J Psycholinguist Res DOI 10.1007/s10936-014-9327-1 Morphological Awareness in Literacy Acquisition of Chinese Second Graders: A Path Analysis Haomin Zhang © Springer Science+Business Media New York 2014 Abstract The present study tested a path diagram regarding the contribution of morphological awareness (MA) to early literacy acquisition among Chinese-speaking second graders (N = 123). Three facets of MA were addressed, namely derivational awareness, compound awareness and compound structure awareness. The model aimed to test a theory of causal order among measures of MA and literacy outcomes. Drawing upon multivariate path analysis, direct and indirect effects of MA were analyzed to identify their role in literacy performance among young children. Results revealed that all three facets of MA made significant contributions to lexical inference ability. In addition, compound awareness showed a unique and significant contribution to vocabulary knowledge. It was also observed that lexical inference ability had a mediating effect predictive of both vocabulary knowledge and reading comprehension. Moreover, vocabulary knowledge mediated the effect of MA on reading comprehension. However, no significant contribution of MA to reading comprehension was found after controlling for lexical inference ability and vocabulary knowledge. Keywords Morphological awareness · Mediating effects · Lexical inference ability · Vocabulary knowledge · Reading comprehension Introduction Morphological awareness (MA) is the knowledge a speaker has about the mapping of sound to meaning in a given language and his/her understanding of the word formation rules that guide the combination of morphemes (Kuo and Anderson 2006). Berko (1958) argues that MA, which is triggered by exposure to oral and written language, begins to develop in Englishspeaking children as young as four. Preschoolers are able to produce inflected words (e.g. plurals) at an early stage. In addition, Carlisle and Fleming (2003) provided some evidence for the development of MA among young elementary-age children. Their findings indicated H. Zhang (B) Department of Modern Languages, Carnegie Mellon University, Pittsburgh, PA, USA e-mail: haominzh@cmu.edu 123 J Psycholinguist Res that first-grade students were able to segment words into meaningful components, such as still-ness, but that children’s ability to extract the meaning of a morphologically-complex word did not develop until the third grade. By third grade, children started to show their awareness of suffixes that change grammatical roles. Numerous studies have explored the role of MA in early literacy or biliteracy acquisition across both alphabetic languages (e.g., Kieffer and Lesaux 2008; Siegel 2008) and morphosyllabic languages (e.g., Ku and Anderson 2003; Pasquarella et al. 2011; Wang et al. 2006). Ku and Anderson (2003) state that children who are knowledgeable about morphology decompose unknown words to meaningful segments—roots, prefixes and suffixes—and then derive the meanings from the integration of those meaningful units. MA has been found to facilitate the development of vocabulary knowledge (e.g., Kieffer and Lesaux 2012; Nagy and Anderson 1984) by enhancing children’s understanding of unknown words, and has also been conjectured to contribute to text comprehension (e.g., Carlisle 2000; Nagy et al. 2006) and reading ability (e.g., Carlisle 1995; Kieffer and Lesaux 2008; Nunes and Bryant 2006). In this study, it is hypothesized that Chinese-specific MA enhances the development of vocabulary knowledge and reading comprehension among Chinese-speaking second graders. Multivariate path analysis was used to test the mediated and unmediated effects of MA on multiple literacy outcomes (vocabulary knowledge and reading comprehension) in a causal order. Conceptualizing Morphological Awareness A morpheme is the smallest grammatical unit in a language. In theoretical linguistics, morphology is a particular sub-discipline devoted to studying a given language’s morphemes and structural units, e.g. affixes; root words. Generally, morphology can be subdivided into inflectional morphology (e.g., past tense -ed, third person singular -s), derivational morphology (e.g., prefixes, in-, un-, im-), and compounding (e.g., class + room). Morphological awareness was defined as “the awareness of morphemic structures of words and the ability to reflect on and manipulate that structure” (Carlisle 1995, p. 194). Mcbride-Chang et al. (2005) conceptualized MA as “the awareness of and access to morphemes, reflected in the ability to apply morphemic knowledge to recognize and create new word forms that are morphologically complex and conform to the structure of a given language” (p. 141). Koda (2000) identified several capabilities that should be included in MA: recognition that words can be segmented to sub-lexical elements; the ability to map those elements onto graphic symbols and use one’s understanding of structural relationships to assemble them. Incorporating these conceptualizations, two particular key constructs of MA are noteworthy: (1) the ability to recognize and extract sub-lexical morphemic information; (2) the ability to understand morphemic structures and their relationships. Therefore, in the current study, MA was defined as sensitivity to the correct extraction of sub-lexical morphological elements from a word and the ability to assemble a word by utilizing one’s understanding of the structural relationships between different morphemes. Morphological Awareness and Literacy Acquisition in Chinese Three types of morphological structures (inflections, derivatives, and compounds) can be found in Chinese morphology (Packard 2000). However, there are few derivational and inflectional affixes in Chinese. Sun (2006) argues that there are only four salient inflection-like 123 J Psycholinguist Res affixes in Chinese: the plural marker for human nominal, –men; the perfective aspect marker, –le; the experiential marker, –guo; the imperfective marker, –zhe. To date, few studies have investigated the relationship between inflectional performance and reading acquisition beyond the second grade (Kuo and Anderson 2006) in either English or Chinese, since acquisition of major inflectional structures is completed by early elementary grades (e.g., Anisfeld and Tucker 1968; Berko 1958). Packard (2000) states that 75 % of Chinese words are compounds formed from two or three morphemes/characters. In addition, Lü (1984) maintains that disyllabic compounds and roots with derivation-like morphemes represent 61 % of the three-thousand most commonly-used Chinese words. Lexical compounding and the addition of derivational morphemes to bound roots are the primary methods of Chinese word formation (Packard 2000). A growing body of recent work has endorsed the important role of lexical compounding knowledge in reading acquisition among Chinese-speaking children (e.g., Chen et al. 2009; Ku and Anderson 2003; Liu and McBride-Chang 2010; Liu et al. 2013; Zhang 2013; Zhang and Koda 2012). Chen et al. (2009) reported that sensitivity to Chinese compounding structure was critical to young children’s literacy acquisition, including character reading and vocabulary development. They considered compound awareness to comprise two layers (1) the ability to identify the head of a compound noun and (2) the ability to construct a new compound word from familiar morphemes. Chinese lexical compounding is vital in chunk meaning construction. The results of Chen et al.’s study demonstrated that Chinese compound awareness developed over time among first-grade and second-grade children and that it made a larger contribution to children’s development of vocabulary knowledge than phonological awareness did. To further uncover the role of compound awareness in literacy acquisition among young children, Liu and McBride-Chang (2010) drew upon a new open-ended lexical compounding task to assess MA among Chinese third graders. The compounding production task that they used identified a unique proportion of variance in both Chinese character reading and vocabulary knowledge by controlling for phonological awareness and non-verbal intelligence. By adding both lexical compounding and derivational morphology as factors in MA tasks, Ku and Anderson (2003) explored the development of MA and reading ability among Englishspeaking and Chinese-speaking elementary-age children. Comparable measures of morphological awareness, including morpheme recognition and discrimination tasks, were adopted to assess children’s awareness of derivational and compounding morphology. Their findings indicated that morphological awareness improved with grade level and it was significantly related to children’s reading ability. More importantly, the study found that Chinese children developed lexical compounding rules earlier than derivational rules, which might imply that lexical compounding is the dominant word formation strategy in Chinese. Questions remain as to whether children’s ability to manipulate Chinese lexical compounding and derivational morphology can be transferred to facilitate reading ability in another language. Zhang and Koda (2014) examined Chinese-English biliteracy acquisition among Chinese fifth and sixth graders. Both compound awareness and derivational awareness independently and significantly predicted bilingual reading comprehension. Evidence of cross-language transfer facilitation indicated that Chinese compound awareness explained a significant proportion of variance in English reading comprehension while English compound awareness was not found to contribute significantly to Chinese reading comprehension. Similarly, Zhang (2013) found that Chinese compound awareness also significantly predicted English compound word meaning inference among fifth-grade and sixth-grade Chinesespeaking EFL learners. Nonetheless, Chinese derivational awareness was not transferred to facilitate the understanding of English derived words’ meaning. 123 J Psycholinguist Res These findings highlighted the role of Chinese lexical compounding in literacy/biliteracy acquisition among Chinese children. In the present study, it is indispensable to further explore and validate the effect of Chinese-specific MA in order to understand what facets of morphological awareness have a significant impact on literacy development among young learners and what metalinguistic tools we can use to boost children’s literacy acquisition. Morphological Awareness, Vocabulary Knowledge and Reading Comprehension: Direct and Indirect Effects of Morphological Awareness Recent studies have probed the causal relationships among morphological awareness, vocabulary knowledge and reading comprehension under the Structural Equation Modeling framework (e.g., Nagy et al. 2006; Kieffer and Lesaux 2012; Zhang and Koda 2012). Given the fact that vocabulary knowledge has a strong relationship with reading comprehension ability (e.g., Anderson and Freebody 1983; Qian 1999), it is predicted that expanding vocabulary knowledge by developing morphological awareness, will strengthen reading comprehension. Nagy et al. (2006) explored the causal order among morphological awareness, vocabulary knowledge and reading comprehension in monolingual English-speaking children. Their findings indicated that both morphological awareness and vocabulary knowledge yielded significant contributions to reading comprehension. In addition, morphological awareness was found to have an indirect effect on reading comprehension via vocabulary knowledge. Similarly, Kieffer and Lesaux (2012) conducted a multiple-group structural equation modeling study in which they tested the direct and indirect contributions of morphological awareness to English reading comprehension among sixth graders from various language backgrounds. Their results showed that morphological awareness made a significant direct contribution to reading comprehension after controlling for reading vocabulary and reading fluency. Furthermore, in line with the results of Nagy et al. (2006), Kieffer and Lesaux (2012) also found that morphological awareness made a significant indirect contribution to reading comprehension via reading vocabulary. According to the results of the two aforementioned studies, vocabulary knowledge can mediate the effect of MA on reading comprehension. This mechanism of developing reading comprehension enriches our understanding about the kinds of linguistic resources we can draw upon to boost reading acquisition in young children. However, it remains to be determined whether the cause-and-effect model can be complicated by other factors and whether there are other path routes connecting MA, vocabulary knowledge and reading comprehension. Koda (2005) maintains that contextual lexical inference creates a pathway between intraword awareness (e.g. morphological awareness) and word-knowledge development, which ultimately contributes to both vocabulary knowledge and reading comprehension. Haastrup (1991) conceptualized lexical inference ability as “making informed guesses as to the meaning of a word, in light of all available linguistic cues in combination with the learner’s general knowledge of the world, her awareness of context and her relevant linguistic knowledge” (p. 40). The ability to apply lexical inferencing largely depends on knowledge of three categories of information, namely word-internal elements (e.g., phonemes and morphemes); contextual information (e.g., syntactic, semantic and pragmatic constraints) and prior knowledge (Wesche and Paribakht 2010). Intraword awareness, especially MA, plays a facilitative role in word learning, which provides insights to assist in the analysis of a word’s internal properties in order to establish the 123 J Psycholinguist Res linkage between newly-encoded word information and preexisting word knowledge (Koda 2005). Considering the role of lexical inference ability in the development of vocabulary and reading, Zhang and Koda (2012) tested the direct and indirect effects of MA on L2 vocabulary knowledge and reading comprehension among college-level EFL learners. The study found MA contributed to L2 vocabulary knowledge directly and indirectly via lexical inference ability. However, no significant direct contribution of morphological awareness to reading comprehension was found. MA was found to contribute to reading comprehension only through vocabulary knowledge and lexical inference ability. Little research has explored the role of lexical inference ability in the development of vocabulary knowledge and reading comprehension among young learners. Zhang and Koda (2012) mention that current research has not provided us with a clear picture of the relationship between lexical inference ability and reading acquisition. More evidence is required to ascertain whether morphological awareness contributes to lexical inference ability, whether lexical inference ability plays a facilitative role in vocabulary knowledge and reading comprehension, and whether lexical inference ability mediates morphological awareness to influence the development of literacy outcomes. Uncertainties remain with respect to the cause-and-effect among those intertwined literacy factors. The Present Study Path Model and Hypotheses Based on the literature and conceptual framework discussed above, the path model shown in Fig. 1 identifies three facets of Chinese-specific morphological awareness (derivational awareness, compound awareness, and compound structure awareness) and hypothesizes that (1) morphological awareness contributes to lexical inference ability; (2) morphological awareness fosters the development of vocabulary knowledge directly or indirectly via lexical inference ability; (3) lexical inference ability and vocabulary knowledge are mediators predicting reading comprehension. More specifically, this study aims to address the following research questions: 1. Does morphological awareness make a significant contribution to lexical inference ability? Does morphological awareness make a significant contribution to vocabulary knowledge directly or indirectly through lexical inference ability? Derivational Awareness Lexical Inference Compound Awareness Reading Comprehension Vocabulary Knowledge Compound Structure Fig. 1 Hypothesized path model 123 J Psycholinguist Res 2. Does morphological awareness significantly contribute to reading comprehension after adjusting for lexical inference ability and vocabulary knowledge? Do lexical inference ability and vocabulary knowledge mediate the effect of morphological awareness on reading comprehension? The two sets of research questions addressed the interconnected relationships among morphological awareness, lexical inference ability, vocabulary knowledge and reading comprehension. Based on the model presented below, it is predicted that Chinese-specific morphological awareness including derivational awareness, compound awareness, and compound structure awareness contributes to vocabulary knowledge and lexical inference ability. Meanwhile, emerging vocabulary knowledge and lexical inference ability consequently predict reading comprehension skill. To put it differently, morphological awareness is predicted to have a significant impact on word-level abilities (vocabulary knowledge and lexical inference) and morphological awareness also predicts reading comprehension skill via the mediation of word-level abilities. Method Participants Participants were one hundred twenty-three Chinese second-grade students (68 boys and 55 girls). The average age was 7.6 years (ranging from 7 years 2 months to 8 years 2 months). Students were from a public school in Suzhou, China. The school is the only public elementary school in the school district. There are around 2000 students and 200 teachers in this public school. Three (out of eight) classes of second graders were randomly selected to participate in this study. The participating students were at the end of the first semester in their second grade. The students had been immersed in intensive literacy training since the first grade. They had two to three Chinese classes every day including oral communication, reading and writing, and exercise classes. They were regular elementary-age children without learning disabilities. Measures and Instruments Derivational Awareness Children’s derivational awareness was measured via a morpheme recognition task and a morpheme discrimination task,1 both of which were adapted from Ku and Anderson (2003) In the morpheme recognition task, children were required to judge the semantic relationship between a derived word and its sub-lexical root morpheme. For instance, children were shown the derived word “ (painter)” and its segmental root morpheme “ (draw/paint)”. They were asked to judge whether the meaning of “ (draw/paint)” is related to the meaning of “ (painter)”. In this case, “ (draw/paint)” is relevant to the meaning of “ (painter)”. In the morpheme discrimination task, children were asked to circle a word with a different morphemic meaning among three disyllabic words. For example, three apparently derived words were presented to children, e.g. (scholar), (reader), (or). All three of these words share one derived morpheme, “ -person”, but the last word is not a derived 1 There were 20 items in the morpheme recognition task and 20 items in the morpheme discrimination task. Concerning the two morphological measures categorized in this study, there were 13 items in the measure of derivational awareness and 27 items in the measure of compound awareness. 123 J Psycholinguist Res Table 1 Compounding structure and examples Compounding structures Examples Subordinate (green-grass) adj_n Coordinate (mountain-water) n_n Subject-predicate (dog-bark) n_v Verb-object (pull-door) v_n word, and does not incorporate the meaning of “person”. There were 13 items in the measure of derivational awareness. The internal consistency of this measure was α = .80. Compound Awareness Children’s compound awareness was measured by the same two tasks discussed above, namely the morpheme recognition task and the morpheme discrimination task. In the morpheme recognition task, children would see one compound-like word “ (happy)” and one of its segmental morphemes “ (tall)”, and were asked to indicate whether the meaning of “ (tall)” is related to the meaning of “ (happy)”. In this case, “ (tall)” and “ (happy) are unrelated given the fact that the meaning of this compound word cannot be derived from the semantic meaning of the segmental morpheme. For the morpheme discrimination task, three compound-like words were presented, e.g. (good at), (left hand), (raise hand), three of which share one morpheme “ hand”. One word—in this case, the first word, “good at”—did not imply the meaning of the common morpheme—in this case, “hand”. The children were instructed to circle whichever word was the “odd man out”. There were 27 items in the measure of compound awareness. The internal consistency of this measure was α = .85. Compound Structure Awareness The compound structure awareness task used was modeled after Nagy et al. (2003), Chen et al. (2009), and Liu and McBride-Chang (2010). Children were asked to choose the most appropriate compound word based on each prompt. The provided compounds were lowfrequent or novel words, which means they were less frequently or not used in real speech, for example, (book-bowl), (frozen-gold). Following the categories identified in Liu and McBride-Chang (2010), four key Chinese compounding structures were tested, as shown below in Table 1. These structures were subordinate, coordinate, subject-predicate, and verb-object. For example, a prompt ? (How would you say “a cat is dancing”?) was presented to children, after which the children needed to choose whether it was (cat-dance) or (dance-cat). There were 20 items in this measure. The internal consistency of this measure was α = .74. Vocabulary Knowledge A “select vocabulary” task (Anderson and Freebody 1983; Meara 1992) was used to measure children’s vocabulary breadth. There were 70 disyllabic words, including 54 real words and 16 non-words. Children were asked to identify the words they knew. Most of the real words were selected from children’s textbooks after consultation with class teachers. Non-words 123 J Psycholinguist Res were comprised of two real characters that the children may or may not be familiar with. For instance, one nonce item was ; the children might know -finish and -cold, but the combination of these two morphemes into a disyllabic word does not exist in Chinese. In this task, the children were asked to circle the words they had known, which means they needed to know the semantic meaning of each printed word rather than phonetic or phonological presentations of the words. Therefore, those words were not read out aloud and the children made judgments about the print lexical items. Before the children started to do the task, test administers read aloud the task description and explicitly told the children to circle the word only if they knew the meaning of the entire compound word. If they only knew one component character in a word, they were not allowed to select the word as known. When the data were analyzed, real words selected as known were coded as “real hits” and pseudowords selected as known were coded as “false alarm”. Based on Signal Detection Theory, their scores were f computed by tr ue h = h− 1− f (h: real hits; f: false alarm). Lexical Inference Ability and Reading Comprehension A reading ability test was used to measure children’s lexical inference ability and reading comprehension skill. The reading test comprised three reading passages with an average length of 220 Chinese characters (roughly 120–150 Chinese words). The reading materials were adapted from extracurricular reading practice books catering for Chinese second-grade students. Each passage consisted of a print narrative describing a number of objects, animals, or a sequential event/story and each reading passage was followed by 6–10 multiple-choices questions. The participating children were read an explicit instruction: (1) please choose the most appropriate meaning of each underlined word. If you do not know the meaning, please read the sentence/passage carefully and look at the two characters in each word to make a guess. (2) After you finish the vocabulary questions, please keep working on other comprehension questions according to each passage. Within these questions, a total of ten disyllabic words from the reading passages were selected to be tested in lexical inference questions. 2 to 4 new words were selected in each passage so that the text coverage was controlled around 95–98 %. The lexical items were finalized in consultation with class teachers so as to ensure that the children knew both component characters in each disyllabic word but had not learned the meaning of the compound word. Based on the rationale of incidental word learning and lexical inference, children were expected to utilize both word-internal (morphological) and word-external (contextual) information to construct the precise meaning of each unknown word. For example, a compound word, “ ”, was underlined in the reading, for which the children knew the morphemic meanings of both characters ( -stop; -rest). To control the amount of contextual cues surrounding an unfamiliar word, “informative context” was used to describe the clues to lexical inference (Cain et al. 2004, 2003). In this example, the unfamiliar word, “ ” was presented in a target sentence (shown in Table 2). The children were supposed to identify the informative context (shown below) in this specific sentence to derive the context-appropriate meaning. Ultimately, by integrating both the morphological information and the contextual cue, children should be able to infer the accurate, contextually appropriate meaning, which is “stop and have a rest”. Based on the stages of reading development (Chall 1996), second-grade and third-grade children should learn to recognize words and read stories composed of increasingly complex lexical items. Young readers in that stage can also take advantage of what is described in a story and relate the information to his or her knowledge and language. In addition, Chinese first and second-grade students should be able to read narrative or expository passages, locate specific information and detect the main idea in the passages (The criteria of 123 J Psycholinguist Res Table 2 Example of text used in lexical inferencing Target sentence: (Its wing has 2 m long. It can fly for hundreds of kilometers continuously and does not need to stop and rest.) Informative context: (It can fly for hundreds of kilometers continuously) Chinese curriculum 2011). Reading comprehension questions, therefore, were designed to direct children to locate/identify textual information and interpret the main idea of each passage. For instance, learners were asked to identify the specific information, such as “ ? What does a cat use to measure objects?” and interpret the main idea like “ ? According to this short passage, what is a cat like?” All test items were multiple-choices questions and each item was worth 1 point. The maximum score of lexical inference ability was 10 points and the maximum score of reading comprehension was 14 points. The internal consistency of these measures were α = .76 (lexical inference) and α = .74 (reading comprehension). Procedure and Data Collection In order to ensure internal validity, separate tasks were administered, in an attempt to rule out priming or carry-over effects from previous tasks. There were five paper-and-pencil tests: morpheme recognition, morpheme discrimination, compound structure, select vocabulary and reading tasks (lexical inference and reading comprehension). Each task was administered to the children in a different week, and tasks were counterbalanced in different classes. The time allotment was 70–80 min for all five tasks. Data collection was completed in 5 weeks. Results Descriptive Statistics and Bivariate Correlations Descriptive statistics pertaining to children’s performance on MA measures and literacy measures are provided in Table 3. There was no ceiling or floor effect. No data transformation was needed. Additionally, there was adequate variability across different measures. As shown in the Table 4 of bivariate correlations, all the variables were significantly correlated. Three measures of MA were significantly intercorrelated, r = .48, p < .001; r = .20, p < 0.05; and r = .36, p < 0.001. Among the correlations of vocabulary and MA measures, a stronger correlation was found between compound awareness and vocabulary knowledge, r = .44, p < .001. Derivational awareness and compound structure awareness were also significantly correlated with vocabulary knowledge, r = .29, p < .01 and r = .24, p < .05. Derivational awareness, compound awareness and compound structure awareness were all significantly correlated with lexical inference ability, r = .40, p < .001; r = .43, p < 0.001; and r = .42, p < 0.001. Furthermore, all the measures of MA were significantly correlated with reading comprehension, r = .36, p < .001; r = .37, p < 0.001; and r = .33, p < 0.001. Finally, lexical inference ability and vocabulary knowledge were significantly correlated with reading comprehension, r = .39, p < .001 and r = .46, p < 0.001, respectively. 123 J Psycholinguist Res Table 3 Descriptive statistics of literacy measures N =123. Numbers in parentheses represent the maximum scores that subjects could get in each measure Measure Minimum Maximum M SD Derivational awareness (13) 5 13 9.83 1.58 Compound awareness (27) 14 26 21.74 2.52 Compound structure (20) 9 20 14.97 2.16 Vocabulary knowledge (100) 22.2 92.1 63 Lexical inference ability (10) 2 10 7.33 15.77 1.65 Reading comprehension (14) 5 14 11.2 2.30 Table 4 Correlations between morphological awareness measures and literacy measures Measure 1 1. Derivational awareness – 2 3 4 5 2. Compound awareness .48*** 3. Compound structure .20* .36*** – 4. Vocabulary knowledge .29** .44*** .24* – 5. Lexical inference ability .40*** .43*** .42*** .38*** – 6. Reading comprehension .36*** .37*** .33*** .39*** .46*** 6 – – ∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .005 1. derivational awareness; 2. compound awareness; 3. compound structure; 4. vocabulary knowledge; 5. lexical inference ability; 6. reading comprehension Derivational Awareness Lexical Inference Compound Awareness Reading Comprehension Vocabulary Knowledge Compound Structure Fig. 2 Observed path model (dotted lines non-significant path routes; solid lines significant path routes) Path Model Testing To explore the causal relationships among multiple variables, a path analysis was conducted to test the path diagram. Starting with the hypothesized model in Fig. 1, the Chi-square goodness-of-fit test, along with other fit indices, was used to determine the consistency between the observed model and the expected model. It was suggested that Comparative Fit Index (CFI) greater than .95 (Bentler 1990) and a Root Mean Square Error of Approximation (RMSEA) less than 0.05 (Browne and Cudeck 1993) would show a good model fit. According to the results of Chi-square goodness-of-fit—χ 2 (3, N = 123) = 5.96, p = .114 (CFI = .979; RMSEA=.04); χ 2 /d f = 1.98 < 2—the χ 2 significance test and model fit indices showed a good fit. In Fig. 2, solid lines represent significant path coefficients between two variables. Table 5 provides standardized parameter estimates for lexical inference ability, vocabulary 123 J Psycholinguist Res Table 5 Standardized regression weights for lexical inference, vocabulary knowledge and reading comprehension VOCK vocabulary knowledge, DERA derivational awareness, COMA compound awareness, COMS compound structure awareness, LEXI lexical inference ability, READ reading comprehension Table 6 Standardized total effects, direct effect, and indirect effect on lexical inference ability ∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .001 Paths SE β̂ C.R. (z) p LEXI ← DERA .262 .092 3.027 .002 LEXI ← COMA .210 .061 2.306 .021 LEXI ← COMS .287 .064 3.472 .000 VOCK ← DERA .068 .010 .704 .481 VOCK ← COMA .355 .007 3.528 .000 VOCK ← COMS .081 .007 .848 .397 VOCK ← LEXI .269 .010 2.670 .008 READ ← LEXI .326 .127 3.538 .000 READ ← VOCK .295 1.338 2.991 .003 Lexical inference ability (R 2 = .325) Direct effect Indirect effect Total effects Derivational awareness Compound awareness .262** – .262** .210* – .210* Compound structure awareness .287*** – .287*** knowledge and reading comprehension. The results did not indicate significant contributions of derivational awareness and compound structure awareness to vocabulary knowledge (standardized β̂ = .068, p = .481; β̂ = .081, p = .397). Since the hypothesized path model had a good model fit, these two non-significant pathways were not removed for model trimming and specification. Testing Direct and Indirect Effects Direct and indirect contributions of MA were tested based on the three structural equations shown below. Table 6 demonstrates that all three facets of MA significantly predicted lexical inference ability (β̂ = .262, p < .01; β̂ = .210, p < .05; β̂ = .287, p < .001). 32.5 % of the total variance of lexical inference ability was explained by MA. Structural equation 1: LEXI = p L E X I D E R A D E R A + p L E X I C O M A C O M A + p L E X I C O M S C O M S + e1 Structural equation 2: VOCK = pV OC K D E R A D E R A + pV OC K C O M A C O M A + pV OC K C O M S C O M S + pV OC K L E X I L E X I + e2 Structural equation 3: READ = p R E AD V OC K V OC K + p R E AD L E X I L E X I + e3 Table 7 addresses the direct and indirect effects of MA on vocabulary knowledge. Among the three MA facets, only compound awareness yielded a significant independent contribution to vocabulary knowledge (β̂ = .355, p < .001). After adjusting for the other effects, it 123 J Psycholinguist Res Table 7 Standardized total effects, direct effect, and indirect effect on vocabulary knowledge ∗ p < .05; ∗∗ p < .01; ∗∗ p < .001 Table 8 Standardized total effects, direct effect, and indirect effect on reading comprehension ∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .001 Vocabulary knowledge (R 2 = .376) Direct effect Indirect effect Total effects Derivational awareness Compound awareness .068 .071 .139 .355*** .057 .412*** Compound structure awareness Lexical inference ability .081 .077 .158 .269** – .269** Reading comprehension (R 2 = .288) Direct effect Indirect effect Total effects Derivational awareness Compound awareness – .140 .140 – .190 .190 Compound structure awareness Lexical inference ability Vocabulary knowledge – .127 .127 .326*** .079 .406*** .295** – .295** was determined that compound awareness had significant combined effects on vocabulary knowledge (β̂ = .412, p < .001). The three facets of MA did not show indirect contributions to vocabulary knowledge (β̂ = .071, β̂ = .057, β̂ = .077). Meanwhile, lexical inference ability significantly predicted vocabulary knowledge (β̂ = .269, p < .01) after the effect of MA was controlled for. MA and lexical inference ability together explained 37.6 % of the variance of vocabulary knowledge. The path diagram could not specifically test the direct effect of MA on reading since these added pathways to reading comprehension would lead to zero degrees of freedom in the Chi-square test of goodness. Instead, partial mediations via lexical inference ability and vocabulary knowledge were tested to see the indirect effects of MA on reading comprehension. Overall, MA, lexical inference ability and vocabulary knowledge explained 28.8 % of the total variance of reading comprehension skill (see Table 8). No facet of MA was found to have a significant impact on reading comprehension after adjusting for the effects of lexical inference ability and vocabulary knowledge (β̂ = .140, β̂ = .190, β̂ = .127). Both lexical inference ability and vocabulary knowledge had significant combined effects on reading comprehension after controlling for the other effects (β̂ = .406, p < .001; β̂ = .295, p < .01). In all, the results demonstrated that lexical inference ability and vocabulary knowledge mediated the effect of MA on reading comprehension. Discussion The goal of this study, building upon previous research, was to extend our understanding of literacy development of young Chinese-speaking children. The main objective was to provide insight into the role of morphological awareness in literacy performance, 123 J Psycholinguist Res including both vocabulary knowledge and reading comprehension skill. The hypothesized path diagram demonstrated a good model fit between the expected model and the observed model, which testified to the mediated and unmediated effects of MA on literacy acquisition. Through different path routes in the model, MA was found to contribute to vocabulary knowledge and reading comprehension directly or via mediating factors. Role of Morphological Awareness in Vocabulary Knowledge The first set of research questions addressed the role of morphological awareness including the awareness of segmental morphological information (derivational and compound awareness) and structural relationships (compound structure awareness) in the development of vocabulary knowledge. Concerning the development of MA, previous empirical evidence demonstrated that Chinese-specific compound awareness started to develop in first-grade students (Chen et al. 2009), second-grade students (Chen et al. 2009; Ku and Anderson 2003) and third-grade students (Liu and McBride-Chang 2010). Additionally, MA was found to improve with grade level (Chen et al. 2009; Ku and Anderson 2003). The present study found that different facets of MA were emergent and developed among Chinese second-grade students. Moreover, prior studies on Chinese vocabulary acquisition indicated that awareness of Chinese lexical compounding predicted vocabulary knowledge (Chen et al. 2009; Liu and McBride-Chang 2010; Liu et al. 2013). However, the aforementioned studies all integrated structure awareness in their measure of compound awareness. In the present study, compound structure awareness was isolated as an independent variable and was found to have no significant contribution to vocabulary knowledge. The measures of compound awareness and derivational awareness adopted in the present study were modeled on tasks designed by Ku and Anderson (2003). Although this study did not trace the developmental trajectory of lexical compounding rules and derivational morphology, it did determine that compound awareness predicted children’s vocabulary over and above derivational awareness. Children’s ability to segment and detect segmental morphological clues enhances their development of vocabulary knowledge. For instance, the morpheme (bag) forms the base for a large number of basic compound words, such as (wallet), (school bag), and (backpack). Elementary-age children who can segment one compound word and extract the key morphological cue (bag) should be able to learn and memorize compound words containing that particular morpheme relatively quickly. Chen et al. (2009) confirmed that the ability to identify the head morpheme of a compound word facilitated Chinese word learning. They argue that this identification skill develops in early elementary school. The results of the current study provided further support for the role of compound awareness in Chinese vocabulary knowledge development (e.g., Chen et al. 2009; Kieffer and Lesaux 2012; Ku and Anderson 2003; Nagy et al. 2006). This study also shed light on the crucial impact of lexical inference ability on the development of vocabulary knowledge. The results presented above demonstrated that the ability to analyze word-internal properties and to detect structural relationships boosted children’s ability to infer the meanings of unfamiliar Chinese compound words. Wesche and Paribakht (2010) found that word morphology was one of the most frequent tools that learners used to infer word meanings. Provided that MA triggers lexical inference ability, this ability would enhance incidental vocabulary learning (Koda 2005) or vocabulary learning on the spot (Nagy 2007). Empirical evidence from the present study revealed that lexical inference ability based on morphological awareness made a significant contribution to vocabulary knowledge, which is consistent with the findings of Zhang and Koda’s (2012) study. 123 J Psycholinguist Res Altogether, morphological awareness and lexical inference ability successfully predicted vocabulary knowledge (37.6 % of variance explained). Chinese compound awareness had a unique and significant effect on vocabulary knowledge among second graders, a fact which underlined the importance of Chinese lexical compounding in the development of young children’s vocabulary knowledge. It was also demonstrated that lexical inference ability was a mediator connecting morphological awareness and vocabulary knowledge. Role of Morphological Awareness in Reading Comprehension The second set of research questions addressed the role of MA in reading comprehension ability. The results of this study revealed that MA did not have significant effects on reading comprehension among Chinese-speaking second graders, which contradicted the findings concerning the relationship between morphological awareness and reading comprehension ability among monolingual English-speaking children (Nagy et al. 2006) and young English language learners (Kieffer and Lesaux 2012). Nagy et al. (2006) and Kieffer and Lesaux (2012) found that both direct and indirect contributions of morphological awareness to reading comprehension were statistically significant. By contrast, the current study found that MA did not contribute significantly to reading comprehension. MA did, however, make a substantial and significant contribution to lexical inference ability, which in turn significantly contributed to reading comprehension skill. Lexical inference ability is therefore a facilitative mediating variable associating morphological awareness and reading comprehension. Zhang and Koda (2012) similarly found that lexical inference ability was the mediating factor which significantly predicted reading comprehension. Pertaining to mediating effects, the present study also showed that vocabulary knowledge mediated the effect of morphological awareness and made a significant contribution to reading comprehension. This result is consistent with previous studies under the multivariate analytical model (Kieffer et al. 2013; Kieffer and Lesaux 2012; Nagy et al. 2006; Zhang and Koda 2012). In order to analyze the discrepancies of results among various studies, it is necessary to investigate reading comprehension measures in different studies. Kieffer et al. (2013) and Kieffer and Lesaux (2012) employed English reading comprehension measures basically to assess learners’ sentence comprehension, short sentence inference and passage comprehension. It is reasonable to speculate that MA at the word level might fill the gaps in wordmeaning knowledge, thus facilitating sentence and passage comprehension. However, Zhang and Koda’s (2012) study measured five sub-skills of reading comprehension including fillin-the-blank (words), conjunction, textual inference, co-reference, and gist detection (main ideas). Conjunction and co-reference questions which focused on organizational knowledge of textual information would possibly benefit less or indirectly from morphological problem solving at the word level. Therefore, the study found that there was a significant indirect contribution of morphological awareness to overall reading comprehension ability while direct contribution of morphological awareness was not significant. The present study focused on reading comprehension measures that emphasized locating specific textual information (80 %) and global comprehension of main ideas (20 %). A large proportion of questions were designed to measure learners’ ability to identify specific textual information, which largely depends on the use of context. By narrowing the focus of the measures in this way, this study uncovered the fact that MA contributed only minimally to reading comprehension ability, while lexical inference ability, which integrated both word-internal (morphology) and word-external (context) information can significantly predict reading comprehension skill. 123 J Psycholinguist Res In other words, lexical inference ability and vocabulary knowledge, both of which developed by MA mediated the impact of MA itself on the path model to influence reading comprehension skill. Limitations and Future Directions This study had several limitations which suggest directions for future research. First of all, this study did not measure children’s current language skills (e.g., decoding and orthographic knowledge) or cognitive ability. Since all tasks involved children’s reading vocabulary knowledge, children’s ability to decode words could have been controlled to reduce the confounding effect. For future studies, it is important to have developmental information (language skills and cognitive development) for individual participants to ensure internal validity. Secondly, as mentioned earlier, the hypothesized path model did not necessarily address the direct causal relationship between morphological awareness and reading comprehension, due to the constraints in the overall model fit. In that sense, we could only see the correlational evidence between morphological awareness and reading comprehension. In addition, reading comprehension measures were not comprehensive enough to incorporate different reading skills, including inferencing skill, comprehension skill and organizational knowledge. More observed variables may be needed to clarify the relationships among morphological awareness and different facets of reading ability. Methodologically, we could address those problems by using a latent variable model. In future research, we hope to include different observed variables (global comprehension, inferencing skill, and organizational knowledge) under a latent variable (reading ability), in order to arrive at a holistic picture of the causal relationship between morphological awareness and reading ability. The third limitation of the current study was its cross-sectionality; a cross-sectional experiment design did not allow us to trace the developmental sequences of morphological awareness and literacy skills among young children. In future research, longitudinal work in this area would allow us to examine the developmental trajectory of literacy skills. Several studies have explored the development of morphological awareness and literacy skills in different time periods (e.g., Chen et al. 2009; Kieffer and Lesaux 2008; Lam et al. 2012), but most have focused on literacy performance within a one-year period. Kieffer and Lesaux (2008) argue that collecting additional data over a longer period would allow us to see whether previously acquired emergent literacy requisites (e.g. morphological awareness) can make contributions to later literacy acquisition. With regard to research methodology, latent growth modeling would be ideal to estimate the growth curve of literacy outcomes. References Anderson, R. C., & Freebody, P. (1983). Reading comprehension and the assessment and acquisition of word knowledge. In B. Huston (Ed.), Advances in reading research (pp. 231–256). Greenwich, CT: JAI Press. Anisfeld, M., & Tucker, G. (1968). English pluralization rules of six-year-old children. Child Development, 38, 1201–1217. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. Berko, J. (1958). The child’s learning of English morphology. Word, 14, 150–177. Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. Sage Focus Editions, 154, 136–136. Cain, K., Lemmon, K., & Oakhill, J. (2004). Individual differences in the inference of word meanings from context: The influence of reading comprehension, vocabulary knowledge and memory capacity. Journal of Educational Psychology, 96, 671–681. 123 J Psycholinguist Res Cain, K., Oakhill, J., & Elbro, C. (2003). The ability to learn new word meanings from context by school-age children with and without language comprehension difficulties. Journal of Child Language, 30, 681–694. Carlisle, J. F. (1995). Morphological awareness and early reading achievement. In L. B. Feldman (Ed.), Morphological aspects of language processing (pp. 189–209). Hillsdale, NJ: Erlbaum. Carlisle, J. F. (2000). Awareness of the structure and meaning of morphologically complex words: Impact on reading. Reading and Writing: An Interdisciplinary Journal, 12, 169–190. Carlisle, J. F., & Fleming, J. (2003). Lexical processing of morphologically complex words in the elementary years. Scientific Studies of Reading, 7, 239–253. Chall, J. S. (1996). Stages of reading development (2nd ed.). Fort Worth, TX: Harcourt Brace. Chen, X., Hao, M., Geva, E., Zhu, J., & Shu, H. (2009). The role of compound awareness in Chinese children’s vocabulary acquisition and character reading. Reading and Writing: An Interdisciplinary Journal, 22, 615–631. Haastrup, K. (1991). Lexical inferencing procedures or talking about words: Receptive procedures in foreign language learning with special reference to English. Tubingen, Germany: Gunter Narr. Kieffer, M. J., & Lesaux, N. K. (2008). The role of derivational morphological awareness in the reading comprehension of Spanish-speaking English language learners. Reading and Writing: An Interdisciplinary Journal, 21, 783–804. Kieffer, M. J., & Lesaux, N. K. (2012). Direct and indirect roles of morphological awareness in the English reading comprehension of native Spanish, Filipino, Vietnamese, and English speakers. Language Learning, 62, 1170–1204. Kieffer, M. J., Biancarosa, G., & Mancilla-Martinez, J. (2013). Roles of morphological awareness in English reading comprehension for Spanish-speaking language minority learners: Exploring partial mediation by vocabulary and reading fluency. Applied Psycholinguistics, 34, 697–725. Koda, K. (2000). Cross-linguistic variations in L2 morphological awareness. Applied Psycholinguistics, 21, 297–320. Koda, K. (2005). Insights into second language reading. New York: Cambridge University Press. Ku, Y. M., & Anderson, R. C. (2003). Development of morphological awareness in Chinese and English. Reading and Writing: An Interdisciplinary Journal, 16, 399–422. Kuo, L. J., & Anderson, R. C. (2006). Morphological awareness and learning to read: A cross-language perspective. Educational Psychologist, 41, 161–180. Lam, K., Chen, X., Geva, E., Luo, Y., & Li, H. (2012). The effects of morphological awareness development on reading achievement in young English Language Learners (ELLs): A longitudinal study. Reading and Writing: An Interdisciplinary Journal, 25, 1847–1872. Liu, P. D., & McBride-Chang, C. (2010). What is morphological awareness? Tapping lexical compounding awareness in Chinese third graders. Journal of Educational Psychology, 102, 62–73. Liu, P. D., McBride-Chang, C., Wong, T.-Y., Shu, H., & Wong, A. M.-Y. (2013). Morphological awareness in Chinese: Unique associations of homophone awareness and lexical compounding to word reading and vocabulary knowledge in Chinese children. Applied Psycholinguistics, 34, 755–775. Lü, S. (1984). Xiandai hanyu dan-shuang yinjie wenti chutan “An initial account of the issues on monosyllabic or disyllabic modern Chinese”. Hanyu yufa lunwenji “ A collection of papers on Chinese grammar”. Beijing: Commerical Press. Mcbride-Chang, C., Wagner, R. K., Muse, A., Chow, B. W., & Shu, H. U. A. (2005). The role of morphological awareness in children’s vocabulary acquisition in English. Applied Psycholinguistics, 26, 415. Meara, P. (1992). EFL vocabulary tests. Swansea: Centre for Applied Language Studies, University of Wales, Swansea. Nagy, W. (2007). Metalinguistic awareness and the vocabulary-comprehension connection. In R. K. Wagner, A. E. Muse, & K. R. Tannenbaum (Eds.), Vocabulary acquisition: Implications for reading comprehension (pp. 52–77). New York, NY: The Guilford Press. Nagy, W. E., & Anderson, R. C. (1984). How many words are there in printed school English? Reading Research Quarterly, 19, 304–330. Nagy, W., Berninger, V. W., & Abbott, R. D. (2006). Contributions of morphology beyond phonology to literacy outcomes of upper elementary and middle-school students. Journal of Educational Psychology, 98, 134. Nagy, W., Berninger, V., Abbott, R., Vaughan, K., & Vermeulen, K. (2003). Relationship of morphology and other language skills to literacy skills in at-risk second-grade readers and at-risk fourth-grade writers. Journal of Educational Psychology, 95, 730. Nunes, T., & Bryant, P. (2006). Improving literacy by teaching morphemes. New York: Routledge. Packard, J. L. (2000). The morphology of Chinese. New York: Cambridge University Press. Pasquarella, A., Chen, X., Lam, K., Luo, Y. C., & Ramirez, G. (2011). Cross-language transfer of morphological awareness in Chinese-English bilinguals. Journal of Research in Reading, 34, 23–42. 123 J Psycholinguist Res Qian, D. D. (1999). Assessing the roles of depth and breadth of vocabulary knowledge in reading comprehension. Canadian Modern Language Review/La revue canadienne des langues vivantes, 56, 282–308. Siegel, L. S. (2008). Morphological awareness skills of English language learners and children with dyslexia. Topics in Language Disorders, 28, 15–27. Sun, C. (2006). Chinese: A linguistic introduction. New York: Cambridge University Press. The Ministry of Education of People’s Republic of China. (2011). The criteria of Chinese curriculum (Yuwen kecheng biaozhun). Beijing: Beijing Normal University Press. Wang, M., Cheng, C., & Chen, S. W. (2006). Contribution of morphological awareness to Chinese-English biliteracy acquisition. Journal of Educational Psychology, 98, 542. Wesche, M., & Paribakht, T. S. (2010). Lexical inferencing in a first and second language: Cross-linguistic dimensions. Bristol, UK: Multilingual Matters. Zhang, D., & Koda, K. (2012). Contribution of morphological awareness and lexical inferencing ability to L2 vocabulary knowledge and reading comprehension: Testing direct and indirect effects. Reading and Writing: An Interdisciplinary Journal, 25, 1195–1215. Zhang, D. (2013). Linguistic distance effect on cross-linguistic transfer of morphological awareness. Applied Psycholinguistics, 34, 917–942. Zhang, D., & Koda, K. (2014). Awareness of derivation and compounding in Chinese-English biliteracy acquisition. International Journal of Bilingual Education and Bilingualism, 17, 55–73. 123