THE LOW EFL READING LEVEL OF GRADE FOUR STUDENTS AND REASONS: WITH PARTICULAR REFERENCE TO STUDENTS IN ABOMSA

The purpose of this study was to explore the EFL reading ability of fourth graders of students in one selected school in Ethiopia. To this end, selected with purposive sampling technique, 118 grade 4 students of Abomsa Primary school in Ethiopia were tested the EFL oral reading test. The results of the data analyzed through frequency and percent, show that the grade four students in the mentioned school were poor readers in English. They were able to name the upper and lower case letters of English language and decode one syllable words with three or four letters. Moreover, the study found the three confusing areas of English letters for these students such as letter shape, letter-sound name and letter similarity with L1 script. Recommendations and implications were provided according to the findings of the study.


681
of Ethiopia. This is because, Ethiopian students are supposed to speak or listen rarely in the English language due to lack of exposure outside classroom (Jha, 2013) even in the classrooms; whereas, they are required to read more because the exposure is ample in both situations. Therefore, in Ethiopia for learning the English language as a subject and other school subjects using as an MOI both at lower and higher grades, the skills of reading is supposed to be the cornerstone (Edosa, 2007).
Regardless of its importance, students in Ethiopia are very poor in their EFL read ing achievement though it often serves as a predictor of future academic success. Research shows that, though Ethiopian students begin learning to read since grade one (Heugh, et al, 2006), most of them are able to read below the expected level not only in the English (Eba, 2014;EGRA , 2012;Michael, 2003), but also in their L1 (EGRA Ethiopia, 2010) even when they complete primary school. The reading achievement results of grade 4 students of Ethiopia within the three National Learn ing Assessment years 2004, 2008 and 2010 were 64.5%, 43.9% and 43.3% respectively (MOE, 2004;2008;2010). The imp lication of this result is the declining of the reading ability of primary school students in the country. Moreover, in Ethiop ia, the cause of grade repetit ion and drop out of students is inability of reading (EGRA Ethiopia, 2010); this in turn resulted in the inability of learn ing both the school subjects and the language itself.
The assessment made by Early Grade Reading Assessment (Ethiopia EGRA, 2010) on the six Ethiopian languages reading skills of g rade 2 and 3 students shows that most students read below the expected level. Based on this alarming finding, grades 2, 3 and 4 students were assessed their English lang uage reading skills by Early Grade Reading Assessment (EGRA Eth iopia, 2012), and the finding indicates that these students were below the standard made by MOE. Using the recommendations made by the two assessments periods, even though the Ethiopian government made a number of attempts how students were taught reading skills of both L1 and English language the problem is still there (Eba, 2014; Gemech is, 2014).In general, the teaching of English language in Ethiopia is daunting (Dereje, 2012). Therefore, conducting research focusing on the EFL reading skills of primary school students is crucially required. The present research sought to answer the following research questions. 1. Do grade 4 students of the participants of the study name all English language letters and sounds? 2. What are the confusing letters of English language for the participants of the study and what are the reasons behind? 3. What types of sight words the participants of the study were able to decode or unable to decode and what were the reasons behind?

Research Methodolog y Populati on, Samples and Sampling Techni ques
The main purpose of this article was to explore the EFL reading skills level of grade four students of Abomsa number one primary school in Eth iopia. Thus, out of 170 students of grade four students who were enrolled in 2016/2017 academic year in the school, 118 (64 male (54.2%) and 54 female (45.8 %) were selected as the participants of the study purposely-leaving mediu m ach ievers but selecting low and high achievers.

Ethical Issue and Data collecting procedures
Informing school director and the homeroom teachers as some of grade four students in the school were going to be selected to participate in the research project, the researcher invited students to be assessed their English language oral reading ability voluntary. It was not required to make parents sign for permission, but through the students, they showed their agreement by sending the children to school against shift for the sake of assessment. Therefore, the students took the test fro m November 15 to 30/ 2016.

Data Anal ysis Procedures
The collected data through oral reading test were corrected and marked by the researcher and the assistant. Due to the objective nature of the test, it didn't need to be marked by two expertise or raters. The obtained res ults were fed into SPSS package 20 and computed applying frequency and percent and displayed using tables and bar -graphs. Each letter and word response of the students were tallied and counted to explore wh ich letters created confusion to name them when reading and to investigate the magnitude of the words being read and misread.
To determine the overall sight word decoding ability level of the part icipants of the study, the benchmark formu lated by EGRA Ethiopia (2012) was applied. Since the number of the words in the test of the current study and the source of the benchmark were different, the following score cut was prepared by the researcher.

Results
Letter Naming Knowledge of the Students As shown in figure 2, most of the students had capital letter naming knowledge. About half 58 (49.1%) o f the students were able to name more than 20 upper case letters of English language when about 24 (20.2%) students were able to say about half of the English language capital letters. This indicates that students were able to say capital letters despite the existence of children who were not able to name most of the letters. As noticed in Figure 2, most of the students performed well on SLNK test too. 21 to 26 small letters were named by 57 (48.4%) of the students when about less than half of small letters (0 -13 letters) were named by 28 (23.5%) students. Therefore, most of the students were able to say both capital and small letters. However, more than half of the students were able say 5 letters wrongly or they did not say.  Figure 3 shows the single English language letter sound naming knowledge of the participants of the study. As shown in the figure, the participants of the study were able to name some of the sounds of single letters of English language. More than one-third 45 (38.1) of the students were able to say only 6 or less sounds of English letters. 68 (57.6%) students read fewer than half of the sounds of the letters when 50 (42.1%) read more than half the letter sounds. Therefore, the participants of the study had low knowledge of naming the sounds of single English language letters.

Sight Word Decoding Ability of the Students
In figure 4, it was indicated that students had the problem of reading or decoding sigh t words they were taught in their previous grades. Most of the students 68 (57.2%) were able to read less than twenty words out of fifty -two words. Similarly, mo re than one-third 44 (37.1) of the students were able to read less than ten words. Surprisingly , 98 (83 %) out of (N=118) of the participants of the study were able to read less than thirty words. Similarly, a number of students were able to read less than ten words. As it was displayed in fig ure 5, twelve students read zero word. About one-fourth of the students 29 (24.5%) were ab le to read less than five words. This number of students read less than ten percent of the sight words they were provided to read. Therefore, their reading or decoding ability is very low.

Figure 5:-Nu mber of Students Read < 10 Words
Generally writ ing, the participants of the study were categorized as below basic level in EFL sight word reading skills. As can be shown on figure 6, out of 118 students participated in the study, 88 (74.3%) of them were p laced under this level of reading. About one tenth 11 (9.4%) of the students were placed under the basic level of reading when good number 16 (13.7%) of students were categorized under proficient level of sight wo rd dec oding ability. Finger counting 3 (2.6%) students were labeled as advanced sight word decoding ability.  As indicated in data collection instrument section, the sight words were grouped into two: category one and category two. Each category was also classified as one-syllable word and more than one syllable word. As can be seen in  To sum up, words such as valve, soldier, drive, rice, fisherman, outside, garlic, maize, remember, puppet, hall, line, quick, May, village, know, and lunch are read by less than 20% of the participants of the study; whereas, words such as ox, cat, egg, job, goat, ill, cook, zebra, week, Monday, feel, and tomato are read by most of the students (<50%). Mango, meet, bowl, meat, sell, farmer, poor, and horse are the words read by the students between of 55% and 60%. Help, dress, finish, boxer, year, angry, different, blow, queen, begin, kitchen, barber, clin ic, library and under are words read moderately by the participants of the students, the range of the students in percent is between 61and 79.
In conclusion, the finding reveals that students had better performance on both capital and small letters naming; on the other hand, they had problems of saying the single letter sound and decoding the sight words taken from their previous class of the English language syllabus.

Discussion:-
The purpose of this study was to exp lore the EFL reading ability level of fourth graders of students in Abomsa Primary schoolin Ethiopia. Having the ability to differentiating English language letters by names, shapes , and linguistic functions help children develop their reading skills performance (Wood and McLemo re, 2011).
The finding of this study evidenced that the participants of the study were weak in their EFL reading skills performance. They were able to read the kindergarten level and below (M ichael, 2003), that is saying letter names and single syllable words.
The huge numbers of the students have the problems of differentiat ing some of the English letters. Research shows that having letter name knowledge at preschool level supports students in their reading achievement when they are at school (Foulin, 2005). Its knowledge also helps students to learn easily letter sounds (Treiman et al., 2008). Students who were taught letter sound name identification skills perform better in decoding words than students who were taught letter name identification skills, the knowledge and the skills of both is imperat ive, however. In the following consecutive sections letter name identification, letter sound recognition and d ecoding sight word were discussed in detail. 686

Letter Name Knowledge
In order to learn reading skills successfully, the knowledge of alphabet is crucially demanding ( Ehri, 2003). The finding of this study reveals that most of the participants of the study were able to name both upper and lower case letters. However, the problem dwelt to identify ing the name of some letters with confusing sounds (Block and Duke, 2015), similar shapes, and similar name with L1 letters.

Letters with Similar Sounds
Except some of English language letters, most of them are created in a combination of different sounds of the language. Some letters are created by combination of vowel sound and consonant sound, some others by consonant sound and the vowel sound; and still, some others in different structures. According to Piasta & Wagner (2010), the structure of the English language letters is grouped into consonant -vowel (CV), vowel consonant (VC) and Letters with no association (NA). Having the knowledge of letter-sound name is considered to be good for students to easily recognize the consonant letters with consonant-vowel (CV) structure than for those with vowel-consonant (VC) structure and with no relat ion to their sounds (Evans et al, 2003) There are ten English letters with the first comb ination. In this structure, ten consonant sounds (/b/, /s/, /d/, /g/, /p/, /t/, /v/) and three vowel sounds (/i/, /e/ and /u/) are constituted. Using the consonant sounds /b/, /s/, /d/, /g/, /p/, /t/, and /v/, as initial and the vowel sounds /i/ as final, letters B, C, D, G, P, T and V are created. Using consonant sounds /j/ and /k/ as initial and vowel sound /e/ as final letters J and K are created; whereas, letter U is the only one which is created by the combination of consonant sound /y/ and vowel sound /u/.
In the second combination-vowel and consonant (VC), some English language letters are created. In such kind of structure two vowel sounds (/e/ and /a/) and eight consonant sounds /f/, /ch/, /l/, /m/, /n/, /s/ /y/ and /r/ are used to create eight English language letters. With the combination of vowel sound /e/ and consonant sounds /f/, /ch/, /l/, /m/, /n/, /s/, letters F, H, L, M, N and S are created when letters I and R are created by combination of vowel sound /a/ and consonant sounds /y/ and /r/ respectively. Some letters have different structures which are named by Piasta and Wagner (2010) as letters without the above two combinations in the researchers' term letters with ‗no association'; however, they have their own association. Some letters such as Y and Z are created using CVC structure /way/ and /zed/. With VCC structure letter X /eks/, and letter W /double-u/ with the init ial sound of /d/ and one vowel sound /e/ and two consonant sounds /bl/ in the middle and one vowel /u/ at the end are created. The rest three letters A, E and O are created differently -each was created using only one vowel sound. Hence, the similarity in sound among lett ers can be the inception of confusion for EFL learners. As a study conducted in the first language context indicates, preschoolers are able to say the letters with the first category (e.g. b) better than the rest category letters (e.g. m) (Treiman et al., 2008). Even though it is difficult teaching children considering such kind of relationship among the letters of English, it would be effective if they are taught by simp lifying based on their age and literacy level. According to Block and Du ke (2015), teachers should consider the confusing nature of English letters, their orthographical complexity and their varied representation of letters to sounds and vice versa when teaching English letter names in relation to their sound names.

Letter Shape Similarity
Some English language letters are also made of similar shapes despite their position difference. Due to their similarity in shape, lower case letters -b‖ and -d‖; as well as, lower case letters -p‖ and -q‖ are the confusing letters of the English language. In this study, most of the students in both groups named these lower case letters interchangeably. Some of the students said letter -b‖ in place of letter -d‖ and letter -p‖ instead of letter -q‖ and vice-versa. Moreover, upper case letter -I‖ was difficult to be differentiated with the lower case letter -l‖.

Mother Tongue Letter name knowledge Interference
Another cause of the deficit of student letter naming knowledge is the interference of script or letter type the students use in their mother tongue. In the Ethiopian educational context, two types of scripts are used for writing and reading-the Saba script and the Latin script. So me of the languages are applying the former while others are using the latter one. For examp le, in the place where this study was taken place both of the scripts are used even in the same school. The participants of this study were fro m both scripts. Though the script of the Latin is similar to English language, it has both negative and positive impact on the studen ts' literacy skills development. Students need to differentiate the two names of one thing as the same time. The same is happening on students of Afan

687
Oro mo mediu m. When they are in Afan Oromo class they are required to name letter -S‖ as /sa/, letter -K‖ as /ka/, letter -C‖ as /cha/, letter -X‖ as -ta‖; however, when he/she is in English classroom, she is required to name the letters as /es/, /ke/ /si/, /eks/ respectively.
Hence, this study divulges that students with the medium of Afan Oro mo were confused about the English language letters due to the influence of the script they use in their L1 literacy skills. Most of these students were in confusion while naming the English language letters such as -e‖ and -i‖. This is because in Afan Oromo, the letter -I‖ is read as the letter -E‖, and the letter -E‖ is read like the letter -A‖. Therefore, most of Afan Oromo students have the limitat ion of naming such English language letters. Beginning naming the letters of the English language correctly; for examp le, -B‖ as /bi/, -F‖ as /ef/, -G‖ as /ji/ …-O‖ as /o/… most of the time, they shift the next letters to Afan Oro mo script such as -R‖ as /ra/ -S‖ as /sa/ instead of saying /ar/ and /es/ respectively. Moreover, the finding of the current study reveals that students have problems of distinguishing letter -W‖ from letter -Y‖. About 70% of the students were not able to say these letters correctly. These amount of students read letter -W‖ as it was letter -Y‖. Moreover, letter -U‖ and letter -W‖ are also read in place of letter -Y‖. Therefore, these letters are found to be confusing for the students. The result of the study theorized that students read letter -W‖ as /way/ due to the /w/ initial sound of letter -Y‖. According to Block and Duke (2015) when children are writing letters, they sometimes write letter -w‖ as letter -y‖; they sometimes write YAT for the word -wait‖ and YD for the word -wide‖. Moreover, Block and Du ke indicate that instead of writ ing letter w children wrote the letter d. However, in this study, no students named these letters in place of the other.
Second, students read the letter -G‖ and letter -J‖ interchangeably. In the study, 30% of the students were not a ble to read these letters. Most of the students read letter -J‖ as letter -G‖. Of these students when two -thirds of them were with Afan Oro mo L1 (one of the Ethiopian reg ional state language, Oro mia) and one-third were with Amharic L1 (the working language of Ethiopian governement). Th is shows that the latter students performed better than the former students by naming these letters. Perhaps, this is occurred due to the support of L1 letters; in the former language, these letters are represented differently with a little shape difference as for example (G=ጂ , J=ጄ), but in the latter language these two letters have totally different names and sounds and shapes; for examp le, in English, when letter G has different sounds in the words (grand and change), letter J has only one sound in all words of the language. However, in A fan Oro mo language both of the letters have single sound each in the words of the language they are used. When the former letter is read as /ga/ like in English word -gar‖ and the latter letter is pronounced as /ja/ like in Eng lish word -jar‖.
In short, the study discovered the following three problemat ic areas of naming the English language lettersinability to identify letters with similar sounds but with different shapes (J and G), with similar shapes but different positions (p and q), and with similar names with L1 letters (i and e). Nevertheless, research declares the importance of identifying the letters by their shapes, by their names and by their sound names in order to acquire literacy skills successfully (Wood and McLemore, 2011).
Successful readers have the ability of looking at all the letters and the words and recognize them when they are reading. In other words, according to Rosenberg, the effective knowledge of letters creates the beneficial mental environment for reading words.
However, the students with this language background performed better on LSNK and decoding sight words than students with Amharic language background. In contrast, students with the latter language background performed better than the students with the former language background in both upper and lower case letters of English language. Therefore, L1 had both positive and negative impact on students' reading ability of English language.

Word Decoding Ability
The decoding ability of familiar words is important for the future success of students reading skills development. If students are able to decode words by using different strategies such as recalling the sound of each letter applying their knowledge of letter-sound correspondence and blending those sounds into words (Ehri, 2003), they are able to read unfamiliar words too. As theorized by Horner(2004), letter name knowledge has positive consequence on the word decoding ability of the students due to two reasons. First, some of the English language words have the pronunciation of the letter names as (e.g. A as in the word ape, B as in the word big, D as in the word dig). Second, the letter names and the sounds of the letters have the logical connectio n between them. Cited in Ho rner (2004), 688 Durrell (1980) discloses that 22 of the English language letters have representation sounds in them. However, the students in this study, although they have the knowledge of naming letters, they were not able to decode most of the words they were provided with.
The participants of the study were weak in word decoding ability. The current study shows that the words such as ox, cat, egg, job, goat, ill, cook, zebra, week, Monday, feel and to mato were read most frequently compared to others by the participants of the study. Most of these words are single syllable and three or four letter words. The finding suggested that students were able to read one-syllable words with only three or four letters. Moreover, they were able to read words with two or more -syllable if they are everyday words like Monday and tomato. They were able to read such kinds of words not by using their linguistic knowledge but by memo rizing them (Ehri, 2005).
On the other hand, the words -valve, soldier, drive, rice, fisherman, outside, garlic, maize, remember, puppet , hall, line, quick, May, village, know, and lunch‖ were decoded less frequently than the rest words in the test. Less than 20% o f the participants of the study were able to read these words. This indicated that students do not have the understanding ability of the different sounds of English language letters. For illustration the vowel letter -i‖ has the different sounds in different words like rice, sol dier, garlic, and maize. In the first word, the letter -i‖ is pronounced as /ay/, in the second as /à/, in the third as /i/, and in the fourth as /y/. Moreover, in the words puppet, outside and under the letter -u‖ is pronounced differently. The same is true for letter -d‖ in the words dri ve and soldier. Therefore, the students may have difficulty with such words due to the absence of informat ion fro m their teachers and/or the textbooks.

Conclusions and Recommendations:-
The EFL reading skills performance of the participants of the study was considered below basic. Most of the students were not able to read most of the words particularly wo rds with more than one syllable. Even though there were students who were not able to identify upper and lower case letters, their achievement on these skills were better in comparison to word decoding ability. In the study, some confusing letters were found which were difficu lt to be named by the grade four students of Abomsa Number One Primary school in Ethiopia. These are -W‖ versus -Y‖; -W‖ versus -U‖; -W‖ versus -V‖; -G‖ versus -J‖; -p‖ versus -q‖; -d‖ versus -b‖; -I‖ versus -e‖, -l‖ versus -I‖. The confusion of these paired letters are sourced from three features of the English alphabet such as having similar sounds, having similar shapes and having similar names in the first language scripts.
Students were able to decode few of the words in the samples taken fro m their previous class. The most frequently decoded words are ox, cat, egg, job, ill and cook. In addition, words such as Monday and tomato were decoded by most of the students. Valve, soldier, drive, rice, f isherman, outside, garlic, puppet and lunch are some of the less frequently read words by the students. Therefore, it can be concluded that students were able to read one syllable words with two or three letters. The multiple sounds of letters (e.g. the so und of letter -I‖ in rice, garlic and soldier; the sound of letter -d‖ in drive and soldier; the sound of letter -u‖ in puppet and under) may affect on the decoding ability of the students.
The finding of this study imp lies that the textbook writers shou ld consider the different sounds of letters in the words while preparing books. EFL Primary school teachers should be trained such kind of reading skills in order to solve the mentioned problems in the find ing. Moreover, EFL teachers who teach students wit h a language in Latin script should show the relationship the letters have in the English langaug and the first language (e.g. Afan Oro mo). Since the nature of English language writ ing system is opaque, it is difficu lt for both teachers and students to teach and to learn. Ch ildren can easily become confused and fail to learn to read. Hence, in EFL teacher education curriculu m, the complex nature of English language should be included and a thorough training should be given for teachers.
Since data were collected fro m only one school only using reading test, the study is not out of limitation. Thus, the finding of the study indicates merely the difficulty area of the read ing failure of the students and minor reasons; it couldn't find the root causes of their underachievement in reading of the English language. Therefore, other studies should be conducted on the nature of the textbooks of grade two and three, the roles of teachers, the roles of parents, the preschool experience of the students and their family educational background.