Content area
The contribution of morphological knowledge to literacy skills has been well-established in previous research. Discrepancies have been detected in how such influence is realized among different populations. This study aimed to examine the applicability of the morphological pathway framework among EFL learners, focusing on morphosemantic knowledge. To achieve this aim, data were collected over the course of three weeks from 101 college-level students taking an intensive English language course. A battery of paper-based and computer-based tests was used to measure learners’ morphosemantic knowledge, morphological-based lexical inferencing, morphological decomposition ability, vocabulary size, and reading comprehension. The study revealed that while morphological-based lexical inferencing greatly mediated the effect of morphosemantic knowledge on reading comprehension, morphological decomposition did not account for any difference. Structural equation modeling analysis uncovered a different processing mechanism, where morphological decomposition facilitates lexical inferencing rather than reading comprehension. The analysis also revealed that the contributions of morphosemantic knowledge and processing were highly dependent on lexical processes, whereby vocabulary mediated the effect of morphosemantic knowledge and significantly assisted in lexical disambiguation. This indicates that L1 and L2 readers follow different morphological processing routes during reading. The findings have implications for morphology and reading instruction in the EFL context.
Introduction
Reading comprehension was defined by Anderson et al. (1985) as the process whereby the reader constructs meaning from a given text. In this view, reading is the product of the smooth interaction of numerous skills and sources of information. There is a general consensus in the literature on the role of morphological awareness (MA) (i.e., morphological knowledge) in reading comprehension among both English L1 (Deacon et al., 2017; Goodwin et al., 2022; James et al., 2021; Kieffer & Lesaux, 2012; Kotzer et al., 2021; Levesque et al., 2017, 2019; Liu et al., 2024; Zhang et al., 2020, 2023) and L2 readers (Alshehri & Zhang, 2022; Jeon, 2011; Yamashita & Kusanagi, 2024; Zhang, 2017, 2021; Zhang & Lin, 2021; Zhang et al., 2020, 2023). The pathways of MA’s contribution to the comprehension process, however, remain blurred. The divergent mechanisms used by researchers focusing on different aspects of MA and following different investigative methods may have contributed to the confounding results. Therefore, this study aimed to assess the role of derivational MA in reading comprehension, focusing largely on semantic knowledge and automatic processes utilized during reading. The morphological pathway framework by Levesque et al. (2021), which builds upon the reading system framework (RSF) by Perfetti and Stafura (2014), depicts a clear mechanism by which morphology can support literacy. Although the framework draws a clear picture of the utilization of MA in reading among native English speakers, its pertinency among English as a foreign language (EFL) learners remains unexamined. This study explored the applicability of the morphological pathway framework in the MA-reading comprehension association by assessing the feasibility of each route and the possibility of a direct contribution of morphosemantic knowledge to reading comprehension.
The MA skills under investigation encompass knowledge of morpheme meaning and processing ability, addressing both morphological decomposition (i.e., morphological segmentation) and morphological-based lexical inferencing (MBLI). The examination of the two processing routes is likely to reveal whether the two processes equally mediate MA and reading comprehension, or whether they act in a complementary manner.
MA and reading comprehension: related theories
Reading is a multifaceted skill that is governed by numerous cognitive processes and sources of information (Foorman et al., 2015). One of the most recognized reading models that captures the complexity of this process is the RSF by Perfetti and Stafura (2014). This framework follows a mental approach towards the reading process. Perfetti and Stafura (2014) believe that comprehension processes are primarily dependent on lexical knowledge and the successful integration of a word’s meaning into the mental representation of the text. Reading comprehension in the RSF involves a series of overlapping processes that start with a word identification system, followed by a comprehension system mediated by the lexicon. These processes draw on three knowledge sources: orthographic knowledge, linguistic knowledge, and general knowledge (Perfetti & Stafura, 2014).
The RSF is the only reading framework that incorporates morphology as an element that contributes to reading comprehension. The role of morphology in the framework is twofold: it contributes directly to the comprehension process and indirectly via lexical processing. The dual placement of morphology within the framework supports the conclusion that different morphological skills contribute to comprehension (Levesque et al., 2019). Focusing on morphology, the morphological pathway framework by Levesque et al. (2021) builds upon the RSF, illustrating the underlying pathways by which morphology can support meaning construction during reading. Unlike many studies (e.g., Alshehri & Zhang, 2022; Jeon, 2011; Wang & Zhang, 2022; Zhang, 2017; Zhang & Koda, 2012; Zhang & Lin, 2021), the framework distinguishes between morphological knowledge, morphological decoding (decomposing multimorphemic words to facilitate word processing), and MBLI, viewing them as three separate constructs that contribute differently to comprehension. The morphological pathway framework proposed by Levesque et al., (2021, Fig. 1), although acknowledging the role of MA and its direct contribution to comprehension, asserts that MA’s indirect contribution through the lexicon is mediated by morpho-orthographic and morphosemantic processing. The framework takes into account morphology within the lexical representation and asserts that morphological decomposition and MBLI are the key mechanisms for lexical access.
The morphological pathway framework draws attention to progressive changes in the morpheme decoding route where, in the early stages of reading, morphological decoding facilitates word reading and reading fluency by focusing on larger orthographic chunks rather than letters. As learners grow older, the readers follow an orthographic route where morphologically complex words are instinctively segmented into morphemic constituents, facilitating access to different aspects of morphological knowledge.
The morphological pathway framework accounts for the influence of inflectional and derivational morphology on reading. A shortcoming of the framework is its lack of specificity regarding the type of morphological awareness that contributes to reading comprehension. Given the multidimensional nature of morphology, this study will focus on derivational morphosemantic knowledge.
The role of MA in reading comprehension in L1 and L2
MA is a metalinguistic skill that refers to the ability to recognize and manipulate morphemes (Carlisle, 1995). Nagy et al. (2014) assert the need to distinguish between three types of morphological knowledge: word form, word meaning, and syntax. Word form knowledge facilitates the spelling and decoding of morphologically complex words. Semantic knowledge facilitates the morphological decomposition and segmentation of multimorphemic words, making it easier to infer their meaning and access the meaning of individual morphemes. The role of syntactic knowledge is simply restricted to enabling learners to infer the part of speech of morphologically complex words. Nagy et al. (2014) also stressed the need to differentiate between MA and morphological processing, as the former refers to knowledge of morphology, whereas the latter refers to the strategic use of such knowledge.
Although the MA-reading comprehension association has been established in the literature, variation can be detected in directionality and mediation patterns. MA was found to directly contribute to reading comprehension. The direct path, however, was mostly observed in cross-sectional studies among L1 children (Deacon et al., 2014; James et al., 2021; Kieffer & Lesaux, 2012) and L2 adults (Alshehri & Zhang, 2022; Zhang & Lin, 2021), when morphological processing was not fully accounted for. A pathway analysis by Levesque et al. (2017) found that MA contributed to reading comprehension both directly and indirectly. The indirect path, manifested through the mediation of morphological decoding and analysis, supports the dual placement of morphology within the RSF. The direct contribution of MA to reading comprehension was found to be more substantial. The results of a later longitudinal examination by Levesque et al. (2019) stand in contrast to previous findings. The results showed that while morphological knowledge plays a significant role in comprehension, it can only be utilized through the mediation of morphological processing.
Examinations among native English-speaking children revealed MA to be a significant predictor of reading comprehension (Deacon et al., 2017; Goodwin et al., 2022; Kieffer & Lesaux, 2012). Weighing different dimensions of morphological knowledge and processing components, Goodwin et al. (2022) revealed asymmetrical contributions of different aspects of morphology where morpheme identification played the most significant role, followed by morphosyntactic knowledge, MBLI, and orthographic-phonological knowledge, respectively, each explaining a significant variance in the analysis. The positive influence of MA on reading comprehension was also detected among older age demographics (Kotzer et al., 2021). The positive association was found to be consistent among participants of different ages (James et al., 2021; Liu et al., 2024) and across various reading abilities (James et al., 2021).
Previous literature on native English speakers reflects a positive association between MA and reading comprehension. The presence of this pattern among native English-speaking populations encourages the comparison and contrast of similar and deviating patterns within the EFL context. A comparison of participants from both L1 and L2 contexts revealed that MA consistently predicted reading comprehension among both populations (Kieffer & Lesaux, 2012; Lee et al., 2023; Zhang et al., 2020, 2023).
Like research among native English speakers, research on EFL learners showed different patterns. While direct and indirect effects were found by Zhang (2017), direct effects were found to be more substantial. The directionality of MA’s influence yet again reveals different mechanisms where the MA processing of different dimensions of morphological knowledge was found to contribute more than others. Mediation patterns were found to be determined by the type of MA knowledge components. Zhang and Lin (2021) found that morphosemantic knowledge was largely mediated by learners’ vocabulary knowledge, whereas the influence of morphosyntactic knowledge was found to be more direct. Distinguishing between morphological knowledge and processing efficacy, Alshehri and Zhang (2022) and Zhang (2021) revealed that readers capitalize on MA knowledge components of meaning, use, and form during reading rather than rapid sub-lexical processing. Zhang (2021) asserts that although morphosemantic knowledge contributes to reading comprehension, such contribution is likely indirect, mediated almost exclusively by meaning-based processes. Similarly, Jeon (2011) revealed that the application of morphosemantic knowledge played a much more substantive role than knowledge of morpheme structure, drawing a clear distinction between morpheme identification and semantic association to facilitate lexical inference. This interpretation, however, should be taken with caution as the assessment of learners’ awareness of morphological structure was based on productive measures rather than receptive measures, which are usually encountered during reading. Productive measures were deemed by Lee et al. (2023) to require a more refined knowledge of morphology, which may have made it more challenging for the participants than the MBLI task used to assess lexical inference.
The investigation of the role of morphology in reading comprehension has been greatly intertwined with vocabulary knowledge: some researchers found it to significantly mediate the contribution of morphology to reading comprehension, whereas others found a more direct effect. Studies focusing on native English speakers found MA to independently influence comprehension among children (Goodwin et al., 2022) and adults (Guo et al., 2011; Kotzer et al., 2021) alike. While these studies affirm a direct contribution of MA and processing to reading, the mediation of vocabulary knowledge was not completely denied. The role of vocabulary knowledge as a mediator in the MA-reading comprehension paradigm can be inferred from the findings of Goodwin et al. (2020) where learners with limited vocabulary knowledge showed restricted morphological processing. Deficient vocabulary knowledge was found to restrict the application of morphosemantic information, thereby making the deduction of the meaning of multimorphemic words challenging during reading.
The mediation of vocabulary knowledge, however, was found to be more robust within the EFL context. Vocabulary knowledge was found to partially mediate the influence of affix meaning and use in reading comprehension (Kieffer & Lesaux, 2012; Yamashita & Kusanagi, 2024). In contrast, Zhang and Koda (2012) found the contribution of MA to reading to be exclusively indirect, mediated either solely by vocabulary knowledge or in combination with learners' lexical inferencing ability. Lee et al. (2023) found age to be a determining factor, where increased exposure through years of education strengthens the link between different lexical knowledge components that facilitate word identification and retention.
The association between MA and vocabulary knowledge was further affirmed by Wang and Zhang (2022), who found that MA correlated positively with vocabulary knowledge. However, others (e.g., Chen, 2019; Goodwin et al., 2020) deemed the relationship complementary. The mediative role of vocabulary knowledge in the MA-reading comprehension relation raises the question of what processing routes are followed to realize this influence. Nagy and Anderson (1984) estimated that for every word a person learns, three other morphologically related words will be comprehensible. The exact number of words was said to be highly variable depending on learners’ ability to perform morphological analysis. The RSF assumes that MA assists directly in reading comprehension and indirectly through the lexicon; however, the morphological pathway framework asserts that MA’s indirect contribution is completely mediated by MBLI and morphological decomposition by facilitating word identification processes. The confounding results in the literature, in addition to the assumptions of the RSF and the morphological pathway framework, necessitate a further look into the contribution of morphosemantic knowledge. Examining the effect of morphosemantic knowledge on meaning construction in the presence of the two processing routes and vocabulary knowledge as a mediator is likely to reveal the true means by which MA facilitates comprehension.
Lexical processes mediating the contribution of MA to reading comprehension
The morphological pathway framework assumes morphosemantic and morpho-orthographic processes to be the only mechanisms by which MA facilitates lexical access. These processes represent parallel pathways within morphology and the lexicon. Given the significant role played by vocabulary knowledge in reading comprehension, morphosemantic analysis of morphologically complex words has been found in L1 and L2 literature to aid in word identification. This can be inferred from studies that have found vocabulary knowledge to correlate positively with MBLI (e.g., Zhang et al., 2022). Deducing meaning from morphemic constituents was found to positively contribute to reading comprehension among native English-speaking children (Deacon et al., 2017) and adolescents (Goodwin et al., 2022). Goodwin et al. (2022) found that MBLI was intertwined with vocabulary knowledge and observed that the application of morphosemantic information was challenging for learners with limited vocabulary knowledge. The struggle to deduce meaning from semantic units among readers with limited vocabulary size/knowledge represents a false dichotomy that is open to more than one interpretation. First, MBLI is highly mediated by vocabulary knowledge, which this particular group lacks. Second, readers’ inability to apply morphosemantic information during reading might be the reason behind their limited vocabulary knowledge.
The ability to deduce meaning from morphologically complex words is comparable among L1 and L2 users (Raudszus et al., 2021). Like L1 literature, MBLI was positively associated with reading comprehension among L2 learners. Zhang and Koda (2012) revealed that MBLI is a consistent mediating route between MA and reading comprehension, with vocabulary knowledge facilitating greatly within the process. The findings were later confirmed by Zhang et al. (2020), whose pathway comparing groups showed similar results. The significant role of MBLI was more evident in Zhang and Shulley (2017), where MBLI significantly predicted reading comprehension even when MA and vocabulary knowledge were controlled for. Thus, superior performance by skilled readers cannot be attributed only to enhanced knowledge but also to effective cognitive processing. Oakhill et al. (2019) stress that skilled and less skilled readers differ not only in the quantity and quality of their lexical representations but also in the ability to utilize linguistic resources to improve impoverished lexical represenations. Inferior performance by poor readers can be ascribed to behavioral patterns, where challenging words are often skipped and no morphological processing is employed even when morphemic constituents are highly frequent (Zhang & Shulley, 2017).
Previous literature leads to the establishment of MBLI as a distinct skill, a form of morphological processing, and not necessarily an entailment of the semantic knowledge of morphemic units. The positive associations deduced in the studies justify its placement within the morphological pathway framework as a separate route for morphological processing that aids in the positive contribution of MA to reading comprehension.
Morphological decomposition, referred to in the morphological pathway framework as morphological decoding, is the second route of morphological processing that occurs through the orthographic system. This process occurs on the word form and provides a pathway from morphological knowledge to the decomposition of multomorphemic words to facilitate lexical access (Levesque et al., 2021). The framework states that in the early stages of language development, morphological decoding facilitates word reading. In later stages, however, word segmentation is primarily orthographic and semantically based. Studies on L1 children have demonstrated morphological decoding to be a significant mediating process facilitating word reading (Deacon et al., 2017; Levesque et al., 2017). Zhang et al. (2023) found that monolinguals' reliance on word decoding was age-dependent. The difference was attributed to learners’ progression from the early years of “learning to read,” focusing mostly on word decoding, to the late stages of “reading to learn,” during which vocabulary knowledge becomes vital. Comparing native and non-native English readers, Zhang et al. (2023) revealed that while young monolinguals relied equally on vocabulary knowledge and word decoding in reading comprehension, bilinguals followed a more meaning-based approach toward reading.
Studies that focused on L2 participants have yielded conflicting results. The confounding results may be attributed to proficiency-based differences between participants. The significance of word reading and decoding was detected only among less proficient learners (Kieffer & Lesaux, 2012; Zhang et al., 2020). The absence of the decoding route among adult native English speakers and highly proficient L2 learners suggests following a semantic-based approach towards reading, relying mostly on vocabulary knowledge and morphological analysis. Kieffer and Lesaux (2012) found the indirect route of MA’s contribution to reading comprehension to be realized primarily via word identification with word recognition (i.e., decoding) having minimal explanatory power. An examination of previous literature concerning the role of decoding in reading comprehension suggests context, age, and proficiency-based differences. Early L1 readers with a strong comprehension system may exclusively rely on decoding for lexical access. Less proficient L2 learners, however, might depend on decoding, among other factors, in an attempt to exploit all available linguistic knowledge—being semantic, orthographic, or phonetic—to achieve lexical access. The increase in age and proficiency would entail improved metalinguistic awareness and enhanced lexical representation, which would lead to more confidence in the reliance on orthographic and semantic routes to achieve comprehension.
Investigations of the contribution of morphological decomposition to reading comprehension among L2 users produced inconsistent results. While some studies detected a positive effect of morphological decomposition on reading comprehension (e.g., Shoeib, 2017; Zhang, 2017), accurate morphological segmentation was found to be futile by others (e.g., Bar-Kochvan & Hasselhorn, 2017; Zhang, 2021; Zhang & Koda, 2012). Zhang (2021) revealed that implicit processing of morpheme recognition and discrimination is ineffective in the reading comprehension process. Bar-Kochvan and Hasselhorn (2017) found that morphological decomposition, which assists in spelling, had no considerable impact on word reading and comprehension. The efficiency of decomposition (focusing on reaction times) was also found irrelevant (Alshehri & Zhang, 2022). The distinction between the accuracy and efficiency of morphological segmentation might be of particular significance within the EFL context, in which morphological processing poses a challenge to EFL learners. Clahsen and Felser (2018) assert that the L2 system is based on whole-word storage rather than morphological components. In addition, morphological analysis amongst non-native English speakers is dependent on declarative rather than procedural memory and is, thus, less sensitive to a word’s internal structure (Ullman, 2006). Collectively, this might explain the variance in research focused on the EFL context. While the role of MBLI has been deemed significant, the question remains whether drawing general semantic associations is sufficient or if accurate, semantic-based deconstruction of words is necessary for lexical access. While these processes were found to work in a parallel manner among native English speakers, a different pattern in which these processes occur successively might be more applicable in the EFL setting. This assumption is prompted by the findings of Zhang et al. (2016), which showed that morphological segmentation significantly assists in MBLI.
The present study
Although the extant literature has attempted to reveal the means by which MA can support literacy, the aforementioned studies have lacked theoretical grounds, focusing on different aspects and processes. Building on prior research, this study aimed to examine the applicability of MA processing routes suggested by the morphological pathway framework focusing in particular on semantic-based mechanisms. In addition, this study intends to show the sufficiency of such processes in accounting for the contribution of morphosemantic knowledge to L2 literacy. This will aid in clarifying the means of meaning construction aided by MA. This will also allow the comparison of possible similarities and differences between native and non-native English speakers in utilizing morphosemantic knowledge in reading. This study aimed to answer the following two research questions.
Do the indirect processes of the morphological pathway framework equally mediate the contribution of morphosemantic knowledge to reading comprehension among EFL learners?
How do the contributional pathways of MA in the morphological pathway framework and readers’ vocabulary knowledge interact to facilitate reading comprehension?
Methodology
Piloting
A pilot study was deemed necessary to determine the feasibility of some of the assessment tools. The main aim of the pilot was to evaluate the materials used, get feedback about the study design from a small sample of the study population, and overcome/minimize potential obstacles in the study experiments. Because of time constraints, not all the tests used in the actual study could be piloted. The selection was based on the need to ensure the reliability of the tool (i.e., MBLI) and the fluent administration of tasks. The MBLI task was piloted on 30 participants to calculate a Cronbach’s alpha which was found to be acceptable (α = .72), requiring no changes to the assessment tool. The reliability of the tool was further affirmed by a focus group of six participants, which was carried out to ensure the suitability of the test to learners’ English proficiency. The feedback received confirmed the assumptions upon which this test was designed.
The computer-based task (separability lexical decision task) was also piloted to ensure the smooth execution of the task. The piloting was done by six students who completed the experiment and then gave feedback on the instructions and words used. The task was found to be acceptable by students, requiring no necessary amendments.
Participants
The study comprised 101 Saudi EFL learners, with 26 male participants and 75 female participants. The students recruited were enrolled in a Saudi University and, at the time of the study, were taking an intensive English language course. Upon entry, learners take a placement test based on which they are placed into one of three proficiency levels, A, B, and C. This study focused on level B students assessed to be of pre-intermediate proficiency upon entry and expected to reach an upper intermediate level upon exit, which was shortly after data collection. These levels correspond to the Common European Framework of Reference levels A2 and B1, respectively. The choice of proficiency level was triggered by the lack of research examining morphological processing among this population. In addition, the choice of a middle category is to avoid having students who might reach a ceiling effect or bottom effect, expected in levels C and A. The placement test is a multiple-choice test designed by the university. The design and format of the placement test are guided by the Cambridge English placement test.
While the researcher preselected a particular proficiency group, convenience sampling was applied. The criterion of inclusion was the accurate execution of tasks, where participants who did not complete all tasks properly were eliminated from the study. The age range of the participants was 17–19 years.
Ethical considerations
In this study, adherence to the university's ethical guidelines was ensured. Given that the data collection was carried out on campus and the target sample consisted of university students, ethical approval was obtained from the university’s institutional review board of the Research Ethics Committee with the reference number (24–037).
After obtaining approval, participants were required to sign a consent form before data collection began. Given the interdependent nature of the knowledge and skills assessed by the tests, students were instructed to use their university ID numbers in all tests to ensure their privacy. Moreover, the participants were assured that their data would be handled with complete anonymity and that the results could not be traced back to them.
Research instruments
The present study employed a quantitative and exploratory methodology to assess the relationship between morphological knowledge, processing, and reading comprehension. With an exclusive emphasis on meaning extraction, this research focused on morphosemantic knowledge, the ability to deconstruct multimorphemic words, and the disambiguation of morphologically complex words—the main components of the morphological pathway framework. Recognizing the pivotal role of vocabulary in reading, the study assessed vocabulary size in addition to reading comprehension.
Morphosemantic knowledge
To measure the participants’ knowledge of morpheme meaning, a word parts level test WPLT by Sasao and Webb (2017) was used; different previous versions of the test were used by Zhang and Lin (2021) and Alshehri and Zhang (2022), which establishes the validity of the research instrument. The exam follows a multiple-choice question format, consisting of 34 items. Each participant was asked to choose the correct meaning of an affix (suffix and prefix) contextualized in two example words to resemble authentic exposure. Given the students' proficiency level, the beginner version was primarily used. Because of redundancy, a few items from the beginner version were replaced with items from the intermediate version. Below is a test item from the assessment tool.
re- (replay; rebuild)
person
again
female
before
The Cronbach’s alpha score of .83 meant that the test was reliable and suitable for the target sample. The test was paper-based and administered on campus.
Morphological decomposition
The investigation of morphological decomposition is particularly significant within the English language context. The English language lacks orthographic semantic consistency, where the relationship between “teach” and “teacher” is semantically transparent, whereas the relationship between “corn” and “corner” is opaque. This establishes the existence of pseudo-complex words, such as “corner” and “witness,” that appear to be morphologically complex but are not (Stevens & Plaut, 2022). This highlights the need, within the EFL context, to measure learners’ ability to distinguish between pseudo-complex and morphologically complex words and properly deconstruct the latter into their morphemic constituents. Morphological ortho-semantic processing was measured using the Koda (2000) separability judgment lexical decision task, previously used by Alshehri and Zhang (2022).
The testing was based on two stimulus conditions where half of the words included in the task were multimorphemic, whereas the remaining 15 words were monomorphemic. The monomorphemic words only included pseudo-complex words that share similar beginnings and endings to morphologically complex words. Word items included pseudo-complex words like ‘awful’ and morphologically complex words like “colorful.” The choice of words was based on the learners’ proficiency level of B1. Only the participants’ scores, but not reaction time (RT), were factored into the analysis. The choice of lexical decision task was made to tap into learners' automatic processing efficacy of the internal structure of the word, rather than explicit knowledge. The test was conducted on an Experiment Builder 2 Gorilla.sc (Anwyl-Irvine et al., 2024). The task started off with two practice items followed by feedback to ensure comprehension and familiarize the subjects with the task format. Each word appeared in the middle of the screen, and participants were asked to respond with yes or no on the keyboard. Students were encouraged to answer as quickly as possible. Each response triggered the onset of the following word. Given the complexity of the task and the redundancy in word beginnings and endings, the task was self-paced rather than time-sensitive. The justification behind such an approach is to mimic natural language processing that is self-paced rather than timed. Given the readers' proficiency range, no cut-off RTs were applied. Cronbach’s alpha was calculated at .65, which is deemed acceptable by Vaske (2008).
Disambiguation of morphologically complex words
Lexical inferencing ability was measured by a researcher-developed task consisting of 30 lexical items presented in isolation along with four options. The selection criterion for the target words was based largely on Deacon et al., (2017), depending mainly on word frequency. The target words were infrequent morphologically complex words constructed from high-frequency morphemic constituents. The morphemes under investigation are derivational, and given the complexity of English language words, 20 words were supplemented with either a suffix or a prefix, whereas the remaining 10 words either have both a prefix and a suffix or two suffixes. The target words chosen are all semantically transparent, but given that the test was presented in writing, phonological transparency was not taken into consideration.
The construction of the options was generally based on Anglin et al.’s (1993) guidelines. Following the criteria, the correct option is either a paraphrase or the exact definition of the word, as obtained from the Merriam–Webster online dictionary (n.d.). The three distractors reflect the meaning of the/one affix, the root word alone, or neither. All three options are of the same length and general structure as the correct answer. As an example of a task item, the correct meaning for the word “agreeable” is “easy to be accepted” while the three distractors are “able to learn,” “having different opinions,” and “having poor taste.” Students were prompted to derive the meaning of the words from their componential makeup. Given that morphosemantic knowledge and vocabulary size were measured separately, MBLI scoring was not ranked but rather marked correct or incorrect.
To ensure the validity of the research tool, the test was assessed by two assistant professors with Ph.D.s in applied linguistics and three language instructors with a master’s degree in applied linguistics and a minimum of four years of experience teaching level B students. The validity of the tool was also assumed based on feedback from the focus group in the pilot study. The reliability of the tool was calculated at (α = .70) among the entire study population. The test was administered in a paper-based format and took nearly 15 min.
Vocabulary knowledge
Given the significant contribution of vocabulary knowledge to reading comprehension, this variable was controlled in the analysis. Learners’ vocabulary size/knowledge was measured using the revised vocabulary levels test VLT by Webb et al. (2017), originally created by Nation (1990).
The test originally measured vocabulary size across five levels, ranging from the 1000-word level to the 5000-word level. Each word level is measured via 10 test items. Due to time constraints and the proficiency level of the students, the 1000-word level was eliminated, and only eight items were used to assess each word level. Each test item consisted of a batch of six words, three of which were target words and three of which were distractors. The participants were provided with three definitions in the form of a grid and asked to check under the correct word for each definition. Table 1 presents the example item provided to students at the beginning of the task.
Table 1. Example from the vocabulary levels test
Game | Island | Mouth | Movie | Song | Yard | |
|---|---|---|---|---|---|---|
A piece of land surrounded by seawater | ✓ | |||||
The part of your body used for eating and talking | ✓ | |||||
A piece of music | ✓ |
The exam was administered in a paper-based format on campus. The validity of the test was established using Messick’s (1989, 1995) validity framework (Webb et al., 2017). The test consisted of 32 subscales. The value of Cronbach's alpha is .86, reflecting good internal consistency.
Reading comprehension
To measure learners’ reading comprehension, a MacGinities–Gates reading test was used. The test levels correspond to American grade expectations. Based on the students' level, a Level 5 test was used. To ensure the suitability of the instrument, the texts were analyzed using Text Inspector, and the analysis revealed that the materials were within the range of the participant’s proficiency, ranging from A2 to B1. Following the methodological approach of Choi (2017), only seven passages were selected out of 11, with a total of 30 questions. Participants were given 45 min to finish the test.
The MacGinities–Gates reading test is norm-referenced based on national standards for American grades. The test is also frequently used by researchers to assess reading comprehension for EFL populations of different proficiency levels. Collectively, this establishes the test as a valid instrument to measure reading comprehension. Furthermore, the internal consistency of the test was found acceptable (α = .71), supporting its suitability for the study population.
Procedure
Data collection took place on campus in a controlled setting within a three-week timeframe. The paper-based tasks were carried out during class time and were allocated the amount of time detailed previously. The computer-based task was conducted in a laboratory on campus. All tasks were graded by the researcher. Given the type of questions used in the assessment tools, one mark was given for each correct answer. Example items were provided for each task to guide learners in answering the questions. Having taken reading exams before, students were not given any example items on the reading test.
Participants’ performance across tasks was linked through their university ID number. The participants belonged to specific sections assigned by the English Language Skills Department based on the desired proficiency. The exams were supervised by the course instructors. Data collection on the female campus was supervised by the researcher. Data collection on the male campus, on the other hand, was managed by a male colleague.
Data analysis
The statistical analysis was performed on RStudio (Version 2024.12.0; Posit Team, 2024). The data analysis was carried out in four steps. First, descriptive analyses and normality tests were conducted. Next, a bivariate correlation analysis was performed to assess the relationship between variables. Then, a hierarchical linear regression was carried out to assess the weighted contribution of each variable to reading comprehension. Finally, structural equation modeling (SEM) was performed to assess possible processing routes. Model selection was based on good model fit. While the literature suggests different cut-off values for fit indices, this research generally adopts the criteria suggested by Kline (2005), CFI and TLI > .95, SRMR < .08, and RMSEA < .05. Chi-square values are interpreted with caution as they may be sensitive to sample size (Kline, 2005). Tabachnick and Fidell (2007) suggest that a non-significant p-value and a value of χ2, normed χ2 of 2 or less, indicate a good fit.
Results
The results of the descriptive analyses and normality testing in Table 2 show that the measurements have a good spread based on standard deviations. This suggests no violation of symmetry and normality.
Table 2. Descriptive statistics of variables
Variable | M | SD | Min | max | α | Skewness | Kurtosis |
|---|---|---|---|---|---|---|---|
MK (34) | 20.25 | 5.795 | 6 | 32 | 0.83 | 0.455 | 2.542 |
MBLI (30) | 13.9 | 4.270 | 6 | 25 | 0.70 | 0.239 | 2.647 |
MD (30) | 21.37 | 3.152 | 14 | 28 | 0.65 | − 0.095 | 2.130 |
VS (96) | 38.19 | 10.700 | 16 | 61 | 0.86 | 0.071 | 2.083 |
RC (30) | 11.45 | 4.710 | 4 | 24 | 0.71 | 0.636 | 2.779 |
The numbers in parentheses indicate the number of items used to measure each variable
MK morphosemantic knowledge, MBLI morphological-based lexical inferencing, MD morphological decomposition, VS vocabulary size, RC reading comprehension
Next, the correlation matrix shown in Table 3 indicates a moderate to weak correlation between the variables in question. Before any statistical analysis was carried out, assumptions of multicollinearity and linearity were checked. The variance inflation factors (VIFs) of variables ranged from 1.1 to 1.8, reflecting no multicollinearity. The non-significant result on the Ramsey reset test (p = 0.19) suggests that the assumption of linearity was met.
Table 3. Bivariate correlations between all measured variables in the study
Variable | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
MK | _ | ||||
MBLI | 0.566*** | _ | |||
MD | 0.311** | 0.343*** | _ | ||
VS | 0.615*** | 0.563*** | 0.216* | _ | |
RC | 0.52 *** | 0.605*** | 0.337*** | 0.601*** | _ |
MK morphosemantic knowledge, MBLI morphological-based lexical inferencing, MD morphological decomposition, VS vocabulary size, RC reading comprehension
*p < .05 **p < .01 ***p < .001
Bivariate correlations revealed a significant moderate correlation between vocabulary knowledge, MBLI, and reading comprehension. Moderate positive correlations were also detected between vocabulary knowledge, MBLI, and morphosemantic knowledge, reflecting a significant relationship between them. This moderate correlation matrix revealed a complex relationship between variables which might be challenging to untangle. To evaluate the proportionate contribution of each variable while controlling for other variables in the model, a hierarchical linear regression was carried out. The entry of variables was not random but theoretically based. Given the significant contribution of vocabulary knowledge to reading comprehension, as established by the RSF, it was entered first into the model. As Table 4 shows, vocabulary knowledge contributed significantly to the reading comprehension process, explaining 36% of the reading comprehension variance. In the next step, morphological awareness was entered into the model, which explained an additional 3.9% variance. In the last model, morphological processing routes were added to the model as a block. Collectively, morphological processing explained a variance of about 8.7% in reading comprehension. Given the significance of MBLI in the model (p < 0.05), it is assumed to be the primary processing route in the framework. The insignificant contribution of morphological decomposition to reading comprehension (p > .05) suggests that semantic-based decomposition is irrelevant to reading comprehension.
Table 4. Results of the linear regression model predicting reading comprehension
Step | Predictors | R2 | Adjusted R2 | ΔR2 | P |
|---|---|---|---|---|---|
1 | Vocabulary size | 0.362 | 0.355 | 0.362 | 0.000 |
2 | Morphosemantic knowledge | 0.401 | 0.389 | 0.039 | 0.0126 |
3 | MBLI Morphological decomposition | 0.488 | 0.467 | 0.087 | 0.001 0.118 |
MBLI morphological-based lexical inferencing
To paint a clear picture of mediation patterns regarding the way different morphological components contribute to reading or whether different morphological processes are interdependent, a structural equation model (SEM) was employed. The design of the SEM was based on the theoretical assumptions of the morphological pathway framework, where processes of MBLI and morphological decomposition are assumed to occur in the lexicon. To reflect these assumptions, morphosemantic knowledge and vocabulary knowledge were placed as initial independent variables. Processes of the morphological pathway framework, however, act as mediating pathways between knowledge components and reading comprehension. The SEM design was also guided by the findings of the linear regression. Upon entry of morphological processing to the linear regression, the significance of morphosemantic knowledge diminished as a contributing factor to reading comprehension. The absence of multicollinearity suggests a mediation effect of MBLI within morphosemantic knowledge and reading comprehension. Another possibility is the mediation of vocabulary knowledge in the contribution of morphosemantic knowledge. Previous literature within the EFL context has consistently found the contributions of MA to reading comprehension to be highly dependent on vocabulary size (Kieffer & Lesaux, 2012; Yamashita & Kusanagi, 2024; Zhang, 2021; Zhang & Koda, 2012; Zhang & Lin, 2021). Consequently, the model assessed the possible mediation of vocabulary knowledge. In the SEM model, vocabulary knowledge acts not only as an initial independent variable but also as a possible mediator within morphosemantic knowledge and reading comprehension.
The model testing the relationship between variables showed χ2(2) = 1.14, P = .248 with good model fit (χ2/df = .57). Supplementary fit indices also suggest a good fit (CFI = 0.999, TLI = 0.991, RMSEA = 0.038, SRMR = 0.014). Given the limited number of participants, a bootstrapping method was employed based on 5.000 resamples. The bias-corrected 95% bootstrap confidence intervals are detailed in Table 5.
Table 5. SEM Results of direct and indirect effects
Effect | Estimate | p-value | Bootstrapped SE | Bootstrapped 95% CI lower, upper) |
|---|---|---|---|---|
MD-RC | 0.204 | 0.055 | 0.107 | [− 0.013, 0.409] |
MD-MBLI | 0.238 | 0.043 | 0.118 | [0.008, 0.467] |
MBLI-RC | 0.382 | 0.000 | 0.101 | [0.171, 0.574] |
VS- RC | 0.166 | 0.000 | 0.033 | [0.102, 0.229] |
MD-MBLI-RC | 0.091 | 0.076 | 0.051 | [0.003, 0.199] |
MK-VS-RC | 0.189 | 0.000 | 0.045 | [0.105, 0.281] |
MK-MBLI-RC | 0.085 | 0.011 | 0.034 | [0.028, 0.158] |
VS-MBLI-RC | 0.052 | 0.007 | 0.019 | [0.018, 0.094] |
MD morphological decomposition, RC reading comprehension, MBLI morphological-based lexical inferencing, VS vocabulary size, MK morphosemantic knowledge
The direct effects of different variables in the model are illustrated in Fig. 1.
[See PDF for image]
Fig. 1
Direct and indirect routes of the contributions of morphosemantic knowledge and processing to reading comprehension. Note MK: morphosemantic knowledge; MD: morphological decomposition; VS: vocabulary size; MBLI: morphological-based lexical inferencing; RC: reading comprehension
The results of the SEM model showed that vocabulary size and MBLI were the only significant predictors of reading comprehension. Morphosemantic knowledge was found to contribute indirectly through the mediation of vocabulary size (β = .189, p < 0.001) and MBLI (β = 0.085, p < 0.05). MBLI's contribution to comprehension was found to mediate vocabulary knowledge and morphosemantic knowledge. The latter was found to contribute the most to lexical inferencing (β = 0.223, p < 0.001). Although less proportionate, vocabulary size was also found to contribute to MBLI (β = 0.135, p < 0.001), leading to an indirect effect of β = 0.052 (p < 0.001). Vocabulary size also provided a direct significant contribution to reading comprehension (β = 0.166, p < 0.001). The most substantial direct contribution to reading comprehension, however, was achieved through MBLI (β = .382, p < 0.001). Promoted by previous findings by Zhang et al. (2016), the SEM model examined the possible effect of morphological segmentation on MBLI. The results revealed that the former contributed significantly to lexical inferencing (β = .238, p < 0.01). The contribution, however, did not extend to affect reading comprehension (β = .090, p > 0.05). Driven only by morphosemantic knowledge, morphological decomposition only assisted in MBLI. Vocabulary size, on the other hand, did not assist in segmentation.
The complex nature of the interaction between the variables, which are all semantically based, implies that the proportionate contribution of vocabulary might be affected by its significant moderate correlation with other independent variables, thereby concealing its precise extent. The correlation between vocabulary size and morphosemantic knowledge (r = .61) reflects a shared variance of 37%, and its correlation with MBLI (r = .56) reflects a shared variance of 31%. This shared variance might have influenced the path coefficients of vocabulary's contribution to reading comprehension. The sum of the two path coefficients of the direct and indirect contributions of vocabulary size on reading comprehension is β = .218, reflecting a moderate contribution to comprehension processes. Previous results of the hierarchical regression and the path coefficients of the contributions of vocabulary knowledge to reading comprehension assert the role of lexical knowledge in comprehension processes, whether direct or indirect. Given that all confidence intervals of the significant path coefficients do not include zero suggests a significant relationship between all the variables included in the model.
Discussion
This study aimed to assess the contribution of morphosemantic knowledge and processing to reading comprehension among intermediate EFL learners. Although the results showed a significant effect of morphosemantic knowledge on reading, morphological processes did not contribute equally to comprehension. An examination of the findings suggests a need for further interpretation of some of the results.
The role of morphological processes in reading comprehension
The linear regression and the SEM were employed to answer the first research question on whether morphological processes of the morphological pathway framework mediate the contribution of morphosemantic knowledge to reading comprehension. The linear regression model revealed that lexical inferencing had a considerable impact on reading comprehension, whereas morphological decomposition did not provide any support. Morphological segmentation, according to the SEM model, is driven by morphosemantic knowledge but is an inefficient processing pathway that does not assist during reading. While the findings stand in contrast to the morphological pathway framework, they are in alignment with previous research within the EFL context where efficient morphological-based decomposition was found ineffective during reading (Alshehri & Zhang, 2022; Zhang, 2021; Zhang & Koda, 2012).
The results of the statistical analysis reflected the weighted contribution of each processing path for morphosemantic knowledge. Lexical inferencing played the most facilitative role in reading. These results are consistent with previous literature (Zhang & Koda, 2012; Zhang & Shulley, 2017; Zhang et al., 2020). A significant correlation and association between MBLI and vocabulary size have been detected by previous scholars (e.g., Goodwin et al., 2022; Zhang et al., 2022; Zhang & Shulley, 2017). MBLI as a skill is dependent on two knowledge components: morphosemantic knowledge and vocabulary knowledge. Nevertheless, the identification of lexical disambiguation as a processing route whereby knowledge components also assist in reading comprehension otherwise asserts its significance. This indicates that lexical disambiguation is not only a facilitative path for knowledge components but also a distinctive skill that assists in meaning construction during reading. The identification of semantic building blocks and the extraction of meaning from semantic units extends beyond knowledge of word components. This skill likely entails metalinguistic awareness of word formation processes and the construction of meaning in the English language, which assist in global text comprehension.
The insufficiency of morphological segmentation as a facilitative skill during reading has already been established in the EFL context (Alshehri & Zhang, 2022; Zhang, 2021; Zhang & Koda, 2012). Differentiation between morphologically complex words and pseudo-complex words was found to be reliant on learners' morphosemantic knowledge, whereas vocabulary knowledge did not have any effect. Based on students’ adequate performance on the task, the inefficiency of the skill as a predictor of reading comprehension cannot be attributed to participants' failure in the task due to whole word storage (Clahsen & Felser, 2018), or learners’ reliance on declarative rather than procedural memory necessary for instant word deconstruction (Ullman, 2006). While ortho-semantic processes have been found to assist in reading comprehension among native English speakers, L2 learners follow different processing routes. This study showed the inapplicability of the morphological pathway framework within the EFL context. This suggests that L2 learners follow different processing mechanisms during reading.
The contributory pathways of morphology to reading comprehension
To answer the second research question, an examination of the interaction between knowledge components and processes can shed light on how morphology can support literacy among EFL learners. While this research and previous literature have established the inefficiency of morphological segmentation as a contributor to reading comprehension (e.g., Alshehri & Zhang, 2022; Zhang, 2021; Zhang & Koda, 2012), its facilitative role in lexical inferencing (Zhang et al., 2016) prompted an investigation of a possible contribution. The study revealed a different outlook on the relationship between the variables. In such a view, morphological decomposition does not necessarily mediate morphosemantic knowledge and reading but rather morphosemantic knowledge and lexical inferencing. This indicates that morphological segmentation might not contribute to reading comprehension directly but rather to the ability to accurately infer the meaning of semantic units, which in turn greatly facilitates text comprehension. Zhang and Koda (2012) stress that the significance of morphological decomposition stems only from its subsequent effect by facilitating lexical inference. Zhang (2017) asserts that morphological decomposition occurs inherently at early stages of MBLI, as without knowledge of morphemic structure, an accurate inference of a word’s meaning cannot be generated. In MBLI tasks, students are commonly encouraged to guess the meaning of the morphologically complex words from their morphemic constituents; however, morphemic boundaries are not marked to assist students. While students are informed about the complexity of the word, they are still required to identify and segment the morphemic units. Focusing on pseudo and morphologically complex words, Rastle and Davis (2008) stress the significance of morphological decomposition to distinguish between the two types of words and segment the latter into their componential parts. They also argue that morphological segmentation is unlikely to be purely orthographic: they highlight that while orthographic-based decomposition might initially take place, semantic-based segmentation occurs on a parallel level.
The findings of this research and previous findings in relation to the utility of morphological decomposition (Alshehri & Zhang, 2022; Bar-Kochvan & Hasselhorn, 2017; Zhang, 2021; Zhang & Koda, 2012) and general word decoding in reading comprehension (Kieffer & Lesaux, 2012; Zhang et al., 2023) among L2 learners question the applicability of the morphological pathway framework within the EFL context. Morphological decoding while found to be effective during reading among native English speakers, it might not be beneficial to EFL learners. The insignificance of ortho-semantic processing in this study and the prior establishment of the futility of decoding among non-native English speakers imply that morphosemantic knowledge among L2 learners is most likely mediated by lexical processes.
The results of the study concerning the contribution of morphosemantic knowledge to reading comprehension revealed it to be exclusively indirect. While lexical inference was found to partially mediate the effect of morphosemantic knowledge on reading comprehension, as predicted by the morphological pathway framework, it was not the only processing route. Morphosemantic knowledge was also found to improve vocabulary knowledge. This effect can be achieved either by enhancing vocabulary growth or by improving learners’ lexical representation. Such knowledge is assumed to aid in general comprehension processes. Identical mediation patterns were detected by previous scholars, specifically among L2 learners (e.g., Kieffer & Lesaux, 2012; Yamashita & Kusanagi, 2024; Zhang, 2021; Zhang & Koda, 2012; Zhang & Lin, 2021). These findings stand in contrast to previous literature among L2 adult learners, where morphosemantic knowledge was found to have a direct impact on reading comprehension (e.g., Alshehri & Zhang, 2022; Yamashita & Kusanagi, 2024). This discrepancy in the findings might be explained by the methodological differences in vocabulary measurements used in the studies. The direct effect of morphosemantic knowledge was detected in studies in which vocabulary knowledge was measured using learners' L1 synonyms. While both bilingual and monolingual versions of the VLT have been found to generally produce approximate results (Karami et al., 2020), measurements of vocabulary size in English might be more suitable in this context. Readers in the EFL context are expected to rely entirely on L2 knowledge during automatic comprehension processes rather than on translation strategies, which might complicate accurate meaning construction. This assumption may be heightened when learners’ vocabulary knowledge is used in conjunction with language-specific features such as morphological knowledge.
The interweaving nature of morphosemantic knowledge and processing with vocabulary seems to be a distinct pattern of MA's contribution to reading within the EFL context. These channels of morphological contribution in support of literacy in this research are in accordance with previous literature (Zhang, 2021; Zhang & Lin, 2021) but stand in opposition to the morphological pathway framework. The divergence is not only detected in the insignificance of morphological deconstruction as a processing pathway but also in the assumptions that only through morphological-based lexical inferencing can morphosemantic knowledge contribute to reading.
The initial hierarchical linear regression affirmed the role of vocabulary knowledge in reading comprehension before accounting for other variables that might mitigate that relationship. This affirms the role of lexical knowledge as a significant predictor that drives reading comprehension and asserts the assumptions of the RSF. The interplay of morphology and vocabulary knowledge in this research is also in alignment with RSF, which contends that enhanced lexical representation drives meaning construction during reading.
Conclusion and limitations
The findings of this study suggest that the morphological pathway framework does not apply to the EFL context. Although lexical inferencing from semantic units has been proven to facilitate reading comprehension by applying morphosemantic knowledge, segmentation of multimorphemic words has not. This study also detangled the means by which morphosemantic knowledge contributes to reading. This revealed that the role of morphosemantic knowledge in reading comprehension is not restricted to facilitating lexical inference but also enriching lexical representations, which in turn aids in text comprehension.
This study has important pedagogical implications. Guided by the processing mechanisms employed by EFL learners, teachers can tailor lessons to better capitalize on available knowledge components and enhance text comprehension. Moreover, the significant findings of this study encourage the incorporation of morphology into lessons. The inclusion of morphology should not be restricted to focusing on its morphosemantic facet, but it should also focus on the instruction of morphological processes that aid in meaning construction. Such processes serve to improve lexical knowledge and comprehend the interaction of morphemic units to disambiguate lexical items. Previous studies on morphological instruction have shown that morphological interventions positively impact learners’ lexical knowledge (Zhang & Zou, 2020) and literacy skills (Amirjalili & Jabbari, 2018). While such studies establish a positive correlation, a more fitted approach based on cognitive processes might be more beneficial.
Several limitations in this study suggest avenues for future research. Although this study only examined ortho-semantic segmentation as a contributor to reading comprehension, future studies might investigate the possible role of morphological decoding as opposed to decomposition. The utilization of decoding skills by intermediate-level EFL learners during reading, as shown in previous studies (Kieffer & Lesaux, 2012; Zhang et al., 2020), further supports this direction.
Moreover, although this study showed the inapplicability of the morphological pathway framework among intermediate EFL learners, future studies might test its applicability among populations of lower or higher proficiency who may share similar processing strategies with native English speakers. Additionally, this study’s unidimensional approach, which focuses entirely on the semantic aspect, might explain the exclusive indirect contribution of morphological knowledge to reading. This shortcoming encourages a more holistic examination, including morphosyntactic knowledge, which has been found to greatly assist and directly impact text comprehension processes (Alshehri & Zhang, 2022; Goodwin et al., 2022; Zhang & Lin, 2021).
While this study shed light on MA’s manner of contribution to reading among EFL learners, interpretations should be made with caution, given the limited number of participants and the gender asymmetry in the study population. This limitation arose from limited access to male participants resulting from procedural restrictions in the study context, where gender segregation is followed. Moreover, within this proficiency level, female participants were more numerous and, therefore, more accessible to the researcher. Thus, the findings of this research serve only as preliminary evidence of the inapplicability of the morphological pathway framework. The findings of this research warrant additional examination with larger samples. A more comprehensive sample is likely to generate more reliable results.
The detection of differences between native and non-native English-speaking populations, regardless of proficiency level, underscores the need to establish models based on EFL learners rather than exclusively on native English speakers. This is heightened by the consistent findings in the literature where vocabulary knowledge plays a pivotal role in the MA-reading comprehension association. Such research will aid not only in the amendment of current morphological models but also in the creation of a more tailored model based exclusively on EFL learners. Additionally, further research is needed in the EFL context to understand the processing mechanisms of learners at different proficiency levels. This is likely to provide an outline of learners’ developmental patterns that guide curriculum design regarding the appropriate time for skill introduction and the manner of instruction to effectively achieve learning outcomes.
Author contributions
The manuscript was written by a single author.
Funding
No funding was received for this research.
Availability of data and materials
The dataset analyzed during the current study is available from the corresponding author on reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Abbreviations
English as a foreign language
Morphological awareness
Morphological-based lexical inferencing
Morphological decomposition
Morphosemantic knowledge
Vocabulary size
Reading comprehension
Reading system framework
Structural equation modelling
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
Alshehri, MG; Zhang, D. The lexical basis of second language reading comprehension: From (sub)lexical knowledge to processing efficiency. Language Learning; 2022; [DOI: https://dx.doi.org/10.1111/lang.12478]
Amirjalili, F; Jabbari, AA. The impact of morphological instruction on morphological awareness and reading comprehension of EFL learners. Cogent Education; 2018; [DOI: https://dx.doi.org/10.1080/2331186X.2018.1523975]
Anderson, RC; Hiebert, EH; Scott, JA; Wilkinson, IAG. Becoming a nation of readers: The report of the commission on reading; 1985; National Institute of Education, U.S. Department of Education:
Anglin, JM; Miller, GA; Wakefield, PC. Vocabulary development: A morphological analysis. Monographs of the Society for Research in Child Development; 1993; 58,
Anwyl-Irvine, AL; Massonnié, J; Flitton, A; Kirkham, N; Evershed, JK. Gorilla in our midst: An online behavioral experiment builder. Behavior Research Methods; 2020; [DOI: https://dx.doi.org/10.3758/s13428-019-01237-x]
Bar-Kochva, I; Hasselhorn, M. The training of morphological decomposition in word processing and its effects on literacy skills. Frontiers in Psychology; 2017; [DOI: https://dx.doi.org/10.3389/fpsyg.2017.01583]
Carlisle, JF. Feldman, LB. Morphological awareness and early reading achievement. Morphological aspects of language processing; 1995; Erlbaum: pp. 189-211.
Chen, T. The role of morphological awareness in L2 Chinese lexical inference: From a perspective of word semantic transparency. Reading and Writing; 2019; [DOI: https://dx.doi.org/10.1007/s11145-018-9916-8]
Choi, Y. (2017). Effects of dimensions of word knowledge and their accessibility on different levels of reading comprehension in adolescent EFL learners (Publication No. 10619869) [Doctoral dissertation, Michigan State University]. ProQuest Dissertations Publishing. https://www.proquest.com/docview/1951788261
Clahsen, H; Felser, C. Some notes on the shallow structure hypothesis. Studies in Second Language Acquisition; 2018; [DOI: https://dx.doi.org/10.1017/S0272263117000250]
Deacon, SH; Kieffer, MJ; Laroche, A. The relation between morphological awareness and reading comprehension: Evidence from mediation and longitudinal models. Scientific Studies of Reading; 2014; [DOI: https://dx.doi.org/10.1080/10888438.2014.926907]
Deacon, SH; Tong, X; Francis, K. The relationship of morphological analysis and morphological decoding to reading comprehension. Journal of Research in Reading; 2017; [DOI: https://dx.doi.org/10.1111/1467-9817.12056]
Foorman, BR; Koon, S; Petscher, Y; Mitchell, A; Truckenmiller, A. Examining general and specific factors in the dimensionality of oral language and reading in 4th-10th grades. Journal of Educational Psychology; 2015; [DOI: https://dx.doi.org/10.1037/edu0000026]
Goodwin, AP; Petscher, Y; Jones, S; McFadden, S; Reynolds, D; Lantos, T. The monster in the classroom: Assessing language to inform instruction. The Reading Teacher; 2020; [DOI: https://dx.doi.org/10.1002/trtr.1870]
Goodwin, AP; Petscher, Y; Reynolds, D. Unraveling adolescent language & reading comprehension: The monster’s data. Scientific Studies of Reading; 2022; [DOI: https://dx.doi.org/10.1080/10888438.2021.1989437]
Guo, Y; Roehrig, AD; Williams, RS. The relation of morphological awareness and syntactic awareness to adults’ reading comprehension: Is vocabulary knowledge a mediating variable. Journal of Literacy Research; 2011; [DOI: https://dx.doi.org/10.1177/1086296X11403086]
James, E; Currie, NK; Tong, SX; Cain, K. The relations between morphological awareness and reading comprehension in beginner readers to young adolescents. Journal of Research in Reading; 2021; 44,
Jeon, EH. Contribution of morphological awareness to second-language reading comprehension. Modern Language Journal; 2011; 95,
Karami, H; Kouhpaee Nejad, M; Nourzadeh, S; Ahmadi Shirazi, M. Validation of a bilingual version of the vocabulary size test: Comparison with the monolingual version. International Journal of Bilingual Education and Bilingualism; 2020; [DOI: https://dx.doi.org/10.1080/13670050.2017.1391744]
Kieffer, MJ; Lesaux, NK. Direct and indirect roles of morphological awareness in the English reading comprehension of native English, Spanish, Filipino, and Vietnamese speakers. Language Learning; 2012; [DOI: https://dx.doi.org/10.1111/j.1467-9922.2012.00722.x]
Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed.). Guilford Press.
Koda, K. Cross-linguistic variations in L2 morphological awareness. Applied Psycholinguistics; 2000; [DOI: https://dx.doi.org/10.1017/S0142716400003015]
Kotzer, M; Kirby, JR; Heggie, L. Morphological awareness predicts reading comprehension in adults. Reading Psychology; 2021; 42,
Lee, J; Wolters, A; Grace Kim, YS. The relations of morphological awareness with language and literacy skills vary depending on orthographic depth and nature of morphological awareness. Review of Educational Research; 2023; [DOI: https://dx.doi.org/10.3102/00346543221123816]
Levesque, KC; Breadmore, HL; Deacon, SH. How morphology impacts reading and spelling: Advancing the role of morphology in models of literacy development. Journal of Research in Reading; 2021; 44,
Levesque, KC; Kieffer, MJ; Deacon, SH. Morphological awareness and reading comprehension: Examining mediating factors. Journal of Experimental Child Psychology; 2017; [DOI: https://dx.doi.org/10.1016/j.jecp.2017.02.015]
Levesque, KC; Kieffer, MJ; Deacon, SH. Inferring meaning from meaningful parts: The contributions of morphological skills to the development of children’s reading comprehension. Reading Research Quarterly; 2019; [DOI: https://dx.doi.org/10.1002/rrq.219]
Liu, Y; Groen, MA; Cain, K. The association between morphological awareness and reading comprehension in children: A systematic review and meta-analysis. Educational Research Review; 2024; [DOI: https://dx.doi.org/10.1016/j.edurev.2023.100571]
Merriam-Webster. (n.d.) Introvert. In Merriam-Webster.com dictionary. Retrieved October 24, 2023, from https://www.merriam-webster.com/
Messick, S. Meaning and values in test validation: The science and ethics of assessment. Educational Researcher; 1989; [DOI: https://dx.doi.org/10.3102/0013189X018002005]
Messick, S. Validity of psychological assessment: Validation of inferences from persons’ responses and performances as scientific inquiry into score meaning. American Psychologist; 1995; [DOI: https://dx.doi.org/10.1037/0003-066X.50.9.741]
Nagy, WE; Anderson, RC. How many words are there in printed school English?. Reading Research Quarterly; 1984; 19,
Nagy, WE; Carlisle, JF; Goodwin, AP. Morphological knowledge and literacy acquisition. Journal of Learning Disabilities; 2014; [DOI: https://dx.doi.org/10.1177/0022219413509967]
Nation, ISP. Teaching and learning vocabulary; 1990; Heinle & Heinle:
Oakhill, J., Cain, K., & Elbro, C. (2019). Reading comprehension and reading comprehension difficulties. In Reading Development and Difficulties: Bridging the Gap Between Research and Practice. https://doi.org/10.1007/978-3-030-26550-2_5
Perfetti, C; Stafura, J. Word knowledge in a theory of reading comprehension. Scientific Studies of Reading; 2014; [DOI: https://dx.doi.org/10.1080/10888438.2013.827687]
Posit Team. (2024). RStudio: Integrated Development Environment for R (Version 2024.12.0). Posit Software, PBC. https://posit.co/products/open-source/rstudio/
Rastle, K; Davis, M. Morphological decomposition based on the analysis of orthography. Language and Cognitive Processes; 2008; [DOI: https://dx.doi.org/10.1080/01690960802069730]
Raudszus, H; Segers, E; Verhoeven, L. Use of morphological and contextual cues in children’s lexical inferencing in L1 and L2. Reading and Writing; 2021; [DOI: https://dx.doi.org/10.1007/s11145-021-10122-z]
Sasao, Y; Webb, S. The word part levels test. Language Teaching Research; 2017; 21,
Shoeib, AF.
Stevens, P; Plaut, DC. From decomposition to distributed theories of morphological processing in reading. Psychonomic Bulletin & Review; 2022; [DOI: https://dx.doi.org/10.3758/s13423-022-02086-0]
Tabachnick, BG; Fidell, LS. Using multivariate statistics; 2007; Pearson Education:
Ullman, MT. The declarative/procedural model and the shallow structure hypothesis. Applied Psycholinguistics; 2006; 27,
Vaske, J. J. (2008). Survey Research and Analysis: Applications in parks, Recreation and Human dimensions. https://ci.nii.ac.jp/ncid/BB11230885
Wang, T; Zhang, H. Contributions of morphological awareness and lexical inferencing ability to L2 vocabulary knowledge among Chinese EFL learners: A structural equation modeling analysis. Applied Linguistics Review; 2022; [DOI: https://dx.doi.org/10.1515/applirev-2022-0022]
Webb, S; Sasao, Y; Ballance, O. The updated vocabulary levels test: Developing and validating two new forms of the VLT. International Journal of Applied Linguistics; 2017; [DOI: https://dx.doi.org/10.1075/itl.168.1.02web]
Yamashita, J; Kusanagi, K. Direct and indirect contributions of three aspects of morphological knowledge to second language reading comprehension. Education Sciences; 2024; [DOI: https://dx.doi.org/10.3390/educsci14030270]
Zhang, D. Derivational morphology in reading comprehension of chinese-speaking learners of English: A longitudinal structural equation modeling study. Applied Linguistics; 2017; [DOI: https://dx.doi.org/10.1093/applin/amv072]
Zhang, D; Ke, S; Mo, Y. Morphology in reading comprehension among school-aged readers of English: A synthesis and meta-analytic structural equation modeling study. Journal of Educational Psychology; 2023; [DOI: https://dx.doi.org/10.1037/edu0000797]
Zhang, D; Koda, K. Contribution of morphological awareness and lexical inferencing ability to L2 vocabulary knowledge and reading comprehension among advanced EFL learners: Testing direct and indirect effects. Reading and Writing; 2012; [DOI: https://dx.doi.org/10.1007/s11145-011-9313-z]
Zhang, D; Koda, K; Leong, CK. Morphological awareness and bilingual word learning: A longitudinal structural equation modeling study. Reading and Writing; 2016; 29,
Zhang, H. The longitudinal effect of morphological awareness on higher-order literacy skills among college L2 learners. Contemporary Educational Psychology; 2021; [DOI: https://dx.doi.org/10.1016/j.cedpsych.2021.101969]
Zhang, H; Lin, J. Morphological knowledge in second language reading comprehension: Examining mediation through vocabulary knowledge and lexical inference. Educational Psychology; 2021; 41,
Zhang, H; Zhang, J; Han, C; Zhou, X; Su, Y; Li, H. Integrating morphological and contextual cues in lexical inferencing of Chinese fourth graders. Applied Psycholinguistics; 2022; [DOI: https://dx.doi.org/10.1017/S0142716422000157]
Zhang, H; Zou, W. Morphological intervention in promoting higher-order reading abilities among college-level second language learners. Sustainability; 2020; [DOI: https://dx.doi.org/10.3390/su12041465]
Zhang, J; Lin, TJ; Liu, Y; Nagy, WE. Morphological awareness and reading comprehension: Differential mediation mechanisms in native English speakers, fluent English learners, and limited English learners. Journal of Experimental Child Psychology; 2020; [DOI: https://dx.doi.org/10.1016/j.jecp.2020.104915]
Zhang, J; Shulley, L. Poor comprehenders in English-only and English language learners: Influence of morphological analysis during incidental word learning. Journal of Research in Reading; 2017; [DOI: https://dx.doi.org/10.1111/1467-9817.12100]
© The Author(s) 2025. This work is published under http://creativecommons.org/licenses/by-nc-nd/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.