G.3
Gençer, Y. ve Çetinkaya, G. (2015). Investigating The
Relationship Between Connectives and Readers’
Reading Comprehension Level. Reading Improvement,
52(2), 39-50. (ERIC tarafından dizinlenmektedir.)
Reading Improvement
PHILLIP FELDMAN, Ed.D.Editor EDITORIAL BOARD RICHARD D. ARNOLD
Purdue University
ANDREA M. KENT
University of South Alabama
RUTH J. BARRON
Edmonton, Canada
JUNE D. KNAFLE
University of Illinois-Chicago
GWYNETH E. BRITTON
Oregon State University
RAYMOND E. LAURITA
Yorktown Heights, New York
DAVID M. BROWN
University of Alabama
A.J. LOWE
University of South Florida
BYRON CALLAWAY
University of Georgia-Athens
RAYMOND F. MORGAN
Old Dominion University-Virginia
SUSIE LAN CASSEL
University of California
BEATRYCE T. NEWTON
University of Alabama-Birmingham
CLYDE G. COLWELL
Kansas State U.-Manhattan Eastern
MARVIN E. OLIVER
Washington State, Washington
THOMAS A. RAKES
The University of Tennessee at Martin
BEVERLY J. DRYDEN
Gwin Oaks Elementary School, GA.
CINDY CHANCE
Georgia Regents University, Augusta
EARL F. RANKIN University of Kentucky ELVA DURAN U. of Texas, El Paso, TX JUDITH C. REIFF University of Georgia MARLOW EDIGER
Northeast Missouri State University
TIM ROBERTS
East Texas State University
ELAINE D. FOWLER University of Texas-Austin MARIAN K. SHAPIRO Lexington, Massachusetts YVON GAUTHIER Laurentian University FRAN SILVERBLANK
New York University
GERALD G. GLASS
Adelphia University, NY
PATRICIA SMITH
West Virginia University, Morgantown
J. DIXON HEARNE
Claremont Grad. School
TROY G. SULLIVAN
University of Alaska
HUGO HARTIG
University of Wisconsin-Oshkosh
FRANCES TRIGS
Committee on Diagnostic Reading Tests, NC
DENNIS J. KEAR
Wichita State University
THOMAS D. YAWKEY
Penn State University, Pennsylvania
Journal Purpose
Reading Improvement publishes reports of investigations and creative theoretical papers dealing with every aspect of reading improvement, and at all different levels of instruc-tion. Preference is given to manuscripts that give promise to better understanding of the teaching of reading and for improving the process. Manuscripts should be submitted through the Project Innovation web site: http://www.pro-jectinnovation.com. Manuscripts and must be prepared to conform to the style and procedures described in the Publi-cation Manual of the American Psychological Association.
Indexed
Articles in Reading Improvement are included in Read-ing Abstracts, Educational Index, and ERIC. Appropriate articles are indexed and abstracted in Language and Lan-guage Behavior Abstracts and Women Studies Abstracts.
Copyrighted Material
Permission must be requested for reprinting of all tables, figures, and text in excess of 200 words (total for one article) in length.
Submissions
Manuscripts must be accompanied by an abstract of 100-120 words and appear at the beginning of the manuscript. It should contain statements of (a) problem, (b) meth-od, (c) results, and (d) conclusions when appropriate. It should provide the reader with an idea of the theme and scope of the article. Manuscript should be double spaced.
Editorial Office
PROJECT INNOVATION INC. P.O. Box 8508 Mobile, Alabama 36608. Subscriber Information US phone: 1-800-633-4931 Non US phone: 205-995-1597 Fax: 205-995-1588 Email: projectinnovation@subscriptionoffice.com Mail: Project Innovation Subscription Office
PO BOX 361
Birmingham, AL 35201-0361 Reading Improvement is published quarterly.
Institutional Subscription (1 year)
US customers
Online Only ...$135 Print and Online ...$190 Canadian subscriptions: Add $15 per year Other international subscriptions: Add $40 per year Microfilmed by University Micro Films, Inc., Ann Arbor, Michigan 48106. Abstracted in The Chicorel Abstracts to Reading and Learning Disabilities.
Printed and circulated by PPF.
© Copyright 2013 by Project Innovation Inc., Mobile, Alabama.
Investigating the Relationship between Connectives and Readers’ Reading Comprehension Level
...Yusuf Gençer, Gökhan Çetiİnkaya 39
“Do I Have to Leave?” Beyond Linear Text: Struggling Readers’ Motivation with an Innovative Musical Program
Susan V. Bennett, Cynthia Calderone, Robert F. Dedrick, AnnMarie Alberton Gunn 51 Creating and Sharing Annotated Bibliographies: One Way to Become Familiar with
Exemplary Multicultural Literature
...Janet C. Richards 61 Listening in the Language Arts
... Marlow Ediger 69
Classic Reprint
Learning Together through Retrospective Miscue Analysis: Salem’s Case Study
Originally published volume 44, Number 3, Fall 2006.
... Karima M. Almazroui 72
39
INVESTIGATING THE RELATIONSHIP BETWEEN
CONNECTIVES AND READERS’ READING
COMPREHENSION LEVEL
Yusuf GençerNigde University, Turkey
Gökhan çetiİnkaYa
Nigde University, Turkey
Abstract
The purpose of this study is to investigate the relationship between connectives in Turkish texts and readers’ reading comprehension. Re-search was conducted with a total of 50 teachers. In the study group, readers’ reading comprehension was determined through 10 descrip-tive texts by using open-ended questions. The results of the analysis revealed that, while there is no statistically significant correlation be-tween the reading comprehension scores of the good readers and the numbers of the connectives such as temporal, causal, adversative and additive in the texts forming the study database, a negative correla-tion was found between the reading comprehension scores of the poor readers and concession connectives at the level of r = -0.805 (p>0.05) and the connectives expressing expansion at the level of r = -0.647 (p>0.05).
Keywords: Coherence, Language education, Discourse markers, Co-hesive markers, Conjunctions
Introduction
In order for the reading process to contin-ue without interruption and result in compre-hension, it is a prerequisite to provide readers a text that is well-formed based on certain criteria. Ebrahimpourtaher (2011) states that the act of reading is an interactive process en-suring the communication between the reader and the writer discursively through a written text. When propositions come together, they form their unique pattern, and a structure having distinguishing characteristics from other texts emerges. “Cohesion” is achieved through cohesive devices that help to ensure semantic fluency, and these devices have an
important role in the reading process. Halli-day and Hassan (1976:3-4) define cohesion as “distinguishing a text from other statements that are not texts, enabling the parts of a text to stick together, and organizing the mean-ing relationships in a text (cited in Coskun 2007:240). Connecting elements are de-scribed as “the elements that connect related words, groups of words, and particularly sen-tences, and link these in terms of meaning and sometimes structure (Atabay et al. 1983:49). Connectives such as “because,” “but” and “after” are means of making connections and help skilled readers in their reading process by emphasizing the relationship between
events (Sanders and Noordman 2000). Con-nectives point to the existence of a relation-ship between certain events and thus, provide opportunities for different formations on how to integrate the information (Gernsbacher 1997; cited in Cain and Nash 2011). In the literature, they are also named as connecting elements, conjunctions, linking element, dis-course markers and connectors.
In the Penn Discourse TreeBank (PDTB) and METU Turkish Corpus projects, discourse connectives are categorized based on both their syntactic and functional characteristics. In terms of syntax, there are three categories including (1) coordinating conjunctions, (2) subordinating conjunctions, and (3) discourse adverbials (Forbes-Riley et al. 2006; Zeyrek and Webber 2008; Kurtul 2011).
Halliday and Hasan (1976) categorized connective elements functionally as follows:
a. Temporal connective elements b. Causal connective elements c. Adversative connective elements d. Additive connective elements
Coordinating conjunctions are divided as simple and correlative coordinating con-junctions. Examples of simple coordinating conjunctions include but, because, and, so. As for correlative conjunctions, the examples include both...and, either....or, neither....nor (Kurtul 2011:60-62).
Subordinating conjunctions are divided as simple and compound subordinating con-junctions. Simple subordinating conjunctions function as connecting the main clause with sub-clauses. Compound subordinating con-junctions are formed with the combination of a suffixal lexical item and nominalizer affix or case affix. While examples of simple sub-ordinating conjunctions include gerund af-fixes, compound subordinating conjunctions in Turkish can be as…, as much as, when…, since…, like…, because of…, after…,
before…, until…, if…, despite… and so on (Kurtul 2011:63-64).
According to Chaudron and Richers (1986), discourse markers are divided into two groups. Macro Discourse Markers determine the direction of a text by emphasizing the im-portant information and listing it or pointing its importance. Micro Discourse Markers are the words showing the relationship between the sentences or filling the gaps in a text. For example, some discourse markers in Turkish include: for this reason, otherwise, unlike, in spite of this, other than this, moreover, firstly, as a result, for instance.
Temporal connectives construct a time and order relationship by showing the events in a text happen before, after or simultaneously with each other (Coskun 2005:83). Consider-ing various classifications, it is clear that they do not differ to a large extent. Researchers making these classifications agree that tempo-ral connectives show order and simultaneity in time (İssever 1995; Balyemez 2010; Kurtul 2011). Halliday and Hasan (1976) exemplify such connectives as follows: after that, at the same time, before this, finally, at first, later, at last, initially, shortly, next time, another time, in the meantime, until ..., so far, briefly etc.
Causal connectives point out the situations where events or cases mentioned in the 1st and 2nd members are causally affected by each oth-er (Kurtul 2011:69). Such connectives make a connection among the units in the regional or holistic level of a text based on causality (Is-sever 1995:101). Halliday and Hasan (1976) exemplify such causal connectives as follows: for this reason, thus, accordingly, because, therefore, as a result, in this case, under these circumstances, otherwise etc. There are several ways of classfying such connective elements.
Kurtul (2011:69-73) examines causal connective elements as four categories in-cluding cause, pragmatic cause, condition and conditional cause. On the other hand, Issever (1995:101) categorizes this type of connective
The Relationship Between Connectives and Reading Comprehension Level / 41 elements based on cause-effect, effect-cause
and purpose relationships.
Adversative connectives are used to em-phasize the difference between two elements, present the cases where one of the two ele-ments is valid, and connect two opposite elements to each other. Halliday and Hasan (1976) exemplify such adversative connec-tives as follows: but, however, if only, still, in spite of this, in fact, yet, on the other hand, at the same time, instead of, at least etc.
Kurtul (2011) describes such connectives in four categories including adverseness, pragmatic adverseness, dissimilarity and pragmatic adverseness.
Additive connective elements are used to expand the discourse, take the topic further (Kurtul 2011), list words, groups of words or sentences with the same function or provide new information (Coskun 2005) and show that two propositions that are linked to each other are in an additive relationship (Issever 1995). In the literature, it can be seen that additive connectives are also named as expansion con-nectives (Issever 1995; Coskun 2005).
Halliday and Hasan (1976) exemplify such additive connectives as follows: and, also, ei-ther...or, or, with, other than this, moreover, for instance, thus, similarly, on the other hand.
Finally, examining the studies revealing the relationship between connectives and readers’ level of reading comprehension, different findings have been reported. While some studies (Sanders and Noordman 2000; Degand and Sanders 2002; Innajih 2006; Wif-ield and Tomitch 2012) found that connective elements had a facilitative effect on reading comprehension, these elements did not have any effect in other studies (Irwin 1982; Geva 1986; Murray 1995).
Significance and Aim of the Study
Reading has an importance place in hu-man life. Comprehension is the prerequisite of reading. If there is no comprehension, it
means that reading is not successful (Ciftci and Temizyurek 2008). Considering that text is in the centre of reading education, the im-portance of the text becomes more clear. In order for the reading comprehension process to be effective and successful, the text should have certain characteristics. One of these characteristics is the role of connectives en-abling the cohesion and coherence of the text in the reading comprehension process. This is where the significance of this study emerges.
Considering the studies in the literature, it is clear that connectives are important vari-ables in predicting reading comprehension (Irwin 1982; Geva 1986; Murray 1995). It can be seen that the studies on connectives in Turkish have been very limited in number and focused mostly on students’ use of connec-tives in their writings. Moreover, no studies on the relationship between connectives and reading comprehension regarding Turkish have been conducted. From this perspective, the significance of this study increases in the sense that it would fill a gap in the literature and provide data regarding the issue.
The findings of this study are expected to point out some characteristics for both au-thors in producing texts for a certain audience and teachers in selecting texts to use in their classrooms.
The primary aim of this study was to ex-amine the relationship between connectives in Turkish texts and readers’ level of reading comprehension.
In this regard, the following research questions guided this study:
1. Is there a relationship between Tem-poral, Causal, Adversative and Ad-ditive connectives and good readers’ level of reading comprehension? 2. Is there a relationship between
Tem-poral, Causal, Adversative and Ad-ditive connectives and poor readers’ level of reading comprehension?
3. Is there a relationship between Sub-ordinating conjunctions, coordinative conjunctions and discourse markers and good readers’ level of reading comprehension?
4. Is there a relationship between Sub-ordinating conjunctions, coordinative conjunctions and discourse markers and poor readers’ level of reading comprehension?
Methodology
This section presents information regard-ing the research design, participants, data gathering tool and data analysis.
Research Design
This study aiming to examine the relation-ship between the connectives in Turkish texts and good and poor readers’ level of reading comprehension separately employed correla-tional survey design. Since it was aimed to describe a case and tried to define the research topic in its own conditions, correlational sur-vey designed seemed suitable for the study.
Correlational survey models are the research designs aiming to determine the existence of variance between two or more variables and/or its extent (Karasar 1994).
Participants
The participants of the study were the teacher candidates studying at the Turkish Language Teaching Department of Nigde University Education Faculty in the spring term of 2012-2013 academic year. Based on the aim of the study, the participants were di-vided into two groups as “good readers” and “poor readers” after a reading comprehension test was administered. Those who got 55 or less points in the test were defined as poor, and those who got 56 or more as good read-ers. The study was conducted with 50 teacher candidates, 25 in the poor readers group and 25 in the good readers group
Data Gathering Tool
To determine good and poor readers in the first phase of the study, a descriptive text titled as “Lion” from the Britannica Basic Education and Culture Encyclopedia (1992) which was suitable for the level of readers in the sample in terms of readability. Readability level and information density were consid-ered in the text selection. In determining the readability level of the text, “Cetinkaya-Uzun Readability Formula” (Cetinkaya 2010) that is a valid evaluation tool in defining and cate-gorizing Turkish texts in terms of readability was used.
In the second phase of the study, 10 de-scriptive texts were employed. These 10 texts whose names are given in Table (3) were taken from the Britannica Basic Educa-tion and Culture Encyclopedia (1992). In the selection of texts that would be used as data gathering tool, two criteria were considered. One of these is the step of identifying the readability level using Cetinkaya-Uzun read-ability formula as mentioned above. Texts which were close to each other in terms of readability were selected. Table 1 presents the “average word length,” “average sen-tence length” and “readability scores” of the texts used in the study.
Table 1. Readability Levels of the Texts Used in the Study
Title AWL ASL RS
Hawk 2.6 15.05 38 The Mediterranean 2.72 14.14 34 Bicycle 2.78 15.81 31 Shoes 2.68 14.57 35 Slang 2.76 14.89 33 Barometer 2.8 15.44 31 The Atlantic 2.71 16.74 32 Ice Age 2.56 15.23 38 Ahi Community 2.78 13.58 33 Bullfighting 2.89 11.18 32
The Relationship Between Connectives and Reading Comprehension Level / 43 As can be seen in Table 1, the average
word length was between 2.56–2.89, the average sentence length was between 11.18– 16.74 and the readability scores were between 31-38.
The second criteria considered in the text selection was that there should not be perfect linear relationship between the independent variables. In this sense, the connectives in the texts were analyzed semantically and structurally (see the section on identifying the variables, Table 2 and Table 3), and the texts having different linear relationship in terms of variables were included in the study. The reason is that the “b” value would be the same between the variables having a perfect linear relationship statistically. In other words, we cannot say which variable is important (Field 2005; cited in Cetinkaya 2010).
Briefly, in this study, two latent variables were identified based on definition and clas-sification of connectives structurally and se-mantically, and the relationship between the variables and the good and poor readers’ level of reading comprehension was investigated. There are also observable variables under the two variables identified. To ensure the reli-ability of the coding, randomly selected two texts out of the 10 texts used were analyzed by two other researchers. The consistency
between the researchers was reliable as it was calculated for the semantic analysis of con-nectives (96%) and the structural analysis of the connectives (100%).
At this step, the following formula was used (Tavsancil and Aslan 2001):
Reliability = Agreement rate/ Agreement + Disagreement rate
The data were analyzed in accordance with the conceptual framework formed be-forehand. The structural analysis of connec-tives was based on the structural categoriza-tion mencategoriza-tioned above, and as for the semantic analysis, the semantic categorization of con-nectives was used.
The number of connectives structurally and semantically with regard to the 10 texts used as the data gathering tool is presented in Table 2 and Table 3.
As can be seen in Table 2, the number of connectives structurally was between 15-26 in subordinating conjunctions, 21-48 in co-ordinating conjunctions, and 1-9 in discourse markers.
As shown in Table 3, the number of connectives semantically was distributed as between 10-25 in temporal connectives, 25-42 in causal connectives, 5-13 in adversative connectives, and 3-13 in additive connectives.
Table 2. Type and Number of Connectives Structurally
Title Subordinating Coordinating Discourse Marker
Hawk 24 44 9 The Mediterranean 16 42 6 Bicycle 22 29 8 Shoes 20 48 3 Slang 26 41 4 Barometer 20 21 4 The Atlantic 15 30 1 Ice Age 25 36 3 Community 20 34 5 Bullfighting 24 39 1
Data Gathering and Analysis
The process of reading each text and an-swering questions related to it was completed in approximately 30 minutes. The implemen-tation was conducted as one text per week and the study was completed in 10 weeks. An instruction on the application was presented before proceeding to the reading comprehen-sion test. This instruction is as follows: a) Please read the whole text before
answer-ing the questions.
b) Please answer all the questions. If you en-counter a difficult question, proceed to the next one, come back to it later and try to answer it.
c) Spelling mistakes will not be evaluated, so do your best.
d) Please write in a readable form. e) Do you have any questions? f) Please start.
The answers of the participants were gath-ered through the answer evaluation form and evaluated based on this form.
The question number column (#) in Table 4 gives the order of each question. The second column indicates the category of questions.
The other three columns are related to scoring procedure and evaluation. Finally, the last row is for writing the total score.
In addition, the correlation between the scoring of the researcher and two field experts was also calculated. The findings of this anal-ysis are presented in Table 5.
As seen in Table 5, the correlation be-tween the scoring of the researcher and the other two raters was positive and high. This finding shows that the scoring reliability is high (Turgut 1977).
The statistical technique used in the analy-sis of the answers is presented below, and the significance level used for rejecting or accept-ing the hypotheses for the research questions was 0.05.
The relationships between the selected variables and the reading comprehension scores were calculated using Pearson Cor-relation Coefficient, and the calculated cor-relation coefficients were tested against the hypothesis that “the correlation coefficient of the universe is equal to zero.”
Results
The aim of this study was to examine the relationship between connectives in Turkish texts and readers’ level of reading
Table 3. Type and Number of Connectives Semantically
Title Temporal Causal Adversative Additive
Hawk 18 38 13 8 The Mediterranean 10 33 13 8 Bicycle 13 34 6 6 Shoes 10 36 12 13 Slang 17 34 9 11 Barometer 10 25 5 5 The Atlantic 12 25 5 4 Ice Age 25 29 7 3 Ahi Community 14 33 5 7 Bullfighting 11 42 5 6
The Relationship Between Connectives and Reading Comprehension Level / 45
Table 4. Answer Evaluation Form
# Name of Reader Text Title
1 Level of Knowledge Definition not written
Definition written partially
Definition written com-pletely
0 point 12 points 25 points 2 Level of interpretation
No interpretation Partial interpretation Full interpretation
0 points 12 points 25 points
3 Level of Using Knowl-edge and Experiences
Not related their knowl-edge and experiences with the text
Partially related their knowledge and experi-ences with the text
Successfully related their knowledge and experiences with the text
0 points 12 points 25 points
4 Level of Evaluation No evaluation Partial evaluation Successful evaluation 0 points 12 points 25 points
Total score 100 points
Table 5. Correlation between the Scoring of the Researcher and the Raters
# Text Title Researcher First Rater Second Rater
1 Hawk .748 .803 2 The Mediterranean .920 .901 3 Bicycle .808 .836 4 Shoes .792 .790 5 Slang .933 .842 6 Barometer .896 .809 7 The Atlantic .702 .746 8 Ice Age .902 .811 9 Ahi Community .742 .703 10 Bullfighting .752 .749
comprehension. In line with this aim, the find-ings revealed in the analysis of the data gath-ered through the tools described in the meth-od section are presented based on the research questions using tables and interpretations.
First Research Question
The first research question of the study was “Is there a relationship between
Temporal, Causal, Adversative and Addi-tive connecAddi-tives and good readers’ level of reading comprehension?.” The findings are
presented in Table 6.
As can be seen in Table 6, there was no significant relationship between the good readers’ level of reading comprehension and “temporal connectives” r= -0.271 (p>0.05), “causal connectives” r= -0.337 (p>0.05),
“adversative connectives” r= -0.523 (p>0.05) and “additive connectives” r= -0.337 (p>0.05).
The findings revealed that there is no sig-nificant relationship between good readers’ level of reading comprehension and temporal, causal, adversative and additive connectives. However, there was a weak, negative relation-ship between the temporal connectives such as but and while “used in situations where
events or cases mentioned in the 1st and 2nd members are followed by each other or over-lapped” (Kurtul 2011), causal connectives
such as because, for this reason and since
“in-dicating that events or cases mentioned in the 1st and 2nd members are causally affected by each other” (Kurtul 2011), and additive
con-nectives such as for instance, thus, either…
or and otherwise which expands the discourse
and takes the topic further. On the other hand, there was a moderate, negative correlation between the good readers’ level of reading
Table 6. The Relationship between Temporal, Causal, Adversative and Additive Connectives and Good Readers’ Level of Reading Comprehension (GRLRC)
1 2 3 4 5 1 GRLRC N=25 r 1 2 Temporal N=25 r -.271 1 3 Causal N=25 r -.337 -.056 1 4 Adversative N=25 r -.523 .033 .363 1 5 Additive N=25 r -.252 -.286 .491 .647(*) 1
* The significance level is .05.
Table 7. The Relationship between Temporal, Causal, Adversative and Additive Connectives and Poor Readers’ Level of Reading Comprehension (PRLRC)
1 2 3 4 5 1 PRLRC N=25 r 1 2 Temporal N=25 r .075 1 3 Causal N=25 r -.525 -.056 1 4 Adversative N=25 r -.805(*) .033 .363 1 5 Additive N=25 r -.647(*) -.286 .491 .647(*) 1
* The significance level is .0 5.
comprehension and adversative connectives such as but and although “making a discourse
relationship to emphasize the differences be-tween the situations pointed by the first and the second members” (Kurtul 2011). Since the
relationships are not statistically significant, interpretations cannot be made regarding these relationships.
Second Research Question
The second research question of the study was “Is there a relationship between Temporal,
Causal, Adversative and Additive connectives and poor readers’ level of reading comprehen-sion?.” The findings are presented in Table 7.
As can be seen in Table 7, there was no significant relationship between the poor readers’ level of reading comprehension and “temporal connectives” r= 0.075 (p>0.05), “causal connectives” r= -0.525 (p>0.05), but a negative, moderate and significant
The Relationship Between Connectives and Reading Comprehension Level / 47 relationship with regard to “adversative
con-nectives” r= -0.805 (p>0.05) and “additive connectives” r= -0.647 (p<0.05).
The findings above show that there is no relationship between the poor readers’ level of reading comprehension and temporal and causal connectives, but there is a negative and significant relationship in terms of adversative and additive connectives. In other words, as the number of adversative connectives such as but and although and additive connectives such as for example, thus, either….or and
otherwise increases, the poor readers’ level of
reading comprehension decreases.
Third Research Question
The third research question of the study was “Is there a relationship between
Sub-ordinating conjunctions, coordinative con-junctions and discourse markers and good readers’ level of reading comprehension?.”
The findings are presented in Table (8) below. As shown in Table 8, there was no signif-icant relationship between the good readers’ level of reading comprehension and
“subordi-nating conjunctions” r= -0.187 (p>0.05), “co-ordinative conjunctions” r= -0.368 (p>0.05),
but a negative, strong and significant relation-ship with regard to “discourse markers” r= -0.760 (p<0.05).
The findings revealed that there is no significant relationship between good read-ers’ level of reading comprehension and
subordinating and coordinative conjunctions. However, there seems to be a strong, negative relationship between the discourse markers such as otherwise, for this reason, for
exam-ple, and for instance and the good readers’
level of reading comprehension.
Fourth Research Question
The forth research question of the study was “Is there a relationship between
Subor-dinating conjunctions, coordinative conjunc-tions and discourse markers and poor read-ers’ level of reading comprehension?.” The
findings are presented in Table 9.
As shown in Table 9, there was no signif-icant relationship between the poor readers’ level of reading comprehension and
“subor-dinating conjunctions” r= 0.096 (p>0.05),
“discourse markers” r= -0.600 (p>0.05), but a negative, strong and significant relationship with regard to “coordinative conjunctions” r= -0.740 (p<0.05).
The findings revealed that there is no sig-nificant relationship between poor readers’ level of reading comprehension and sub-ordinating and coordinative conjunctions. However, there is a moderate, negative relationship between the discourse markers such as also, in fact, before, after and both….
and and the poor readers’ level of reading
comprehension.
Table 8. The Relationship between Subordinating Conjunctions, Coordinative Conjunctions and Discourse Markers and Good Readers’ Level of Reading Comprehension (GRLRC) 1 2 3 4 1 GRLRC N=25 r 1 2 Subordinating N=25 r -.187 1 3 Coordinating N=25 r -.368 .227 1 4 Discourse Marker N=25 r -.760(*) .148 .089 1
Discussion
The finding that there is no significant re-lationship between the connective elements categorized semantically and the good read-ers’ level of reading comprehension can be interpreted as that the readers were able to make the connection between propositions through the connective elements. When com-pared with the findings of other studies in the literature, this case seems to be parallel with Haberlandt’s view (1982) that conjunctions makes it faster to comprehend sentences by meeting the readers’ expectations regarding the relationship between two sentences. On the other hand, the findings regarding the relationship between the poor readers’ level of reading comprehension and adversative and additive connectives do not support this view of Haberlandt’s. In the study of Mey-er et al. (1980), those having poor reading skills are defined as students lacking the knowledge of high level text structures that could be used to comprehend and remember a text. At the same time, based on the find-ings of their study, these researchers stated that students with poor reading skills did not have text knowledge that would lead them to identify and use conjunctions (cited in Ben-Anath 2005).
Examining the findings regarding the re-lationship between the connective elements categorized structurally and good and poor
readers’ levels of reading comprehension, discourse markers such as in addition, firstly,
as a result decreased the good readers’ level
of comprehension significantly, and coor-dinative conjunctions such as on the other
hand, and, or, or, both…and and neither…. nor decreased the poor readers’ level of
comprehension significantly. The fact that the coordinative conjunctions used in Turk-ish are borrowed from foreign languages and the relationships between propositions in Turkish are originally achieved through gerunds can be seen as a reason that such connective elements negatively affected reading comprehension. As for the subordi-nating conjunctions, they positively affected the poor readers’ comprehension, but were observed not to be effective in the good read-ers’ comprehension. This case can be due to the fact that a sub-clause is an element of a main clause through a set of affixes added to a verb, and consequently, the processing be-comes easier as a result of the simplification of the syntactic structure of the main clause. Arya et al. (2011) stated that in their study, the syntactic structure did not affect the students’ reading comprehension either negatively or positively. On the other hand, the findings of Cain and Nash (2011) showed that the con-nective elements helped developing readers to process the text, but did not have any pos-itive effects on developed readers.
Table 9. The Relationship between Subordinating Conjunctions, Coordinative Conjunctions and Discourse Markers and Poor Readers’ Level of Reading Comprehension (PRLRC) 1 2 3 4 1 PRLRC N=25 r 1 2 Subordinating N=25 r .096 1 3 Coordinating N=25 r -.740* .227 1 4 Discourse Marker N=25 r -.600 .148 .089 1
The Relationship Between Connectives and Reading Comprehension Level / 49 As a result, it was found that there was
a negative, significant relationship between adversative connectives, additive connectives and discourse markers and the poor readers’ level of reading comprehension while there was a negative, significant relationship be-tween discourse markers and the good read-ers’ level of reading comprehension.
According Cetinkaya’s etc. (2014) re-search findings; ; the students’ connectives knowledge test points average who never or rarely read is lower than the students who read more frequently. Similarly, the stu-dents’ average with low writing frequency was found lower than students with more frequent writing.
Conclusion
In this study aiming to investigate the re-lationship between the connective elements in Turkish texts and readers’ level of reading comprehension, the relationship between both the good readers’ and the poor readers’ levels of reading comprehension was exam-ined. The level of reading comprehension was determined by means of 10 descriptive texts and open-ended questions. On the other hand, the connective elements in the texts used for determining the readers’ level of reading comprehension were identified, and analyzed structurally and semantically. At this step, based on the syntactic criteria, the connectives in the Turkish texts were cate-gorized as “subordinating conjunctions,” “coordinative conjunctions” and “discourse markers.” Moreover, in terms of semantics, the connective elements were categorized based on being “temporal,” “causal,” “ad-versative” and “additive.”
The analysis revealed that while there was no significant relationship between the good readers’ level of reading comprehension and temporal connectives, causal connectives, adversative connectives and additive con-nectives, there was a negative, moderate and
significant relationship with regard to adver-sative connectives, r= -0.805 (p>0.05) and additive connectives, r= -0.647 (p<0.05).
Furthermore, based on the structural variables, there was a negative, significant relationship between the good readers’ lev-el of reading comprehension and discourse markers, r= -0.760, as there was also a nega-tive, significant relationship between the poor readers’ level of reading comprehension and subordinating conjunctions, r= -0.740.
Recommendations
In accordance with the results of the research, it would be reasonable for the teachers or text writers that would choose a reading text for a certain reader group to choose the texts that would be suitable for the levels of reading skills of readers in terms of structural features of the text because the texts with high reading-difficulty levels, or outside of the area of interest of readers not only would not improve the reading motiva-tion and reading habit of students, but also would cause students to have a negative atti-tude towards reading.
Author’s Note
This study is complied from master the-sis titled as “Investigating The Relationship Between Connectives In Turkish Texts and Readers Reading Comprehension Level” Which was defended at the Institude of Ed-ucational Sciences of Nigde University, 2013 by the supervision of Assist. Prof. Dr. Gokhan Cetinkaya.
References
Ana Yayincilik (1992). Temel Britannica Temel Egitim ve Kultur Ansiklopedisi. Istanbul: Ana Yayincilik. Arya, D. J., Hiebert, E. H. & Pearson, P. D. (2011). The
effects of syntactic and lexical complexity on the comprehension of elementary science texts. Interna-tional Electronic Journal of Elementary Education, 4 (1): 107-125.
Atabay, N., Ozel, S. & Kutluk, I. (2003). Baglac. In: Do-gan Aksan (Ed.): Sozcuk Turleri. Istanbul: Papatya Yayincilik, pp. 125-154.
Balyemez, S. (2010). Bagdasiklik ve dede korkut hikâye-leri’nde bagdasiklik gorunumleri. Turkish Studies, 5 (4): 133-173.
Ben-Anath, D. (2005). The role of connectives in text comprehension. Columbia University Working Pa-pers in Tesol and Applied Linguistics, 5 (2): 1-27. Cain, K.& Nash, H. (2011). The influence of connectives
on young readers processing and comprehension of text. Journal of Educational Psychology, 103: 429-441.
Cetinkaya, G., Ulper, H. & Bayat, N. (2014). Baglayici testinin gelistirilmesi ve ogrencilerin baglayici bilg-isinin cesitli degiskenlere gore incelenmesi. Interna-tional Journal of Language Academy, 2(3): 88-98. Cetinkaya, G. (2010). Turkce Metinlerin Okunabilirlik
Duzeylerinin Tanimlanmasi ve Siniflandirilmasi. (Ph. D. Thesis, Unpublished). Ankara: Ankara University. Chaudron, C., Richards J. C. (1986). The effect of dis-course marker on the comprehension of lectures. Applied Linguistics, 7: 113-127.
Ciftci, O., Temizyurek, F. (2008). Ilkogretim 5. sinif ogrencilerinin okudugunu anlama becerilerinin ol-culmesi. Mustafa Kemal Universitesi Sosyal Bilimler Enstitusu Dergisi, 5: 109-129.
Coskun, E. (2007). Turkce Ogretiminde Metin Bilgisi. In: Ahmet Kirkkilic, Hayati Akyol (Eds.): Ilkogretimde Turkce Ogretimi. Ankara: Pegem Yayincilik, pp. 233-279.
Coskun, E. (2005). Ilkogretim Ogrencilerinin Oykuleyi-ci Anlatimlarinda Bagdasiklik, Tutarlilik ve Metin Elementleri. (Ph. D. Thesis, Unpublished). Gazi University, Ankara.
Degand, L., Sanders, T. (2002). The impact of relational markers on expository text comprehension in L1 and L2. Reading and Writing, 15 (7-8): 739-758. Ebrahimpourtaher, A. (2011). Awareness of Referential
Devices in Text and Reading Comprehension. Paper presented in 1st International Conference on Foreign Language Teaching and Applied Linguistics, Saraje-vo, 2011.
Forbes-Riley, K., Webber, B. & Joshi, A. (2006). Com-puting discourse semantics: The predicate-argument
semantics of discourse connectives in D-LTAG. Journal of Semantics, 23: 55-106.
Geva, E. (1986). Reading comprehension in a second language; The role of conjunctions. TESOL Canada Journal, Special Issue1 (NOV): 85-96.
Haberlandt, K. (1982). Reader Expectations In Text Comprehension. In. J-F Le Ny, W Kintsch (Eds.): Language and Comprehension. Amsterdam: North-Holland, pp. 239-249.
Halliday, M., Hasan, R. (1976). Cohesion in English. New York: Longman Group UK. Limited.
Innajih, A. (2006). The impact of textual cohesive con-junctions on the reading comprehension of foreign language students. ARECLS e-journal, 3: 1-20. Irwin, J. R. (1982). The effect of coherence explicitness
on college readers’ prose comprehension. Journal of Reading Behaviour, 14 (3): 275-284.
Issever, S. (1995). Turkce Metinlerdeki Baglanti Ogeleri-nin Metinbilim ve Kullanimbilim Acisindan Islevleri. (Master Thesis, Unpublished). Ankara University, Ankara.
Karasar, N. (1994). Bilimsel Arastirma Yontemi: Ka-vramlar, Ilkeler, Teknikler. Ankara: 3A Arastirma Egitim Danismanlik Ltd.
Kurtul, K. (2011). Turkce ve Ingilizce’deki Baglaclarin Yazili Metinlerde Kullanimi. (Ph. D. Thesis, Unpub-lished). Ankara University, Ankara.
Murray, J. D. (1995). Logical Connectives and Local Co-herence. In: RF Lorch Jr, EJ O’Brien (Eds.): Sources of Coherence in Reading. Hillsdale, NJ: Lawrence Erlbaum, pp. 107-125.
Sanders, T. J. M., Noordman, L. G. M. (2000). The role of coherence relations and their linguistic markers in text processing. Discourse processes, 29 (1): 37-60. Tavsancil, E., Aslan, A. E. (2001). Sozel Yazili ve Diger
Materyaller Icin Icerik Analizi ve Uygulama Or-nekleri . Istanbul: Yeni Cizgi Yayin Dagitim. The Rand Study Group (2002). Reading for
understand-ing. Toward an R and D program in reading compre-hension. Santa Monica, CA: Rand.
Turgut, M. F. (1977). Egitimde Olcme ve Degerlendirme Metotlari. Ankara: Nuve Matbaasi.
Wifield, C. M., Tomitch, L. M. B. (2012). Revising the influence of conjunctions in L2 reading comprehen-sion. Revista Intercâmbio, 26: 16-41.
Zeyrek, D., Webber, B. (2008). A Discourse Resource for Turkish: Annotating Discourse Connectives in the METU Corpus. The 6th Workshop on Asian Language Resources, The Third International Joint Conference on Natural Language Processing, (IJN-LP). 65-71.
