Readability characteristics of texts in middle school Turkish textbooks

Turkish textbooks are important tools for gaining skills and competencies identified in curriculum. Textbooks should be designed to contribute to students’ basic language skills as well as high level thinking skills in alignment with curriculum. The texts to be selected for textbooks should be prepared in alignment with the targeted skills and learning outcomes in terms of visuals and content. Textbooks kids encounter during elementary education play an important role in their upbringing as individuals with reading habits in later phases of their education. Therefore, texts to be included in textbooks should be selected carefully to reflect the best examples of text types, to be compatible with the child’s language and meaning universe, and should be understandable. In this study, the readibility levels of texts in Turkish course textbooks were analyzed. Readibility levels reveal the comprehensibleness of texts. Readability depends on the number of words in a sentence and syllables in a word. First the average word and sentence length were calculated and then the readibility scores of texts were calculated. Two formulas adapted to Turkish were used in identifying the readibility levels of texts. The findings were analyzed comparatively. The results showed that the readability levels of narrative texts are easier compared to informing texts.


Introduction
The purpose of Turkish education is to improve communication skills of students and to internalize the language use rathen than the language within the scope of comprehension and narrative skills so that they can contribute in all kinds of written and oral communication environments competently (Dilidüzgün, 2017). In alignment with this goal, the purpose if to improve reading, listening, talking, and writing skills through learning outcomes and activities in educational environment.
Various materials are used in gaining basic language skills in Turkish education. Textbooks are among these materials as an essential component of education (Cemiloğlu, 2015, p. 221).
However, one of the most important issues of Turkish education is to select a right and effective text, and to teach the text in class effectively and efficiently (Yalçın, 2018, p. 164). Without a doubt, texts presented to students in Turkish education are of high importance. In order to like a language, the need to learn it and the services provided in teaching it are very important (Baki & Karakuş, 2012, p. 6).
One of the fundamental conditions of raising language awareness and sensitivity in students it to provide them with dfferent kinds of literary texts that are compatible with students' levels, prepared with an artistic sensitivity. Thus, the reflection, correct and effective use of Turkish word existence in texts should be adopted as a main principle. Texts should be able to give students the responsibility of hearing and thinking rather than the purpose of teaching (Sever, 2013, p. 22).
In selecting texts for Turkish textbooks, not only children's reality, but also literariness criteria should be considered. This is because regardless how much a text reflects children's reality, if children can't participate in the fictional world created in the text, or can't find appropriate clues for the meaning of the text, and if children pay attention to teachings and tenets rather than anticipation and perception, and can't fill in the semantic gaps between word values, then the text is not compatible for children as it can't meet the literariness criteria (Çer, 2016, p. 1403).
Texts in Turkish textbooks play an important role in developing an interest and habit of reading as well. Reading skills include a multi-dimensional and complex process such as seeing, perceiving, vocalizing, and structuring in the brain (Güneş, 2014, p. 128;Karatay, 2014, p. 8;Stauffer, 1969, p. 5;Sever, 2004, p. 14;Yalçın, 2018, pp. 119-130). Günay (2013, p. 13) evaluates reading within the frame of interaction with text and considers reading as an intellectual activity performed by readability of Turkish texts and this study still holds an important place. Similarly, the formula developed by Çetinkaya-Uzun (2010) is used by researchers in identifying the readability level of Turkish texts.
The field literature includes studies on readabillity in Turkey. Temur (2002) compared the texts in 5th grade Turkish textbooks with student compositions from the aspect of readability levels.
The calculations done by using formulas developed by Ateşman (1997) did not show a significant difference in the level of readability between textbook texts and student compositions. Çiftçi, Çeçen, & Melanlıoğlu (2007) examined the texts in 6th grade Turkish textbooks by using Ateşman's (1997) formula. Significant differences were found between narrative texts and informative texts in readability levels. While 35% of informative texts were either "difficult" or "very difficult", only 3% of narrative texts were either "difficult" or "very difficult." 62% of narrative texts were "easy" or "very easy" while 18% of informative texts were "easy" or "very easy." Zorbaz (2007) studied the readability of texts and the change of word or sentence lengths in tales in Turkish textbooks according to grades by using the Ateşman formula. The results of t-test performed showed a significant difference only between grades in sentence length and readability. Çetinkaya's (2010) study focused on identifying the relationships between the readability levels of texts and linguistic characteristics such as word difficulty and syntactic complexity. With this purpose, participants of the study engaged in an inductive reading activity and the relationship between inductive score and linguistic characteristics of the texts was analyzed. Statistical analysis revealed the readability score of Turkish texts and with the regression formula to determine the readability score of Turkish texts, the Çetinkaya-Uzun readability formula was developed which is an important scale to identify the structural difficulty of reading materials to be distributed to students in class and to match the readability level of the text with the reading level of targeted reader group.
Okur & Arı (2013) examined texts in 15 different Turkish texts books between 2010 and 2011 by using Ateşman (1997) and Çetinkaya-Uzun (2010) readability formula. The results showed that the narrative texts were easier than informative texts and as the grade level increased the readability level of texts increased. In a study focusing on readability of texts in 8th grade Turkish textbooks, Bağcı & Ünal (2013) found that the the difficulty level of texts was medium. In a study conducted by Durukan (2014) to identify the relationship between readability levels of texts in Turkish textbooks and students' reading speed and comprehension levels of these texts revealed a significant difference between readability levels and student's reading speed and comprehension levels. Baş & İnan Yıldız (2015) focused on the readability of texts in 2nd grade Turkish textbooks and found that the readability scores of narrative texts were higher than informative texts.
A study by Iskender (2013) focused on identifying the relationship between the word and sentence structures of texts in Turkish textbooks and readability levels. In order to identify the readability levels of texts, Ateşman's (1997) formula wsa used. The results showed that texts in 5th, 6th, 7th, and 8th grade Turkish textbooks have an average readability level of medium difficulty.
There was no significant difference between the class levels and word and sentence length of texts in Turkish textbooks. The average length of word and sentence do not significantly differ from 5th grade through 8th. Bolat (2016) examined the readability levels of texts in 5th-8th grade Turkish textbooks based on length of worda and the frequency of word use in texts. By using cloze readability test in texts, the readability of texts were identified based on text type, length of word and sentences, and grade and gender variables. The results indicated that 5th-8th grade students read narrative texts more easily than informative texts. When examined according to the grade levels, the texts in the 6th grade textbooks were found to be easier to read than other textbooks. A significant difference was found between the word frequency and readability level. Also, the texts consist of common words were easier to comprehend. Another finding was that when the sentence length of texts are more than 12 words, the readability score was significantly lower which indicated that the average sentence length was the variable that impacts the readability of texts the most.
When the studies in the literature were examined, it was found that the majority of the studies focused on different levels of textbooks and students. In today's information and technology world, Turkish curricula are being updated and improved in order to keep up with change and innovation. With this update and improvement, the contents of Turkish textbooks are being improved and updated as well. Within this scope, the current study aims to identify the readability levels of middle school Turkish textbooks prepared in alignment with the updated 2018 Turkish Course Curricula. This study is significant as it focuses on evaluating texts in Turkish textbooks that were prepared in alignment with the new curriculum, and provides an opportunity to compare with previous studies.
identifying readability characteristics of texts in middle school Turkish textbooks, this study is guided by the following research questions: 1. What is the average difficulty level of texts in 5th, 6th, 7th, and 8th grade Turkish textbooks?
2. What is the readability level of narrative texts in 5th, 6th, 7th, and 8th grade Turkish textbooks?
3. What is the readability level of informative texts in 5th, 6th, 7th, and 8th grade Turkish textbooks?

4.
What is the average length of word and sentence in texts in 5th, 6th, 7th, and 8th grade Turkish textbooks?

Methodology
The current study with the purpose of identifiying readability levels of texts in middle school Turkish textbooks is a descriptive survey research. "Descriptive studies describe a given situation as accurate and comprehensive as possible" (Büyüköztürk, Kılıç Çakmak, Akgün, Karadeniz & Demirel, 2015, p. 22).
This study is limited to informative and narrative texts in the Turkish textbook that was used during the 2018-2019 academic year. Listening/watching texts that are in themes and free reading texts were not included in the study. As readability formulas are based on syllable, word, and sentence count, comprehensiveness was not included in the scope of the study. There are differences in classification of text types in the literature. In the current study, the classification listed in the Turkish Course Curriculum for the year of 2018 was used in identifying the text types.

Data Collection and Analysis
Two formulas used in this study to identify the readability levels of texts in Turkish text books (Ateşman, 1997;Çetinkaya-Uzun, 2010). The researcher identified the numbers of syllables, words, and sentences of texts. Every component in between two gaps was considered as a word. Each unit that is thought to be independent of other sentence or objects grammatically was considered as a sentence. Period (.), question mark (?), colon (:), and two parantheses ( ) were considered as a finished sentence. Syllables were counted as they are phonated. Symbols and shapes were counted as they are phonated. Then, the average word length was calculated by dividing the total number of syllables by total number of words (X) and the average sentence length was calculated by dividing the total number of words by total number of sentences (Çetinkaya & Uzun, 2014, p. 148). Following the calculations of average word length and average sentence length, data were placed in the formula to calculate the readability scores. The formulas used in this study are explained in the next section.
Ateşman formula is applied as follows: RS= 198,175 syllables while the sentence length is 9-10 sentences (as cited in Ateşman, 1997, p. 73). The characteristics of the most difficult and the easiest texts in Ateşman's (1997)   The Easiest Text 2,2 4 The Most Difficult Text 3,0 30

Findings and Discussion
In this section, the findings related to the readability levels, average word and sentence lengths and sub-types of narrative and informative texts in Turkish textbooks are presented. 1 represents narrative texts and 2 represents informative texts in the tables provided. ASL= average sentence length, A RS = Ateşman readability score, ÇU RS= Çetinkaya-Uzun readability score and the findings listed in the level column represent readability scores according to two formulas (Ateşman, Çetinkaya-Uzun) . The text titled "Karagöz & Hacivat-Kindness" is at the very easy level while texts titled "Toy", "I was a Plane Tree", "Is There Anyone who Doesn't Know?," "Dove", "Tip", "Forsa"

Readability Levels of Texts in 5th Grade Turkish Textbooks
and "The Journey of a Snowflake" are at the easy level. According to the Çetinkaya-Uzun formula, the scores range between 41.37 (educational) and 55.05 (independent). The text titled "Karagöz & Hacivat-Kindness" is at the independent level while "Toy"," "I was a Plane Tree," "Is There Anyone who Doesn't Know?", "Dove", "Tip", "Forsa" and "The Journey of a Snowflake" are at the educational level. According to both formulas, it was found that the "Karagöz & Hacivat-Kindness" text was easier than other texts.
According to the findings presented in the (independent). According to the Ateşman formula, texts titled "15th of July", " A Day in Space", "Pasta with Garlic Sauce", "Talking to a Genius", "My Reading Books", "Anatolia's Javelin Game" and "Prescription" are at the level of medium while the texts titled "Tambour Player Mr. Cemil Explains Instruments" and "I'm Living Healthy" are at the level of easy. According to the Çetinkaya-Uzun formula, texts titled " A Day in Space" and "Talking to a Genius" are at the frustration level while "15th of July", "Pasta with Garlic Sauce", "Tambour Player Mr. Cemil Explains Instruments", "My Reading Books", "Anatolia's Javelin Game", "I'm Living Healthy" and Prescription" are at the level of educational. and 99.06 (very easy). The texts titled "Riddle" and "Nasreddin Hodja Jokes" are very easy, "Forsa," "Boiled Seed", "Those who Helped the Gazelle", "An Empty Coop, A Few Full Hearts", "Who Am I?", "Life-Draining", "Energy with Robotics" and "Newton's Apple" are easy, "Atatürk Forest Farm" are at the medium level. According to the Çetinkaya-Uzun formula, readability scores vary between 40.03 (educational) and 56.55 (independent). According to the Çetinkaya-Uzun formula, "Forsa", "Boiled Seed", "Those Who Help The Gazelle", "An Empty Coop, A Few Full Hearts", "Who Am I?"

Readability Levels of Texts in 6th Grade Turkish Textbooks
"Life-Draining", "Energy With Robotics" and "Newton's Apple" were found to be at the educational level. "Nasreddin Hodja Jokes" and "Riddle" texts were the easiest according to both formulas.
The findings presented in the table show that the average word length of informative texts vary between 2.34 and 2.63 and the average sentence length is between 5.50 and 17.82. According to the Ateşman formula, the readability scores of texts vary between 33.95 (difficult) and 71.03 (easy).
According to this formula, "On Smiling Face and Smiling" text is wasy, "Ataturk Had Written A Geometry Book", "Technological Celebrations of Holidays", "Petition Without a Stamp", "Our Planet is Warming", "Far Lands" texts are medium, and the texts titled "Newruz" and "Recycling" are at the difficult level. According to the Çetinkaya-Uzun formula, the readability scores vary between 24.95 (frustration) and 42.27 (educational). According to this formula, "Ataturk Had Written A Geometry Book", "On Smiling Face and Smiling", "Technological Celebrations of Holidays", "Petition Without a Stamp", "Our Planet is Warming" and "Far Lands" are educational, while "Newruz" and Recycling" are at the frustration level. Texts titled "Newruz" and "Recycling" are found to be difficult according to both formulas. According to the Ateşman formula, readability scores of the texts vary between 50.63 (medium) and 82.63 (easy). Texts titled "My Dear Bookshelf", "The Old Grandmother", "Aziz Sancar", "Eating, Drinking and Digesting", " 10 Questions and 10 Answers About Obesity", and "On Friendship" are easy, "This Is My Story", "15th of July", "How did People Use to Measure Time?", "Technology Addiction", "Things We Are Curious About", "Afyon", "Water Pollution", "The Story of Tarhana", "Time of Bicycle" texts are at the medium level. The readability scores according to the Çetinkaya-Uzun formula vary between 29.45 (frustration) and 47.74 (educational). According to this formula, "15h of July", "Technology Addiction" and "Water Pollution" are found to be at the frustration level while others are at the educational level. According to Çetinkaya-Uzun formula, the text titled "Snowman with Green Eyes" is at the independent reading level while others are at the educational level. This text is also at an easier level according to both formulas compared to other texts.

Readability Levels of Texts in 7th Grade Turkish Textbooks
The table shows that the average word length is between 2.41 and 3.00 while average sentence length is between 6.58 and 38.75 in informative texts. According to Ateşman formula, the readability score of informative texts are between -23.62 (very difficult) and 71.28 (easy) while they are in the range between 2.73 (frustration) and 41.19 (educational) according to Çetinkaya-Uzun formula. According to the Ateşman formula, the text titled "Our Responsibilities" is easy, "Life Without Gravity", "The Man Who Makes the World Laugh", "The Personality and Characteristics of Ataturk", "Sufficient and Balanced Eating", "Traditional Hand Crafts Bazaar", "Towards Invention" and "The Media Does Not Consist of Only Television and Newspaper" texts are at the medium level while "The Names of Martyrs of 15th of July Live in Schools" text is at the very difficult level.
According to Çetinkaya-Uzun formula, "Life Without Gravity", "The Man Who Makes the World Laugh", "Our Responsibilities", "Sufficient and Balanced Eating" and "Towards Invention" texts are educational while "Traditional Hand Crafts Bazaar", "The Media Does Not Consist of Only Television and Newspaper", "The Personality and Characteristics of Ataturk", and "The Names of Martyrs of 15th of July Live in Schools" texts are at the frustration level. "The Names of Martyrs of 15th of July Live in Schools" text is a more difficult text according to both formulas.  54.14 (independent). When the Ateşman formula is applied, "Piri Reis" is found to be medium, "Inspiration of Bees", "Munise", "My Left Foot", "I, Mimar Sinan", "The Story of Divanu Lügat-Türk", "An Elephant in Aksehir" texts are found to be easy while "Friendship" text is very easy. 78.57 (easy). "Armies! Your first target", "Ask for Forgiveness from Your Teacher", "The Last Letter of a Martyr", "Reading is a Privilege, Not Everyone Can Read" and "First Snow" texts are easy while "Barış Manço", "The Art of Getting Along with People", "Coffee in the Turkish Cuisine Culture", "New World", "A Day in the Life in 2100", "On Reading" and "Moena, A Turkish Village" texts are at the medium level and the text title "The Village Smelled Like Lavender" is at the difficult level.
According to the Çetinkaya-Uzun formula, "New World" and "The Village Smelled Like Lavender" texts are at the frutration level while other texts are at the educational level. The text "The Village Smelled Like Lavender" is a at the difficult level according to both formulas.  (independent). According to the Ateşman formula, the text titled "Reading is for Thinking" is at the easy level; "Wooden Bicycle", "Pushing Towards Success or Pulling Success?", "Rug in Anatolia", "I had Decided to Become a Soccer Player", "Âşık Veysel Şatıroğlu" and "Dipper Carved Out of Tree" are easy level. "Kaleidoscope", "The Function of Reading", "What is Karagöz", "Two-wheel Freedom", "Be A Friend of Snowy Mountains" and "Ebru: A Traditional Turkish Art" texts are at the medium level while "The Kid Whose Name Is Written in the Sky" text is at the difficult level. The text "Reading is for Thinking" is independent and "The Kid Whose Name Is Written in the Sky" is at the frustration level according to the Çetinkaya-Uzun formula. Other texts are found to be at the educational level. According to both formulas, the text titled "Reading is for Thinking" is at the level easy and the text titled "The Kid Whose Name Is Written in the Sky" is at the difficult level.  is at the frustration level while the others are at the educational level according to the Çetinkaya-Uzun formula.  texts were identified in the 6th grade textbooks while no texts were identified at the very easy or very difficult levels. There were 1very easy level (2.78%), 12 easy (33.33%), 20 medium (55.56%), 2 difficult (5.56%), and 1 very difficult (2.78%) level texts were identified in the 7th grade textbooks.

Findings on the Readability Level of Informative Texts
Among the texts in 8th grade Turkish textbooks, 2 easy (22.22%) and 7 medium (77.78%) level texts were identified. There were no very easy, difficult and very difficult level texts were found. The table shows that only 1 text was very easy and 1 text was very difficult in the 7th grade Turkish textbooks amonf all grade levels. From a readability level perspective, the table shows that the most accumulation is at the level of medium while the least amount of accumulation is at the very easy and very difficult levels. Table 11. Readability levels of informative texts according to the Çetinkaya-Uzun formula Table 11 shows that there are 7 educational (77.78%), 2 frustration levels of texts were identified in the informative texts in the 5th grade Turkish textbooks while 18 educational level (78.26%) and 5 frustration (21.74%) texts were identified in the 6th grade textbooks. Among the 7th grade Turkish textbooks, 1 independent (2.78%), 28 educational (77.78%) and 7 engelli (19.44%) level texts were identified while 8 educational (88.89%) and 1 engelli (11.11%) level of texts were identified in the 8th grade Turkish textbooks. The highest level of accumulation was at the educational level while the lowest accumulation was at the independent level. From an all grade levels perspective, only 1 text was identified as independent level. In table 12, according to the Ateşman formula, 1 very easy (12.50%), 7 easy (87.50%) texts in 5th grade Turkish textbooks were identified while 2 very easy (12.50%), 12 easy (81.25%) and 1 medium level (6.25%) texts were identified in the 6th grade textbooks. Among the 7th grade Turkish textbooks, 1 very easy (4.17%), 21 easy (87.50%), 2 medium (8.33%) level texts were identified while 7 easy (87.50%) and 1 medium (12.50%) level narrative texts were identified in the 8th grade Turkish textbooks. When all levels are considered, there were no texts identified at the difficult or very difficult levels. It's seen that the texts are accumulated at the easy level.      (Ateşman, 1997, p.73). The average word lengths of the texts analyzed in this study are close to this statistical rate. However, it was seen that the average word lengths of some informative texts were above this rate.

General Readability Levels of Texts in Turkish Textbooks
The average sentence lengths of narrative texts analyzed were found to be between 4.15 and 6.79 in 5th grade textbooks, between 3.14 and 8.53 in 6th grade textbooks, between 4.72 and 7.55 in 7th grade textbooks and between 5.71 and 10.30 in 8th grade textbooks. The average sentence lengths of informative texts ranged between 4.07 and 10.71 in 5th grade textbooks, between 3.90 and 17.82 in 6th grade Turkish textbooks, between 4.94 and 26.14 in 7th grade textbooks and between 7.06 and 14.41 in 8th grade textbooks. The average sentence length in Turkish texts is 9-10 words (Ateşman, 1997, p.73). According to the findings, the average sentence lengths of certain informative texts are above this number.
It is seen that the majority of the texts analyzed in Turkish textbooks are at the level of easy.
According to the Ateşman readability formula, 52.94% of the texts in 5th grade textbooks, 51.28% of the texts in 6th grade books, 55% in 7th grade books and 52.94% of the texts in 8th grade books are at the easy level. According to the Çetinkaya-Uzun forumla, the accumulation occurs at the educational level. The lowest accumulation according to the Ateşman formula is at the level of very difficult while according to the Çetinkaya-Uzun formula, the accumulation occurs at the independent level.
Okur and Arı (2013) identified in their study that the highest accumulation at the medium level and the lowest accumulation at the very difficult level in 6th and 7th grade while at the very easy level in the 8th grade. Bağcı and Ünal (2013) conducted a study on 8th grade Turkish textbooks and identified that the majority of the texts are at the medium level. When compared to the current study, the texts analyzed in the current study are compatible with the students' level in terms of readability.
When the examined Turkish textbooks are analyzed in terms of text types, informative texts are accumulated at the medium level while narratie texts are accumulated at the easy level. When analyzed across all grade levels, only 1 text was found at the very easy level in informative texts while there are 4 texts in narrative texts. There are 4 difficult and 1 very difficult level texts found in informative texts while there are no texts identified at these levels in narrative texts. When the data obtained through the Çetinkaya-Uzun formula, the accumulation is at the educational level. There are narrative texts. In light of the findings obtained through both formulas, informative texts are more difficult than narrative texts. There are similar results found in the literature (Bağcı & Ünal, 2013;Okur & Arı, 2013;Özdemir, 2016).
In a study conducted on 5th grade Turkish textbooks by Mirzaoğlu & Akın (2015), the readability scores of narrative texts were 74.86 (easy) while informative texts had a score of 71 (easy).
In a study conducted by Özdemir (2016), 71.4% of narrative texts are easy level and 28.6% are medium level while 16.7% of informative texts are easy, 58.3% are medium and 25% are at the difficult level. 46% of all the texts are easy, 42.3% are medium and 11.5% are at the difficult level. In the current study, 12.50% of narrative texts are very easy, 87/50% are easy level. 5.88% of all the texts are easy, 52.94% are easy and 41.18% are medium level. While there are differences between the findings of the current study and the findings of the study conducted by Özdemir (2016), there is a similarity as well.
Departing from the current study and the literature in the field, a few recommendations can be made: 1. It is seen that the average word and sentence lengths of informative texts in Turkish textbooks are longer than narrative texts. This needs to be taken into consideration when selecting texts for textbooks in terms of comprehensibility of texts.
2. In selecting texts for Turkish textbooks, texts that are created with an artist's sensitivity and are compatible with children's meaning universe should be preferred.
3. Readability levels of texts should be compatible with the students' levels. Thus, necessary examinations should be performed in selecting texts for textbooks.