demonstrated the activities, and then conducted activities with the children, and finally made the children do the activities themselves. Various multisensory activities (e.g. games, tongue twisters, and songs) were used to model and practice activities. At the beginning of each session, goals were set and sentences, rimes and keywords from the previous session were analyzed orally. Rimes and session keywords were introduced, their pronunciation modeled, a tongue twister sentence presented and segmented into words, and the word’s onset, rime, and vowel(s) identified to describe the word parts. Also, rhyming words with the keywords were generated and listed, keywords were selected and analyzed with a word analysis template, and spelling of words was practiced with the Elkonin box.

3.7. Reliability and validity

The tests, training and teaching materials were made available to the research project supervisors, an educational psychologist, a primary school English teacher and two experienced EFL teachers from the university. They were informed about the objectives of the study and received background information about the study participants. All expert comments were duly noted and taken into account to improve the tools and materials prior to conducting the study. The consistency of assessors’ ratings (N = 2) was checked through inter-rater reliability check. The data for the inter-rater reliability check were gathered during the assessors’ training. First, the assessors were asked to observe and rate the responses of a colleague the researcher was testing. The researcher’s and assessors’ score sheets were then compared to see how closely they agreed in their assessment of the same observation. There were slight discrepancies between assessments, particularly on Assessor-1. Accordingly, discussions were held and appropriate feedback was given on how the assessment performance of the assessors could be improved. Finally, both assessors received new score sheets and were invited to individually score the researcher’s responses while he was being tested by a peer. The scoring results were then served to compute the agreement of the scores given by each assessor. The researcher then used Pearson’s product-moment correlation coefficient to examine the extent to which the assessors’ ratings correlated and found a very strong correlation between the ratings.

3.8. Fidelity to the intervention program

Treatment adherence was assessed by observing randomly chosen sessions and employing an independent rater observing the sessions with the researcher. A total of 14 sessions were observed. Of these, seven sessions were observed with an independent rater. Researchers-developed checklist was used to monitor and assess sessions. The checklist was developed to measure the extent to which the teacher implemented the session plan. The checklist had two sections. The first section contained a list of 26 main activities for the teacher to undertake in each of the four phases of the intervention lesson. The second section had a yes-no subsection. Observers were expected to tick yes if they observed the teacher’s specific activity and tick no if they did not. The average scores from researchers and independent raters were used to determine teacher’s ability to deliver the training-based intervention lessons according to the session plans. Accordingly, the average overall adherence to treatment was found to be 92% ranging from 85% to 98%.

3.9. Ethical issues

Informed consent to conduct this study was obtained from the city administration education department, and informed consent to participate in the study was obtained from teachers, children, and children’s parents or guardians. Prior to the test, participating children were given orientations about the purpose of the study and the type and procedures of the test. The researchers and the school principal confirmed to the children that the result of the test will not affect their school evaluation and will only be used confidentially for research purposes. The students were also told that if they did not want to take the test they had the absolute right not to take it and could even interrupt during the test. Finally, each child was given an identification code and told to write her/his code and tell it to the assessor before the start of the test. The test was conducted twice: before and after the intervention. It was carried out in one-to-one basis with instructions in Amharic or Sidamuu Affoo, the children’s mother tongue. Both the pre- and posttests were conducted in two separate classrooms (one for each assessor) by university English instructors (N = 2) who received hands-on training from the researcher in administering the tests.

All of the items in the test were administered orally, and the children did not see the copy of the score sheet for the data. A sample item was given before each test to ensure the child understood what was expected of her/him. The child was then given a sentence, word or phoneme and asked to segment, produce, blend, delete or substitute them verbally. The assessor recorded the scores for each item and finally provided the score sheets to the researcher for further statistical analysis.

3.10. Methods of data analysis

One point was given to one item and the maximum score for each subtest was seven and the total score for the test was 77. The scores for 67 students were used for the analysis purpose since the rest of students did not take the post test. The results obtained were analyzed with the SPSS software package version 26. Prior to analyses, the distributions of all component subtests were assessed for normality. The onset-rime segmentation and rhyme production subtests at pre-test had two scores which were outliers and caused a right-skewed distribution. The phoneme segmentation subtest at pre-test had one score which was an outlier and caused a right-skewed distribution. These scores were replaced with a value that was closer to the distribution while retaining the order of values to meet the assumption of normality. The remaining component subtests all produced a normal distribution at both time points. Due this reason, a parametric test was chosen to analyze the results. First, an independent samples t-test was conducted to examine if there were group differences at pre-test or post-test. Effect sizes were also calculated for all subtests and the total test score using Cohen’s d with pooled standard deviation. The criteria as that defined by Cohen (1988) is being used in which d ≥ 0.2 is a small effect, d ≥ 0.5 is a medium effect, and d ≥ 0.8 is a large effect. Then Repeated measures ANOVA was used to compare the effects of time (change from pre-test to post-test), group (experiment versus control), and time* group interaction on the scores (group differences in change). The results are presented below.

4. Results and discussion

The results obtained from pre and post-tests for all component subtests and the total PAT in both groups is presented and discussed in this section.

4.1. Results

4.1.1. Pre-test and post-test group comparisons

The pre-test and post-test means and standard deviations in the two groups, as well as group comparison results for each component subtest and the total PAT are reported in Table 1 below.

Table 1. Descriptive statistics and group comparisons

PA-Subtests	Test category	Experimental M (SD)	Control M(SD)	t	Group effect	Time effect	Interaction effect
Sentence segme.	Pretest	3.20(.71) 5.06(1.02)	3.34(.96) 3.55(1.07)	t(65)=-.70 t(65)=5.83***	F(1,65)=9.70**	F(1,65)=94.88***	F(1,65)=65.09***
	Posttest
Rhyme recognition	Pretest	3.0(.68) 4.63(.96)	2.57(.83) 2.73(.66)	t(65)=1.09 t(65)=9.42***	F(1,65)=42.70***	F(1,65)=49.81***	F(1,65)=61.81***
	posttest
Rhyme production	Pretest	2.54(.61) 4.60(.74)	2.57(.69) 2.38(6.69)	t(65)=-.18 t(65)=13.88***	F(1,65)=78.45***	F(1,65)=73.15***	F(1,65)=103.13***
	Posttest
Onset-rime segme.	Pretest	2.62(.64) 4.57(.70)	2.37(.49) 2.44(6.69)	t(65)=1.87 t(65)=4.17***	F(1,65)=93.45**	F(1,65)=168.71***	F(1,65)=149.43***
	Posttest
Onset-rime blending	Pretest	2.74(.44) 5.30(4.92)	2.54(.61) 2.61(.65)	t(65)=1.56 t(65)=3.10**	F(1,65)=10.35**	F(1,65)=9.94**	F(1,65)=9.07**
	Posttest
Phoneme Segme.	Pretest	3.85(.91) 4.75(1.45)	3.71(1.15) 3.70(.97)	t(65)=.57 t(65)=3.48**	F(1,65)=4.67*	F(1,65)=19.28***	F(1,65)=24.78***
	Posttest
Phoneme blending	Pretest	3.00(.90) 5.39(1.76)	2.97(1.01) 2.94(1.59)	t(65)=.12 t(65)=5.96***	F(1,65)=16.88***	F(1,65)=57.20***	F(1,65)=63.11***
	Posttest
Initial phon. Del	Pretest	2.42(1.83) 4.63(1.72)	1.85(.61) 1.97(1.38)	t(65)=1.43 t(65)=6.90***	F(1,65)=7.64**	F(1,65)=132.01***	F(1,65)=111.77***
	Posttest
Final phon. Del	Pretest	2.42(1.83) 4.63(1.72)	1.85(1.47) 1.97(1.38)	t(65)=1.06 t(65)=6.96***	F(1,65)=18.06***	F(1,65)=121.10***	F(1,65)=103.02***
	Posttest
Initial phon.Sub.	Pretest	2.11(1.81) 4.51(1.82)	1.68(1.49) 1.85(1.47)	t(65)=1.08 t(65)=6.57***	F(1,65)=15.76***	F(1,65)=96.89***	F(1,65)=79.38***
	Posttest
Final phon. Sub	Pretest	2.00(1.32) 4.81(1.82)	2.02(1.46) 2.29(1.58)	t(65)=-0.8 t(65)=6.02***	F(1,65)=12.10**	F(1,65)=121.86***	F(1,65)=86.88***
	Posttest
Total PA	Pretest	29.62(8.61) 52.60(12.71)	28.02(8.24) 28.76(8.21)	t(65)=.79 t(65)=9.14***	F(1,65)=45.25***	F(1,65)=51.69***	F(1,65)=51.69***
	Posttest

Note: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

The results (see Table 1) showed that there were no pre-test significant group differences in the total phonological awareness scores and the scores for each component subtest. At post-test, group differences in favor of the experimental group were significant for all the component subtests and the total PAT score and effect sizes were also large for all component subtests and the total score; sentence segmentation (t(65) = 5.83, p = 0.000, Cohen’s d = 1.45), rhyme recognition (t(65) = 9.42, p = 0.000, Cohen’s d = 1.30), rhyme production (t(65) = 13.88, p = 0.000, Cohen’s d = 1.65), onset-rime segmentation (t(65) = 4.17, p = 0.000, Cohen’s d = 1.50), onset-rime blending (t(65) = 3.5, p = 0.004, Cohen’s d = 0.96), phoneme segmentation (t(65) = 3.48, p = 0.001, Cohen’s d = 0.85), phoneme blending (t(65) = 5.96, p = 0.000, Cohen’s d = 1.45), initial phoneme deletion (t(65) = 6.90, p = 0.000, Cohen’s d = 1.32), final phoneme deletion (t(65) = 6.96, p = 0.000, Cohen’s d = 1.70), initial phoneme substitution (t(65) = 6.57, p = 0.000, Cohen’s d = 1.60), final phoneme substitution (t(65) = 6.02, p = 0.000, Cohen’s d = 1.47) and total phonological awareness score (t(65) = 9.14, p = 0.000, Cohen’s d = 2.18).

4.1.2. Group comparisons of development from pre-test to post-test

The total phonological awareness scores at pretest were 29.62 and 28.02 for the experimental and control groups respectively (see Table 1). By posttest the mean score in the PAT was 52.60 for the experimental group and 28.76 for the control group. A repeated measure ANOVA was carried out on the component subtests and the total score of the PAT. There was a significant main effect of time, a significant main effect for group, as well as a significant interaction effect, for all the component subtests and the total PAT with the experimental group showing significantly higher scores and faster development than the control group over time.

4.2. Discussion

The present study examined the effects of 12-week basic reading skills instruction with an explicit rime-based phonics method on the phonological awareness outcomes of Ethiopian primary school EFL students.

The results of the data analysis led to the following main findings. First, instruction was found to facilitate the acquisition and development of phonological awareness in the experimental group. It should be noted that in the control group there was very little emphasis on reading words by identifying and manipulating the phonological and orthographic components of words, instead the emphasis was more on learning to read words by copying the orthographic image of whole word and memorizing its pronunciation and relying on letter cues, which is the typical beginning English reading instruction in Ethiopia.

It should be underscored that the extent of growth in the experimental group was the result of the intervention, which lasted two hours per week for 12 weeks. The control group students showed very little progress on various phonological awareness tasks and almost none on phonemic awareness. This suggests that existing primary school beggining reading instruction in Ethiopia, with its emphasis on exposing students to copying orthographic images and memorizing print pronunciations and letter names, may not allow children to develop adequate phonological awareness to support English word reading. According to Huo and Wang (2017) most EFL children have low phonological and/or phonemic awareness and delayed word recognition development compared to English speaking children. Therefore, these children need to be explicitly taught the phonological structures of words and given opportunities to practice manipulating the structures and decoding unfamiliar words (Tan et al., 2007).

An important implication of the present study is that young Ethiopian EFL children are able to learn phonological awareness quickly through a brief and direct beginning reading instruction with an explicit rime based phonics method. The current findings are generally consistent with past studies of rime based phonics instruction delivered to ESL/EFL children (e.g., Ng & Yiakoumetti, 2010; Zuriyatiaslina et al., 2018) to enhance beginning word reading skills. Given that this was, to the best of the researcher’s knowledge, the first training based intervention study to target primary school EFL children in Ethiopia and the findings contribute to the empirical literature that beginning reading instruction with an explicit rime based phonics method and provided by a class teacher in primary school settings is beneficial to the acquisition and development of EF phonological awareness in students from non- English speaking context.

Another important finding from this study is that English reading instruction emphasizes the teaching of letter names and whole word recognition is helping little to improve students’ phonological awareness and participants in this research generally had poor phonological awareness, in particular phoneme awareness. This is in contrast to the view that knowledge of letters alone is a sufficient condition for phoneme awareness to develop (Castle & Coltheart, 2004). Rather, the current findings support that both phonological awareness instruction and the development of letter knowledge promote reading ability (Caravolas et al., 2001).

The findings support the idea that teachers’ training experiences influence teachers’ classroom instructional practices and enhance the course of students’ reading development (Seymour et al., 2003; Vellutino et al., 2004). During the classroom observation it was noticed that the control group teacher encouraged students to copy the orthographic image of words and memorize the words’ pronunciation through repetition. The teacher frequently reminded students to remember the names of the letters from the English alphabet and motivated them to guess the pronunciation of new words relying on letter cues. The teacher did not teach phonological awareness and orthographic skills to facilitate students’ word reading. This is a typical method of teaching reading in transparent Ethiopian languages. Different local studies (see: Almaz, 2015; Belilew, 2016; Negede, 2012; Endalkachew, 2017) revealed that most teachers in Ethiopia were not taught the basic elements of reading and how to teach these skills to students. Teachers in these studies reported not having any in-service and pre-service training in teaching children how to read in English. The experimental group teacher, in this study received an intensive training on explicit rime based phonics method and got school based professional support from the researcher. This helped the teacher change her instructional practices and improve students’ phonological awareness outcomes.

5. Conclusions

In summary, the results demonstrated that the experimental group and the control group were at similar levels of phonological awareness performances at the pretest phase of the intervention. Further, the results showed that the experimental group made greater gains over the course of the intervention programme than the control group in all component subtests and total PAT.

Overall, the results indicated that a twelve-week explicit rime-based phonics training based intervention, can accelerate the development of phonological awareness of Ethiopian primary school EFL children to a greater degree than the regular classroom beginning reading instruction. Furthermore, the result also provided preliminary evidence that a beginning reading instruction with an explicit rime-based phonics method can promote the phonological awareness outcomes of EFL children with from non-English speaking context whose first language (L1) is significantly different from English. We hope that future studies with larger sample sizes, longer intervention periods, standard measurements, and more effective interventions will provide a solid generalization of the effectiveness of such an instruction.

Availability of data and materials

The data and materials for the current study are available from the corresponding author on request.

Authors’ contributions

All authors contributed to this study. Material preparation, data collection and analysis were performed by Abiy Zewdu. The first draft of the manuscript was also written by Abiy Zewdu and then commented and improved by all authors.

Acknowledgments

The authors acknowledge Hawassa University for sponsoring this research project and the Hawassa City Government Education Department for allowing the schools to participate in the study, principals and participating teachers and children for their willingness to participate in this challenging experimental research project.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The study was supported by Hawassa University

Notes on contributors

Abiy Zewdu Agegnehu

Abiy ZewduAgegnehu is a Lecturer in the Department of EnglishLanguage and Literature at Hawassa University. He received his BEd degree in English from Dilla University and his MA degree in TEFL from Addis Ababa University. He is currently pursuing his PhD in ELT at Hawassa University. His main research interests include early childhood education, development of child literacy skills, improving reading performance in poor and disabled readers. He has published his research in international journals such as the Journal of Language Teaching and Research, The African Journal of Teacher Education and The International Journal of Research in Teacher Education.

Mebratu Mulatu Bachore

Mebratu Mulatu Bachore graduated from Addis Ababa University with a B.A. in English Language and Literature, MA and PhD in TEFL. He is currently Associate Professor of TEFL at Hawassa University. He has been teaching various undergraduate and graduate courses, and supervising masters and doctoral students at various universities in Ethiopia. He has more than 18 publications in renowned international journals.

Zeleke Arficho Ayele

Zeleke Arficho Ayele received his B.Ed. Degree in English from Debub University and his MA and PhD in TEFL from Addis Ababa University. He taught English in high schools. He has taught English at Hawassa University for over a decade. He also served as Director for Academic Affairs Directorate at Hawassa University. He is an Associate Professor of TEFL and currently Dean of the College of Social Sciences and Humanities at Hawassa University. He has published over 25 original research articles in prestigious peer-reviewed journals.