The impact of rapid automatized naming and phonological awareness on the reading fluency of a minority student population.
Keywords: achievement, phonological awareness, assessment, early literacy, intervention, reading fluency, childhood educators, reading intervention, double-deficit
Reading is a critical skill that children living in a literate and technologically advanced society need to acquire. However, not all students acquire reading skills at the same rate or trajectory. It is estimated that about one in three students experiences difficulty acquiring reading skills and approximately one in five students experiences significant difficulties (Lyon & Moats, 1997; Shaywitz, Escobar, Shaywitz, Fletcher, & Makuch, 1992). Unfortunately, the highest proportion of students experiencing difficulty acquiring functional literacy skills is from disadvantaged backgrounds and educated in inner-city schools (Knapp, Shields, & Turnbull, 1992). Nevertheless, without early intervention, all students experiencing difficulty acquiring reading skills in the early grades may never read adequately.
Projections suggest that a student who was a poor reader in 1st grade would also be a poor reader in 4th grade (Jeul, 1988). Similarly, a student who experienced difficulty reading grade-level material in 3rd grade most likely would not make significant gains in basic reading skills by the end of high school (Torgesen & Burgess, 1998). Not only do these students experience persistent reading difficulty, they also tend to fall further behind their more literate grade-level peers in all academic areas (Chall, Jacobs, & Baldwin, 1990). The poor prognosis for students experiencing difficulties acquiring early literacy skills has led educators and legislators to place greater emphasis on early intervention (e.g., Individuals With Disabilities Education Act, 2004; No Child Left Behind, 2002). These early intervention initiatives include prevention, intervention, data-based decision making, and greater accountability for teachers.
Such legislative initiatives place great emphasis on the early identification of students who are at risk for reading problems and the importance of providing at-risk students with explicit and systematic reading instruction and intervention. The process of early identification and early intervention often falls on the classroom teacher. Therefore, two important activities for the classroom teacher include identifying students at-risk for reading difficulties and identifying the poor readers' specific areas of reading weakness to target intervention.
In its comprehensive review of reading research, the National Reading Panel (2000) identified five basic skills that are important and necessary for early literacy development: phonological awareness, phonics, fluency, vocabulary development, and comprehension. The development of reading begins at the grapheme to phoneme level and progresses to word decoding, fluency, and comprehension. In the early grades, students learn how to read; in the later grades, students read to learn (Vacca et al., 2003). Thus, much intervention in the early grades targets students' acquisition of early reading skills.
RAPID AUTOMATIZED NAMING
Research has identified two key components critical to the development of word decoding: phonological awareness (PA) and rapid automatized naming (RAN). RAN is broadly defined as how fast the student is able to identify simple stimuli, such as letters, digits, or pictures. For the classroom teacher, the advantage of RAN, in the identification of at-risk students, is the ease of test administration. RAN assessments generally take only about one minute to administer. In a typical RAN assessment, a student is presented with a single page containing rows of printed letters (or digits or pictures) in random order. The student is asked to read as many letters as possible within 30 seconds. The student's score is the number of letters identified correctly. RAN has demonstrated statistical significance in the prediction of reading achievement after controlling for letter knowledge, speed of processing, and phonological awareness (Bowey, McGuigan, & Ruschena, 2005; Kirby, Parrila, & Pfeiffer, 2003; Leopla, Poskiparta, Laakkonen, & Niemi, 2005).
The National Early Literacy Panel (2008) identified RAN (naming of digits or letters, and of objects or colors) as one of the five key variables correlating with future reading that also had predictive validity. The other variables were alphabet knowledge, phonological awareness, writing or writing one's name, and phonological memory. RAN's predictive validity for future reading acquisition makes it an excellent and economical task to identify students at-risk for reading difficulty at an early age.
PA is the ability to detect, manipulate, or analyze components of spoken words. This involves many skills, including identifying, blending, segmenting, reversing, and adding or deleting phonemes to form words. Also included under the umbrella of PA are the detection of common onsets between words, identification of common rime units, and the ability to combine syllables and phonemes to form words (Armbruster, Lehr, & Osborn, 2001; National Early Literacy Panel, 2008). Among the most important of these PA skills are blending and segmenting (Armbruster, 2002; Shaywitz, 2003); when students first receive reading instruction, they are taught blending and segmenting (Armbruster, 2002; National Reading Panel, 2000; Shaywitz, 2003).
Classroom-based interventions focusing on the acquisition of (PA) skills are particularly important for at-risk readers. A wide body of research supports PA skills as a prime target of early intervention for students experiencing reading difficulties (e.g., National Early Literacy Panel, 2008; Snow, Burns, & Griffin, 1998). RAN is thought to have a moderate correlation with PA (r = .34 - .44), because recognizing and identifying phonemes are precursors to pronouncing them (Share, 1995; Torgesen, Wagner, Rashotte, Burgess, & Hecht, 1997). The relationship between RAN and PA is a topic of debate. RAN is thought to be due to RAN's task demand on phonological coding skills (Share, 1995; Torgesen et al., 1997; Wagner & Torgesen, 1987). In contrast, other research identified RAN and PA as two distinct constructs. This view, which is the foundation of the double-deficit hypothesis, is supported by research demonstrating RAN's ability to predict reading above and beyond PA (Manis, Doi, & Bhadha, 2000; Wolf & Bowers, 1999).
PURPOSE OF THE STUDY
Contemporary research has identified RAN and PA screening instruments as valid predictors of students at-risk for future reading failure. A strong body of research has identified the acquisition of PA skills as a cornerstone in the development of early literacy skills (Institute for the Development of Educational Achievement, 2008; National Reading Panel, 2000). Some debate exists in the literature regarding the relationship between RAN and PA (e.g., Torgesen et al., 1997; Wolf & Bowers, 1999). The true relationship between these two constructs and the portion of variance each contributes to reading is a topic of current research.
Although much research has investigated the relationship between RAN and PA, little research has explored the relationship between the two constructs in homogenous minority populations. Within minority populations, PA skills are believed to develop during prekindergarten and continue through 3rd grade (Snow et al., 1998; Thomas, Washington, & Edwards, 2004; Whitehurst & Lonigan, 2001). The process by which RAN contributes to the reading skills of minority populations is less known. The purpose of this study was to investigate the relative contribution of RAN and PA on the RF of students from traditionally underrepresented, minority backgrounds.
All participants attended an inner-city charter school. A total of 86 students attending 1st through 4th grade participated in the study (16 first-, 36 second-, 23 third-, and 11 fourth-graders). Students in 1st through 4th grades were selected for participation in the study for two reasons. First, research suggests that minority populations' PA skills develop through 3rd grade (Snow et al., 1998; Thomas et al., 2004; Whitehurst & Lonigan, 2001). The second reason was to obtain as large a sample as possible within a school that only educated about 125 students between 1st and 5th grade.
Participant ages ranged from 6 to 10 years of age (M = 8.15, SD = 1.0). There were 48 females (56%) and 38 males (44%). Approximately 97% of the participants were African American, and all of the participants were from economically disadvantaged families, as evidenced by their receiving free or reduced-price lunch. The charter school followed the same curriculum, and received the same per-pupil funding as other schools within the same district. A key difference between the charter school and other district-based schools was that parents chose to have their children attend the charter school.
The Comprehensive Test of Phonological Processing (C-TOPP; Wagner, Torgesen, & Rashotte, 1999) was used to assess students' PA skills and RAN skills. The C-TOPP has a mean of 10 and a standard deviation of 3. The authors reported that test-retest coefficients range from .70-.92 and alternate form reliability or internal consistency reliability coefficients are both in excess of .80 (Wagner et al., 1999).
The Woodcock-Johnson III: Tests of Achievement (WJ-III ACH; Woodcock, McGrew, & Mather, 2001) was used to measure participants' reading fluency. The WJ-III ACH has a mean of 100 and a standard deviation of 15. Reported test-retest reliability coefficients for examinees within the age group of the present study ranged from .89 to .90 (McGrew & Woodcock, 2001).
Measures of phonological awareness skills. PA was assessed using the Segmenting Nonwords and Blending Nonwords tests from the C-TOPP. On the Blending Nonwords test, participants were required to form individual phonemes into whole words. In contrast, the Segmenting Nonwords test required students to break individual sounds into their constituent phonemes.
Measures of rapid automatized naming skills. RAN was assessed using the Naming Letters and Naming Digits subtests from the C-TOPP. On the Naming Letters subtest, participants were asked to read as many letters as they could within 30 seconds. The letters were presented randomly. The Naming Digits subtest was exactly the same as the Naming Letters subtest, with the exception that participants read random digits ranging from 0 to 9, in contrast to letters.
Measure of reading fluency skills. The Reading Fluency (RF) test from the WJ-III ACH measures an individual' s ability to read simple sentences quickly and respond if the statement is true or not. The test requires participants to read and rate as many sentences as possible within 3 minutes.
After obtaining parent permission and child assent, participants were tested over a 5-day period. All tests were individually administered to participants by graduate students. All 12 graduate students were in their second year of study and had completed a semester-long course on standardized test administration. All standardized administration procedures, as identified within each instrument's technical manual, were followed. All of the graduate students were also supervised by university-based faculty during practice test administrations, which were conducted to ensure each graduate student's competency to administer all of the instruments. A university-based faculty member also provided on-site supervision during all school-based assessments.
All of the participants were assigned a code number; thus, neither participants' names nor any other identifying information were used. Participant codes along with their corresponding standard scores were entered for data analysis.
Structural equation modeling (SEM) was used to test the relative contribution of the independent PA and RAN variables on RF via the AMOS statistical program (Arbuckle & Wothke, 2004). Two models were tested. The first model, Model 1, tested the relative contribution of PA and RAN on RF. Model 2 investigated whether the skill represented in RAN task performance was also a skill within PA; in other words, whether RAN and PA were two independent constructs (Manis, Doi, & Bhadha, 2000; Wolf & Bowers, 1999) or whether RAN was a form of PA (Share, 1995; Torgesen et al., 1997; Wagner & Torgesen, 1987).
Although multiple regression (MR) provides statistical analyses to examine the relations between variables, the use of SEM in this study goes beyond much previous research by offering several advantages over MR as well as other statistical analyses. One advantage of SEM is that the latent variables are cleansed of error. A second advantage is that SEM allows for the simultaneous estimation of direct and indirect effects between variables (see Keith, 2006). This was most important in Model 2, where the direct effect of PA on RF was examined, as was the indirect effect of RAN on RF through the PA factor (this also provided a test of the relationship between RAN and PA) within one model. Neither factor analysis nor MR allows this type of analysis.
The goal of SEM is to find the common variance between variables, not the unique contribution of each variable. Direct effects are the direct influence of one factor on another factor, whereas an indirect effect is the influence of one factor through another factor to a target factor. Model 1, as presented in Figure 1, has only direct effects, whereas Model 2, displayed in Figure 2, has direct and indirect effects. Direct effects are indicated by single-headed arrows in Figures 1 and 2.
This study investigated the effect of key components of RAN and PA on the RF of 86 students from traditionally underrepresented backgrounds, attending an inner-city charter school. Further, this study examined the relationship between RAN, PA, and RF.
Descriptive statistics associated with all assessments are presented in Table 1. The correlation between all independent and dependent variables is presented in Table 2. Cohen's (1988) strength of relationship guidelines suggest that correlations between .10 and .29 are small, correlations between .30 and .49 are medium, and correlations between .50 and 1.0 are large. As indicated in Table 2, statistically significant correlations were found between the PA variables r = .467 (p < .01) as well as the RAN variables r = .805 (p < .01). A statistically significant relationship was also found between RAN variables and RF, as well as blending and RF; however, the relationship between segmenting and RF was not statistically significant.
Two different SEM models were tested. The first model, Model 1, investigated the direct effect of the two independent variables (RAN, PA) on RF. The numbers above the arrows, in Figure 1, are referred to as standardized path coefficients and are analogous to regression coefficients or standardized beta weights in multiple regression (Keith, 2006). The value of the path coefficient is the direct effect. Effect sizes of .05 and above can be considered small effect sizes. Effect sizes above. 15 are moderate, and effect sizes above .25 may be considered large (Keith, 1999, 2006; Pedhazur, 1997).
In Figure 1, the rectangular boxes represent participant test scores. In SEM nomenclature, rectangles are called observed or exogenous variables. The ellipses or ovals are non-measured variables (factors) also called latent variables or endogenous factors. The path or single-headed arrow between blending and PA and segmenting and PA represent the direct effect of each variable on PA. Similarly, the single-headed arrow between PA and RF represents the direct effect of PA on RF. The correlation between segmenting and blending is indicated by the double-headed arrow between the two variables in Figure 1.
SEM provides an opportunity to evaluate the fit of a model to the data through fit indices. Table 3 presents the fit indices associated with all models. The purpose of fit indices in SEM is to provide an indication of how well a model approximates reality. An examination of the fit indices in Table 3 indicates that the comparative fit index (CFI) of Model 1 was .996.
Rules of thumb suggest that CFI values over .90 indicate adequate fit, whereas values above .95 represent a good fit of the model to the data (Hu & Bentler, 1999). Rules of thumb for the root mean square error of approximation (RMSEA) suggest that values below .08 indicate an adequate fit and values below .05 suggest a close fit of the model in relation to degrees of freedom. Thus, the RMSEA of .036 suggests a close fit of the model to the data. The chi-square of 5.57 is not statistically significant. The nonstatistically significant chi-square suggests that the model and data are consistent with each other, again indicating the model provides a good fit to the data.
Table 4 presents the direct effects within the models tested in this study. In Model 1, the direct effect of PA on RF was .24. The direct effect or effect size between RAN and RF was .41. Thus, PA had a moderate-to-large effect on RF, whereas, RAN had a large effect on RF.
Model 2, presented in Figure 2, was similar to Model 1, with the exception of moving the direct effect of RAN from RF to PA. This provided an estimation of the shared variance between RAN and PA in the prediction of RE Within Model 2, the effect of RAN on RF was indirect, meaning that RAN's effect on RF was through the PA factor. In this model, the PA factor represents the shared variance between the PA tests and RAN.
In Model 2, the direct effect of RAN on PA was .85, a large effect. This supports that RAN is a member of the phonological awareness family--in other words, RAN and PA are not independent constructs. The indirect effect of RAN on RF can be calculated by multiplying the path coefficient between RAN and PA by the path coefficient between PA and RF. The indirect effect of RAN on PA was .40 after rounding (.85 x .47 = .3999). Thus, RAN (as a PA construct) had a large, indirect effect on RF.
This study investigated the relationship between PA and RAN, as well as their impact on the RF of students from traditionally underrepresented backgrounds attending 1st through 4th grade at an inner-city charter school. Two models were tested. The first model, Model 1, tested the direct effects of PA and RAN on RE The results from this analysis suggest that RAN is a better predictor of RF than PA--the direct effects are .41 and .24, respectively. The second analysis investigated the direct effect of RAN on PA and PA on RF as well as the indirect effect of RAN on RF (i.e., RAN as a PA construct). The direct effect of RAN on PA was .85. This surprisingly large, direct effect of RAN on PA suggests that RAN shares a large portion of variance with PA.
Also of interest was the decrease in the direct effect of blending on PA, between Models 1 and 2, .88 and .39, respectively (see Table 4). This suggests that when the direct effect of RAN was added to PA, the portion of variance accounted for by blending decreased, or was better accounted for by RAN. Together, this may mean that RAN, although related to PA, shares less variance with segmenting than blending. Additionally, when the direct effect of RAN was added to PA, the effect of PA on RF increased (.47 versus .24 and .41, Table 4). Thus, Model 2--the model that specified RAN as having a direct effect on PA and an indirect effect on RF--resulted in more variance accounted for when compared to Model 1.
In summary, when testing the direct effect of RAN and PA (Model 1) on the RF of students from traditionally underrepresented minority populations, RAN had a larger direct effect on RF than PA (.41 vs. .24, Table 4). Thus, RAN was a good and valid predictor of reading.
Although RAN demonstrated a large direct effect on RF, this model did not test the independence of RAN from PA in the prediction of RE The results from Model 2, testing the relationship between RAN and PA, found that RAN had a large direct effect on PA (.85, Table 4).
As indicated above, the relationship between RAN and PA has been the topic of some research and debate. The present study adds to this body of research by (1) employing SEM in an effort to move beyond multiple regression in the investigation of the relationship between these two constructs, (2) investigating the direct and indirect effects of RAN on PA and RF, and (3) conducting analyses using data representing a traditionally underrepresented population. The finding in this study that RAN and PA are significantly related lends support to earlier research suggesting that RAN is a part of the PA family (e.g., Share, 2008; Torgesen et al., 1997; Wagner & Torgesen, 1987). Thus, the results from the present study do not support the double-deficit hypothesis (e.g., Manis et al., 2000; Wolf & Bowers, 1999).
IMPLICATIONS FOR INSTRUCTION
Clarification of the relationship between RAN and PA is very important to the classroom teacher and all educators. This is because if RAN is related to PA, then at-risk students with low RAN scores should benefit from interventions designed to improve PA skills. In contrast, if RAN is found to be independent of PA, then at-risk students who demonstrate low RAN scores may need RAN-based intervention.
The finding of a strong relationship between RAN and PA in this study may suggest that for traditionally underrepresented students, RAN is a member of the PA family, and the two constructs are not independent. By extension, these results provide support for the use of RAN tests to screen for students who are at-risk for reading problems. After identification, however, the results suggest that providing explicit and systematic instruction in the development of PA skills should benefit students, when compared to interventions designed to increase scores on RAN tasks (i.e., naming letters or numbers faster).
The findings from this study may be limited by the relative homogeneity of participants who were all economically disadvantaged and primarily of African American descent (97%). The study also was limited by the relatively small sample size, which was in part a result of using individually administered, norm-referenced assessments. Although the homogeneity of the participants may be a limitation of the study, it is also one of its many strengths, because the current study examined a population considered to be underinvestigated in social science research (Sue, 1999). Although this study used several widely accepted and administered assessments, it is recommended that future studies include additional measures and predictors of early reading skills within traditionally underrepresented populations.
Arbuckle, J. L., & Wothke, W. (2004). Amos 5.0 user's guide. Chicago, IL: Smallwaters.
Armbruster, B. (2002). Research-based instruction in reading. Retrieved September 1, 2004, from www.ed.gov/admis/ lead/read/rb/edlite-slide001.html
Armbruster, B. B., Lehr, E, & Osborn, J. (2001). Put reading first: The research blocks for teaching children to read. Washington, DC: U.S. Department of Education.
Bowey, J. A., McGuigan, M., & Ruschena, A. (2005). On the association between serial naming speed for letters and digits and word-reading skill: Towards a developmental account. Journal of Research in Reading, 28(4), 400-422.
Chall, J. S., Jacobs, V., & Baldwin, L. (1990). Learning to read: The great debate. New York, NY: McGraw-Hill.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.
Individuals With Disabilities Education Act (IDEA) of 2004. (2004). 20 U.S.C. 33 Section 1400 et seq. Retrieved from http://idea.ed.gov/explore/home
Institute for the Development of Educational Achievement. (2008). Big ideas in beginning reading. Retrieved January 15, 2008, from http://dibels.uoregon.edu/index.php
Jeul, C. (1988). Learning to read and write. A longitudinal study of 54 children from first through fourth grades. Journal of Educational Psychology, 80, 437-447.
Keith, T. Z. (1999). Effects of general and specific abilities on student achievement: Similarities and differences across ethnic groups. School Psychology Quarterly, 14, 239-262.
Keith, T. Z. (2006). Multiple regression and beyond. Boston, MA: Allyn & Bacon.
Kirby, J. R., Parrila, R., & Pfeiffer, S. (2003). Naming speed and phonological awareness as predictors of reading development. Journal of Educational Psychology, 95, 453-464.
Knapp, M. S., Shields, P. M., & Turnbull, B. J. (1992). Academic challenge for the children of poverty. Menlo Park, CA: SRI International.
Leopla, J., Poskiparta, E., Laakkonen, E., & Niemi, P. (2005). Development of and relationship between phonological and motivational processes and naming speed in predicting word recognition in grade 1. Scientific Studies of Reading, 9, 367-399.
Lyon, G. R., & Moats, L. C. (1997). Critical conceptual and methodological considerations in reading intervention research. Journal of Learning Disabilities, 30, 578-588.
Manis, F. R., Doi, L. M., & Bhadha, B. (2000). Naming speed, phonological awareness, and orthographic knowledge in second graders. Journal of Learning Disabilities, 33, 325-333.
McGrew, K. S., & Woodcock, R. W. (2001). Technical manual. Woodcock-Johnson III. Itasca, IL: Riverside Publishing.
National Early Literacy Panel. (2008). Developing early literacy: Report of the National Early Literacy Panel. Washington, DC: National Institute for Literacy.
National Reading Panel. (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implication for reaching instruction. Reports of the subgroups. Washington, DC: National Institute of Child Health and Development.
No Child Left Behind. (2002). Washington, DC: U.S. Department of Education. Retrieved from http://www2.ed.gov/ policy/elsec/leg/esea02/index.htm
Pedhazur, E. J. (1997). Multiple regression in behavioral research. New York, NY: Wadsworth.
Share, D. L. (1995). Phonological recoding and self-teaching: Sine qua non of reading acquisition. Cognition, 55, 151-218.
Share, D. L. (2008). Orthographic learning, phonological reading, and self-teaching. In R. V. Kall (Ed.), Advances in child development and behavior (pp. 31-82). San Diego, CA: Elsevier Academic Press.
Shaywitz, S. E. (2003). Overcoming dyslexia: A new and complete science-based program for overcoming reading problems at any level. New York, NY: Alfred Knopf.
Shaywitz, S. E., Escobar, M. D., Shaywitz, B. A., Fletcher, J. M., & Makuch, R. (1992). Evidence that dyslexia may represent the lower tall of the normal distribution of reading disability. New England Journal of Medicine, 326, 145-150.
Snow, C. E., Burns, M. S., & Griffin, P. (1998). Preventing reading difficulties in young children. Washington, DC: National Academy Press.
Sue, S. (1999). Science, ethnicity, and bias: Where have we gone wrong? American Psychologist, 54, 1070-1077.
Thomas, S., Washington, J., & Edwards, J. (2004). Standardized assessment of phonological awareness skills in low- income African American first graders. American Journal of Speech-Language Pathology, 13, 182-190.
Torgesen, J. K., & Burgess, S. R. (1998). Consistency of reading related phonological processes throughout early childhood: Evidence from longitudinal-correlational instructional studies. In J. Metsala & L. Ehri (Eds.), Word recognition in beginning reading (pp. 161-188). Hillsdale, NJ: Lawrence Erlbaum.
Torgesen, J. K., Wagner, R. K., Rashotte, C. A., Burgess, S., & Hecht, S. (1997). Contributions of phonological awareness and rapid automatic naming ability to the growth of word-reading skills in second- to fifth-grade children. Scientific Studies of Reading, 1, 161-185.
Vacca, J. L., Vacca, R. T., Gove, M. K., Burkey, L., Lenhart, L. A., & McKeon, C. (2003). Reading and learning to read (5th ed.). New York, NY: Allyn and Bacon.
Wagner, R. K., & Torgesen, J. (1987). The nature of phonological processing and its causal role in the acquisition of reading skills. Psychological Bulletin, 101, 192-212.
Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1999). Comprehensive test of phonological processing. Austin, TX: Pro-Ed.
Whitehurst, G. J., & Lonigan, C. J. (2001). Emergent literacy: Development from prereaders to readers. In S. B. Neuman & D. K. Dickinson (Eds.), Handbook of early literacy research (Vol. 1, pp. 11-29). New York, NY: Guilford.
Wolf, M., & Bowers, P. G. (1999). The double-deficit hypothesis for the developmental dyslexias. Journal of Educational Psychology, 91, 415-438.
Woodcock, R. W., McGrew, K. S., & Mather, N. (2001). Woodcock-Johnson III. Itasca, IL: Riverside Publishing.
Gordon E. Taub and Judit Szente
University of Central Florida, Orlando, Florida
Submitted September 8, 2010; accepted May 9, 2011.
Address correspondence to Gordon E. Taub, University of Central Florida, P.O. Box 161250, Orlando, FL 32816. E-mail: email@example.com
TABLE 1 Descriptive Statistics From the C-TOPP and WJ-III ACH M SD Range C-TOPP Segmenting Nonwords 8.36 1.84 3-13 Blending Nonwords 8.45 2.13 4-14 Letter Naming 10.62 2.08 5-16 Digit Naming 10.73 2.10 5-18 WJ III ACH Reading fluency 102.88 11.092 76-148 Note. C-TOPP = Comprehensive Test of Phonological Processing; WJ III ACH = Woodcock-Johnson III: Tests of Achievement. N = 86. TABLE 2 Correlations Between the Independent and Dependent Variables Segmenting Blending Digit Letter Blending .467 ** Digit .163 .044 Letter .086 .016 .805 ** Reading .199 .263 * .403 ** .413 ** Note. Digit = digit naming; Letter = letter naming; Reading = reading fluency. N = 86. * p<.05. ** p<.01. TABLE 3 Fit Statistics for Two Structural Equation Models Model [chi square] df RMSEA CFI Model 1 5.55 5 .036 .996 Model 2 6.00 5 .049 .992 Note. RMSEA = root mean square error of approximation; CFI = comparative fit index. * p < .05. TABLE 4 Standardized Direct Effects of Two Structural Equation Models Standardized Effects Model 1 Model 2 To PA From blending .88 .39 From segmenting .22 .22 From RAN .85 To RAN From letter .58 .75 From digit .47 .30 To RF From PA .24 .47 From RAN .41 Note. PA = phonological awareness; RAN = rapid automatized naming; Letter = letter naming; Digit = digit naming; RF = reading fluency.
Please note: Some tables or figures were omitted from this article.
|Printer friendly Cite/link Email Feedback|
|Author:||Taub, Gordon E.; Szente, Judit|
|Publication:||Journal of Research in Childhood Education|
|Date:||Oct 1, 2012|
|Previous Article:||Storybook-induced arousal and preschoolers' empathic understanding of negative affect in self, others, and animals in stories.|
|Next Article:||Assessing the phonological skills of bilingual children from preschool through kindergarten: developmental progression and cross-language transfer.|