Lexical stress awareness and orthographic stress in Spanish

Learning and Individual Differences 45 (2016) 144–150

Contents lists available at ScienceDirect

Learning and Individual Differences journal homepage: www.elsevier.com/locate/lindif

Lexical stress awareness and orthographic stress in Spanish Nicolás Gutiérrez-Palma a,⁎, Sylvia Defior b, Gracia Jiménez-Fernández b, Francisca Serrano b, Mª. Carmen González-Trujillo b a b

Department of Psychology, University of Jaen, Spain Department of Developmental and Educational Psychology, University of Granada, Spain

a r t i c l e

i n f o

Article history: Received 24 February 2015 Received in revised form 14 July 2015 Accepted 27 November 2015 Keywords: Reading acquisition Prosody Lexical stress Lexical stress awareness

a b s t r a c t This work aims to study whether prosody is related to reading acquisition. In particular, the study focuses on lexical stress, and on the awareness of lexical stress measured by the ability to detect the loudest syllable in a pseudoword. The hypothesis is that stress awareness may play a role in the acquisition of word reading in Spanish. A cross-sectional study was carried out with 233 children from 3rd to 6th grades. Cognitive and phonemic awareness skills were tested. A lexical stress awareness task and a reading aloud task were also used. Results highlight the relevance of prosodic knowledge in learning how to read the Spanish written words. In all grades, lexical stress awareness appears to be related to stress reading errors. On the contrary, phonemic awareness is not related to stress errors, except in 4th grade. These data are discussed considering that, apart from phonological awareness, prosodic (lexical stress) awareness would also be relevant in reading acquisition. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Most models of reading acquisition consider that word reading entails using different kind of information, mainly of grammatical, semantic, orthographic, and phonological nature. However, in alphabetic systems, phonological processing plays a particularly important role especially in early stages of reading and writing acquisition (Caravolas et al., 2012; Defior, Martos, & Cary, 2002; Share, 2004). Most studies on the influence of phonological awareness on reading acquisition have focused on the importance of phonemes or other sublexical units such as syllables or intrasyllabic units. In contrast, other aspects of phonology necessary to correctly talk, write, and read, as stress or intonation have received much less attention. The present study focuses on these other aspects of phonology, namely prosody, and on its relationship with reading acquisition in Spanish. In the light of previous research on the role of phonological awareness (henceforth PA), this study argues that awareness of prosodic aspects of language may also be a relevant factor for reading acquisition. 1.1. Defining prosody There is not a single definition of prosody. Some proposals refer to prosody as an abstract organization of speech, while others emphasize the phonetic realization of suprasegmental features (Cutler, Dahan, & Donselaar, 1997). According to the first approach, prosody can be ⁎ Corresponding author at: Universidad de Jaén, Departamento de Psicología, Campus las Lagunillas s/n, edificio C5, 23071 Jaén, Spain. E-mail address: [email protected] (N. Gutiérrez-Palma).

http://dx.doi.org/10.1016/j.lindif.2015.11.026 1041-6080/© 2015 Elsevier Inc. All rights reserved.

defined as a phonological subsystem that represents speech into hierarchically arranged units (Nespor & Vogel, 2007). This prosodic hierarchy includes units such as the syllable, the phonological word, or the utterance (Hayes, 1995). The second approach refers to prosody as the realization of prosodic features, such as stress, intonation, and timing (Dowhower, 1991). Prosody is also referred as suprasegmental phonology – opposite of segmental phonology – as prosodic units and features are realized across more than one single phoneme. Stress is one of the main prosodic features. Two varying perspectives are lexical and metrical stress. Lexical stress refers to the relative prominence of syllables within a word, while metrical stress refers to sequences of strong and weak syllables, i.e., the rhythmic pattern that occurs across multiple syllables. In stress-timed languages, as English, stress is also the rhythmic unit. In syllable-timed languages, as Spanish, the rhythmic unit is the syllable. However, it has been proposed that stress is the basis of rhythm in all languages, and that there is a continuum from more to less stress based languages (Arvaniti, 2009; Dauer, 1983). If stress were necessary for rhythm perception in all languages it would have the potential to affect literacy acquisition (see the next section). This research focuses on lexical stress in Spanish. In some orthographies, such as Greek or Spanish, lexical stress is represented by a stress mark (Defior, Jiménez-Fernández, & Serrano, 2009). This mark consists of an acute accent placed over the stressed vowel (e. g., cajón [drawer]). There are clear rules1 for the use of this stress mark in Spanish: The mark should be on all words stressed on the antepenultimate syllable or proparoxytones (e.g., lá-ti-go [whip]),

1

There are exceptions and particularities (Real Academia Española, 2009).

N. Gutiérrez-Palma et al. / Learning and Individual Differences 45 (2016) 144–150

on those stressed on the penultimate syllable or paroxytones when the word ends in a consonant other than n or s (e.g., más-til [flagpole]), and on words stressed on the last syllable or oxytones when the word ends in a vowel or in consonants n or s (e.g., ru-bí [ruby]; ca-jón [drawer]) (Real Academia Española, 1999). In reading, words without the stress mark can be read following general implicit rules: Words should be read as oxytones when ending in a consonant other than n or s, and as paroxytones in all other cases. This study analyses the role of lexical stress awareness in reading acquisition and in learning this stress mark. 1.2. Prosody and reading acquisition Wood, Wade-Woolley, and Holliman (2009) proposed that sensitivity to prosody is connected to literacy acquisition through several possible pathways. One of them considers that children are born with a periodicity bias (Cutler & Mehler, 1993), which helps them to learn the speech rhythm. Once learnt, children can use rhythm for segmenting the speech stream and learning vocabulary (Nazzi, Bertoncini, & Mehler, 1998), which in turn promotes the development of PA. Therefore, the children more sensitive to the speech rhythm would have more PA skills and then better literacy acquisition. In a second and third pathways, speech rhythm sensitivity (henceforth, SRS) is directly linked to phoneme and rhyme awareness. One of the rhythm's prosodic features, stress (see the previous section), may facilitate the identification of phonemes (Kitzen, as cited by Holliman et al., 2014) and then phoneme awareness. Moreover, rhythm perception requires awareness of the peak of loudness associated with the vowel, which is the natural boundary between the onset and the rime. As a consequence, rhythm would facilitate the isolation of the rime, and then rhyme awareness, which is important for reading in English. In a fourth pathway, SRS is related to morphological awareness in at least two possible ways: On one hand, rhythm may be used as a clue that indicates whether a word is a compound word or two separate words (e.g., snowman; snow, man); on the other hand, the stress component may help children to learn morphological rules for stress assignment — for example, some suffixes of morphologically complex words indicate a change in the stress position (e.g., grave [serious], gravedad [seriousness]). In Spanish, lexical stress can also be used to differentiate similar words that are grammatically unequal (e.g., bebe [he/she drinks], bebé [baby]). Further to phonological or morphological awareness, Wood et al. (2009) considered the possibility that SRS can explain additional variance “…linked to the need of lexical stress to be assigned during the reading of polysyllabic words” (p. 19). More recently, Holliman et al. (2014) tested Wood et al.'s (2009) model in children aged 5 to 7. They found that the model fitted much better if additional links were included: between vocabulary and morphological awareness, rhyme and phoneme awareness, and phoneme and morphological awareness. Moreover, the first step in the model was slightly different, as it was proposed that prosody exerts its role via vocabulary, and rhyme awareness. Links between prosody and phoneme awareness, and between prosody and morphological awareness were not significant. Interestingly, although a direct link between prosody and literacy was proposed in the original Wood et al.'s (2009) model, it was not included in the new version. However, prosody is necessary for reading aloud, and it may actually be directly represented in the written script. This research further explores this possibility. 1.3. Empirical evidence Wood and Terrell (1998) carried out the first study relating SRS with reading – using a metrical stress measure – in children with reading difficulties. Results showed that children with reading difficulties had problems on SRS compared to a control group of the same chronological

145

age and another control group of the same reading level and therefore younger. Later studies with typically developing readers have replicated the relationship between metrical stress and reading ability (Clin, Wade-Woolley, & Heggie, 2009; Whalley & Hansen, 2006). At the word level, research suggests that there is a relationship between lexical stress sensitivity (henceforth, LSS) and literacy acquisition. Wood (2006) explored this relationship in English with a sample of 5 to 7 years-old children that had to recognize mispronounced (reversed stress) words (e.g., sofa as opposed to sofa). Results showed that LSS explained a significant part of writing performance after controlling for age, PA, or vocabulary. Later studies have shown a similar relationship with reading. Holliman, Wood, and Sheehy (2008) found that LSS predicted reading ability in children aged 5 and 6 years, regardless of variables such as age, vocabulary, or PA. More recently, Holliman, Wood, and Sheehy (2010) used an improved version of the task developed by Wood (2006); they also found that LSS predicted word reading regardless of variables such as age, vocabulary, PA, short-term memory, and non-linguistic rhythmic skills. However, when the sample was composed of children with reading difficulties, Holliman, Wood, and Sheehy (2012) found a different pattern or results. Once vocabulary and PA were both controlled for there were no LSS differences between a group of children with reading difficulties and a chronological control group (10 years old). Clearly, more research is needed in order to determine whether or not LSS makes a unique contribution to literacy acquisition. Opposite to English, some languages like Spanish or Greek have an orthographic mark that explicitly indicates stress. These languages provide an excellent opportunity to study the role of LSS, since prosody is inherent to their orthographies. For this reason, it is possible that LSS makes a unique contribution just in these languages, and this may not be generalized to other orthographies. In any case, it would be a relevant finding that would contribute to better understand literacy development across languages (Share, 2008). Some studies carried out in Spanish have suggested that LSS is linked to reading ability. Gutiérrez-Palma and Palma-Reyes (2007) measured LSS in Spanish children in 1st and 2nd grade of Primary Education. They used the task designed by Dupoux, Peperkamp, and SebastiánGallés (2001), which ask to discriminate between minimum pairs of pseudowords that include phonemic (kúpi vs. kúti) or prosodic contrasts (mípa vs. mipá). Children performing at a high level in the prosodic contrast task read more fluently and assigned stress more accurately in a pseudoword reading task than children with low performance. Afterward, Gutiérrez-Palma, Raya, and Palma (2009) partially replicated these results controlling phonological awareness. That study's results showed that, after controlling the influence of phonological awareness, participants performing at a high level at prosodic contrasts detection read texts faster and made fewer mistakes when assigning stress to pseudowords than children with low performance. The above studies were limited to children up to 2nd grade of Primary Education, who are still in the initial stages of reading acquisition. Besides, only pseudoword reading was examined and no real-word reading aloud was assessed. More recently, Defior, Gutiérrez-Palma, and Cano-Marín (2012) have used a sample of more advanced readers (5th grade) to examine the relationship of prosodic skills and word reading and writing. Moreover, they have used a different task, which consisted of listening to three-syllable pseudowords (e.g., páfica, nipora, zirotal) and indicating the stressed syllable. This task involves explicitly thinking in word's stress, and then it could be considered a way to measure lexical stress awareness (henceforth, LSA) rather than LSS. In LSS tasks participants may process the overall stress pattern at the wholeword level without detecting the stressed syllable. Defior et al.'s (2012) results showed that LSA was related to reading and writing accuracy, regardless phonological awareness. Furthermore, reading and spelling errors were analyzed. Result showed that in reading, LSA was higher related to stress errors (e.g., apóstol [apostle] written as apostól) than to phoneme-to-grapheme errors (e.g., apóstol written as abóstol).

146

N. Gutiérrez-Palma et al. / Learning and Individual Differences 45 (2016) 144–150

This pattern of results shows that LSA is relevant not only for reading, but also for writing acquisition, although results seems clearer for reading than for writing. If LSA is related to reading it is expected that children with reading difficulties have poor LSA skills. This is exactly what JiménezFernández, Gutiérrez-Palma, and Defior (2015) found when compared 31 Spanish children with dyslexia against a chronological group of 3rd graders. They found that participant with dyslexia performed worse than controls in a LSA task very similar to that used by Defior et al. (2012). They used two versions of this task: With words and pseudowords stimuli. In the word task, differences between the two groups disappeared once phonemic awareness was controlled for. Moreover, differences (reaction times) between words and pseudowords were not significant for the group of dyslexia, suggesting that they had difficulties to use their knowledge about stress. These studies suggest a significant relationship between stress and reading in Spanish, but this evidence is mainly correlational. More recently, Calet, Gutiérrez-Palma, Simpson, González-Trujillo, and Defior (2015) followed up children from the end of kindergarten until the end of grade 2. They found that even controlling for nonverbal intelligence, vocabulary, PA, and reading level, LSA2 at the start of grade 1 predicted reading at the end of this year. During grade 1 is when children are learning to decode words; as stress is inherent to word's phonology LSA may have a significant role on this learning. Except for the Defior et al.'s study (2012), there is no evidence about the role of LSA when reading words with different stress. If LSA facilitates the stress assignment process a significant relationship is expected between LSA and word reading. 1.4. Study approach This study extends previous results and explores the role of LSA in reading acquisition. In particular, it includes 3rd and 4th grades, when children are supposed to be taught in the use of the stress mark. The main hypothesis is that LSA development will be related to word reading acquisition, in particular to lexical stress assignment. This hypothesis is tested on an orthographic code – Spanish – in which stress is represented by a written mark. Additionally, a second aim is to find out whether LSA is directly connected to reading ability, independently of PA. It is predicted that LSA should be more related with stress errors than with grapheme–phoneme conversion errors. In contrast, as PA does not entail thinking about the relationship between sounds but about sounds themselves, it is expected to be more related to grapheme–phoneme conversion errors than to stress errors. In short, the study tries to test the following predictions: (a) LSA would predict a significant part of the variance in word reading; (b) in particular, LSA would mainly predict stress errors, while PA would mainly predict grapheme–phoneme conversion errors. 2. Method 2.1. Participants Participants were 233 children (123 girls) from 3rd to 6th grade of Primary Education (mean age in months in each grade: 103.1, SD = 3.3; 115.18, SD = 3.9; 126.7, SD = 3.5; 138.1, SD = 3.3, respectively). Twenty-three participants were excluded from the analyses because they scored below the percentile 10 or above the percentile 90 in the RAVEN test, and then were considered outliers (8, 4, 6, and 5 participants, respectively). The rest of participants were typically developing children, had Spanish as their first language, and were of medium socio-economic status. They were recruited at four schools through 2 They used the term LSS, but the task was very similar to that used here and explicitly required to think in the stress.

letters sent to parents by the local school system. Parental informed consent was obtained and verbal assent from each child before the testing session was required. 2.2. Instruments and measures Some standardized tests and some tests designed ad hoc were used. 2.2.1. Cognitive skills Raven's Progressive Matrices — SPM (Raven, 1995) was used and the number of correct responses, with a maximum possible score of 60, was registered. 2.2.2. Phonological awareness A counting sound task was used. The task was similar to the counting sound task of the THM battery (Test de Habilidades Metalingüísticas [Test of Methalinguistic Skills]) (Gómez, Valero, Buades, & Pérez, 2005), except for the response mode and for the stimuli. Participants were asked to listen to words and write the number of sounds down on the answer sheet (e.g., tostada [toast], 7 sounds). This task response mode – written – had the advantage of allowing collective testing, using small groups of participants. Words were presented through loudspeakers, assuring adequate testing conditions. The task was composed of 14 trisyllabic words plus 3 practice items. Items included several types of syllables (V, CV, VC, CVV, CCV, CVC, and CVCC), then covering main syllabic structures in Spanish. Sounds by word range was 5–9. Six items included a consonant cluster either in onset or coda position (e.g., preciso [precise], constante [constant]), thus increasing phonological processing demands in the task. One word (anoraks) had to be eliminated due to a corrected itemtotal correlation of 0. Moreover accuracy for this word was very low (19%). Thus, only 13 items were analyzed. Number of correct responses was counted, with a maximum possible score of 13. Test's reliability (Cronbach's alpha) was .68. 2.2.3. Lexical stress awareness A test composed of 18 trisyllabic pseudowords plus 3 practice tests was used. Pseudowords were used to eliminate the influence of lexical knowledge in detecting the stressed syllable. Item's structure was CVCV-CV or CV-CV-CVC; 6 were proparoxytones (e.g., bápujo), 6 were paroxytones (e.g., cibape) and 6 were oxytones (e.g., birogán). Children were asked to indicate the syllable sounding the strongest by pressing the keys b, n, or m, depending on whether it was the first, the second, or the third syllable, respectively. The test was carried out using a computer and headphones, so assuring adequate quality in items listening. The number of correct responses (a maximum possible of 18) was counted. Test's reliability (Cronbach's alpha) was .86. 2.2.4. Word reading This test is composed by 115 medium frequency words with different length (disyllable and trisyllable words), stress pattern (proparoxytone, paroxytone, and oxytone), and stress mark (present or absent). Words were presented on a computer screen and children had to read them aloud. The number of correct responses (a maximum possible of 115) was counted. Reliability (Cronbach's alpha) was .93. Errors were registered, but only phoneme and stress errors were further analyzed. Phoneme errors were those concerning to word grapheme decoding; the following error types were observed: replacing one grapheme by other – e.g., atóstol instead of apóstol (apostle); altering order inside the word – e.g., débli instead of débil (weak); addition – e.g., mudéjara instead of mudéjar (mudejar); and omission – e.g., mátil instead of mástil (flagpole). Stress errors were registered when reading words stressed in the wrong place, either if the word had a stress mark (e.g., ápostol instead of apóstol) or it did not (e.g., fórmal instead of formal [formal]). When both phoneme and stress errors were found within

N. Gutiérrez-Palma et al. / Learning and Individual Differences 45 (2016) 144–150

the same word (e. g., matil instead of mástil), both kinds of errors were registered. 2.3. Procedure Children were assessed at the end of the first quarter of the school year. Two individual sessions and two collective sessions were performed with an interval of approximately one week among sessions. The reading test was carried out in individual sessions 1 and 2; the LSA test was carried out in individual session 2. The counting sounds test took part in collective session 1; the Raven test was part of collective session 2. Collective sessions were carried out with small groups of children, with no more than 7 participants in each group. 3. Results Correct responses were registered for all tests. Additionally, phoneme and stress errors in the reading test were counted. Table 1 shows descriptive statistics for all variables by school grade. An initial correlation analysis was performed to get a general idea of the relationship among the variables included in the study. It was done by combining data from all the school grades and computing partial correlations controlling for Grade (see Table 2). All measures were significantly correlated, except PA with LSA and with phonological errors. The lack of relationship between PA and LSA is an interesting finding as it may suggest some kind of independence between both variables. This matches our first hypothesis — that is, LSA would predict reading performance independently of other variables. However, the finding of no relationship between PA and phonological errors is surprising. A possible explanation could be that PA may be linked to reading in early grades only (Defior et al., 2002). To further analyze the relations between all these variables several regressions were carried out. Cases with residuals greater than ±2.5 standard deviations were excluded from these analyses. Predictor variables were Grade, as a dummy variable, PA and LSA. Variables were included in the following order: first, Grade, as a measure of reading experience; second, PA, given its role in reading acquisition; third, LSA. For word reading accuracy, results showed that both, PA and LSA, were significant predictors (see Table 3). Therefore, as expected, LSA accounted for unique variance in word reading. A more detailed analysis for each school grade showed that LSA was a significant predictor of reading in all grades except grade 6th. PA was significant predictor in 3rd and 4th grades (see Table 4). As regards the second hypothesis – that is, PA would be mainly related to phonological errors while LSA would be related to stress-related errors –, a global hierarchical regression analysis was performed including Grade as a dummy variable. The predictor variables were included in the same order as in the previous analysis. Results showed that LSA, but not PA, was related to phonological errors. Regarding stress errors both, PA and LSA, were significant predictors, although LSA seemed to explain a higher percentage of the stress errors' variance. Moreover, LSA

147

Table 2 Partial correlations controlling for Grade.

1. PA 2. LSA 3. R 4. Rpe 5. Rse

1

2

3

4

– .12 .21⁎⁎ −.08 −.18⁎

– .36⁎⁎ −.22⁎⁎ −.42⁎⁎

– −.71⁎⁎ −.9⁎⁎

– – −.59⁎⁎

Note. PA = phonological awareness; LSA = lexical stress awareness; R = reading accuracy; Rpe = reading phonological errors; Rse = reading stress errors. ⁎ p b 0.05 (two-tailed). ⁎⁎ p b 0.01 (two-tailed).

explained a higher percentage of stress errors than of phonological errors (see Table 5). Detailed analyses by Grade indicated that PA approached to be a significant predictor in 3rd grade for both, phonological (p = .07) and stress errors (p = .06), and that it is significant predictor of stress errors in 4th grade. However, LSA was a significant predictor in all grades, but only for stress errors (see Table 6).

4. Discussion The main objective of this study was to determine the role of Lexical Stress Awareness (LSA) in reading in a transparent orthographic system such as Spanish, whose characteristics include the use of a written mark to represent word stress. The first hypothesis was that, regardless other variables, LSA would predict reading ability. This result seems to be confirmed in all the school grades except in 6th grade. In contrast, the influence of phonological awareness (PA) on reading accuracy is less evident and only appears in 3rd and 4th grades. It is well-known fact that grapheme–phoneme conversion rules are learnt relatively quickly in Spanish, due to the orthographic code transparency (Caravolas, Lervåg, Defior, Seidlová-Málková, & Hulme, 2013; Defior et al., 2002; Serrano, Genard, Sucena, Defior, & Alegria, 2011; Seymour et al., 2003). Consequently, word decoding processes are mastered earlier than in other languages. Because of this, the influence of PA would be no longer relevant after the initial phases of reading acquisition (Defior, 2008). This finding agrees with other studies showing that the PA skills role in reading and writing acquisition seems to differ depending on the level of the code transparency (Vaessen & Blomert, 2010); however, this is still an issue under discussion. Studies in English, whose orthographic code is highly opaque, show that PA is a stronger and longer lasting predictor of reading accuracy and fluency than it is in more transparent languages (Defior, 2008; de Jong & van der Leij, 1999, 2002; Landerl & Wimmer, 2008). Studies in transparent languages suggest that the influence of PA decreases quickly as reading experience increases. Therefore, it is not surprising to find that phonemic awareness keeps having influence in 3rd and 4th grades, but not in later grades. Nevertheless, the combined analysis of all the school grades shows PA to be a significant reading predictor.

Table 1 Descriptive statistics – mean and (standard deviation) – for all variables by Grade. Grade

CS

PA (Max = 13)

LSA (Max = 18)

R (Max = 115)

Rpe

Rse

3rd N = 49 4th N = 55 5th N = 53 6th N = 53

31.6 (6.7)

6.5 (2.7)

9.2 (3.8)

85.6 (12.3)

2.6 (2.4)

33.7 (6.8)

7.1 (2.9)

12.8 (4.5)

95.2 (11.4)

1.8 (2.3)

38.3 (6.6)

7.3 (1.7)

14.2 (3.5)

100.7 (10.2)

1.2 (2.5)

9.4 (6.6)

43 (4.9)

8.1 (2.4)

15.5 (2.9)

105.6 (7)

1 (1.1)

6.2 (5.5)

21 (8.9) 13.2 (8)

Note. CS = cognitive skills; PA = phonological awareness; LSA = lexical stress awareness; R = reading accuracy; Rpe = reading phonological errors; Rse = reading stress errors; N = number of participants; Max = maximum possible score.

148

N. Gutiérrez-Palma et al. / Learning and Individual Differences 45 (2016) 144–150

Table 3 Hierarchical linear regression analysis predicting reading accuracy. R2 change

Table 5 Hierarchical linear regression analyses predicting phonological and stress errors. R2 change

Final β

G

PA

LSA

G

PA

LSA

.39⁎⁎

.04⁎⁎

.09⁎⁎

.42⁎⁎

.17⁎⁎

.35⁎⁎

Note. G = grade; CS = cognitive skills; PA = phonological awareness; LSA = lexical stress awareness. ⁎⁎ p b 0.01.

Overall, these results suggest that LSA makes a unique contribution to reading acquisition, as shown by other studies (Gutiérrez-Palma et al., 2009; Holliman et al., 2010; Whalley & Hansen, 2006). It has been argued that LSA is in fact another type of PA (Goodman, Libenson, & Wade-Woolley, 2010). If this was so, on a theoretical level it would be also relevant to prove that there is another level of PA related to suprasegmental information that should be considered in models of reading acquisition. The results of the present study may partially support this idea: When all the school grade results are combined, LSA predicts 3.4% of phonological errors. Moreover, the capacity to think about the stress may help to identify some sounds (Wood et al., 2009), as phoneme identification is easier in stressed than in unstressed syllables. When the school grades are separately analyzed the general pattern is that PA approaches to predict phonological errors in 3rd grade, and predicts stress errors in 4th grade. On the contrary, LSA predicts stress errors in all grades, but not phonological errors. This may suggest some independence between LSA and PA, however the relationship between PA and phonological errors was not significant. Probably, this could be due to the fact that the amount of phonological errors in word reading was extremely low, which makes difficult to find significant relations. In the case of stress the amount of errors was much higher, and then it was much easier to find significant relations. A possible explanation of PA playing a role in stress-related errors in 4th grade is that some variance is shared by PA and LSA in reading, as some rules for assigning the stress mark specifically refer to the last letter of words. In these cases, the ability to think of these letters and/or the sounds associated to them should make it easier to correctly assign the stress mark. Findings also show that LSA predicts a high percentage of the variance in word reading ability (accuracy) in grades from 3rd to 5th. On the contrary, PA no longer plays a significant role from 5th grade onwards. As suggested by Whalley and Hansen (2006), prosodic skills may play a more relevant role in more advanced stages of expert reading acquisition, including the achievement of reading fluency (Benjamin & Schwanenflugel, 2010; Cowie, Cowie-Douglas, & Wichmann, 2002; Kuhn, Schwanenflugel, & Meisinger, 2010; Miller & Schwanenflugel, 2008). LSA could be considered to be relevant for the automation and fluent reading of words given that assigning stress is necessary for reading words in Spanish. LSA can be expected to help automate this process, just as PA helps automate the use of grapheme–phoneme conversion rules. After certain levels of automation have been achieved, the role of LSA may be less relevant. This may be the reason why LSA is

Rpe Rse

PA

LSA

G

PA

LSA

.09⁎⁎ .38⁎⁎

.01 .03⁎⁎

.03⁎⁎ .11⁎⁎

−.18⁎ −.4⁎⁎

−.08 −.14⁎⁎

−.22⁎⁎ −.38⁎⁎

Note. G = grade; PA = phonological awareness; LSA = lexical stress awareness; Rpe = reading phonological errors; Rse = reading stress errors. ⁎ p b 0.05. ⁎⁎ p b 0.01.

not related to the reading ability in 6th grade. Moreover, standard deviation diminishes in 6th grade, and reading accuracy approaches to a ceiling effect (see Table 1). Another possibility is to explore the role of LSA in text reading, where demands on word automation are presumably more exigent. These results add evidence and complement previous studies in Spanish (Calet et al., 2015; Defior et al., 2012; Gutiérrez-Palma & Palma-Reyes, 2007; Gutiérrez-Palma et al., 2009). As commented before, those studies do not explore the role of LSA when reading aloud words of different stress. Defior et al.'s study (2012) does, but just with 5th grade children. This study extends the sample from 3rd up to 6th grade. Moreover, as the current study suggests that prosodic skills are related to reading acquisition, it would be expected that children with reading acquisition difficulties would have poor prosodic skills, as it has been confirmed in other studies (Goswami et al., 2002; Holliman et al., 2012; Jiménez-Fernández et al., 2015; Muneaux, Ziegler, Truc, Thomson, & Goswami, 2004; Wood & Terrell, 1998). In summary, results show that LSA accounts for a considerable part of performance in reading independently from other variables. This predictive value is mainly associated to errors in stress assignment. This would suggest that LSA may play a direct role in reading acquisition; awareness of certain suprasegmental features, such as stress, seems to facilitate the learning of the stress marks that represent them in the written code. Although this role may be more important in orthographic systems like Spanish or Greek, it is not necessarily limited to them. In systems that do not use these marks, such as English, there are also stress regularities (Kelly, 2004; Kelly, Morris, & Verrekia, 1998; Rastle & Coltheart, 2000), which readers can implicitly learn and in which LSA may play a significant role. A limitation of this study may be that the PA task had less variation (only 13 items) than the LSA task, which may have limited finding significant relations. However ceiling effects were not found, which suggests that PA variability might be enough. The counting task, which is indeed a particular type of segmenting task that uses counting as response measuring, has proven to be a useful for participants with high reading ability (Serrano, González-Trujillo, & Defior, 2002; Serrano, Defior, & Martos, 2003). It was used as a difficult PA task, to avoid

Table 6 Hierarchical linear regression analyses predicting phonological and stress errors by grade. R2 change Grade

Table 4 Hierarchical linear regression analyses predicting reading accuracy by Grade. 2

Rpe

Final β

R change Grade

PA

LSA

PA

LSA

3 4 5 6

.14⁎⁎ .13⁎⁎ .0 .08⁎

.16⁎⁎ .08⁎ .2⁎⁎ .02

.35⁎⁎ .31⁎ −.02 .25

.4⁎⁎ .29⁎ .45⁎⁎ .15

Note. PA = phonological awareness; LSA = lexical stress awareness. ⁎ p b 0.05. ⁎⁎ p b 0.01.

Final β

G

Rse

3 4 5 6 3 4 5 6

Final β

PA

LSA

PA

LSA

.07 .02 .01 .02 .07 .12⁎⁎

.04 .05 .07 .0 .19⁎⁎ .16⁎⁎ .21⁎⁎ .09⁎

−.25 −.11 .07 −.14 −.23 −.27⁎

−.2 −.23 −.27 .01 −.44⁎⁎ −.41⁎⁎ −.46⁎⁎ −.3⁎

.03 .09⁎

.13 −.23

Note. PA = phonological awareness; LSA = lexical stress awareness; Rpe = reading phonological errors; Rse = reading stress errors. ⁎ p b 0.05. ⁎⁎ p b 0.01.

N. Gutiérrez-Palma et al. / Learning and Individual Differences 45 (2016) 144–150

obtaining ceiling effects in older children, because we aimed to measure PA until 6th grade, a relatively late grade. It is noteworthy that the present results are correlational, which could be considered as exploratory. In order to study the causal relations between LSA and reading training studies are required. Despite this limitation, this study contributes improving the understanding of which cognitive variables of literacy acquisition are universal or specific, thus bringing us closer to a global science of literacy acquisition (Share, 2008). Moreover, there are some implications for educational practice as well. Indeed, it is suggested that, apart from paying attention to the development of phonological awareness, education professionals should promote the development of other forms of metalinguistic awareness such as stress awareness to improve the acquisition of written language. These findings may also have implications for diagnosis and intervention in children with reading acquisition difficulties, at least in systems that represent stress in their orthography. Acknowledgments This study was partially funded through projects, PSI2011-29155 and PSI2010-21983-C02-01 of the Spanish Ministry of Science and Innovation. It also was supported by the research groups HUM-820 and HUM-883, of the Regional Government of Andalusia, in Spain. References Arvaniti, A. (2009). Rhythm, timing and the timing of rhythm. Phonetica, 66, 46–63. http://dx.doi.org/10.1159/000208930. Benjamin, R.G., & Schwanenflugel, P.J. (2010). Text complexity and oral reading prosody in young readers. Reading Research Quarterly, 45, 388–404. http://dx.doi.org/10.1598/ RRQ.45.4.2. Calet, N., Gutiérrez-Palma, N., Simpson, I., González-Trujillo, M.C., & Defior, S. (2015). Suprasegmental phonology development and reading acquisition: A longitudinal study. Scientific Studies of Reading, 19, 51–71. http://dx.doi.org/10.1080/10888438.2014. 976342. Caravolas, M., Lervåg, A., Mousikou, P., Efrim, C., Litavský, M., Onochie-Quintanilla, E., ... Hulme, C. (2012). Common patterns of prediction of literacy development in different alphabetic orthographies. Psychological Science. http://dx.doi.org/10.1177/ 0956797611434536 Advance online publication. Caravolas, M., Lervåg, A., Defior, S., Seidlová-Málková, G., & Hulme, C. (2013). Different patterns, but equivalent predictors, of growth in reading in consistent and inconsistent orthographies. Psychological Science. http://dx.doi.org/10.1177/0956797612473122 Advance online publication. Clin, E., Wade-Woolley, L., & Heggie, L. (2009). Prosodic sensitivity and morphological awareness in children's reading. Journal of Experimental Child Psychology, 101, 197–213. http://dx.doi.org/10.1016/j.jecp.2009.05.005. Cowie, R., Cowie-Douglas, E., & Wichmann, A. (2002). Prosodic characteristics of skilled reading: Fluency and expressiveness in 8–10 year old readers. Language and Speech, 45, 47–82. http://dx.doi.org/10.1177/00238309020450010301. Cutler, A., Dahan, D., & Donselaar, W. (1997). Prosody in the comprehension of the spoken language: A literature review. Language and Speech, 40, 141–201. Cutler, A., & Mehler, J. (1993). The periodicity bias. Journal o Phonetics, 21, 103–108. Dauer, R.M. (1983). Stress-timing and syllable-timing reanalyzed. Journal of Phonetics, 11, 51–62. Defior, S. (2008). ¿Cómo estimular el aprendizaje inicial de la lectoescritura? Papel de las habilidades fonológicas [How to encourage the initial acquisition of reading and writing? The role of phonological skills]. Infancia y Aprendizaje, 31(3), 333–345. http://dx. doi.org/10.1174/021037008785702983. Defior, S., Gutiérrez-Palma, N., & Cano-Marín, M.J. (2012). Prosodic awareness skills and literacy acquisition in Spanish. Journal of Psycholinguistic Research, 41, 285–294. http://dx.doi.org/10.1007/s10936-011-9192-0. Defior, S., Jiménez-Fernández, G., & Serrano, F. (2009). Complexity and lexicality effects in spelling acquisition in Spanish. Learning and Instruction, 19, 55–65. http://dx.doi.org/ 10.1016/j.learninstruc.2008.01.005. Defior, S., Martos, F., & Cary, L. (2002). Differences in reading acquisition development in two shallow orthographies: Portuguese and Spanish. Applied PsychoLinguistics, 23, 135–148. http://dx.doi.org/10.1017/S0142716402000073. Dowhower, S. (1991). Speaking of prosody: Fluency's unattended bedfellow. Theory Into Practice, 30, 165–175. http://dx.doi.org/10.1080/00405849109543497. Dupoux, E., Peperkamp, S., & Sebastián-Gallés, N. (2001). A robust method to study “stress deafness”. Journal of the Acoustic Society of America, 110, 1606–1618. http://dx.doi. org/10.1121/1.1380437. Gómez, P.F., Valero, J., Buades, R., & Pérez, A.M. (2005). THM: Test de habilidades meatlingüísticas [Test of methalinguistic skills]. Madrid: EOS. Goodman, I., Libenson, A., & Wade-Woolley, L. (2010). Sensitivity to linguistic stress, phonological awareness and early reading ability in preschoolers. Journal of Research in Reading, 33, 113–127. http://dx.doi.org/10.1017/S0142716400000424.

149

Goswami, U., Thomson, J., Richardson, U., Stainthorp, R., Hughes, D., Rosen, S., & Scott, S.K. (2002). Amplitude envelope onsets and developmental dyslexia: A new hypothesis. Proceedings of the National Academy of Sciences, 99, 10911–10916. http://dx.doi.org/ 10.1073/pnas.122368599. Gutiérrez-Palma, N., & Palma-Reyes, A. (2007). Stress sensitivity and reading performance in Spanish: A study with children. Journal of Research in Reading, 30, 157–168. http:// dx.doi.org/10.1111/j.1467-9817.2007.00339.x. Gutiérrez-Palma, N., Raya, M., & Palma, A. (2009). Detecting stress patterns is related to children's performance on reading tasks. Applied PsychoLinguistics, 30, 1–21. http:// dx.doi.org/10.1017/S0142716408090012. Hayes, B. (1995). Metrical stress theory: Principles and case studies. Chicago: University of Chicago Press. Holliman, A., Critten, S., Lawrence, T., Harrison, E., Wood, C., & Hughes, D. (2014). Modeling the relationship between prosodic sensitivity and early literacy. Reading Research Quarterly, 49, 469–482. http://dx.doi.org/10.1002/rrq.82. Holliman, A., Wood, C., & Sheehy, K. (2008). Sensitivity to speech rhythm explains individual differences in reading ability independently of phonological awareness. British Journal of Developmental Psychology, 26, 357–367. http://dx.doi.org/10.1348/ 026151007X241623. Holliman, A., Wood, C., & Sheehy, K. (2010). The contribution of sensitivity to speech rhythm and non-speech rhythm to early reading development. Educational Psychology, 30, 247–267. http://dx.doi.org/10.1080/01443410903560922. Holliman, A., Wood, C., & Sheehy, K. (2012). A cross-sectional study of prosodic sensitivity and reading difficulties. Journal of Research in Reading, 35, 32–48. http://dx.doi.org/10. 1111/j.1467–9817.2010.01459.x. Jiménez-Fernández, G., Gutiérrez-Palma, N., & Defior, S. (2015). Impaired stress awareness in Spanish children with dyslexia. Research in Developmental Disabilities, 37, 152–161. http://dx.doi.org/10.1016/j.ridd.2014.11.002. de Jong, P.F., & van der Leij, A. (1999). Specific contributions of phonological abilities to early reading acquisition: Results from a Dutch latent variable longitudinal study. Journal of Educational Psychology, 91, 450–476. http://dx.doi.org/10.1037/0022-0663.91.3.450. de Jong, P.F., & van der Leij, A. (2002). Effects of phonological abilities and linguistic comprehension on the development of reading. Scientific Studies of Reading, 6, 51–77. http://dx.doi.org/10.1207/S1532799XSSR0601_03. Kelly, M.H. (2004). Word onset patterns and lexical stress in English. Journal of Memory and Language, 50, 231–244. http://dx.doi.org/10.1016/j.jml.2003.12.002. Kelly, M.H., Morris, J., & Verrekia, L. (1998). Orthographic cues to lexical stress: Effects on naming and lexical decision. Memory & Cognition, 26, 822–832. http://dx.doi.org/10. 3758/BF03211401. Kuhn, M.R., Schwanenflugel, P.J., & Meisinger, E.B. (2010). Aligning theory and assessment of reading fluency: Automaticity, prosody, and definitions of fluency. Reading Research Quarterly, 45, 230–251. http://dx.doi.org/10.1598/RRQ.45.2.4. Landerl, K., & Wimmer, H. (2008). Development of word reading fluency and spelling in a consistent orthography: An 8-year follow-up. Journal of Educational Psychology, 100, 150–161. http://dx.doi.org/10.1037/0022-0663.100.1.150. Miller, J., & Schwanenflugel, P.J. (2008). A longitudinal study of the development of reading prosody as a dimension of oral reading fluency in early elementary school children. Reading Research Quarterly, 43(4), 336–354. http://dx.doi.org/10.1598/ RRQ.43.4.2. Muneaux, M., Ziegler, J.C., Truc, C., Thomson, J., & Goswami, U. (2004). Deficits in beat perception and dyslexia: Evidence from French. Neuroreport, 15, 1255–1259. http://dx. doi.org/10.1097/01.wnr.0000127459.31232.c4. Nazzi, T., Bertoncini, J., & Mehler, J. (1998). Language discrimination by newborns: Toward an understanding of the role of rhythm. Journal of Experimental Psychology: Human Perception and Performance, 24, 756–766. http://dx.doi.org/10.1037/0096-1523.24.3.756. Nespor, M., & Vogel, I. (2007). Prosodic phonology: With a new foreword. Berlin: Walter and the Gruyter GmbH & Co. KG. Rastle, K., & Coltheart, M. (2000). Lexical and nonlexical print-to-sound translation of disyllabic words and nonwords. Journal of Memory and Language, 42, 342–364. http:// dx.doi.org/10.1006/jmla.1999.2687. Raven, J. (1995). Matrices progresivas de Raven. Escala SPM [Raven's progressive matrices. SPM scale]. Madrid: TEA. Real Academia Española (2009). Ortografía de la lengua Española [Spanish orthography]. Madrid: Espasa Calpe. Serrano, F., Defior, S., & Martos, F. (2003). To be or not to be phonologically aware: A reflection about metalinguistic skills of student of teacher. In R.M. Joshi, C.K. Leong, & B.L.J. Kaczmarek (Eds.), Literacy acquisition: The role of phonology, morphology and orthography (pp. 209–215). Amsterdam, NL: IOS Press. Serrano, F., Genard, N., Sucena, A., Defior, S., Alegria, J., Mousty, A., ... Seymour, P.H.K. (2011). Variations in reading and spelling acquisition in Portuguese, French and Spanish: A cross-linguistic comparison. Journal of Portuguese Linguistics, 10, 183–204. Serrano, F., González-Trujillo, M.C., & Defior, S. (2002). Phonological awareness differences between children and adults. In M.N. Trindade (Ed.), Literaçias e Cidadania. Convergencias e interfaces (pp. 1–7). Évora. Universidade de Evora. Seymour, P., Aro, M., Erskine, J., Wimmer, H., Leybaert, J., Elbro, C., ... Olofsson, Å. (2003). Foundation literacy acquisition in European orthographies. British Journal of Psychology, 94, 143–174. Share, D.L. (2004). Knowing letter names and learning letter sounds: A causal connection. Journal of Experimental Child Psychology, 88, 213–233. http://dx.doi.org/10.1016/j. jecp.2004.03.005. Share, D.L. (2008). On the anglocentricities on current reading research and practice: The perils of overreliance on an “outlier” orthography. Psychological Bulletin, 134, 584–615. http://dx.doi.org/10.1037/0033-2909.134.4.584. Vaessen, A., & Blomert, L. (2010). Long-term cognitive dynamics of fluent reading development. Journal of Experimental Child Psychology, 105(3), 213–231. http://dx.doi.org/ 10.1016/j.jecp.2009.11.005.

150

N. Gutiérrez-Palma et al. / Learning and Individual Differences 45 (2016) 144–150

Whalley, K., & Hansen, J. (2006). The role of prosodic sensitivity in children's reading development. Journal of Research in Reading, 29(1), 288–303. http://dx.doi.org/10.1111/ j.1467–9817.2006.00309.x. Wood, C. (2006). Metrical stress sensitivity in young children and its relationship to phonological awareness and reading. Journal of Research in Reading, 29(3), 270–287. http://dx.doi.org/10.1111/j.1467–9817.2006.00308.x.

Wood, C., & Terrell, C. (1998). Poor readers' ability to detect speech rhythm and perceive rapid speech. British Journal of Developmental Psychology, 16, 397–413. http://dx.doi. org/10.1111/j.2044-835X.1998.tb00760.x. Wood, C., Wade-Woolley, L., & Holliman, A.J. (2009). Phonological awareness: Beyond phonemes. In C. Wood, & V. Connelly (Eds.), Contemporary perspectives on reading and spelling (pp. 7–23). London: Routledge.