Bayburt University Journal of Education Faculty. Bayburt Üniversitesi. Eğitim Fakültesi Dergisi. Sahibi: Rektör Prof. Dr. Murat MOLLAMAHMUTOĞLU

Transkript

1 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Bayburt University Journal of Education Faculty Sahibi: Rektör Prof. Dr. Murat MOLLAMAHMUTOĞLU Editör: Yrd. Doç. Dr. Esen TAŞĞIN Yayın Kurulu: Prof. Dr. Ali Rıza AKDENİZ (Rize Üniversitesi) Prof. Dr. Ali Paşa AYAS (Karadeniz Teknik Üniversitesi) Prof. Dr. Hüseyin AYDIN (Atatürk Üniversitesi) Prof. Dr. Uğur ÇEVİK (Karadeniz Teknik Üniversitesi) Prof. Dr. Turan GÜVEN (Gazi Üniversitesi) Prof. Dr. Kamil KAYGUSUZ (Karadeniz Teknik Üniversitesi) Prof. Dr. Ali YILDIRIM (Atatürk Üniversitesi) Doç. Dr. Mehmet YALÇIN (Bayburt Üniversitesi) Editör Yardımcısı: Yrd. Doç. Dr. Fatih YALÇIN Bölüm Editörleri: Esra ÖZKAN Figen ÇAM Sakıp KAHRAMAN Tuğba AYDIN Bayburt Üniversitesi Eğitim Fakültesi Dergisi (BUEFAD) yılda iki kez yayımlanan hakemli bir dergidir. Sosyal bilimler, fen bilimleri ve güzel sanatlar ile bu alanların eğitimi ve öğretimine ilişkin her türlü makale, bilimsel yayın kurallarına uygun olmak koşulu ile yayımlanır. İletişim: Bayburt Üniversitesi Bayburt Eğitim Fakültesi / BAYBURT e-posta: befdergi@bayburt.edu.tr Owner: Rector: Prof. Dr. Murat MOLLAMAHMUTOĞLU Editor: Assist. Prof. Dr. Esen TAŞĞIN Editorial Board: Prof. Dr. Ali Rıza AKDENİZ (Rize University) Prof. Dr. Ali Paşa AYAS (Karadeniz Technical University) Prof. Dr. Hüseyin AYDIN (Atatürk University) Prof. Dr. Uğur ÇEVİK (Karadeniz Technical University) Prof. Dr. Turan GÜVEN (Gazi University) Prof. Dr. Kamil KAYGUSUZ (Karadeniz Technical University) Prof. Dr. Ali YILDIRIM (Atatürk University) Doç. Dr. Mehmet YALÇIN (Bayburt University) Assistant Editor: Assit. Prof. Dr. Fatih YALÇIN Section Editors: Esra ÖZKAN Figen ÇAM Sakıp KAHRAMAN Tuğba AYDIN Bayburt University Journal of Education Faculty (BUJEF) is a refereed journal that is published two times a year. It publishes scientific papers related to the follo wing fields: social sciences, basic sciences, fine arts, social sciences teaching, basic sciences teaching, and fine arts teaching. Contact: Bayburt University Bayburt Education Faculty / BAYBURT befdergi@bayburt.edu.tr Dizgi - Tasarım - Baskı Design - Printing Yağız Matbaacılık Yağız Matbaacılık Kadızade Mah. Koçoğlu Çarşısı No:6 BAYBURT Kadızade Mah. Koçoğlu Çarşısı No:6 BAYBURT Tel: Tel:

2 BU SAYININ HAKEMLERİ Doç. Dr. Ahmet Arıkan Doç. Dr. Çetin Doğar Doç. Dr. Fatoş Silman Doç. Dr. İdris Kadıoğlu Doç. Dr. Mehmet Yalçın Doç. Dr. Mustafa AKÇAY Doç. Dr. Salih Sabri Yavuz Doç. Dr. Süleyman Solak Yrd. Doç. Dr. Erdal Bay Yrd. Doç. Dr. H. İsmail Arslantaş Yrd. Doç. Dr. Gülden Tüm Yrd. Doç. Dr. Güney Haciömeroğlu Yrd. Doç. Dr. Kerim Gündoğdu Yrd. Doç. Dr. Mümin Topçu Yrd. Doç. Dr. Oktay Yağız Yrd. Doç. Dr. Ramis Bayrak Yrd. Doç. Dr. Sare Şengül Yrd. Doç. Dr. Taner Yıldız Yrd. Doç. Dr. Tevfik İşleyen Gazi Üniversitesi Erzincan Üniversitesi Yakın Doğu Üniversitesi Dicle Üniversitesi Bayburt Üniversitesi Sakarya Üniversitesi Rize Üniversitesi Selçuk Üniversitesi Gaziantep Üniversitesi Kilis 7 Aralık Üniversitesi Çukurova Üniversitesi Çanakkale Onsekiz Mart Üniversitesi Atatürk Üniversitesi Dicle Üniversitesi Atatürk Üniversitesi Bayburt Üniversitesi Marmara Üniversitesi Karadeniz Teknik Üniversitesi Atatürk Üniversitesi

3 BAYBURT ÜNİVERSİTESİ EĞİTİM FAKÜLTESİ DERGİSİ Cilt: 5 Sayı: I-II İÇİNDEKİLER Adem S. TURANLI Öğrencilerin Yazılı Çalışmalarına İlişkin Öğretmenlerin Değerlendirmelerini Etkileyen Etmenler Ayhan Çinici ve Yavuz Demir İşbirlikli Tahmin-Gözlem-Açıklama (TGA) Aktivitelerinin 9. Sınıf Öğrencilerinin Demokratik Tutumları Üzerine Etkisi Birol TEKİN ve Alper Cihan KONYALIOĞLU Trigonometrik Fonksiyonların Toplam ve Fark Formüllerinin Ortaöğretim Düzeyinde Görselleştirilmesi... Hasan Kurt Cumhuriyet Dönemi İlmihal Kitaplarında İman Esaslarına Genel Bakış ( ) Kerim Gündogdu, Zehra Sedef Coşkun, Canan Albez ve Erdal Bay Sınıf Öğretmeni Adaylarının Öğretmenlik Uygulamasına İlişkin Görüşleri ve beklentileri Kürşat Kara Bayburt Halk Şiirinde Kop Savunması Süleyman YAMAN ve Seher TEKİN Öğretmenler İçin Hizmet-İçi Eğitime Yönelik Tutum Ölçeği Geliştirilmesi Yasemin Koç, Ramis Bayrak, A.Cihan Konyalıoğlu ve Abdullah Kaplan Fen Eğitiminde Kavram Yanılgıları, Grafikler ve Matematik Öğretimi Habib Özgan, Çiğdem Celik ve Fatih Bozbayındır İlköğretim Okulu Müdürlerinin Müzakere Becerilerinin İncelenmesi Nevzat AYDIN Bir Âlimi Doğduğu Topraklardan Günümüze Taşımak Anısına: Ekmelüddin Bâbertî Sempozyumu (28-30 Mayıs 2010,

4 Volume: 5 BAYBURT UNIVERSITY JOURNAL OF EDUCATION FACULTY CONTENTS Number: I-II Adem S. TURANLI Factors Influencing Teachers' Quality Judgments Of Students' Written Work Ayhan Çinici ve Yavuz Demir The Effects of cooperative Predict-Observe-Explain (POE) tasks on 9th grade students' democratic attitudes... Birol TEKİN ve Alper Cihan KONYALIOĞLU Trigonometrik Fonksiyonların Toplam ve Fark Formüllerinin Ortaöğretim Düzeyinde Görselleştirilmesi... Hasan Kurt An Overview on Islamic Creeds in Catechetical Books (ilmihal) Puplished in the beginnig of Turkish Republic Kerim Gündogdu, Zehra Sedef Coşkun, Canan Albez ve Erdal Bay The Opinions and Expectation of Teacher Trainee About Teaching Practice... Kürşat Kara The Kop Resistence in Bayburt Folk Poetry Süleyman YAMAN ve Seher TEKİN Developıng An Attıtude Scale Towards Inservıce Educatıon For Teachers Yasemin Koç, Ramis Bayrak, A.Cihan Konyalıoğlu ve Abdullah Kaplan Misconceptions, Graphs and Teaching Mathematics in Science Education Habib Özgan, Çiğdem Celik ve Fatih Bozbayındır Investigation Of Negotiation Skills of Elementary School Principals Nevzat AYDIN Bir Âlimi Doğduğu Topraklardan Günümüze Taşımak Anısına: Ekmelüddin Bâbertî Sempozyumu (28-30 Mayıs 2010, Bayburt)

5 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) 1 Factors Influencing Teachers' Quality Judgments of Students' Written Work 1 Adem S. TURANLI Abstract This study aimed at identifying the relationships between teachers' quality judgments about student written work and some related variables. 48 teachers working for a language school at tertiary level constituted the participants. The researcher himself wrote two essays and set up a design that would provide the required data. The study indicated that assessment strategy and the adjacency of essays assessed are likely to affect teachers' judgments of students' written work. Similar conclusions were found when the contribution of assessment strategy is controlled. The perceived quality of the essay (assigned scores) are correlated with the amount of the writing. Keywords: holistic evaluation, analytic evaluation, writing in a foreign language, reliability in assessment 1 Yrd. Doç. Dr., Erciyes Üniversitesi, Eğitim Fakültesi turanli@erciyes.edu.tr

6 2 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) Öğrencilerin Yazılı Çalışmalarına İlişkin Öğretmenlerin Değerlendirmelerini Etkileyen Etmenler Özet Bu araştırma, öğrencilerin yazılı çalışmalarına ilişkin öğretmenlerin değerlendirmeleriyle ilintili bazı değişkenler arasındaki ilişkileri irdelemeyi amaçlamıştır. Çalışma grubunu, üniversite düzeyinde bir dil okulunda çalışmakta olan 48 öğretmen oluşturmuştur. Araştırmacının kendisi, araştırma sorularına koşut olarak, iki makale (yazı-essay) yazmış ve gerekli verileri temin edecek araştırma desenini oluşturmuştur. Çalışma, ölçme yaklaşımı ve daha iyi olan yazının önce ya da sonra okunmasının, öğrencilerin yazılı çalışmalarına ilişkin öğretmenlerin değerlendirmelerini etkileyebileceğini göstermiştir. Ölçme yaklaşımının etkisi denetlenince de benzer sonuçlar bulunmuştur. Ayrıca, yazının algılanan niteliği (verilen notlar); okunan metnin uzunluğuyla ilişkili bulunmuştur. Anahtar sözcükler: bütüncül değerlendirme, çözümleyici değerlendirme, yabancı dilde yazma, ölçme güvenirliği 1. Introduction In order to discover students' abilities in productive skills, open-ended questions are more favored because they force students to produce, which helps teachers with their judgments about the examinee's language skills (Genesee & Upshur, 1996). Teachers generally prefer traditional evaluation methods in identifying students' achievements and many teachers complain that they cannot use alternative strategies because classes are too crowded, they lack time either for preparation or for evaluation (Gelbal & Kelecioğlu, 2007). Weigle (2002) defines reliability as a consistency of measurement across different characteristics or facets of a testing situation. The measurement errors of any kind deriving from the factors other than the answer itself decrease the reliability (Tan, 2005). Identifying the factors likely to affect teachers' judgments of the quality of student work may create awareness and sensitivity about the sources of potential errors. However, it is a complex process to identify the criteria and the standards to be used in assessing student work in the target language in terms of its authenticity and academic efficiency (Kroll, 1990). Various contextual factors may influence the assessors' mood at the time of scoring, causes significant systematic differences among the assigned scores (Tekin, 1984). Wragg (2001) claims that even the teachers using double blind assessment procedure tend to approximate the score to the average and, thereby, decrease the reliability of the examination. As an alternative to traditional holistic assessment, analytic assessment was developed to eliminate systematic errors in assessment. In holistic judgment, different assessors may focus on different aspects of written work although in analytic scoring, focusing on several aspects of the work separately may be time consuming and often difficult (Davies et al., 1999). Therefore, holistic assessment is a more natural process than analytic one (White, 1995). Nakamura (2002) claims that asking the raters to assess analytically will not ensure that they will really score analytically, since they may tend to adjust analytic scores to match their holistic impressions (Weigle, 2002). Especially in holistic assessment, some external features such as the gender of the examiner or the examinee may affect assessors (Goddard-Spear, 1983; Gipps & Murphy, 1994). Several other studies have also dealt with other factors which may contaminate true scores such as the length or quantity of the essay produced predicts holistic scores (Grobe, 1981; Breland & Jones, 1984), syntactic maturity, usage, mechanics and vocabulary (Grobe, 1981), the context position of an essay in the order of assessment (Hales & Tokar, 1975; Hughes et al., 1980; Daly & Dickson-Markman, 1982), the degree to which the essay content fulfills the expectations (Tedick

7 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) 3 & Mathison, 1995), the first impressions (Vaughan, 1992) and assessors' expectations (Stock & Robinson, 1987). Likewise, the teacher's prejudices about the student may contaminate their judgments (Wragg, 2001). It is difficult to bring assessors to a complete agreement about their quality judgments, even if a temporary agreement on common standards can be reached through training (Weigle, 2002). Weigle suggests investigating how some variables interact to influence assessors' judgments rather than dealing with individual variables. Accordingly, this study was conducted to identify how assessors' quality judgments are related to some contextual variables. The variables which are included in the study are assessment strategy, location of mistakes, the length of the language course allegedly completed, the adjacency of essays assessed (whether a better essay is assessed earlier or later), the length (amount) of the text considered before scoring. Therefore, the research questions are: 1. Do the assessors' quality judgments interact with their assessment strategy, the location of grammatical or spelling mistakes, the length of the language course allegedly completed, the adjacency of essays assessed (whether a better essay is assessed earlier or later), the length (amount) of the text considered before scoring? 2. When the assessment strategy is controlled, do the assessors' quality judgments interact with the abovementioned variables? 2. Method 2.1. Context This study aimed to identify the factors that interact with assessors' judgments in ELT writing. In order to answer this research question, the scores assigned by the assessors (teachers) for two essays were used as dependent variables while assessment strategy, location of mistakes, the length of the language course completed, the adjacency of essays assessed (whether a better essay is assessed earlier or later), the length of the text considered before scoring were included as independent variables. The study was carried out in a school at tertiary level, where as a regulation, students who have gained a place in it attend an eight-month Basic English program through upper-intermediate level before they start their majors. Students follow an intensive program for eight months and were taught by 58 non-native English teachers, 48 of whom constituted the participants. The independent variables were included, mainly because the school administrators and program planners believed that these factors contaminated scores Data gathering First, the researcher invited all the teachers working at the school to take part in the study, 52 of whom accepted to participate. Afterwards, the researcher gave them the two essays and asked them to assess the essays as they were instructed on the cover page of the essays. The essays had been prepared by the researcher and specifically equipped with some features so as to help answer the research questions. Within the same day, the researcher gathered the essays with the ascribed scores. However, during the data entry, it was discovered that four teachers had not followed the instructions properly. These teachers, who would assess the essays analytically, had assigned holistic scores. Consequently, they were eliminated, and the data from the 48 teachers were used for the analysis Instrumentation Since it was not possible to find a context that could naturally present enough data in line of the research questions, the researcher purposefully prepared the essays. Initially, the researcher prepared two essays, with some mistakes purposefully included (see Appendix). The essays were intended to be at different levels (one better than the other in terms of the vocabulary and structures used). Despite the mistakes which were purposefully placed in both essays, they were named as E- Essay (elementary) and I-Essay (intermediate). E-Essay was a description of a boy's life and his

8 4 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) family, and I-Essay was about the importance of learning a foreign language. Ten of the teachers working for the school were asked to judge their appropriateness for the stated levels. They all commented that students of the stated levels could also write similar essays with similar syntactical and spelling mistakes. However, this was of limited concern whether since existence of the difference itself in level was not considered as a factor. Additionally, the average scores calculated showed that the essays were of different levels. The mean score for E-Essay was calculated to be 64,5 while the mean for I-Essay was 73,0. By paired-samples t-test, the difference was found statistically significant (t=4,37; p<0,01). Also, the big standard deviations in the scores imply that the teachers had highly differing quality judgments. For each essay the researcher created two versions with slight differences (Version A and Version B in terms of the mistakes purposefully located). Each version had a unique purpose so as to provide data for a different research question. First, in order to investigate how the length of the text and the assigned scores interacted, each essay was divided into four equal parts in terms of the number of the words. The first part was named 'text 1', the first and second parts were called 'text 2', the first, second and third parts were names as 'text 3', and finally the whole essay was called 'text 4'. Table 1 Distribution of Assessors According to Independent Variables Essays Versions Course Length (hours) E-Essay (Elementary) I-Essay (Intermediate) Version A (mmistakes) Increasing numbers of mistakes Version B (Mmistakes) Decreasing numbers of mistakes Version A (mmistakes) Increasing numbers of mistakes Version B (Mmistakes) Decreasing numbers of mistakes Assessment Strategy Dispersion of mistakes: For each essay, two versions were constructed which were the same in essence but different in terms of the dispersion of the mistakes: in Version A, 'texts' 1 through 4 had increasing numbers of mistakes while in Version B, 'texts' had decreasing numbers of them (See Appendix). Namely, in Version A, 'texts' 1 through 4 had increasing numbers of spelling and syntactical mistakes (1, 3, 3 and 5 respectively) and in Version B, they had 5, 3, 3 and 1 mistakes and Versions A and B were were also named as mmistakes and Mmistakes respectively for simplicity and clarity) Length of the course allegedly completed: In order to identify how the length of the allegedly-completed course and assessors' judgments interact, each of the abovementioned versions was affixed a note of 'The student has written this essay after the completion of a course of N 150 Holistic 6 Analytic Holistic 7 Analytic Holistic 7 Analytic Holistic 6 Analytic Holistic 6 Analytic Holistic 7 Analytic Holistic 7 Analytic Holistic 6 Analytic 6

9 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) (or 300) hours'. It was because 150- and 300-hour courses were referred to as 'elementary' and 'pre-intermediate' levels respectively at the school where the study was conducted. Assessment strategy: In order to test how the assessment strategy and assessors' judgments interact, half of the teachers were asked to assess the essays holistically while the other half were to assess them analytically. In order to avoid ambiguity and confusion, the analytic and holistic assessment procedures were used in the same way they were carried out for the examinations at the school. The holistic assessors were asked to consider the percentages of the allotted time to the writing components in the course (given below) while assigning a score and assign just one score to each 'text'. On the other hand, the analytic rubric was quite simple and the assessors would ascribe separate scores between 0 and 100 to each of the three components for every 'text'. Next, these scores were multiplied by the corresponding weights and the results were added to find a composite score for each 'text'. Organization and mechanics : %25 Vocabulary and language use : %40 Coverage of content : %35 Adjacency of essays (order of assessment): In order to explore how assessors' judgments interact with the level of the previously adjacent paper, the two essays (E-Essay and I-Essay) were clipped together so that half of the assessors started with E-Essay while the rest graded I-Essay first. Length of the text: This study also aimed at identifying how the length of the 'text' interacts with assessors' judgments. Therefore, the scores assigned to 'texts' 2 and 4 were compared. 'Text' 2 covered the first half of the essay and 'text' 4 constituted the whole of it. 'Texts' 1 and 3 were not added into comparisons, just to keep analyses simpler. Table 1 displays the distribution of the 48 assessors (teachers). Although at the phase of planning, an equal distribution had been planned, four teachers did not assess the essays properly. At the phase of assessment, the assessors were asked to assign scores for 'texts' 1 through 4 and not to do any changes backwards in the previous grades that they had already assigned, even if they thought they needed to. 3. Findings In order to test whether there were differences between the subgroups of each independent variable, independent samples t-test was used as a statistical tool. The interactions between the pairs of independent variables were descriptively investigated, cautiously discussed, and generalizations were avoided. The mean scores and standard deviations were used for the analyses and the findings are presented below. Location of mistakes: Readers' first impressions are commonly believed to be shaped right after starting to read an essay and not to change afterwards. In order to test this, the teachers assessed either Version A or Version B of either E-Essay or I-Essay. The assessors had not been informed about the frequency or the location of the mistakes. The difference between the two versions for either essay was insignificant (t=0,14; p>0,05 for E-Essay and t=1,45; p>0,05 I- Essay). However, the high dispersions of the scores (from 12,4 to 14,7) imply that assessors have different tendencies towards mistakes. Length of the course completed: This study also aimed to identify whether the scores changed according to the length of the course presumably done. The mean scores for E-Essay were x=70,0 and =59,0 after the completion of 150 and 300 hour courses, respectively. This difference was found to be statistically significant (t=3,06; p<0,01). However, this was not the case for I- Essay (t=0,30; p>0,05); the means for I-Essay was found to be =73,6 and =72,4 after the alleged

10 6 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) completion of the 150 and 300 hour courses, respectively. Table 2 Scores According to Independent Variables Included (Means and SDs) Features Compared Means and SDs for E-Essay Means and SDs for I-Essay x SD x SD N Overall 64,5 13,5 73,0 14,0 48 t (46) = 4,37; p<=0,05 Assessment Strategy Holistic 67,9 14,7 76,5 13,7 26 Analytic 60,5 10,8 68,9 13,5 22 t (46) = 1,94; p>0,05 t (46) = 1,95; p>0,05 Location of Mistakes A (mmistakes) 64,3 12,4 75,9 14,1 24 B (Mmistakes) 64,8 14,7 70,1 13,6 24 t (46) = 0,14; p>0,05 t (46) = 1,45; p>0,05 Course Length 150 hours 70,0 12, hours 59,0 12,1 72,4 13,0 24 t (46) = 3,06; p<0,05 t (46) = 0,30; p>0,05 Adjacency of Essays (Order of Assessment) E-Essay First 66,2 9,5 74,1 12,1 25 I-Essay First 62,7 16,8 71,9 16,0 23 t (46) = 0,92; p>0,05 t (46) = 0,54; p>0,05 Length of the Text Text 2 (Half) 60,3 15,1 71,2 15,0 48 Text 4 (Whole) 64,5 13,5 73,0 14,0 48 t (47) = 2,43; p<0,05 t (47) = 0,98; p>0,05 Adjacency of essays (order of assessment): In order to answer whether teachers are affected by the quality of the previously assessed paper, half of the teachers assessed E-Essay first and I-Essay second; the other half assessed I-Essay first. For E-Essay, the mean was calculated as x=66,2 when it was assessed first (before I-Essay) and =62,7 when it was scored after I-Essay. This difference of 3,5 was not found statistically significant (t=0,92; p>0,05). The case was quite similar for I-Essay; when E-Essay was assessed first and I-Essay second, the mean of I-Essay was calculated to be somewhat higher, although not statistically significant (t=0,54; p>0,05). In other words, the means of both E-Essay and I-Essay were higher when E-Essay was assessed first than when it was assessed second. Another point worth consideration was big standard deviations in the scores of both E-Essay (SD=16,8) and I-Essay (SD=16,0) when I-Essay was assessed first. On the other hand, the standard deviation of E-Essay was calculated the smallest (SD=9,5) of all. This finding indicates that when an essay of a lower level is assessed after one of a higher level, assessors' judgments may be contaminated by the quality of the previous one and also, raters have more difficulty complying with the assessment criteria in scoring the higher level text.

11 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) 7 Length of texts: In order to answer whether assessors' quality judgments about an essay chang while reading through it, each essay was divided into four equal parts, and for simplicity the scores assigned for 'text' 2 (half of the essay) and 'text' 4 were compared. Since the assessors had already been informed about the length of the complete essays, their quality judgments were expected not to change only due to the length of the 'texts'. However, the means of 'text' 2 (=60,3) and 'text' 4 (=64,5) significantly differed for E-Essay (t=2,43; p<0,05). However, the difference between 'text' 2 (=71,2) and 'text' 4 (=73,0) for I-Essay was smaller and statistically insignificant. This means that the length of the text influences assessors' judgments for students' work within elementary level and this may arise from assessors' tendency to be affected by students' efforts considered as 'good' work at earlier levels. Interactions: This study also aimed at investigating the interactions between the assessment strategy (holistic or analytic assessment) and the other variables included in the study. In order to avoid overgeneralization which may derive from the limitations of the study, the interactions were investigated descriptively and accordingly, a further in-depth study was believed to be needed for statistical comparisons. Table 3 Interactions of Variables Included: A Comparative Exploration Assessment E-Essay I-Essay Strategy x SD x SD Location of Mistakes Version A H 66,2 13,7 73,1 11,1 (mmistakes) A 62,0 10,7 66,6 15,9 Version B H 69,6 16,0 80,0 15,5 (Mmistakes) A 59,1 11,2 71,1 11,0 Adjacency of Essays (Order of Assessment) E-Essay H 69,6 8,2 78,2 9,6 Assessed First A 63,6 10,0 70,9 13,3 I-Essay Assessed First H 66,7 18,3 75,3 16,3 A 55,1 10,7 65,4 14,1 Course Length 150 hours H 76,5 9,7 80,0 16,8 A 62,3 11,6 74,3 11,2 300 hours H 59,2 14,0 73,1 9,1 A 58,8 10,2 63,5 14,0 Assessment strategy and location of more mistakes: When they were sorted out according to the assessment strategy and the location of more of the mistakes for either essay, there did not appear to be big differences between the groups However, high dispersions were observed in the scores. First, the range of the means for E-Essay (10,5) was smaller than that of the means for I-Essay (13,4), indicating that for I-Essay, the interaction between the assessment strategy and the location of the mistakes in the essay was bigger, even though not statistically significant. The mean of the analytic scores for Version A of I-Essay ( x=66,6) was much smaller than that (=80,0) of holistic scores for Version B. Although a similar pattern was observed for E-Essay, the lowest mean (=59,1) was calculated for analytic scores for Version B. Assessment strategy and adjacency of essays (order of assessment): The analysis revealed that when the scores were grouped according to the assessment strategy and the assessment order

12 8 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) (whether E-Essay or I-Essay was assessed first), there appeared to be some noteworthy points related to the means and standard deviations. In both cases the teachers assigned higher holistic scores to both essays than when they assessed them analytically. Second, while holistic scores showed a difference of around 3 for both cases, the differences were bigger for the analytic scores. For the analytic scores, the differences were 8,5 for E-Essay and 5,5 I-Essay, and in both of the cases the scores were calculated to be higher when E-Essay was assessed first. Also, when the means were considered as a whole, for E-Essay, the means ranged from =55,1 (analytic scores when the essay was assessed after I-Essay) to =69,6 (holistic scores when it was assessed first). Similarly, for I-Essay, the means ranged from =65,4 (analytic scores when it was assessed before E-Essay) to 78,2 (analytic scores when E-Essay was assessed first). These differences in means (14,4 for E-Essay and 12,8 for I-Essay) and standard deviations (from 8,2 to 18,3 for E-Essay and from 9,6 to 16,3 for I-Essay) show that scores spread much according to the assessment strategy and the adjacency of assessment. Assessment strategy and length of course completed: When the scores were grouped according to the assessment strategy and the length of the course allegedly completed, considerable patterns were observed. Unlike after the completion of a 150 hour course, in which holistic scores were higher than their counterparts, the holistic and analytic scores for E-Essay were fairly close to each other (a difference of 0,4) after the completion of 300 hour course. However, for E-Essay again, the difference between holistic and analytic scores were very big (with a difference of 14,2) after the completion of 150 hour course. Besides, for I-Essay, for both courses, analytic scores were calculated smaller than their holistic counterparts (a difference of 6,7 for the 150 course and 9,4 for the 300 hour course). Also, the distributions of the mean scores for the 150 and 300 hour courses looked different. Correlations: The scores assigned for E-Essay were found significantly correlated with those for I-Essay (r=0,52; p<0,01), meaning that the teachers assessed both essays either more positively or more negatively. A teacher who assigned a higher score for one essay also assigned a higher score for the other. This indicates that teachers are inclined to assign either higher or lower scores to students' work regardless of its quality. Furthermore, with the assessment strategy being controlled, the correlation was found statistically insignificant between the analytic scores for the essays (r=0,32; p>0,05) but statistically significant for holistic ones (r=0,59; p<0,00). Given that the assessors who participated in the study had to assess both essays within a short time, the analytic assessment strategy might have helped them eliminate the subjectivity arising from the adjacency of papers. 4. Discussion And Implications It is much harder to achieve reliability in writing tests than any other examinations. It results from the multitude of subjective elements in what is assessed (Çetin & Kelecioğlu, 2004) This study provided insights about the factors that shape assessors' quality judgments of student written work in ELT settings despite the limitation that the researcher himself prepared the essays. First, the scores show a big dispersion when the task requires students to write an essay for the development of productive skills. Therefore, teachers are to consider the cons and pros that open-ended questions might involve. Holistic and analytic scores show considerable differences in favor of holistic ones. Nevertheless, when the length of the course taken is used as a controlling variable, the differences get bigger in favor of holistic scores. Also, no difference was found between the holistic and analytic scores ascribed after the completion of the longer course, but holistic scores were higher for the shorter course. So it can be concluded that assessors' analytic and holistic judgments are differently shaped by the length of the course that students do. This may lead to deceptive conclusions about the effectiveness of the program and, consequently, may delude decision makers. Therefore, the institutional priorities and the curriculum implemented should have a key role determining the appropriate assessment strategy. Since both holistic and

13 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) 9 analytic assessments have cons and pros, teachers should be flexible in their preferences. Mainly, holistic scales are found faster and more efficient, and analytic scales are somewhat more reliable; however, analytic assessment will provide better feedback for students, because students are informed about their respective strengths and weaknesses in different aspects of writing (Bauer, 1981; Weigle, 2007). When other effects are ignored, assessors' judgments do not seem to be affected by the location where mistakes abound in an essay. However, when there appear more mistakes at the beginning of the essay, assessors assign lower analytic scores. Analytic assessment may limit assessors' subjectivity by being more rigid in criteria and may prevent assessors from revising their first impressions, even if the quality of the text gets better or worse. Therefore, assessors must read the examination paper thoroughly before they assign a score and a second reading may help to decrease assessment errors. When a student paper is assessed after a better one, the score assigned for the second one tends to decrease, and assessors' judgments disperse more than when the better work is evaluated as second. Besides, analytic scores considerably decrease when the lower-level student work is assessed first. This indicates that with a big difference between the essays in quality, the judgments about the earlier paper may shape the judgments about the following; the quality of the first paper is unconsciously taken as a point of reference and the next paper is compared to that, which may or may not be in favor of the second paper, which is in line with Hales & Tokar (1975), Hughes et al. (1980), Daly & Dickson-Markman (1982). They maintain that when a medium quality essay is read after a number of high quality essays, it will likely receive a low score or it will be given a high score if it is assessed after a number of high quality essays. In order to avoid this tendency, after first reading, teachers may need to sort out the papers according to their first impressions of the quality and then read them thoroughly and more carefully. Although assessors consider the length of the course taken by the student whose essay is assessed, the scores are not proportional with the lengths of the courses. The assessors expect proportionally much more from the students who have finished a shorter course than those who have finished twice as long as the other. When students take up learning a foreign language, they often have unrealistically high expectations. This is also how some teachers think; they may have too high expectations from their students. The small difference, even though statistically significant, between the scores of the two groups of students who have finished courses of different lengths implies that teachers' expectations become more realistic as students reach higher levels. Since unrealistically high goals which are set in order to motivate students may lead to disappointment, both teachers and students should build realistic expectations at the very beginning of the course in order to avoid frustration. The study shows that the scores gradually increase in accordance with the length of student work, when the other variables are controlled. This shows that teachers' judgments are affected by the amount of text on the paper regardless of its quality. Therefore, teachers should review their expectations before they start assessing the papers and eliminate the factors leading to systematic measurement errors. It was also observed that the more mistakes assessors observe in the successive parts of the essay while reading through the paper, the lower scores they assign. However, the first and final judgments for the essay with fewer mistakes in the second part do not show any considerable change. This indicates that assessors' first impressions at the beginning of the student work hardly change. Thus, it can be said that any work that builds positive first impressions ends positively, even if its quality gets poorer. The teacher should be helped to overcome some of the problems related to reliability; teaching them how to use rubrics will pay off (Çetin & Kelecioğlu, 2004). Alderson (2001) suggests that examiners should understand the principles behind the rating scales with which they work. Despite these findings and their implications, the results should be carefully dealt with

14 10 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) because of the limitations of the study. First, it was carried out with teachers at the tertiary level. Second, the number of the teachers was limited to 48, which made the means vulnerable to measurement errors. Therefore, the findings related to the analyses of significance need to be cautiously accepted. Another limitation was with the number of essays used; two purposefully invented essays were used to depict teachers' attitudes in assessment. Even if the number of the essays (two) may be assumed to be sufficient for the aim of this study, further studies with more deliberate designs (both qualitative and quantitative) may provide further information on how assessors are influenced by various factors. References Alderson, J. C.; Clapham, C. & Wall, D. (2001). Language test construction and evaluation. Cambridge: Cambridge University Press. Bauer, B. A. (1981). A study of the reliabilities and cost-effeciencies of three methods of assessment for writing ability (ERIC Document Reproduction Service No. ED ). Breland, H. M. & Jones, R. J. (1984). Perception of writing skills. Written Communication 1 (1), Daly, J. A. & Dickson-Markman, F. (1982). Contrast effects in evaluating effects, Journal of Educational Measurement, 19 (4) Çetin, B & Kelecioğlu, H. (2004). The relation between scores predicted from structured features of essay and scores based on scoring key and overall impression in essay type examinations. Hacettepe Üniversitesi Eğitim Fakültesi Dergisi (H. U. Journal of Education), 26, (2004), Davies, A., Brown, A., Elder, C., Hill, K., Lumley, T. & McNamara, T. (1999). Dictionary of language testing. Cambridge: Cambridge University Press. Gelbal, S. & Kelecioğlu, H. (2007). Teachers' proficiency perceptions of the measurement and evaluation techniques and the problems they confront. Hacettepe Üniversitesi Eğitim Fakültesi Dergisi (H. U. Journal of Education), 33, Genesee, F. & Upshur, J. A. (1996). Classroom-based evaluation in second language education. Cambridge: Cambridge University Press. Gipps, C. & Murphy, P. (1994). A fair test? Assessment, achievement and equity. Milton Keynes, Open University Press. Goddard-Spear, M. (1983). Sex bias in science teachers' ratings of work. Contribution to the second GASAT Conference, Oslo, Norway. Grobe, (1981). Syntactic maturity, mechanics, and vocabulary as predictors of writing quality. Research in the Teaching of English, 15, Hales, L. W. & Tokar, E. (1975). The effects of quality of preceding responses on the grades assigned to subsequent responses to an essay question. Journal of Educational Measurement, 12, Hughes, D. E; Keiling, B. & Tuck, B. F. (1980). The influence of context position and scoring method on essay scoring. Journal of Educational Measurement, 17, Kroll, B. (1990). Second language writing: Research insights from the classroom. Cambridge: Cambridge University Press.

15 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) 11 Nakamura, Y. (2002). A comparison of holistic and analytic scoring methods in the assessment of writing. Retrieved from the Web Stock, P. L. & Robinson, J. L., Taking on testing: Teachers as researchers. English Education, 19, Tan, Ş. (2006). Öğretimi planlama ve değerlendirme, Ankara: Pegem Yayıncılık. Tedick, D. J. & Mathison, M. A. (1995). Holistic scoring in ESL writing assessment: What does an analysis of rhetorical features reveal? In: Academic writing in a second language: Essays on research and pedagogy. Norwood: Ablex Publishing Corporation, Tekin, H. (1984). Eğitimde ölçme ve değerlendirme. Ankara: Has-soy Matbaası. Vaughan, C. (1992). Holistic assessment: What goes on in the rater's mind? In: HampLyons, L., Editor, Assessing second language writing in academic contexts, Norwood: Ablex, NJ, Weigle, S. C. (2002). Assessing writing. Cambridge: Cambridge University Press. Weigle, S. C. (2007). Teaching writing teachersaboutassessment. Journal of Second Language Writing, 16 (3), White, E. M. (1995). An apologia for the timed impromptu essay test. College Composition and Communication, 46 (1), Wragg, E. C. (2001). Assessment and learning in the secondary school. Florence, KY, USA: Routledge.

16 12 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) Appendix (Slashes show where the essays were divided into 'texts' and did not exist in the copies given to the teachers) E-Essay (Essay at Elementary Level) Version A My name is Ali and I am a university student. I want to be an economist when I will finish school. I sometimes go to the cinema at weekends. I like sciyence fiction and fantastic comedies. / My father is a teacher and is working in a pravite school for about 15 years. He likes write poems and reading boks. My mother is a housewife and cooks really good. She often watches TV and plays the piyano. / My sister 16 years ould and she is a student at a high school. She intend to be an electronic engineer. She can speak English fluently althouh she makes meny mistakes. She solves often puzzles. / We live in a larg flat at the three floor. Our flat faces a beatiful park and I like it very. We often go out to a walk after dark at Frydays. I think we are a heppy family. E-Essay (Essay at Elementary Level) Version B My name is Ali and I am an university student. I wont be an economist when I will finish schol. I go sometimes to the sinema in weekends. I like sciyence fiction with fantastic comedies. / My father is a teacher and worked in a pravite school for about 15 years. He likes write poems and reading boks. My mother is a housewife and cooks really good. She often watches TV and plays the piyano. / My sister 16 years ould and she is a student at a high school. She intend to be an electronic engineer. She can speak English fluently althouh she makes meny mistakes. She solves often puzzles. / We live in a large flat on the three floor. Our flat faces a beautiful park and I like it very much. We often go out for a wolk after dark on Fridays. I think we are a happy family. I-Essay (Essay at Intermediate Level) Version A Learning a foreign language has many avantages. If you know a foreign language, it is easier to find a job in an international organization. In almost all paper, you can see job advertisements for people who know at least a foreign language. / In adition, one who has learn English can read books in English to learn more about his fild of work. Since many of professional books are writen in English, he can improve heself. / Moreover, he can read valuble literary works writing in English for pleasur. Many people travel aboard. It is know that almost everywhere in the world, you can meet people which can speak English. / Therefore, English is enoug to communicates with the nativ peoples of the country. Last but not the least; learn the stracture of the English language is similar to solving a large puzzle. It give you pleasur and improves your mind. I-Essay (Essay at Intermediate Level) Version B Learning a foreign language have many avantages. If you know foreign language, it is easyer find a job in an international organization. In olmost all paper, you can see job advertisments for people whom know at leest a foreign language. / In adition, one who has learn English can read books in English to learn more about his fild of work. Since many of professional books are writen in English, he can improve heself. / Moreover, he can read valuble literary works writing in English for pleasur. Many people travel aboard. It is know that almost everywhere in the world, you can meet people which can speak English. / Therefore, English is enoug to communicate with the native people of the country. Last but not the least; learn the structure of the English language is similar to solving a large puzzle. It gives you pleasure and improves your mind.

17 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) 13 İşbirlikli Tahmin-Gözlem-Açıklama (TGA) Aktivitelerinin 9. Sınıf Öğrencilerinin Demokratik Tutumları Üzerine Etkisi 1 2 Ayhan Çinici ve Yavuz Demir Özet Bu çalışmada işbirlikli tahmin-gözlem-açıklama (TGA) etkinliklerinin 9. sınıf öğrencilerinin demokratik tutumları üzerine olan etkisi incelenmiştir. Çalışmada tek gruplu ön-test son-test deneysel modeli kullanılmıştır. Çalışmanın örneklem grubunu Erzurum Anadolu Lisesinden 30 (13 kız; 17 erkek), dokuzuncu sınıf öğrencisi oluşturmuştur. Araştırmanın verileri Demokratik Tutum Ölçeği kullanılarak toplanmıştır. Elde edilen veriler t-testi ile analiz edilmiş ve tanımsal istatistiksel verilerle birlikte sunulmuştur. Çalışmadan elde edilen sonuçlar, öğrencilerin demokratik tutum ön- ve son-test puanları arasında ( ön-test=112.56; son-test=116.80) istatistiksel olarak anlamlı bir fark olduğunu göstermiştir (t(29) =2.08; p<.05). Yani, küçük işbirlikli gruplarda uygulanan TGA etkinliklerinin öğrencilerin demokratik tutum gelişimleri üzerinde olumlu etkisi olmuştur. Anahtar sözcükler: demokratik tutum, işbirlikli öğrenme, TGA etkinlikleri. 1 Dr., Erzurum Anadolu Lisesi, e-posta: ayhancinici@gmail.com. 2 Prof. Dr., Atatürk Üniversitesi, e-posta: ydemir_409@yahoo.com.

18 14 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) th The Effects of cooperative Predict-Observe-Explain (POE) tasks on 9 grade students' democratic attitudes Abstract In this study was investigated the effects of a of cooperative Predict-Observe-Explain (POE) th activities on 9 grade students' democratic attitude. In the study was used one-group pre-test postth test experimental model. The sample for study was composed of 30 (13 girls; 17 boys) 9 grade students from the Erzurum Anatolian High School. The data were collected with, Democratic Attitude Scale. The data were analyzed by t-test and presented together with descriptive statistics. Results of the study showed a statistically significant difference (t(29) =2.08; p<.05) between students' democratic attitude pre- and post-test mean scores ( pre-test=112.56; post- test=116.80). Namely, the POE tasks which were implemented in small cooperative groups make a positive contribution to the development of students' democratic attitudes. Keywords: democratic attitude, cooperative learning, POE tasks. Giriş Demokrasi, en genel anlamıyla halk iradesi, halkın yönetime etkin katılımı olarak tanımlanmaktadır. Halk iradesi, sadece halkın yöneticilerini seçmesi olarak değil, aynı zamanda toplumun tüm birey ve kurumlarıyla demokrasiyi yaşamasını gerektirmektedir. Aynı zamanda demokrasi, bir değerler sistemidir. Bu değerler; dürüstlük, insan hakları, yönetime etkin katılım, sosyal ve ekonomik adalet, eşitlik, güvenlik, sağlık, eğitim ve hoşgörüdür (Köse 2009). Demokratik toplumlarda yaşayan bireylerin temel özelliklerinden biri açık görüşlü olmalarıdır. Açık görüşlülük, tolerans ve kabullenme, çeşitli konulardaki problemlerin çözümlenmesi için etkin bir iletişimin sağlanmasını gerektirir (Genç ve Kalafat, 2007). İnsanların yoğun kaygı ve stres duyguları yaşadığı günümüz dünyasında problemlere neden olan birçok olayın kişilerin birbirlerine karşı gösterdiği olumsuz davranışlardan ve etkili iletişim kuramamalarından kaynaklandığı belirtilmektedir. Bu davranışları yok etmenin en iyi yolu ise bireylere başkalarını anlama ve başkalarına demokratik tutumlar gösterebilme becerilerini kazandırmaktan geçmektedir (Genç ve Kalafat, 2008). Harber (2002), demokratik tutum ve değerlerin, kalıtsal olmadığını ve öğrenilebilir olduğunu vurgulamaktadır. Bireyler küçük yaşlardan itibaren, eğer demokratik bir çevrede yetişirlerse, uygulamalı olarak demokratik tutumlar gösterebilme bilincini kazanırlar. Bu noktada aile ve okul, çocuğun şekillenmesinde en etkili iki kurumdur. Özellikle anne-baba, ailenin diğer üyeleri, akranlar, öğretmenler vb. demokrasi bilincinin kazanılmasında çocuk üzerinde oldukça önemli bir etkiye sahiptirler. Günlük yaşam içerisinde çocuk, aile ve okul ortamında diğer bireylerle ilişkilerinde demokratik veya demokratik olmayan çokça davranışla karşılaşmaktadır. Yani, çocuğa temelde aile, yakın çevre ve sonra da okul ortamında gösterilen davranışlar ve sunulan eğitim ile demokratik tutum gelişimi arasında, üst düzeyde bir ilişki bulunmaktadır (Köse, 2009). Eğitim kurumlarının ve eğitim sistemlerinin temel görevi bireylerin bağımsız düşünme becerilerini, demokratik yaşama kültürlerini, tutumlarını ve değerlerini geliştirmek olarak tanımlanmaktadır (Davis, 2003). Levin (1998) okulların sağlıklı bir eğitimin parçası olarak demokratik prensipleri yansıtması gerektiğini vurgulamıştır. Eğitim; demokrasinin birey, toplum

19 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) 15 ve devlet bazında yaşaması ve benimsenmesi için önemli bir önkoşul; demokrasi ise, eğitimden beklenen kalitenin artırılması ve yetişen bireylerin kendilerine ve tüm insanlığa faydalı olabilmeleri için yine önemli bir ön koşuldur (Yeşil, 2002). Tüm kurumlarda olduğu gibi, eğitim kurumlarında da demokrasi kültürünün oluşturulmasında ve geliştirilmesinde demokrasinin temel ilkelerinin iyi anlaşılması önemli bir rol oynamaktadır (Karadağ, Baloğlu ve Yalçınkayalar, 2006). Okullar ve sınıflar, öğrencilerin katılımcı demokrasiye hazırlanabilecekleri ve demokratik tutumlarını geliştirebilecekleri sosyal ortamlar şeklinde organize edilmelidir. Okullarda kazandırılabilecek demokratik tutumlardan kastedilen; eşitlik, katılım, çoğulculuk, uzlaşma, hoşgörü, sorumluluk alma, fikirleri serbestçe tartışma, özgür ve bilimsel düşünme ve işbirliği yapma gibi ilkeler ve niteliklerdir (Duman ve Şahiner, 2008). Ancak okullarda demokratik değer ve davranışların ayrı bir ders olarak öğrencilere verilmesinin, demokratik tutumların davranış haline getirilmesini yeterince sağlamadığı sonucuna varılmıştır (Kuzgun, 2000). Singleton (2010) pratikte uygulama sahası sağlanmadıkça düşüncelerin bilgiye dönüşmeyeceğini belirtmiştir. Singleton (2010)'a göre, uygulamaya dökülmedikçe demokrasi öğretiminin bir yolu yoktur. Benzer şekilde Gerzon (1997), demokrasi öğretimiyle ilgili farklı metotların olabileceğini, ancak Toplumsal bir yapısı olan okulda demokratik uygulamaların (saygı, hoşgörü ve başkalarından bir şeyler öğrenmeye gönüllü olma gibi) modellenmesi prensibinin değişmemesi gerektiğini ifade etmiştir. Sonuç olarak, toplumda eksikliği hissedilen demokratik davranışların genç kuşaklara kazandırılması için, okullarda bununla ilgili bir ders konulmasından ziyade, okulların demokratik kültürün yaşanarak öğrenildiği ortamlar haline getirilmesinin daha etkili olacağı vurgulanmaktadır. Yani okulların demokratik okul olması beklenmekte ve bunun için de öğrencilerin bütün derslerde, kendi öğrenmelerinden sorumlu olmaya, diğer arkadaşlarıyla etkin bir iletişime girmeye, işbirliği yapmaya, paylaşmaya, üretmeye, bağlanmaya ve sevmeye bırakılmaları önerilmektedir (Okutan, 2010). Osborne (1998) görev yaptığı ilköğretim okulunda birinci ve üçüncü sınıf öğrencilerine bazı fen konularını (yerçekimi ve enerji) öğretirken, demokratik tutum ve anlayışları da nasıl geliştirmeye çalıştığını betimlemiştir. Osborne (1998), yürüttüğü öğretim etkinliklerini ayrıntılı olarak sunduğu çalışmasında, belirli örnek olaylar üzerinde öğrencileri kendi fikirlerini paylaşmaya, karşıt fikirleri ise saygılı bir şekilde dinlemeye ve eleştirmeye cesaretlendirmiştir. Böylece bazı demokratik tutumları öğretim sürecine taşıyarak, bu tutumları erken yaşlarda kazandırmaya çalışmıştır. Güleç ve Balçık (2009), öğretmenlerin eğitim-öğretim sürecinde, öğrencilerin demokratik kazanımlarının farkına varabilmelerine olanak sağlayacak etkileşimli ortamlar hazırlamaları gerektiğini vurgulamışlardır. Duman ve Şahiner (2008), öğretmenlerin demokratik tutumları geliştirmek için, öğrencilere eğitim-öğretim sürecinde daha fazla sorumluluk vermelerini, sınıfta bir saygı ve güven ortamı yaratmalarını önermişlerdir. Bunun için, öğretmenlerin okul ve sınıf ortamında demokratik kültür dokusunu yapılandıran davranışların kazandırılmasına yönelik model olmaları ve bunu sağlayacak eğitimöğretim yöntemlerini kullanmaları gerekmektedir (Dilekmen, 1999). Fakat böyle demokratik ve paylaşımcı bir eğitim ortamın oluşturulması ve başarının artırılması için, öğretmenlerin de demokrasiye inanmış, demokrasiyi bir yaşam biçimi haline getirmiş bireyler olmaları son derece önemlidir (Çankaya ve Seçkin, 2004). Öğretmenlerin ders işleme ve soru sorma biçimi, öğrencilerle ilişkilerinde takındığı tavır ve onların bir şeyler üretme çabalarını cesaretlendirmesi ya da tam tersi tutumlarının, öğrencilerin psikolojik sağlığını dahi etkilediği bilinmektedir (Dilekmen, 1999; Küçükahmet, 1989). Son zamanlarda yürütülen bir çalışmada (Duman, 2010), öğretmen adaylarında demokratik tutumların gelişmişlik düzeyi arttıkça modern eğitim felsefelerine bakışlarının da geliştiği rapor edilmiştir. Modern eğitim felsefelerinde, öğretmenin daha demokratik, insan haklarına saygılı ve eğitim-öğretim sürecinde de öğrenciyi merkeze alıp ona rehberlik eden ve ona cesaret veren, bir rol oynadığı vurgulanmaktadır (Duman, 2010). Modern eğitim felsefelerini benimseyen eğitimciler; öğrencilerin aktif katılımına dayalı işbirlikli, projeye dayalı, probleme dayalı, sorguya dayalı gibi farklı öğretim stratejilerini kullanabilirler (Şimşek, Doymuş ve Karaçöp, 2009). Sonuç olarak derslerde kullanılan öğretim yöntemleriyle demokratik sınıf ortamı arasında

20 16 Bayburt Üniversitesi Eğitim Fakültesi Dergisi Cilt 5, Sayı I-II (2010) sosyal yeterlik, sorumluluk ve güven duygusuna, empatiye ve iletişim becerilerine sahip bireyler olurlar. Yukarda özetlenen çalışmalarda (Erikson, 1968; Karahan, Sardoğan, Özkamalı ve Dicle, 2006) belirtilen sonuçlar dikkate alınırsa, işbirlikli öğrenme etkinliklerinin, kişilik ve tutum gelişimi üzerinde oldukça faydalı olduğu söylenebilir. Bu çalışmada dokuzuncu sınıf biyoloji dersinde işbirliğine dayalı TGA etkinliklerinin, öğrencilerin demokratik tutumlarına olan etkisinin cinsiyet değişkeni de göz önünde bulundurularak incelenmesi amaçlanmıştır. III. Yöntem Örneklem Bu çalışmada tek gruplu ön-test ve son-test modeli kullanılmıştır (Mcmillan & Schumacher, 2001). Çalışmanın örneklemini, eğitim öğretim yılında Erzurum Anadolu Lisesinde dokuzuncu sınıfta öğrenim gören 17'si erkek ve 13'ü kız olmak üzere toplam 30 öğrenci oluşturmaktadır. Veri toplama aracı Araştırma verileri Gömleksiz (1993) tarafından geliştirilen Demokratik Tutum Ölçeği (DTÖ) kullanılarak toplanmıştır. Ölçek likert tipi, her biri beş alternatiften oluşan (tamamen katılıyorum, katılıyorum, kararsızım, katılmıyorum ve hiç katılmıyorum) toplam 29 tutum maddesinden oluşmaktadır. Bu maddeler ise; eşitlik, bilimsellik, açıklık, düşünceye saygı, birlikte çalışma, katılı, sınıf oyuna başvurma ve azınlık hakları olarak sekiz alt bölüm içinde sınıflandırılmıştır. Ölçekte yer alan olumlu ifadeler, hiç katılmıyorum seçeneğinden başlanarak 1'den 5'e, olumsuz ifadeler ise aynı seçenekten başlayarak 5'den 1'e doğru puanlanmıştır. Ölçekten alınabilecek en düşük puan 29, en yüksek puan ise 145'tir. Ölçme aracının güvenilirlik katsayısı (Cronbach Alpha) bu çalışmada 0.72 olarak hesaplanmıştır. Uygulama Öğrencilerin işbirlikli çalışmalara alıştırılması ve pilot uygulama 2007/2008 Eğitim Öğretim yılı başladıktan 6 hafta sonra araştırmanın örneklem grubunu oluşturacak olan sınıf dört ayrı dokuzuncu sınıf arasından rastgele şekilde belirlenmiştir. Belirlenen sınıfın işbirlikli öğrenme etkinliklerine alıştırılması amacıyla sırasıyla şu uygulamalar yapılmıştır: 1. Öğrencilerin işbirlikli öğrenme çalışmalarıyla tanışmalarını sağlamak ve bu süreçle ilgili başlangıç kaygılarını en aza indirebilmek amacıyla hazırlanan Powerpoint gösterisi öğrencilere sunulmuştur. Bu sunuda işbirlikli öğrenmenin ne olduğu, genel özellikleri, küme çalışmalarından farkları, yararlarının neler olduğu ve araştırmada kullanılacak Öğrenci Takımları Başarı Bölümleri (ÖTBB) yönteminin ve TGA tekniğinin aşamaları hakkında bilgiler verildi. 2. Daha sonra öğrencilere, bir sonraki ders anlatılan konuları içeren bir sınav yapılacağından çalışarak gelmeleri söylenmiştir. 3. Sonraki ders sınav ve demokratik tutum ölçeği (ön-test) uygulanmış, ertesi gün sınav sonuçları ve cinsiyetler göz önünde bulundurularak, bu özellikler bakımından heterojen yapıya sahip 5 kişilik 6 takım oluşturulmuştur. 4. Takımlar oluşturulduktan sonra her bir öğrenci grubundan takımları için birer isim bulmaları istenmiştir. Takım isimlerini belirleyerek takım ruhunu oluşturmaya başlayan öğrencilerle işbirlikli uygulamalara başlanmıştır. 5. Pilot çalışma için Canlıların Temel Bileşenleri ünitesinde yer alan bazı konular (karbonhidratlar, yağlar, proteinler, enzimler, nükleik asitler, vitaminler ve ATP) hazırlanan TGA etkinlikleri içeren çalışma yaprakları kullanılarak işbirlikli gruplarda 4 hafta boyunca ve haftada