understanding of sentences or passages, which readers understand texts based on exact words and might rapidly fade out from memory. Propositional model or situation model refers to the psychology reality of proposition, in which readers store the meaning of sentences in mind. Sometimes making reading propositions requires readers to fill in the gap between the lines with common sense that author suppose readers would know.
Namely, readers extend beyond the explicit content of texts to gain propositional model when they connect and make sense of chunks across the whole readings. Examples suggested by Carroll (2007) are (1) Paul walked into the room and (2) Paint was all over his shirt. With regard to inferential comprehension, it is reasonable or effortless for proficient readers to infer the possible fact that the paint might have been placed on the top of the door and fallen down while Paul was walking into the room. For literal comprehension, sentences (1) and (2) might represent two separate anecdotes. In other words, successful inferential RC or situational model of comprehension requires readers to connect the proposition of sentences and conjecture the missing information so as to generate a global and comprehensive understanding, or the mental representation about the reading texts.
To sum up, literal RC or surface model is the literal or linguistic understanding of the texts while inferential RC or propositional/situation model requires the readers to fill in between sentences with background knowledge and also to retain certain significant details or the meaning of presented information. Moreover, at the inferential
comprehension level, readers not only maintain the proposition of information from the linguistic information of texts but also have to be aware when and how to draw inferences between the lines or across the whole passages.
Procedure
42
To relate participants’ WMC measured with a recall-RST to their RC, 104
participants from 4 classes took the recall-RST to gain the WMC result and completed a RC test which contained four passages with five literal and five inferential RC questions.
To relate participants’ WMC measured with a recognition-RST to their RC, the rest of 86 participants from 3 classes took the recognition-RST to gain the WMC result and like the recall-RST group, finished a test of RC. Both the recall-RST and the recognition-RST were displayed with a PowerPoint file on the computer monitors in front of the participants (See the clear procedure flow in fig.3.1).
The experimental procedure was undertaken around the end of the semester, and therefore some of the participants were absent for personal reasons, athletic competitions and city band practice for instance. Those absent from experiments did not show up at school till the end of the semester, which made it difficult to replenish the missing
unfinished data. In the beginning, there were 112 participants in the recall-RST group and 87 in the recognition-RST. However, in the recall-RST group, 8 out of the participants could not attend the test of comprehension test. 1 of the 87 participants from the recognition-RST group missed the comprehension test.
Figure 3.1 Procedure Flow Chart
Data Analysis
The purpose of this research study was to understand the relationship between WM 104 participants
recall-RST
a reading test
(literal and inferential questions)
86 participants
recognition-RST
a reading test
(literal and inferential questions)
43
and RC. A Pearson production correlation was applied to understand the correlation between performance of literal and inferential reading comprehension.
To answer the first research question,” Is there a relationship between L2 WM recall reading span score and English comprehension accuracy in terms of L2 literal and inferential reading?,” a correlation analysis was used to determine the relationship participants’ recall-RST WMC and their literal and inferential RC performances. A p-value of < 0.05 was used to determine statistical significance of correlations between variables.
To answer the second research question, “Is there a relationship between L2 WMC recognition reading span score and English comprehension accuracy in terms of L2 literal and inferential reading?, ”a correlation analysis was used to determine the relationship participants’ recognition-RST WMC and their literal and inferential RC performances.
A p-value of < 0.05 was used to determine statistical significance between variables.
The correlation coefficients, usually denoted by r, were used to demonstrate the strength of linear or near-linear relationship between two variables. The value of correlation falls between +1 and -1: the former indicates positive relationship, which means one variable increase along with the increase of the other; the latter indicates negative relationship, which means one variable decreases along with the increase of the other. When it comes to the significance value, p-value of < 0.05 was used to determine statistical significance. Only when the correlation coefficient reaches the level of
significance would make the relationship between variables significant and meaningful in terms of statistical significance.
To answer the last research questions, “Will the influence of L2 WM on L2 RC be different if WMC is measured via these two types of RST, recall-RST and
recognition-RST?” a regression analysis was applied. Regression analysis was used to evaluate the interrelationship among variables. The strength of predictive power and the
44
relationship was demonstrated via the value of R2 (coefficient of determination, R2). In this study, participants’ L2 literal and inferential RC performances were dependent variable, while participants’ L2 recall-RST and recognition-RST WMC results measured were independent variables.
45