View in EDS HTML Full Text PDF Full Text

Morphological Decomposition Supports Word Recognition in Primary School Children Learning to Read: Evidence from Masked Priming of German Derived Words

Saved in:
Bibliographic Details
Title: Morphological Decomposition Supports Word Recognition in Primary School Children Learning to Read: Evidence from Masked Priming of German Derived Words
Language: English
Authors: Fleischhauer, Elisabeth (ORCID 0000-0003-1033-3207), Bruns, Gunnar (ORCID 0000-0001-6143-3596), Grosche, Michael (ORCID 0000-0001-6646-9184)
Source: Journal of Research in Reading. Feb 2021 44(1):90-109.
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 20
Publication Date: 2021
Document Type: Journal Articles
Reports - Research
Education Level: Elementary Education
Descriptors: Morphology (Languages), Language Processing, Word Recognition, Elementary School Students, German, Reading Processes, Semantics, Orthographic Symbols, Priming, Morphemes
DOI: 10.1111/1467-9817.12340
ISSN: 0141-0423
Abstract: Background: When reading a word, skilled adult readers automatically decompose the word into its separate morphemes by processing the word's morpho-orthography. In children, however, it still remains unclear when and how they start to automatically decompose words into morphemes. Methods: To better understand how primary school children learn and integrate automatic morphological processes into their reading, we conducted a masked priming experiment with n = 218 first to fourth graders and a control group of 36 adult readers. Participants saw prime words on a computer screen for 67 ms, followed by a lexical decision task. For each target word, we constructed three prime words: an unrelated control prime, a semantic control prime and a test prime. The test prime was either morpho-semantically, morpho-orthographically or purely orthographically related to the target word. We analysed error rates and reaction times with linear mixed-effects models and linear combinations. Results: The error analysis revealed one significant interaction in the morpho-semantic condition for fourth graders. The reaction time analysis revealed different priming effects depending on age group: first and second graders showed no priming effects, while third graders showed priming in the morpho-semantic condition, and fourth graders and adults showed priming both in the morpho-semantic and morpho-orthographic condition. Conclusions: We conclude that (i) fourth graders automatically decompose written words into morphemes based on the words' morpho-orthography, (ii) third graders automatically detect stems in morphologically complex forms but do not rely on morpho-orthographic representations while (iii) first and second graders show no indication of either of these processes. Our results are theoretically consistent with the 'semantic view' within a localist account and can be explained by edge-aligned embedded word activation.
Abstractor: As Provided
Entry Date: 2021
Accession Number: EJ1282991
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwGUjYBRfex5MXa5gcHWLpY5AAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDPXbIb09DEbG2sV-rgIBEICBm8TcM-lLLRozxX2z2YYLZMARGyK4CefNNMeg53UiEil8IsbfCSxaZ2-xRIB5unM6fwmIXDQ8Wcjqv1gjyutCZ7lix1izxfyTkz5LxOZJFd6mZWiDbKcLXP8o94O3BaMaopXtVJhBwBcb7HrdQh2mjF4nd2Fv-Vd4fLZnRyWllLjJxUUGM5fbuW2ODzfSbNO1gvPvn_j_YanLPHza
Text:
  Availability: 1
  Value: <anid>AN0148363453;d8c01feb.21;2021Jan30.02:50;v2.2.500</anid> <title id="AN0148363453-1">Morphological decomposition supports word recognition in primary school children learning to read: evidence from masked priming of German derived words </title> <p>Background: When reading a word, skilled adult readers automatically decompose the word into its separate morphemes by processing the word's morpho‐orthography. In children, however, it still remains unclear when and how they start to automatically decompose words into morphemes. Methods: To better understand how primary school children learn and integrate automatic morphological processes into their reading, we conducted a masked priming experiment with n = 218 first to fourth graders and a control group of 36 adult readers. Participants saw prime words on a computer screen for 67 ms, followed by a lexical decision task. For each target word, we constructed three prime words: an unrelated control prime, a semantic control prime and a test prime. The test prime was either morpho‐semantically, morpho‐orthographically or purely orthographically related to the target word. We analysed error rates and reaction times with linear mixed‐effects models and linear combinations. Results: The error analysis revealed one significant interaction in the morpho‐semantic condition for fourth graders. The reaction time analysis revealed different priming effects depending on age group: first and second graders showed no priming effects, while third graders showed priming in the morpho‐semantic condition, and fourth graders and adults showed priming both in the morpho‐semantic and morpho‐orthographic condition. Conclusions: We conclude that (i) fourth graders automatically decompose written words into morphemes based on the words' morpho‐orthography, (ii) third graders automatically detect stems in morphologically complex forms but do not rely on morpho‐orthographic representations while (iii) first and second graders show no indication of either of these processes. Our results are theoretically consistent with the 'semantic view' within a localist account and can be explained by edge‐aligned embedded word activation.</p> <p>Keywords: literacy development; morphological decomposition; priming; word reading</p> <hd id="AN0148363453-2">Highlights</hd> <p></p> <hd id="AN0148363453-3">What is already known about this topic</hd> <p></p> <ulist> <item> When reading a word, adults quickly and implicitly decompose words into anything that looks like a morpheme, even if it is not (<emph>corner</emph>  = <emph>corn</emph>  + ‐ <emph>er</emph>).</item> <p></p> <item> During reading development, children start processing the morphological structure of words quickly and implicitly.</item> </ulist> <hd id="AN0148363453-4">What this paper adds</hd> <p></p> <ulist> <item> We tested younger readers than previous studies on highly productive derivational suffixes in the morphologically rich and productive language German to detect the earliest signs of primary school children quickly and implicitly processing morphological structure during reading.</item> <p></p> <item> Third graders do exhibit early signs of the ability to quickly and implicitly detect larger units within words. Fourth graders show quick and implicit decomposition of words into anything that looks like a morpheme – the same process as in adults.</item> <p></p> <item> Children start forming mental representations of morphemes when they detect words within words (<emph>teach</emph> in <emph>teacher</emph>) and represent the orthographic rest of the word as a morphological entity (‐ <emph>er</emph>).</item> </ulist> <hd id="AN0148363453-5">Implications for theory, policy or practice</hd> <p></p> <ulist> <item> To be able to tailor reading instructions to the abilities of the children, it is important to know if and when primary school children start (implicitly) detecting morphemes. Our results suggest that children learning to read a morphologically rich and productive language would particularly benefit from morpho‐semantic reading training beginning in the second grade.</item> </ulist> <p>The majority of words in many languages consists of two or more morphemes (<emph>teach +</emph> ‐<emph>er</emph>). A considerable body of research suggests that skilled adult readers process morphemes of words during word recognition to efficiently understand the word form (<emph>teacher</emph>). In the first automatic, that is, early and implicit, step in this reading process, skilled readers detach orthographic material from a word if it has the form of a morphological ending, regardless of whether this makes sense semantically or not (e.g., <emph>‐er</emph> from <emph>teacher</emph> and <emph>corner</emph>, Rastle, Davis, Marslen‐Wilson, & Tyler, 2000). One major question is when and how children learning to read start automatically processing morphemes of complex words (<emph>teach</emph> + ‐<emph>er</emph>) and if they use morpheme‐specific ways of representing and processing these words. Answering these questions would be invaluable for designing and implementing reading instruction in primary school.</p> <p>Numerous studies have shown that children are aware of morphological structure in spoken language before they enter school and that morphological awareness is conducive to reading development (Kuo & Anderson, 2006, for a review, see Pacton & Deacon, 2008, Rabin & Deacon, 2008). Evidence from lexical decision studies underlines the importance of morphological knowledge in early reading development in different languages (for Italian: Burani, Marcolini, & Stella, 2002; French and English: Casalis, Quémart, & Duncan, 2015; German: Hasenäcker, Schröter, & Schroeder, 2017). Furthermore, morphological knowledge affects word reading in older students (Carlisle & Stone, 2005 on English second graders and older students; Roman, Kirby, Parrila, Wade‐Woolley, & Deacon, 2009; Deacon, Campbell, Tamminga, & Kirby, 2010 on English fourth, sixth and eighth graders). However, all these studies used techniques that do not allow to draw conclusions on the early automatic process during word reading.</p> <p>One of the most established research methods on adults' automatic access to morphemes during word recognition is the masked priming technique. In masked priming tasks, prime words are presented on a computer screen very briefly (about 40–70 ms; Beyersmann, Castles, & Coltheart, 2012), each followed by a target word (prime: <emph>teacher</emph> – target: <emph>teach</emph>). Participants are then asked to respond to the target word, for example, by making a lexical decision. Because the presentation duration of the prime words is too short to consciously recognise them, the effect of the prime words on the processing of target words provides insight into early, automatic processes in visual word recognition. If prime and target words share an element at some level of mental representation, the activation of that element by the prime word (<emph>teacher</emph>) will facilitate recognition of the same element in the target word (<emph>teach</emph>). In adults, morpho‐semantically related word pairs (<emph>teacher – teach</emph>) and morpho‐orthographically related word pairs (<emph>corner – corn</emph>) are commonly found to facilitate lexical decision while purely orthographically related word pairs do not (<emph>dragon</emph> – <emph>drag</emph>; Rastle, Davis, Marslen‐Wilson, & Tyler, 2000).</p> <p>Two theoretical frameworks offer different explanations of morphological facilitation during word recognition and, consequently, how this morphological facilitation develops in young readers.</p> <p>The 'distributed account' suggests that readers do not form mental representations of morphemes but form connections between different orthographical, phonological and semantic representations of a word (Seidenberg & Gonnerman, 2000); similar words activate similar connections and representations. In masked priming, facilitation for a target (<emph>teach</emph>) is derived from the preactivation of the same orthographical, phonological and semantic units by the prime (<emph>teacher</emph>). From a developmental viewpoint, children continuously strengthen orthographic, phonological and semantic representations and their connections through exposure to written words. In masked priming, children should show facilitation effects when they activate the same connections and representations for two words. The strength of facilitation should depend on the level of shared form and meaning, no matter whether prime and target share a morpheme or not.</p> <p>According to the 'localist account', readers form mental representations of morphemes and activate these during word recognition (Schreuder & Baayen, 1995; Rastle & Davis, 2008). Representations of morphemes are 'localised' (rather than 'distributed') as discrete morphological units, and words sharing morphemes activate similar morphological representations. In masked priming, facilitation for a target (<emph>teach</emph>) is derived from preactivation of the same morpheme by the prime (<emph>teach</emph> in <emph>teacher</emph>). There are two main theoretical positions within the 'localist account' on what drives children to start forming morphemic representations.</p> <p>Following the 'orthographic view', children identify morphemes based on form properties, for example, high frequency of co‐occurring letter sequences marking a morpheme. Consequently, all entities with the orthographic form of a morpheme within a word are decomposed, for example, <emph>‐er</emph> in <emph>teacher</emph> but also in <emph>corner</emph> (Carlisle & Fleming, 2003; Rastle & Davis, 2008). Following the 'semantic view', children start forming mental representations of morphemes when they detect the systematic relationship between semantic meaning and their orthographic consistency in different words, for example, in <emph>teach</emph> and <emph>teacher</emph> but not in <emph>corn</emph> and <emph>corner</emph> (Schreuder & Baayen, 1995; Rabin & Deacon, 2008).</p> <p>Studies using the masked priming technique to examine which of these views best accounts for the patterns exhibited by children in reading development in different languages have yielded mixed results. Some found priming in children solely for morpho‐semantically related prime‐target pairs, supporting the semantic view (English: Beyersmann, Castles, & Coltheart, 2012; artificial language learning: Merkx, Rastle, & Davis, 2011; German: Hasenäcker, Beyersmann, & Schröder, 2020; Hebrew: Schiff, Raveh, & Fighel, 2012), others in morpho‐semantic and morpho‐orthographic conditions, supporting the orthographic view (Quémart, Casalis, & Cole, 2011; Quémart & Casalis, 2015). Evidence in favour of the distributed account comes from a cross‐modal priming study by Quémart, Gonnerman, Downing, and Deacon (2017), reporting graded priming effects as a function of prime‐target overlap of form and meaning, irrespective of morphological structure.</p> <p>Priming studies have yielded equally mixed results regarding the time when children start showing morpho‐orthographic priming effects similar to adults, if at all. Some suggest that morpho‐orthographic structure affects automatic word recognition already in grade three (French: Quémart, Casalis, & Cole, 2011), others in grade four (German: Hasenäcker, Beyersmann, & Schröder, 2020; French: Casalis, Dusautoir, Colé, & Ducrot, 2009) or, by contrast, not even in grade five (English: Beyersmann, Castles, & Coltheart, 2012; French: Beyersmann, Grainger, Casalis, & Ziegler, 2015; German: Hasenäcker, Beyersmann, & Schroeder, 2016).</p> <p>One reason for these discrepancies could lie in the selected age groups, not allowing to discover automatic decomposition during word recognition in younger children (from grade two: Hasenäcker, Beyersmann, & Schröder, 2020; grade three: Quémart, Gonnerman, Downing, & Deacon, 2017). Another reason could lie in the different languages tested. The richness of a language's morphological system (Casalis, Quémart, & Duncan, 2015) and its morphological productivity (Beyersmann et al., 2020) affect readers' morphological decomposition efficiency. Detecting morphemes in a morphologically productive and rich language may be easier, and accessing morphemes may be particularly efficient during word recognition. Consequently, young readers of morphologically rich languages like German (80% multimorphemic words; Segbers & Schroeder, 2017) and French may use automatic morphological decomposition earlier compared with young readers of morphologically poorer languages like English (multimorphemic words in French: 75%, in English: 55%; Casalis, Quémart, & Duncan, 2015). Indeed, previous studies have reported the earliest automatic decomposition in French and German. Hasenäcker, Beyersmann and Schröder (2020) examined children's reading skills and their masked priming patterns longitudinally in grades two, three, and four. They tested priming in German word material (<emph>kleidchen – KLEID</emph> 'small dress – dress') and contrasted this to priming in nonword material, that is, suffixed nonwords (<emph>kleidtum – KLEID</emph>), nonsuffixed nonwords (<emph>kleidekt – KLEID</emph>) and unrelated words (<emph>träumerei – KLEID</emph>). Results showed priming in all conditions for third graders, indicating that third graders automatically identify the stem, but not the affix as in morpho‐orthographic processing that is the focus of the current study. Fourth graders showed priming in suffixed words and nonwords combined with no priming for nonsuffixed nonwords, indicating morpho‐orthographic decomposition. Priming from suffixed words was stronger than from nonword primes.</p> <p>If we wish to detect the earliest signs of automatic processing, it seems useful to study real words, possibly finding earlier indications of automatic processing. Furthermore, it seems not fully clear whether the priming effects that Hasenäcker, Beyersmann and Schröder (2020) observed in grade three were due to the suggested recognition of stems within the prime word, as a localist account would suggest, or due to shared whole‐word properties of <emph>kleidchen</emph> and <emph>KLEID</emph>, as a distributed account would suggest. A semantic control condition is necessary to differentiate between these options, as only the distributed account but not the localist account would predict priming effects in a semantic control condition.</p> <p>This underlines the need for further research to find the beginning of automatic morphological processing during word recognition and to study how children start representing morphemes. Such research could help decide which theoretical accounts and views provide a better explanation of reading development. To this end, we used masked priming in the morphologically rich and productive language German, included the youngest school children available, tested real words instead of nonwords and used a semantic control condition.</p> <hd id="AN0148363453-6">Research questions and hypotheses</hd> <p>Our first research question is whether, and if so in which grade, primary school children show priming effects indicating automatic morpho‐orthographic decomposition during reading. We expect fourth graders, but not younger students, to decompose words based on morpho‐orthographic information automatically, reflected by significant priming effects in the morpho‐semantic and the morpho‐orthographic condition (Hasenäcker, Beyersmann, & Schröder, 2020). We further expect the control group of adult readers to show significant priming effects in the morpho‐semantic and the morpho‐orthographic condition as in previous studies (Rastle, Davis, Marslen‐Wilson, & Tyler, 2000).</p> <p>Second, we ask whether priming effects in younger and older children speak in favour of one of the developmental views on how children start automatically processing morphemes of complex words, that is, the distributed account, the orthographic view or the semantic view. We therefore compare priming effects in younger children with those in older children – but not with those in adults.</p> <p>If the distributed account holds true, the strongest priming effects are expected for morpho‐semantically related prime‐target pairs (<emph>teacher – teach</emph>) and smaller effects for morpho‐orthographically (<emph>corner – corn</emph>) and orthographically related prime‐target pairs (<emph>dragon</emph> – <emph>drag</emph>). Crucially, the effects for morpho‐orthographically and orthographically related prime‐target pairs should be similar as they both show the same level of prime‐target overlap: no semantic and similar orthographical prime‐target overlap (cf. Quémart, Gonnerman, Downing, & Deacon, 2017).</p> <p>If the orthographic view of the localist account holds true, priming effects in younger readers should be observed in the morpho‐semantic (<emph>teacher – teach</emph>) and the morpho‐orthographic condition (<emph>corner – corn</emph>, cf. Quémart & Casalis, 2015), but not in the orthographic condition (<emph>dragon</emph> – <emph>drag</emph>).</p> <p>If the semantic view of the localist account holds true, we expect priming in younger children, namely, third graders, only in the morpho‐semantic condition (<emph>teacher – teach</emph>) before older children, namely, fourth graders, show priming effects in the morpho‐semantic and the morpho‐orthographic condition (<emph>corner – corn</emph>). First and second graders are not expected to show automatic priming effects (Beyersmann, Castles, & Coltheart, 2012; Hasenäcker, Beyersmann, & Schröder, 2020).</p> <hd id="AN0148363453-7">Methods</hd> <p></p> <hd id="AN0148363453-8">Participants</hd> <p>We tested 218 German‐speaking children between 6.8 and 12.3 years; of these 48 were first graders (mean age 7.86, <emph>SD</emph> = 0.42), 44 were second graders (mean age 8.80, <emph>SD</emph> = 0.48), 51 were third graders (mean age 9.83, <emph>SD</emph> = 0.39) and 75 were fourth graders (mean age 10.52, <emph>SD</emph> = 0.56). The children all attended a primary school in a rural area in Northern Germany, and their parents spoke German as first language. Parents or legal guardians provided written consent on behalf of their children. According to the curricula, all primary school students are taught the same morphological topics as, for example, 'morphological analysis: deducing from words' and 'recognising high‐frequency morphemes at first glance', deepening their knowledge as they proceed to higher grades (Ministry of Education in Lower Saxony, 2017). In addition, 36 adult participants between 19.40 and 39.89 years of age (mean age 22.42, <emph>SD</emph> = 3.37) participated in the experiment. Adult participants had received at least 12 years of formal education and spoke German as their first language. They received a gift certificate of EUR 5.00 for their voluntary participation. All participants had normal or corrected‐to‐normal vision and no history of reading or spelling disabilities.</p> <hd id="AN0148363453-9">Procedure</hd> <p>Participants were tested individually in a quiet room. The masked priming experiment was presented using the DMDX software (Forster & Forster, 2003) on a DELL Latitude E7440 with a refresh rate of 60 Hz, a resolution of 1,920 × 1,080 and a screen size of 14 inch. Participants sat at a distance of 60 cm from the screen. Target words were presented in white letters against a black background in 36‐point <emph>Norddruck Druckschrift</emph>, the block‐letter font typically used in German primary school textbooks. Prime words were presented in <emph>Norddruck Schreibschrift</emph>, the typical script font of German primary school textbooks, to reduce visual prime‐target overlap (Heyer & Clahsen, 2015).</p> <p>Participants saw 54 experimental prime‐target word pairs, 99 filler prime‐target word pairs and 153 word‐nonword filler pairs. For each target word, participants were asked to decide whether or not it was an existing word in German and to indicate their response by pressing a 'yes' or a 'no' button as quickly and accurately as possible (lexical decision task). Each trial followed the same sequence (Heyer & Clahsen, 2015): first, a forward mask – a white string of hash marks in the centre of a black computer screen – appeared for 500 ms. Second, the prime word appeared on the screen for 67 ms. Third, the target word appeared on the screen until participants made a lexical decision or until the maximum presentation duration of 5,000 ms expired, followed by the next trial. In order to increase the probability that the youngest children could also process the prime (Casalis, Dusautoir, Colé, & Ducrot, 2009 used a prime duration of 75 ms for this reason with fourth graders), we used a 67 ms prime duration, which is at the upper limit of the priming technique, but still produces semantically blind morpho‐orthographic priming (Rastle, Davis, Marslen‐Wilson, & Tyler, 2000; Casalis, Dusautoir, Colé, & Ducrot, 2009; Beyersmann, Castles, & Coltheart, 2012; Heyer & Clahsen, 2015). A pilot run with five adults and five first graders showed that the tested adults were not able to name any prime words and that the tested first graders were able to complete the task.</p> <hd id="AN0148363453-10">Materials</hd> <p>The test conducted here examined the morphological type of derivation. Derivation is the most prominent morphological category in German children's morphologically complex lexical entries in first and second grade (Segbers & Schroeder, 2017). We chose the highly productive suffixes <emph>‐er</emph> and <emph>‐ung</emph> with orthographically and semantically fully transparent word forms to increase the probability that the youngest children were able to access these during word recognition (Diependaele, Duñabeitia, Morris, & Keuleers, 2011). Both suffixes operate on verb stems and result in nominalised forms. The <emph>‐ung</emph> nominalisation denotes actions: <emph>Lösung</emph> (<emph>Lösung</emph> = <emph>Lös‐</emph> 'solve' + <emph>‐ung</emph>) is 'the act of solving'. The <emph>‐er</emph> suffix denotes actors: <emph>Lehrer</emph> (<emph>Lehrer</emph> = <emph>Lehr‐</emph> 'teach' + <emph>‐er</emph>) is 'somebody who teaches' (Fleischer & Barz, 2012, p. 201 for <emph>‐er</emph>; Fleischer & Barz, 2012, p. 225 for <emph>‐ung</emph>). The <emph>‐ung</emph> nominalisations are always, the <emph>‐er</emph> nominalisation mostly, phonologically transparent and easily decomposable into stem and affix. Nominalisations with phonological change were not included in the materials (<emph>backen</emph> – <emph>Bäcker</emph> 'bake – baker').</p> <p>The 54 prime‐target word pairs consisted in equal parts of <emph>‐er</emph> and <emph>‐ung</emph> suffixed nouns as prime words and their infinitive verb forms as target words. In a 3 × 3 experimental design, we manipulated the morphological relationship between primes and targets in three <emph>test conditions</emph>: (i) a morpho‐semantic condition (<emph>Leser</emph> 'reader' = <emph>lesen</emph> 'read' + <emph>‐er</emph>); (ii) a morpho‐orthographic condition (<emph>Messer</emph> 'knife' ≠ <emph>mess</emph> 'measure' + <emph>‐er</emph>) and (iii) an orthographic condition (<emph>Nagel</emph> 'nail' ≠ <emph>nag</emph> 'nibble' + <emph>‐el</emph>). Only in the morpho‐semantic condition could the meaning of the prime be deduced from stem and affix. For each test condition, we used (i) a semantic control condition to assess the impact of whole‐word semantic activation on word recognition (<emph>Zeitung</emph> – <emph>lesen</emph> 'newspaper – read'), and (ii) a morphologically and semantically unrelated control condition to assess the priming effect in the test conditions (<emph>Ärger</emph> – <emph>lesen</emph> 'trouble – read'). Table 1 provides an example stimulus set. The complete list of primes and targets can be found online (Table A1, https://osf.io/ck93z).</p> <p>1 TableExample stimulus set for experimental conditions following a 3 × 3 design.</p> <p> <ephtml> <table><tbody valign="top"><tr><td /><td /><td>Primes</td><td /><td>Target</td></tr><tr><td>Test</td><td>Semantic</td><td>Unrelated</td><td /></tr><tr><td>Morpho‐semantic</td><td>Leser 'reader'</td><td>Zeitung 'newspaper'</td><td>Ärger 'trouble'</td><td>lesen '(to) read'</td></tr><tr><td>Morpho‐orthographic</td><td>Messer 'knife'</td><td>Lineal 'ruler'</td><td>Keller 'basement'</td><td>messen '(to) measure'</td></tr><tr><td>Orthographic</td><td>Nagel 'nail'</td><td>Biber 'beaver'</td><td>Segler 'sailor'</td><td>nagen '(to) nibble'</td></tr></tbody></table> </ephtml> </p> <p>The prime‐target word pairs in the three test conditions (morpho‐semantic vs. morpho‐orthographic vs. orthographic) differed regarding their shared information in prime and target, allowing us to distinguish between orthographic, morpho‐orthographic and morpho‐semantic sources of priming. If priming effects are due to purely orthographic prime‐target overlap, we expect similarly strong priming effects in all test conditions ('orthographic priming'). If priming effects are due to morpho‐orthographic decomposition, we expect no priming effect in the orthographic condition and equally strong priming effects for the morpho‐semantic and morpho‐orthographic condition ('morpho‐orthographic priming'). If priming effects are due to morpho‐semantic decomposition, we expect priming effects in the morpho‐semantic condition, and no priming in the morpho‐orthographic, the orthographic condition and the semantic control condition ('morpho‐semantic priming').</p> <p>The target and prime words were matched for lemma frequency, word‐form frequencies, length in letters and syllables (<emph>childLex</emph> database; Schroeder, Würzner, Heister, Geyken, & Kliegl, 2014) and formal overlap of related primes and targets (SOLAR overlap measure Match Calculator; Davis, 2000; Table 2 for characteristics, all <emph>F </emph>< 1.1, all <emph>ps</emph> > .35). The three prime types (related vs. semantic control vs. unrelated control) within test conditions (morpho‐semantic, morpho‐orthographic and orthographic) were matched for lemma frequency and word‐form frequencies, length in letters and length in syllables (all <emph>t</emph> < 1.8, <emph>p</emph> ≥ .07). Finally, the semantic prime‐target overlap values (extracted from Latent Semantic Analysis Web facility [<ulink href="http://lsa.colorado.edu">http://lsa.colorado.edu</ulink>]; Landauer & Dumais, 1997), as expected, were significantly higher in the morpho‐semantic condition compared with the morpho‐orthographic and orthographic condition (all <emph>t </emph>> 5.1, all <emph>p</emph>s < .001) but did not differ significantly between the morpho‐orthographic and orthographic items (<emph>t </emph>= 0.09, <emph>p </emph>= .46).</p> <p>2 TableLinguistic properties means and SD of experimental items.</p> <p> <ephtml> <table><thead><tr><th /><th /><th align="center">Frequency</th><th align="center">Length</th><th align="center">Neighbourhood</th><th align="center">Overlap</th></tr><tr><th align="left">Lemma</th><th align="left">Form</th><th align="left">Letters</th><th align="left">Syllables</th><th align="left">Coltheart</th><th align="left">Formal</th><th align="left">Semantic</th></tr></thead><tbody valign="top"><tr><td>Morpho‐semantic</td><td>target</td><td>111.89 (145.06)</td><td>35.12 (41.08)</td><td>6.11 (.96)</td><td>2.00 (.00)</td><td>183.06 (509.85)</td><td>–</td><td>–</td></tr><tr><td>Test</td><td>17.83 (26.83)</td><td>15.71 (24.09)</td><td>6.67 (1.18)</td><td>2.06 (.23)</td><td>23.41 (57.44)</td><td>.75 (.03)</td><td>.54 (.29)</td></tr><tr><td>Semantic</td><td>21.64 (25.45)</td><td>17.69 (20.00)</td><td>7.11 (2.82)</td><td>2.11 (.83)</td><td>21.52 (60.80)</td><td>–</td><td>.26 (.14)</td></tr><tr><td>Unrelated</td><td>28.07 (55.64)</td><td>22.73 (40.96)</td><td>6.89 (1.60)</td><td>2.28 (.46)</td><td>40.08 (84.56)</td><td>–</td><td>.10 (.06)</td></tr><tr><td>Morpho‐orthographic</td><td>target</td><td>130.29 (176.73)</td><td>22.89 (36.51)</td><td>6.00 (.59)</td><td>2.00 (.00)</td><td>222.53 (292.89)</td><td>–</td><td>–</td></tr><tr><td>Test</td><td>23.88 (38.79)</td><td>21.56 (35.39)</td><td>6.56 (.92)</td><td>2.17 (.38)</td><td>46.36 (88.30)</td><td>.74 (.02)</td><td>.13 (.09)</td></tr><tr><td>Semantic</td><td>82.87 (227.90)</td><td>57.55 (151.70)</td><td>6.44 (1.58)</td><td>2.06 (.54)</td><td>189.80 (449.75)</td><td>–</td><td>.27 (.13)</td></tr><tr><td>Unrelated</td><td>21.63 (17.57)</td><td>20.11 (17.30)</td><td>6.94 (1.35)</td><td>2.17 (.38)</td><td>12.83 (23.89)</td><td>–</td><td>.08 (.05)</td></tr><tr><td>Orthographic</td><td>target</td><td>125.82 (142.99)</td><td>19.79 (22.07)</td><td>6.11 (1.07)</td><td>2.00 (.00)</td><td>287.87 (606.15)</td><td>–</td><td>–</td></tr><tr><td>Test</td><td>27.23 (91.59)</td><td>12.94 (36.26)</td><td>6.56 (1.46)</td><td>1.94 (.63)</td><td>10.69 (23.79)</td><td>.73 (.07)</td><td>.13 (.17)</td></tr><tr><td>Semantic</td><td>74.44 (230.89)</td><td>53.28 (154.10)</td><td>7.00 (2.28)</td><td>2.22 (.64)</td><td>91.60 (236.23)</td><td>–</td><td>.37 (.23)</td></tr><tr><td>Unrelated</td><td>22.46 (72.61)</td><td>12.98 (38.00)</td><td>7.22 (1.35)</td><td>2.28 (.46)</td><td>3.11 (7.19)</td><td>–</td><td>.07 (.06)</td></tr></tbody></table> </ephtml> </p> <p>Ninety‐nine filler prime‐target word pairs were added to conceal the purpose of the experiment (proportion of related prime‐target pairs < 25%). In addition, 153 word‐nonword filler pairs were added to make sure that a 'yes' response in the lexical decision ask was as likely as a 'no' response. The filler prime‐target pairs were picked from the same range of linguistic parameters as the experimental prime‐target pairs. An additional set of five prime‐target pairs served as practice items before the experiment.</p> <p>Prime‐target pairs were distributed over three counterbalanced experimental lists so that each participant saw each target word only once. The prime‐target pairs in each list were pseudo‐randomised so that no more than two items for the same condition occurred in sequence and subsequent target words had different onsets. Each of the three lists was reversed for half of the participants to counterbalance fatigue and training effects.</p> <hd id="AN0148363453-11">Data preparation and analysis</hd> <p>Error rates and reaction times were analysed separately. Error rates were calculated as the proportion of incorrect responses in relation to all responses from a particular participant. A response was incorrect when the participant categorised an existing word as a nonword or vice versa.</p> <p>The full data set contained <emph>N</emph> = 13,716 observations, that is, 54 target responses from each of the 254 participants. The responses of four participants (all from grade 1) were removed due to an individual accuracy rate below chance level (≤ 50%), leaving <emph>n</emph> = 13,500 observations. Further, 54 outlier responses below 200 ms were removed, as we assumed these responses to originate from premature button presses. Finally, one response slower than 2,000 ms in the adult group was removed (Felser & Roberts, 2007), because extremely long reaction times might involve strategic thinking or distraction during processing. We adapted the cut‐off for children's age, as reaction times decrease exponentially with age (cut‐off = 2,000 ms + 5.16e<sups>‐0.21age</sups>; Kail, 1991 cited by Weiler, Forbes, Kirkwood, & Waber, 2003). For 6‐year‐olds: 5,000 ms (no observations removed, as 5,000 ms was equivalent to time‐out), 7‐year‐olds: 4,400 ms (42 observations removed), 8‐year‐olds: 4,000 ms (98 observations removed), 9‐year‐olds: 3,600 ms (55 observations removed), 10‐year‐olds: 3,300 ms (29 observations removed), 11‐year‐olds: 3,000 ms (35 observations removed) and 12‐year‐olds: 2,800 ms (5 observations removed). Together with the extremely fast responses below 200 ms, 319 (2.3%) observations were removed, leaving <emph>n</emph> = 13,181 observations. Of these, 11,614 were correct, and 1,567 were incorrect responses.</p> <p>Reaction times were calculated from the moment that the visually presented target words appeared on screen to the participant's button press. All remaining 1,567 incorrect responses were removed for the reaction time analysis, leaving a data set of <emph>n</emph> = 11,614 observations. The total amount of removed observations was <emph>n</emph> = 2,102 (15.33%).</p> <p>To answer the research questions of whether primary school children show automatic morphological decomposition and how children establish morphological representations, we analysed the error rates in the children groups and the log‐transformed reaction times (as indicated by the Box–Cox procedure; Box & Cox, 1964) to correct for distribution skews in the adult and children groups with (i) linear mixed‐effects models (Baayen, Davidson, & Bates, 2008) and (ii) linear combinations (Hothorn, Bretz, & Westfall, 2008). For all analyses, we used the software 'R', version 3.4.4 (R Core Team, 2018) with <emph>lme4</emph> (Bates, Maechler, Bolker, & Walker, 2015) and <emph>multcomp</emph> packages (Hothorn, Bretz, & Westfall, 2008). Using a 3 × 3 × 5 design, linear mixed‐effects models included the experimental factors 'test condition' (morpho‐semantic vs. morpho‐orthographic vs. orthographic), 'prime type' (related vs. semantic control vs. unrelated control) and 'group' (first through fourth graders and adults); in all factors, the latter condition was the reference level for dummy‐coding. We also included the full interactions between test conditions and prime types. In addition, we included by‐subject and by‐item random intercepts, accounting for variation within the samples of both participants and words. The following centred continuous covariates were included as fixed factors, controlling for possible confounding variables: word form frequency, length, position in the experiment, neighbourhood size (log‐transformed), and prime‐target overlap.</p> <p>Because of the dummy‐coding, differences are only tested in relation to the reference conditions (here the orthographic condition, the unrelated control prime and the adult group) but not between the other conditions. Therefore, we conducted subsequent analyses of linear combinations to directly test whether priming main effects for each group and condition were significantly different from zero. Linear combination effects can be understood as simultaneous inferences similar to planned contrasts (Hothorn, Bretz, & Westfall, 2008). For example, to test whether third graders show a priming effect that is different from zero in the morpho‐orthographic condition, the equation is specified as follows: <emph>y</emph> = <emph>b</emph><subs>0</subs> + <emph>b</emph><subs>1</subs> × <emph>PrimeTest</emph> + <emph>b</emph><subs>8</subs> × <emph>PrimeTest</emph> × <emph>MorphoOrth</emph> + <emph>b</emph><subs>12</subs> × <emph>PrimeTest</emph> × <emph>Grade</emph>3 + <emph>b</emph><subs>16</subs> × <emph>PrimeTest</emph> × <emph>MorphoOrth</emph> × <emph>Grade</emph>3.</p> <hd id="AN0148363453-12">Results</hd> <p>Table 3 presents descriptive statistics (means and <emph>SD</emph>) for error rates and reaction times per group (between‐subject) and per relation and prime type (both within‐subject). The full linear mixed‐effects model on <emph>error rates</emph> revealed main effects of age groups, indicating that first (<emph>p </emph>< .001), second (<emph>p </emph>< .001) and fourth graders (<emph>p </emph>= .032) had significantly higher error rates than adults. No other main or interaction effect was significant except for one significant three‐way interaction indicating an attenuated priming effect in the morpho‐semantic condition for fourth graders; this effect occurred for the test prime in the morpho‐semantic condition (<emph>b</emph> = 1.6435, <emph>p</emph> = .04, while all other 40 effects' <emph>p</emph>s were >.081). In comparison with a reduced model without interaction effects of age group with prime type and test condition, the full model with the group interactions did not show a significantly better model fit, Δχ<sups>2</sups>(<reflink idref="bib32" id="ref1">32</reflink>) = 35.82, <emph>p </emph>= .294. All linear combinations also proved to be nonsignificant (all <emph>p</emph>s ≥ .067).</p> <p>3 TableReaction time means and SD and error rates percentages per group and conditions.</p> <p> <ephtml> <table><thead><tr><th /><th align="center">Test prime</th><th align="center">Semantic prime</th><th align="center">Control prime</th></tr><tr><th /><th align="center">Mean RT</th><th align="center">Error rates</th><th align="center">Mean RT</th><th align="center">Error rates</th><th align="center">Mean RT</th><th align="center">Error rates</th></tr></thead><tbody valign="top"><tr><td>Adult</td><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /><td align="left" /></tr><tr><td>Orthographic</td><td>632 (180)</td><td>10.19 (13.38)</td><td>637 (192)</td><td>7.87 (10.9)</td><td>656 (236)</td><td>6.48 (9.98)</td></tr><tr><td>Morpho‐orthographic</td><td>573 (164)</td><td>4.17 (8.33)</td><td>603 (149)</td><td>3.70 (8.08)</td><td>608 (149)</td><td>4.17 (7.32)</td></tr><tr><td>Morpho‐semantic</td><td>543 (124)</td><td>1.85 (6.64)</td><td>590 (134)</td><td>4.17 (9.24)</td><td>591 (132)</td><td>3.24 (7.79)</td></tr><tr><td>Fourth grade</td><td /><td /><td /><td /><td /><td /></tr><tr><td>Orthographic</td><td>1,179 (510)</td><td>12.27 (14.79)</td><td>1,154 (493)</td><td>12.31 (15.72)</td><td>1,192 (506)</td><td>12.76 (14.85)</td></tr><tr><td>Morpho‐orthographic</td><td>1,139 (545)</td><td>7.02 (11.13)</td><td>1,186 (512)</td><td>8.36 (11.02)</td><td>1,195 (505)</td><td>9.24 (13.99)</td></tr><tr><td>Morpho‐semantic</td><td>1,051 (457)</td><td>7.07 (11.32)</td><td>1,150 (503)</td><td>4.44 (8.81)</td><td>1,111 (463)</td><td>4.04 (8.23)</td></tr><tr><td>Third grade</td><td /><td /><td /><td /><td /><td /></tr><tr><td>Orthographic</td><td>1,331 (573)</td><td>10.98 (13.14)</td><td>1,326 (592)</td><td>12.16 (15.66)</td><td>1,302 (576)</td><td>10.95 (14.58)</td></tr><tr><td>Morpho‐orthographic</td><td>1,337 (635)</td><td>8.24 (11.78)</td><td>1,340 (605)</td><td>10.20 (13.42)</td><td>1,342 (593)</td><td>9.41 (12.75)</td></tr><tr><td>Morpho‐semantic</td><td>1,172 (545)</td><td>6.14 (10.70)</td><td>1,230 (525)</td><td>5.95 (9.41)</td><td>1,240 (531)</td><td>6.21 (11.53)</td></tr><tr><td>Second grade</td><td /><td /><td /><td /><td /><td /></tr><tr><td>Orthographic</td><td>1903 (793)</td><td>18.11 (18.03)</td><td>1,745 (744)</td><td>22.88 (20.48)</td><td>1,777 (731)</td><td>21.52 (18.71)</td></tr><tr><td>Morpho‐orthographic</td><td>1,772 (778)</td><td>22.20 (25.64)</td><td>1,804 (801)</td><td>22.80 (21.08)</td><td>1793 (814)</td><td>17.88 (18.64)</td></tr><tr><td>Morpho‐semantic</td><td>1,748 (772)</td><td>12.42 (15.13)</td><td>1,719 (819)</td><td>9.66 (13.34)</td><td>1,708 (764)</td><td>12.80 (14.75)</td></tr><tr><td>First grade</td><td /><td /><td /><td /><td /><td /></tr><tr><td>Orthographic</td><td>2,306 (867)</td><td>23.03 (18.91)</td><td>2,301 (872)</td><td>26.86 (23.35)</td><td>2,356 (886)</td><td>25.95 (24.50)</td></tr><tr><td>Morpho‐orthographic</td><td>2,317 (827)</td><td>24.58 (22.50)</td><td>2,204 (757)</td><td>28.64 (19.98)</td><td>2,283 (854)</td><td>23.48 (19.48)</td></tr><tr><td>Morpho‐semantic</td><td>2,175 (875)</td><td>19.24 (19.06)</td><td>2,142 (857)</td><td>20.42 (18.49)</td><td>2,113 (815)</td><td>16.59 (20.30)</td></tr></tbody></table> </ephtml> </p> <p>1 <emph>Note</emph>. Reaction times (RT) in ms. Error rates in italics as percentage of incorrect responses across all subjects in the respective group and condition. Standard deviations are shown in parentheses.</p> <p>Table 4 presents the significant effects of the full linear mixed‐effects model on <emph>reaction times</emph>. In comparison with a reduced model without interaction effects of age group with prime type and test condition, the full model showed only a trend towards a better model fit, Δχ<sups>2</sups>(<reflink idref="bib32" id="ref2">32</reflink>) = 44.01, <emph>p </emph>= .077. However, because it appears promising to test specific interactions and linear combinations to reveal possible differential patterns for the groups, we decided in favour of the full model.</p> <p>4 TableOutput from the linear mixed‐effects model on log‐transformed reaction time data, displaying only significant effects.</p> <p> <ephtml> <table><thead><tr><th>Full model</th><th align="center">Estimate</th><th align="center">SE</th><th align="center"><italic>df</italic></th><th align="center"><italic>t</italic> value</th><th align="center"><italic>p</italic> value</th></tr></thead><tbody valign="top"><tr><td>(Intercept)</td><td>6.4426</td><td>0.0538</td><td>348.95</td><td>119.84</td><td><.001</td></tr><tr><td>Group (first vs. adult)</td><td>1.2783</td><td>0.0632</td><td>405.15</td><td>20.23</td><td><.001</td></tr><tr><td>Group (second vs. adult)</td><td>1.0020</td><td>0.0628</td><td>395.14</td><td>15.96</td><td><.001</td></tr><tr><td>Group (third vs. adult)</td><td>0.6708</td><td>0.0603</td><td>382.71</td><td>11.12</td><td><.001</td></tr><tr><td>Group (fourth vs. adult)</td><td>0.5968</td><td>0.0562</td><td>383.03</td><td>10.62</td><td><.001</td></tr><tr><td>Target‐frequency (centred)</td><td>−0.0011</td><td>0.0003</td><td>52.39</td><td>−3.59</td><td>.001</td></tr><tr><td>Target‐length (centred)</td><td>0.0230</td><td>0.0126</td><td>53.13</td><td>1.82</td><td>.074</td></tr><tr><td>Relation (morpho‐orthographic vs. orthographic) * Group (third Grade vs. adult)</td><td>0.0713</td><td>0.0399</td><td>11309.80</td><td>1.79</td><td>.074</td></tr><tr><td>Prime (test vs. unrelated) * Relation (morpho‐semantic vs. orthographic) * Group (first vs. adult)</td><td>0.1215</td><td>0.0608</td><td>11313.20</td><td>2.00</td><td>.046</td></tr></tbody></table> </ephtml> </p> <p>2 <emph>Note</emph>. The linear mixed‐effects model includes by‐subject and by‐item random intercepts. Reaction times are log‐transformed. Reference categories for the dummy‐coded categorical variables are <emph>unrelated</emph> prime type, <emph>orthographic</emph> test relation and <emph>adult</emph> group. Only significant effects (with <emph>p</emph>  < .10) are displayed.</p> <p>Word frequency of target words (as control variable) was significantly negatively associated with reaction times, that is, lexical decisions were faster for words with higher frequencies. Word length of target words was positively associated with reaction times, that is, lexical decisions were slower for longer words. All other effects of possible confounding word variables were not significant.</p> <p>We found significantly slower reaction times for lexical decisions in all children groups than in the adult group (all <emph>p</emph>s < .001). There was a trend towards an interaction between the morpho‐orthographic condition and third graders (<emph>p</emph> = .074). This means that in relation to the adult reference group, the benefit for morpho‐orthographically related control primes (as found in adults) was slightly attenuated in third graders. A three‐way interaction was found between the test prime, the morpho‐semantic condition and first graders, which was significant (<emph>p</emph> = .046). This means that the benefit of a morpho‐semantically related test‐prime (as found in adults) was attenuated in first graders. However, interactions in the full model cannot be interpreted as straight‐forward, since the categorical variables are dummy‐coded and comparisons are only made in contrast to the reference group. Therefore, research hypotheses were tested via the analyses of linear combinations based on the full model.</p> <p>Table 5 presents results from the linear combinations on priming effects of test primes to unrelated control primes per age group in each condition. First graders did not show any priming effects (all <emph>p</emph>s ≥ .324), the same holds for second graders (all <emph>p</emph>s ≥ .248). Third graders showed significant priming only in the morpho‐semantic condition (<emph>p</emph> = .014), but not in the morpho‐orthographic and the orthographic conditions (both <emph>p</emph>s ≥ .36). Fourth graders showed significant priming in both the morpho‐semantic (<emph>p</emph> = .018) and the morpho‐orthographic condition (<emph>p</emph> = .008), but not in the orthographic condition (<emph>p</emph> = .425). This pattern is also found in adults, where morpho‐semantic (<emph>p </emph>= .005) and morpho‐orthographic (<emph>p </emph>= .025) priming can be found, but not in the orthographic condition (<emph>p </emph>= .523). Thus, primes do not significantly affect first and second graders' word reading at all, third graders' reading is facilitated by morpho‐semantic or semantic information only and fourth graders use morpho‐orthographic information to decompose written words while reading.</p> <p>5 TableOverview of linear combinations to test for main priming effects per group and condition.</p> <p> <ephtml> <table><thead><tr><th /><th align="center">Test prime vs. unrelated control prime in ...</th></tr><tr><th align="center">Morpho‐semantic condition</th><th align="center">Morpho‐orthographic condition</th><th align="center">Orthographic condition</th></tr></thead><tbody valign="top"><tr><td>Adults</td><td>p = .005</td><td>p = .025</td><td>p = .523</td></tr><tr><td>Fourth graders</td><td>p = .018</td><td>p = .008</td><td>p = .425</td></tr><tr><td>Third graders</td><td>p = .014</td><td>p = .432</td><td>p = .360</td></tr><tr><td>Second graders</td><td>p = .293</td><td>p = .635</td><td>p = .248</td></tr><tr><td>First graders</td><td>p = .363</td><td>p = .621</td><td>p = .324</td></tr></tbody></table> </ephtml> </p> <p>3 <emph>Note</emph>. Linear combinations test whether reliable priming effects that are different from zero occur in the respective group and condition. Only primes of type 'test' are contrasted to the 'control' condition. Significant effects are bold printed.</p> <p>To control for semantic relationships between prime and target words, we tested priming effects of semantic primes (<emph>Zeitung</emph> 'newspaper' – <emph>lesen</emph> 'read') on the unrelated control primes in the morpho‐semantic condition in each age group with linear combinations. None of the groups showed significant priming effects in the semantic control condition for targets of the morpho‐semantic condition (adults: <emph>p =</emph> .998, fourth graders: <emph>p =</emph> .171, third graders: <emph>p =</emph> .965, second graders: <emph>p =</emph> .759, first graders: <emph>p =</emph> .203). Thus, words sharing purely semantic information with the target word do not facilitate recognition of the target word.</p> <p>Readers interested in details on data analysis and results may access the Open Science Framework (https://osf.io/ck93z/) for the code of the lmer()‐function (Table A2), the equation of the full model (Table A3), an output of all effects of the full linear mixed model on reaction times (Table A4) and the model coefficients of the linear combinations for each comparison, per group and condition (Table A5).</p> <hd id="AN0148363453-13">Discussion</hd> <p>The present study examined when and how children start to automatically process morphemes of complex words (<emph>teach</emph> + ‐<emph>er</emph>) and if they activate mental representations of morphemes during word recognition or, rather, use orthographical, phonological and semantic representations of the full word form. To systematically study the earliest effects of morphological structure possible, we tested young children attending primary school using two of the most productive derivational suffixes in German, <emph>‐er</emph> and <emph>‐ung</emph>, in orthographically and semantically transparent word forms in the morphologically rich and productive language German on real words.</p> <p>In adults, we found priming effects in the morpho‐semantic condition (<emph>teacher – teach</emph>) and the morpho‐orthographic condition (<emph>corner</emph> – <emph>corn</emph>) in the absence of semantic priming and orthographic priming. Our results indicate that adults automatically access the same mental representation of two words only if these are morphologically related, but not if they share purely orthographic or semantic information. This replicates findings from studies on adults (e.g., Rastle & Davis, 2008; Hasenäcker, Beyersmann, & Schroeder, 2016) and has been explained by semantically blind automatic decomposition based on morpho‐orthographic information, that is, readers detach orthographic material from a word if it has the form of a morphological ending, regardless of whether this makes sense semantically or not (e.g., ‐<emph>er</emph> is detached from <emph>corner</emph>). The pattern of priming in the child groups differ in one aspect: only the oldest group, fourth graders, show priming in the morpho‐semantic and the morpho‐orthographic condition. Third graders show priming only in the morpho‐semantic condition, and first and second graders show no priming at all. We therefore discuss the fourth graders' results first, followed by the younger age groups.</p> <hd id="AN0148363453-14">When do children start to automatically process morphemes of complex words?</hd> <p>Regarding our first research question, whether, and if so in which grade, primary school children start to automatically process morphemes of complex words. We found a significant morpho‐orthographic priming effect, indicating automatic morpho‐orthographicv decomposition during reading, in fourth graders but in no younger age group. This result is in consistency with our predictions, suggesting that fourth graders use semantically blind automatic decomposition based on morpho‐orthographic information. Our results are in line with previous studies reporting morpho‐orthographic priming in developing readers in French (Casalis, Dusautoir, Colé, & Ducrot, 2009; Quémart, Casalis, & Cole, 2011) and German (Hasenäcker, Beyersmann, & Schröder, 2020). Hasenäcker et al. (2020) observed morpho‐orthographic priming effects in fourth grade in stronger <emph>and</emph> weaker readers. These observations strengthen the assumption that automatic morpho‐orthographic decomposition in German children can be robustly located in fourth grade, that is, that fourth graders decompose written words into morphemes while reading.</p> <p>Meanwhile, the observed effect contradicts the findings of studies that did not detect morpho‐orthographic priming in primary school children but morpho‐semantic priming (English: Beyersmann, Castles, & Coltheart, 2012) or stem priming (French: Beyersmann, Grainger, Casalis, & Ziegler, 2015; German: Hasenäcker, Beyersmann, & Schroeder, 2016). Some of the reasons could lie in different age ranges and lower grades of morphological richness in the tested languages. However, it seems more difficult to explain the diverging results on French and German from the current results. One possible explanation could be that both studies focus on nonwords, contrasted to morpho‐semantic prime‐target word pairs. The studies report robust facilitation for words but varying results for nonwords. Hasenäcker, Beyersmann and Schröder (2020) also observe this tendency and suggest that morphological segmentation is affected not only by morphological structure but additionally by the words' full‐form semantic properties (Quémart, Gonnerman, Downing, & Deacon, 2017).</p> <hd id="AN0148363453-15">How do children start to automatically process morphemes of complex words?</hd> <p>The second research question was how primary school children start to automatically process morphemes of complex words. First and second graders did not show any priming effects at all. Third graders did not show morpho‐orthographic priming (<emph>corner – corn</emph>) as fourth graders did. Instead, the pattern of our results, that is, priming effects in the morpho‐semantic condition (<emph>teacher – teach</emph>) in the absence of priming in the morpho‐orthographic condition (<emph>corner – corn</emph>), orthographic and semantic control condition, suggests morpho‐semantic priming in third graders but gives no indication of automatic morpho‐orthographic decomposition. This morpho‐semantic priming in young readers has also been found in studies on English (Beyersmann, Castles, & Coltheart, 2012), Hebrew (Schiff, Raveh, & Fighel, 2012) and German (Hasenäcker, Beyersmann, & Schröder, 2020). In consistency with Hasenäcker, Beyersmann and Schröder (2020), our results support the assumption that morphological decomposition can be observed in third grade but not earlier. One reason might be that first and second graders mostly convert letters and syllables to sounds based on grapheme–phoneme correspondences during reading and do not yet decode lexical‐semantic units as morphemes (Hasenäcker & Schroeder, 2017).</p> <p>Our findings are consistent with predictions from the semantic view (Rabin & Deacon, 2008), supporting the hypothesis that children form automatically accessible representations when they recognise that semantic entities take the same orthographic form in different words (<emph>teacher</emph>, <emph>teaching</emph>; Schreuder & Baayen, 1995). Thus, the current results are not consistent with the orthographic view, which predicts that morpho‐orthographic priming effects will emerge earliest in development.</p> <p>Our results further do not support the 'distributed account', which predicts priming in all related conditions, including the semantic control condition, depending on the amount of orthographic, phonological and/or semantic prime‐target overlap, irrespective of morphological relations. One might argue that results in third graders are consistent with the distributed account if semantic or orthographic priming has occurred but was too small to be detected in our sample. However, following the distributed account, priming should be strongest in the morpho‐semantic condition, and, crucially, similarly strong in the morpho‐orthographic and orthographic condition. That is not the case in the current study, as we observe significant priming in the morpho‐orthographic condition in the absence of priming in the orthographic condition in fourth graders and adults. The distributed view can thus not fully account for the present results. Instead, our results support the semantic view within the localist assumption, suggesting that children represent morphemes on a specific level of morphological and not purely semantic or orthographic representation.</p> <hd id="AN0148363453-16">Theoretical implications</hd> <p>These observations and explanations raise other questions: Why do children detect morphemes in morpho‐semantically transparent word forms (<emph>teacher</emph>) but not in semantically opaque forms (<emph>corner</emph> and <emph>dragon</emph>) and what drives them to establish morphological representations of semantically less informative entities as affixes, to eventually rely on morpho‐orthographic decomposition later in reading development? The semantic view does not fully answer these, but the theoretical framework of 'edge‐aligned embedded word activation' by Grainger and Beyersmann (2017) does. Both accounts are compatible in their assumption that children's first step into morphological processing is that they recognise semantically and orthographically consistent parts within words. According to the semantic view, these parts are morphemes, according to the edge‐aligned embedded word activation account, these parts are free stems, that is, words. Grainger and Beyersmann (2017) assume that the reading system develops by activating links between morpho‐semantically related lexical entries and inhibitory links between orthographically related, but semantically distinct lexical entries, such as morpho‐orthographically related words (<emph>corn</emph> and <emph>corner</emph>) and orthographically related words (<emph>drag</emph> and <emph>dragon</emph>). Consequently, and in consistency with the present results for third graders, priming arises for morpho‐semantically but not for morpho‐orthographically and orthographically related prime‐target pairs. With growing reading experience, young readers start processing the rest of the orthographical material in addition to the stem (affixes such as <emph>‐er</emph> in <emph>teacher</emph>), leading them to form separate morphological representations of these (<emph>‐er</emph>). Only then do young readers access affix representations during word recognition and, as the fourth graders in our study do, produce morpho‐orthographic priming effects. Our results can thus be explained within the framework of edge‐aligned embedded word activation and support the assumption that children start forming mental representations of morphemes when they detect meaningful parts within words (<emph>teach</emph> in <emph>teacher</emph>) and represent the orthographic rest of the word as a morpho‐orthographic entity (<emph>‐er</emph>).</p> <p>If reading in an experimental setting, like the present one, involves the same processes as reading under usual circumstances in school, our study may add to the discussion on how and when morphological instruction may be most efficient for young children's reading development in a morphologically rich language like German. Bowers, Kirby and Deacon (2010) reported that morphological instruction was already effective for young English readers in grade one and two (Weiss, Grabner, Kargl, Purgstaller, & Fink, 2010 for German). Our results support the assumption that morphological training on derived words may already be effective from grade two onwards, focusing on morpho‐semantic aspects of the language (e.g., finding 'words within words') and shifting to more formal aspects of word structure in third and fourth grade, for example, affixes (Deacon, Campbell, Tamminga, & Kirby, 2010).</p> <hd id="AN0148363453-17">Limitations</hd> <p>The material selection was restricted in two respects. First, our focus on the highly productive suffixes <emph>‐er</emph> and <emph>‐ung</emph> restricts the interpretation of our results on these suffixes. Future research will need to focus on other suffixes to find out whether processing of these can equally be located in grades three and four. Second, it was not possible to restrict stimuli to a certain range regarding their linguistic properties. Even though differences between conditions were not significant, we would like to explain some trends. The test items in the orthographic condition had on average fewer syllables than unrelated items (Table 2). Test items in the morpho‐orthographic condition had a lower lemma frequency than semantic control items, and test items in the orthographic condition had a lower word form frequency than in the semantic control condition. We are still confident that these did not affect our results, as reaction times in the test conditions were generally faster than in the semantic control condition.</p> <p>Also, our participants' young age could have been better considered in the experimental design and analysis. Regarding the experimental design, an identity condition would be recommended in future studies to assess whether null results could be explained by first and second graders' limited processing capacities. Regarding the analysis, another approach to reaction time cut‐offs in studies on samples with high variability would be based on individual reaction times, rather than age, reflecting individual processing speed differences without a developmental focus.</p> <hd id="AN0148363453-18">Conclusion</hd> <p>Our study set out to detect the earliest signs of morphological structure affecting reading of derived forms in German children. We conclude that in learning to read, (i) fourth graders automatically decompose written words into morphemes based on the words' morpho‐orthography, as the adult control group did, and (ii) third graders automatically detect stems in morphologically complex forms but do not rely on morpho‐orthographic representations. First and second graders show no indication of either of these processes. Our results are consistent with the semantic view within a localist account of morphological processing and can be explained by edge‐aligned embedded word activation (Grainger & Beyersmann, 2017).</p> <hd id="AN0148363453-19">Acknowledgement</hd> <p>The authors thank the wonderful students and teachers of the Clemens‐Schule Hornburg in Lower Saxony, Germany, for their eager support of this study. Open access funding enabled and organized by Projekt DEAL.</p> <hd id="AN0148363453-20">Data availability statement</hd> <p>The data of this study are available on request from the corresponding author.</p> <ref id="AN0148363453-21"> <title> Footnotes </title> <blist> <bibl id="bib1" type="bt">1</bibl> <bibtext> Present addressElisabeth Fleischhauer, Franz Carl Achard Primary School, Adolfstraße 25, 12621 Berlin, Germany.</bibtext> </blist> </ref> <ref id="AN0148363453-22"> <title> References </title> <blist> <bibtext> Baayen, R., Davidson, D. & Bates, D. (2008). Mixed‐effects modeling with crossed random effects for subjects and items. Journal of Memory and Language, 59 (4), 390 – 412. https://doi.org/10.1016/j.jml.2007.12.005</bibtext> </blist> <blist> <bibl id="bib2" type="bt">2</bibl> <bibtext> Bates, D., Maechler, M., Bolker, B. & Walker, S. (2015). Fitting linear mixed‐effects models using lme4. Journal of Statistical Software, 67 (1), 1 – 48. <ulink href="http://doi.org/10.18637/jss.v067.i01">http://doi.org/10.18637/jss.v067.i01</ulink></bibtext> </blist> <blist> <bibl id="bib3" type="bt">3</bibl> <bibtext> Beyersmann, E., Castles, A. & Coltheart, M. (2012). Morphological processing during visual word recognition in developing readers: Evidence from masked priming. The Quarterly Journal of Experimental Psychology, 65 (7), 1 – 21. https://doi.org/10.1080/17470218.2012.656661</bibtext> </blist> <blist> <bibl id="bib4" type="bt">4</bibl> <bibtext> Beyersmann, E., Grainger, J., Casalis, S. & Ziegler, J. (2015). Effects of reading proficiency on embedded stem priming in primary school children. Journal of Experimental Child Psychology, 139, 115 – 126. https://doi.org/10.1016/j.jecp.2015.06.001</bibtext> </blist> <blist> <bibl id="bib5" type="bt">5</bibl> <bibtext> Beyersmann, E., Mousikou, P., Javourey‐Drevet, L., Schroeder, S., Ziegler, J. & Grainger, J. (2020). Morphological processing across modalities and languages. Scientific Studies of Reading. Advance online publication, 24 (6), 500 – 519. https://doi.org/10.1080/10888438.2020.1730847</bibtext> </blist> <blist> <bibl id="bib6" type="bt">6</bibl> <bibtext> Bowers, P., Kirby, J. & Deacon, S. (2010). The effects of morphological instruction on literacy skills: A systematic review of the literature. Review of Educational Research, 80 (2), 144 – 179. https://doi.org/10.3102/0034654309359353</bibtext> </blist> <blist> <bibl id="bib7" type="bt">7</bibl> <bibtext> Box, G. & Cox, D. (1964). An analysis of transformations. Journal of the Royal Statistical Society: Series B: Methodological, 26 (2), 211 – 243. https://doi.org/10.1111/j.2517‐6161.1964.tb00553.x</bibtext> </blist> <blist> <bibl id="bib8" type="bt">8</bibl> <bibtext> Burani, C., Marcolini, S. & Stella, G. (2002). How early does morpholexical reading develop in readers of a shallow orthography? Brain and Language, 81 (1–3), 568 – 586. https://doi.org/10.1006/brln.2001.2548</bibtext> </blist> <blist> <bibl id="bib9" type="bt">9</bibl> <bibtext> Carlisle, J. & Fleming, J. (2003). Lexical processing of morphologically complex words in the elementary years. Scientific Studies of Reading, 7 (3), 239 – 253. https://doi.org/10.1207/S1532799XSSR0703_3</bibtext> </blist> <blist> <bibtext> Carlisle, J. & Stone, C. (2005). Exploring the role of morphemes in word reading. Reading Research Quarterly, 40 (4), 428 – 449. https://doi.org/10.1598/RRQ.40.4.3</bibtext> </blist> <blist> <bibtext> Casalis, S., Dusautoir, M., Colé, P. & Ducrot, S. (2009). Morphological effects in children word reading: A priming study in fourth‐graders. British Journal of Developmental Psychology, 27 (3), 761 – 766. https://doi.org/10.1348/026151008X389575</bibtext> </blist> <blist> <bibtext> Casalis, S., Quémart, P. & Duncan, L. (2015). How language affects children's use of derivational morphology in visual word and pseudoword processing: Evidence from a cross‐language study. Frontiers in Psychology, 6, 452. https://doi.org/10.3389/fpsyg.2015.00452</bibtext> </blist> <blist> <bibtext> Davis, C. (2000). Match calculator. Retrieved from: <ulink href="http://www.pc.rhul.ac.uk/staff/c.davis/">http://www.pc.rhul.ac.uk/staff/c.davis/</ulink></bibtext> </blist> <blist> <bibtext> Deacon, S.H., Campbell, E., Tamminga, M. & Kirby, J. (2010). Seeing the harm in harmed and harmful: Morphological processing by children in grades 4, 6 and 8. Applied PsychoLinguistics, 31 (4), 759 – 775. https://doi.org/10.1017/S0142716410000238</bibtext> </blist> <blist> <bibtext> Diependaele, K., Duñabeitia, J., Morris, J. & Keuleers, E. (2011). Fast morphological effects in first and second language word recognition. Journal of Memory and Language, 64 (4), 344 – 358. https://doi.org/10.1016/j.jml.2011.01.003</bibtext> </blist> <blist> <bibtext> Felser, C. & Roberts, L. (2007). Processing wh‐dependencies in a second language: A cross‐modal priming study. Second Language Research, 23 (1), 9 – 36. https://doi.org/10.1177/0267658307071600</bibtext> </blist> <blist> <bibtext> Fleischer, W. & Barz, I. (2012). Wortbildung der deutschen Gegenwartssprache [word formation in the German contemporary language]. Berlin : De Gruyter.</bibtext> </blist> <blist> <bibtext> Forster, K. & Forster, J. (2003). DMDX: A windows display program with millisecond accuracy. Behavior Research Methods, Instruments, & Computers, 35 (1), 116 – 124. https://doi.org/10.3758/BF03195503</bibtext> </blist> <blist> <bibtext> Grainger, J. & Beyersmann, E. (2017). Edge‐aligned embedded word activation initiates morpho‐orthographic segmentation. Psychology of Learning and Motivation, 67, 285 – 317. <ulink href="http://doi.org/10.1016/bs.plm.2017.03.009">http://doi.org/10.1016/bs.plm.2017.03.009</ulink></bibtext> </blist> <blist> <bibtext> Hasenäcker, J., Beyersmann, E. & Schröder, S. (2020). Morphological priming in children: Disentangling the effects of school‐grade and reading skill. Scientific Studies of Reading, 1 – 16. Advance online publication. https://doi.org/10.1080/10888438.2020.1729768</bibtext> </blist> <blist> <bibtext> Hasenäcker, J., Beyersmann, E. & Schroeder, S. (2016). Masked morphological priming in German‐speaking adults and children: Evidence from response time distributions. Frontiers in Psychology, 7. https://doi.org/10.3389/fpsyg.2016.00929</bibtext> </blist> <blist> <bibtext> Hasenäcker, J. & Schroeder, S. (2017). Syllables and morphemes in German reading development: Evidence from second graders, fourth graders, and adults. Applied PsychoLinguistics, 38 (3), 733 – 753. https://doi.org/10.1017/S0142716416000412</bibtext> </blist> <blist> <bibtext> Hasenäcker, J., Schröter, P. & Schroeder, S. (2017). Investigating developmental trajectories of morphemes as reading units in German. Journal of Experimental Psychology: Learning, Memory, and Cognition, 43 (7), 1093 – 1108. https://doi.org/10.1037/xlm0000353</bibtext> </blist> <blist> <bibtext> Heyer, V. & Clahsen, H. (2015). Late bilinguals see a scan in scanner AND in scandal: Dissecting formal overlap from morphological priming in the processing of derived words. Bilingualism: Language and Cognition, 18 (3), 543 – 550. https://doi.org/10.1017/S1366728914000662</bibtext> </blist> <blist> <bibtext> Hothorn, T., Bretz, F. & Westfall, P. (2008). Simultaneous inference in general parametric models. Biometrical Journal, 50 (3), 346 – 363. https://doi.org/10.1002/bimj.200810425</bibtext> </blist> <blist> <bibtext> Kail, R. (1991). Developmental change in speed of processing during childhood and adolescence. Psychological Bulletin, 109 (3), 490 – 501. https://doi.org/10.1037/0033‐2909.109.3.490</bibtext> </blist> <blist> <bibtext> Kuo, L. & Anderson, R. (2006). Morphological awareness and learning to read: A cross‐language perspective. Educational Psychologist, 41 (3), 161 – 180. https://doi.org/10.1207/s15326985ep4103_3</bibtext> </blist> <blist> <bibtext> Landauer, T. & Dumais, S. (1997). A solution to Plato's problem: The latent semantic analysis theory of acquisition, induction, and representation of knowledge. Psychological Review, 104 (2), 211 – 240. https://doi.org/10.1037/0033‐295X.104.2.211</bibtext> </blist> <blist> <bibtext> Merkx, M., Rastle, K. & Davis, M. (2011). The acquisition of morphological knowledge investigated through artificial language learning. Quarterly Journal of Experimental Psychology, 64 (6), 1200 – 1220. https://doi.org/10.1080/17470218.2010.538211</bibtext> </blist> <blist> <bibtext> Ministry of Education in Lower Saxony. (2017). Kerncurriculum für die Grundschule Schuljahrgänge 1–4: Deutsch [Core Curriculum for Primary School Grades 1–4]. Retrieved from: <ulink href="http://www.cuvo.nibis.de">http://www.cuvo.nibis.de</ulink></bibtext> </blist> <blist> <bibtext> Pacton, S. & Deacon, S. (2008). The timing and mechanisms of children's use of morphological information in spelling: A review of evidence from English and French. Cognitive Development, 23 (3), 339 – 359. https://doi.org/10.1016/j.cogdev.2007.09.004</bibtext> </blist> <blist> <bibtext> Quémart, P. & Casalis, S. (2015). Visual processing of derivational morphology in children with developmental dyslexia: Insights from masked priming. Applied PsychoLinguistics, 36 (2), 345 – 376. https://doi.org/10.1017/S014271641300026X</bibtext> </blist> <blist> <bibtext> Quémart, P., Casalis, S. & Cole, P. (2011). The role of form and meaning in the processing of written morphology: A priming study in French developing readers. Journal of Experimental Child Psychology, 109 (4), 478 – 496. https://doi.org/10.1016/j.jecp.2011.02.008</bibtext> </blist> <blist> <bibtext> Quémart, P., Gonnerman, L., Downing, J. & Deacon, S. (2017). The development of morphological representations in young readers: A cross‐modal priming study. Developmental Science, 21 (4), e12607. https://doi.org/10.1111/desc.12607</bibtext> </blist> <blist> <bibtext> R Core Team (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Retrieved from: https://www.R‐project.org/</bibtext> </blist> <blist> <bibtext> Rabin, J. & Deacon, H. (2008). The representation of morphologically complex words in the developing lexicon. Journal of Child Language, 35 (2), 453 – 465. https://doi.org/10.1017/S0305000907008525</bibtext> </blist> <blist> <bibtext> Rastle, K. & Davis, M. (2008). Morphological decomposition based on the analysis of orthography. Language & Cognitive Processes, 23 (7–8), 942 – 971. https://doi.org/10.1080/01690960802069730</bibtext> </blist> <blist> <bibtext> Rastle, K., Davis, M., Marslen‐Wilson, W. & Tyler, L. (2000). Morphological and semantic effects in visual word recognition: A time‐course study. Language & Cognitive Processes, 15 (4–5), 507 – 537. https://doi.org/10.1080/01690960050119689</bibtext> </blist> <blist> <bibtext> Roman, A., Kirby, J., Parrila, R., Wade‐Woolley, L. & Deacon, S. (2009). Towards a comprehensive view of the skills involved in word reading in grades 4, 6, and 8. Journal of Experimental Child Psychology, 102 (1), 96 – 113. https://doi.org/10.1016/j.jecp.2008.01.004</bibtext> </blist> <blist> <bibtext> Schiff, R., Raveh, M. & Fighel, A. (2012). The development of the Hebrew mental lexicon: When morphological representations become devoid of their meaning. Scientific Studies of Reading, 16 (5), 383 – 403. https://doi.org/10.1080/10888438.2011.571327</bibtext> </blist> <blist> <bibtext> Schreuder, R. & Baayen, R. (1995). Modeling morphological processing. In L.B. Feldman (Ed.), Morphological aspects of language processing, (pp. 131 – 154). Hillsdale, NJ : Erlbaum.</bibtext> </blist> <blist> <bibtext> Schroeder, S., Würzner, K.‐M., Heister, J., Geyken, A. & Kliegl, R. (2014). childLex: A lexical database of German read by children. Behavior Research Methods, 47 (4), 1085 – 1094. https://doi.org/10.3758/s13428‐014‐0528‐1</bibtext> </blist> <blist> <bibtext> Segbers, J. & Schroeder, S. (2017). How many words do children know? A corpus‐based estimation of children's total vocabulary size. Language Testing, 34 (3), 297 – 320. https://doi.org/10.1177/0265532216641152</bibtext> </blist> <blist> <bibtext> Seidenberg, M. & Gonnerman, L. (2000). Explaining derivational morphology as the convergence of codes. Trends in Cognitive Science, 4 (9), 353 – 361. https://doi.org/10.1016/S1364‐6613(00)01515‐1</bibtext> </blist> <blist> <bibtext> Weiler, M., Forbes, P., Kirkwood, M. & Waber, D. (2003). The developmental course of processing speed in children with and without learning disabilities. Journal of Experimental Child Psychology, 85 (2), 178 – 194. https://doi.org/10.1016/S0022‐0965(03)00034‐1</bibtext> </blist> <blist> <bibtext> Weiss, S., Grabner, R.H., Kargl, R., Purgstaller, C. & Fink, A. (2010). Behavioral and neurophysiological effects of morphological awareness training on spelling and reading. Reading and Writing, 23 (6), 645 – 671. https://doi.org/10.1007/s11145‐009‐9177‐7</bibtext> </blist> </ref> <aug> <p>By Elisabeth Fleischhauer; Gunnar Bruns and Michael Grosche</p> <p>Reported by Author; Author; Author</p> <p></p> <p>Dr. Elisabeth Fleischhauer is a post doc in psycholinguistics and special education. Her work addresses written language processing in educational contexts. She works as a special education teacher.</p> <p>Gunnar Bruns is a PhD student in special education at University of Wuppertal. He is interested in verbal and memory processes in learning disabilities.</p> <p>Dr. Michael Grosche is a professor in special learning needs at University of Wuppertal. His research focuses on cognitive processes as well as assessment and intervention in learning disabilities.</p> </aug> <nolink nlid="nl1" bibid="bib32" firstref="ref1"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1282991
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Morphological Decomposition Supports Word Recognition in Primary School Children Learning to Read: Evidence from Masked Priming of German Derived Words
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Fleischhauer%2C+Elisabeth%22">Fleischhauer, Elisabeth</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-1033-3207">0000-0003-1033-3207</externalLink>)<br /><searchLink fieldCode="AR" term="%22Bruns%2C+Gunnar%22">Bruns, Gunnar</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-6143-3596">0000-0001-6143-3596</externalLink>)<br /><searchLink fieldCode="AR" term="%22Grosche%2C+Michael%22">Grosche, Michael</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-6646-9184">0000-0001-6646-9184</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Journal+of+Research+in+Reading%22"><i>Journal of Research in Reading</i></searchLink>. Feb 2021 44(1):90-109.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 20
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2021
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Elementary+Education%22">Elementary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Morphology+%28Languages%29%22">Morphology (Languages)</searchLink><br /><searchLink fieldCode="DE" term="%22Language+Processing%22">Language Processing</searchLink><br /><searchLink fieldCode="DE" term="%22Word+Recognition%22">Word Recognition</searchLink><br /><searchLink fieldCode="DE" term="%22Elementary+School+Students%22">Elementary School Students</searchLink><br /><searchLink fieldCode="DE" term="%22German%22">German</searchLink><br /><searchLink fieldCode="DE" term="%22Reading+Processes%22">Reading Processes</searchLink><br /><searchLink fieldCode="DE" term="%22Semantics%22">Semantics</searchLink><br /><searchLink fieldCode="DE" term="%22Orthographic+Symbols%22">Orthographic Symbols</searchLink><br /><searchLink fieldCode="DE" term="%22Priming%22">Priming</searchLink><br /><searchLink fieldCode="DE" term="%22Morphemes%22">Morphemes</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1111/1467-9817.12340
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0141-0423
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Background: When reading a word, skilled adult readers automatically decompose the word into its separate morphemes by processing the word's morpho-orthography. In children, however, it still remains unclear when and how they start to automatically decompose words into morphemes. Methods: To better understand how primary school children learn and integrate automatic morphological processes into their reading, we conducted a masked priming experiment with n = 218 first to fourth graders and a control group of 36 adult readers. Participants saw prime words on a computer screen for 67 ms, followed by a lexical decision task. For each target word, we constructed three prime words: an unrelated control prime, a semantic control prime and a test prime. The test prime was either morpho-semantically, morpho-orthographically or purely orthographically related to the target word. We analysed error rates and reaction times with linear mixed-effects models and linear combinations. Results: The error analysis revealed one significant interaction in the morpho-semantic condition for fourth graders. The reaction time analysis revealed different priming effects depending on age group: first and second graders showed no priming effects, while third graders showed priming in the morpho-semantic condition, and fourth graders and adults showed priming both in the morpho-semantic and morpho-orthographic condition. Conclusions: We conclude that (i) fourth graders automatically decompose written words into morphemes based on the words' morpho-orthography, (ii) third graders automatically detect stems in morphologically complex forms but do not rely on morpho-orthographic representations while (iii) first and second graders show no indication of either of these processes. Our results are theoretically consistent with the 'semantic view' within a localist account and can be explained by edge-aligned embedded word activation.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2021
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1282991
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1282991
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1111/1467-9817.12340
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 20
        StartPage: 90
    Subjects:
      – SubjectFull: Morphology (Languages)
        Type: general
      – SubjectFull: Language Processing
        Type: general
      – SubjectFull: Word Recognition
        Type: general
      – SubjectFull: Elementary School Students
        Type: general
      – SubjectFull: German
        Type: general
      – SubjectFull: Reading Processes
        Type: general
      – SubjectFull: Semantics
        Type: general
      – SubjectFull: Orthographic Symbols
        Type: general
      – SubjectFull: Priming
        Type: general
      – SubjectFull: Morphemes
        Type: general
    Titles:
      – TitleFull: Morphological Decomposition Supports Word Recognition in Primary School Children Learning to Read: Evidence from Masked Priming of German Derived Words
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Fleischhauer, Elisabeth
      – PersonEntity:
          Name:
            NameFull: Bruns, Gunnar
      – PersonEntity:
          Name:
            NameFull: Grosche, Michael
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 02
              Type: published
              Y: 2021
          Identifiers:
            – Type: issn-print
              Value: 0141-0423
          Numbering:
            – Type: volume
              Value: 44
            – Type: issue
              Value: 1
          Titles:
            – TitleFull: Journal of Research in Reading
              Type: main
ResultId 1