Nahatame, S. (2016, August). Text coherence perceived by Japanese learners of English: The role of sentence connectivity and language proficiency. Poster presented at the Annual Conference of the Euro Second Language Association, Jyväskylä, Finland.

1. Text Coherence Perceived by Japanese Learners of English:
The Role of Sentence Connectivity and Language Proficiency
Shingo Nahatame
Kyoei University, Japan
Horiba, Y. (1996). The role of elaborations in L2 comprehension: The effect of encoding task on recall. The
Modern Language Journal, 80, 151–164. doi: 10.1111/j.1540-4781.1996.tb01155.x
Landauer, T. K., Foltz, P. W., & Laham, D. (1998). An introduction to Latent Semantic Analysis. Discourse Processes,
25, 259–284. doi: 10.1080/01638539809545028
Nahatame, S. (2014, March). Effects of semantic and causal relatedness on text memory and coherence judgment
by Japanese EFL learners. Poster presented at the Annual Conference of the American Association for
Applied Linguistics, Portland.
Todaro, S., Millis, K., & Dandotkar, S. (2010). The impact of semantic and causal relatedness and reading skill on
standards of coherence. Discourse Processes, 47, 421–446. doi: 0.1080/01638530903253825
Wolfe, M. B. W., Magliano, J. P., & Larsen, B. (2005). Causal and semantic relatedness in discourse understanding
and representation. Discourse Processes, 39, 165–187. doi: 10.1080/0163853X.2005.9651678
References
Introduction.
• Successful text comprehension requires not only word recognition and syntactic
parsing, but also an understanding of connectivity between sentences.
• Studies suggest two important types of sentence connectivity for
comprehension (e.g., Todaro et al., 2010; Wolfe, 2005):
 Causal Relatedness (CR) or cause-and-effect relations
 Semantic Relatedness (SR) or semantic overlap between propositions
e.g., Mary could not find anything to read in the library. She went to the bookstore
to get new books. (This text is high in CR and SR.)
• L1 readers perceive both CR and SR between sentences; however, skilled
readers emphasize CR more in their perception, whereas less skilled readers
emphasize SR more in their perception (Todaro, et al., 2010). 1
• Previous studies found L2 learners perceive CR between sentences (Horiba,
1996); however, few studies have explored whether and how they perceive SR.
Research Questions (RQs).
RQ1: How do CR and SR between sentences influence L2 learners’ perception of
text coherence?
RQ2: How do the effects of CR and SR differ according to learners’ L2 proficiency?
• The present study is an approximate replication of the author’s previous study
(Nahatame, 2014).
Introduction
• The main effects of CR and SR were found, with higher ratings for sentences
high in relatedness than those low in relatedness (see Table 3 in Appendix)
• A significant interaction between CR and SR was found, indicating that CR
effects were more prominent for SR-low than SR-high, F(1, 53) = 181.49, p < .001,
ηG
2 = .61; F(1, 47) = 34.07, p < .001, ηG
2 = .25, whereas SR effects were more
prominent for CR-low than CR-high, F(1, 100) = 168.38, p < .001, ηG
2 = .63; F(1, 100) =
92.14, p < .001, ηG
2 = .48
 CR and SR had both independent and interacting influences on judgments.
• A significant interaction between CR and L2 proficiency was found, indicating
that CR more greatly affected judgments made by the higher than the lower
proficiency groups, F(1, 55) = 191.56, p < .001, ηG
2 = .67; F(1, 45) = 36.44, p < .001, ηG
2
= .45.
• A significant interaction between SR and L2 proficiency was found, indicating
that SR more greatly affected judgments made by the lower than the higher
proficiency groups , F(1, 45) = 95.16, p < .001, ηG
2 = .51; F(1, 55) = 72.38, p < .001, ηG
2
= .36.
 CR had a greater impact on judgments made by proficient learners,
whereas SR had a greater impact on judgments made by less proficient
learners.
Methods
• Participants were classified into either a higher or lower proficiency group
according to scores of the EFL reading test widely used in Japan (EIKEN).
• A mixed three-way ANOVA was conducted on coherence ratings, with SR as a
within-participants variable, and CR and proficiency as between-participants
variables. Effect sizes are discussed along with statistical significance claims.
Analysis • Japanese EFL learners’ perception of text coherence varies according to not only
CR but also SR between sentences.
• All learners do not similarly perceive coherence of the same texts; there exist
individual differences in the perception of text coherence
 CR is more important for proficient learners’ coherence perception, and SR is
more important for coherence perception in less proficient learners.
Conclusions
Participants.
• 104 Japanese university EFL students (A1 to B2 levels in CEFR)
Materials.
• 20 experimental passages of two-sentence texts varying in CR and SR (adapted
from Wolfe et al., 2005; see Table 1).
 Four types of CR–SR combinations in the first sentence to complement the
second sentence
• CR was verified by subjective ratings in the preliminary study. 2
• SR was calculated by a corpus-based technique called Latent Semantic Analysis
(Landauer et al., 1998). 3
Design and Procedure.
• CR was manipulated between participants, whereas SR was manipulated within
each participant (cf. Nahatame, 2014). 4
z
• Participants were told to read the texts for comprehension and then judged the
coherence of the texts on a 5-point scale. (1 = not coherent at all; 5 = very coherent)
Results & Discussion
Table 1. Example of an Experimental Passage
Figure 1. Boxplots of coherence rating data.
Table 2. Means, 95% CIs, and SDs for the Coherence Judgment Task
SR-High SR-Low
Higher (n = 57) M 95% CI SD M 95% CI SD
CR-High (n = 29) 4.53 [4.32, 4.74] 0.54 4.11 [3.83, 4.39] 0.74
CR-Low (n = 28) 3.28 [3.00, 3.56] 0.72 1.79 [1.58, 2.01] 0.55
Lower (n = 47)
CR-High (n = 20) 4.62 [4.41, 4.83] 0.45 3.38 [2.90, 3.86] 1.04
CR-Low (n = 27) 3.55 [3.29, 3.81] 0.66 1.93 [1.70, 2.17] 0.60
Total
CR-High (n = 49) 4.56 [4.42, 4.71] 0.50 3.81 [3.54, 4.08] 0.94
CR-Low (n = 55) 3.41 [3.22, 3.60] 0.70 1.86 [1.71, 2.02] 0.58
1. Mary could not find anything to read in the library. (CR-High / SR-High)
2. Mary wanted to look for recipes for her dinner party. (CR-High / SR-Low)
3. Mary went to the library to look for something to read. (CR-Low / SR-High)
4. Mary was having a dinner party for her office. (CR-Low / SR-Low)
Target sentence: She went to the bookstore to get new books.
Shingo Nahatame，Kyoei University, Japan
Email: nahatame@kyoei.ac.jp
Website: https://sites.google.com/site/snahatame
Twitter ID: @Nahacchi
Contact
This research was supported by Grants-in-Aid for Research Activity Start-up (15H06571)
from the Japan Society for the Promotion of Science.
Acknowledgement