Abstract
We evaluated the narrative and expository writing samples of 12 adolescents with language-learning disabilities (LLD) in Grades 6 to 12 for elements of microstructure (e.g., productivity, grammatical complexity) and macrostructure (genre-specific text structure elements) using an experimental measure. Writing samples were elicited with genre-specific prompts via paper and pencil and transcribed according to Systematic Analysis of Language Transcripts (SALT) conventions. Wilcoxon signed ranks tests indicate that levels of productivity and grammatical complexity were significantly greater in the narrative genre than in the expository genre. However, participants’ writing samples demonstrated equally impoverished text structure for both genres. Positive correlations were found between microstructure and macrostructure performance. Findings confirm the effects of discourse genre on measures of microstructure and further elucidate the use of microstructure and macrostructure elements in the writing of adolescents with LLD. Future research, comprehensive writing assessment, and interventionists should consider direct measurement of both microstructure and macrostructure components across genres for this population.
The ability to write in multiple forms and styles, or discourse genres (Graham & Harris, 2003), is linked to academic success (Berman & Nir-Sagiv, 2007; Scott, 2009; Singer, 2007). Knowledge of discourse genres is typically acquired in a developmental progression and is related to reading comprehension and writing achievement (Berninger, Garcia, & Abbott, 2009; Englert & Thomas, 1987). Narrative and expository genres are the two most commonly encountered genres in the school environment. Narrative discourse involves telling a story, often about personal events or other life experiences (e.g., novels, personal letters, and short stories). Awareness and use of narrative discourse in written language begin to develop early in life, often through storytelling experiences (Liles, 1993; Nelson, Bahr, & Van Meter, 2004). In contrast to narrative discourse, expository discourse involves conveying facts or describing procedures, sharing basic information, and relating cause–effect relationships (e.g., essays). The structure of expository discourse is typically introduced later in the school years and follows a longer path to achieve mastery of complex expository structures; as a consequence, expository text is more difficult to produce and comprehend for many students (Berman & Verhoeven, 2002).
Writing Challenges of Students With Language-Learning Disabilities (LLD)
Previous work highlights the difficulties that students with LLD experience in written expression (Mackie & Dockrell, 2004; Scott & Windsor, 2000; Singer & Bashir, 2004). Among the high cognitive demands facing these students in the writing process is the need to consider the global requirements of a particular genre (i.e., macrostructure), while negotiating the semantic and syntactic demands of the language (i.e., microstructure). Multiple linguistic skills must be coordinated to generate a written product, which can be problematic for students with LLD when the language system becomes overloaded due to deficiencies in one or more underlying areas (e.g., phonological, morphological, and orthographic awareness, and lexical and syntactic processes; Berninger et al., 2009; Mackie & Dockrell, 2004; Singer & Bashir, 2004).
Previous investigations of the written compositions of students with LLD have focused on microstructure and/or macrostructure in either narrative (Fey, Catts, Proctor-Williams, Tomblin, & Zhang, 2004; Gillam & Johnston, 1992; Mackie & Dockrell, 2004; McFadden & Gillam, 1996) or expository writing samples (Puranik, Lombardino, & Altmann, 2007). For example, in a study designed to compare analytic measures of complexity to holistic judgments of imaginative narrative discourse quality, McFadden and Gillam examined multiple elements of narrative microstructure and macrostructure. The overall quality (e.g., holistic score) of spoken and written imaginative narratives produced by students with LLD was lower than that of their age-matched peers but similar to their language- and reading-matched peers. The researchers concluded that the structural deficits evident in the narrative discourse of students with LLD may have negatively influenced judgments of quality. McFadden and Gillam’s findings are applicable to imaginative narratives, a subtype of narrative genre; however, further investigation is warranted to examine similar aspects within additional subtypes of narrative and/or expository compositions. Additional measures of narrative macrostructure also would be beneficial to provide a more detailed analysis.
Puranik et al. (2007) examined levels of productivity, syntactic complexity, lexical diversity, and spelling accuracy in the expository writing of 15 students with LLD, 17 students with dyslexia, and 15 typically developing, chronological-age (CA)-matched peers with a text-retell procedure. Participants’ ages ranged from 11 to 21 years. The CA-matched group outperformed groups of students with LLD and dyslexia on all measures. Participants with dyslexia outperformed those with LLD on four measures (number of words, number of ideas, number of T-units, number of different words [NDW]) but were similar in terms of spelling accuracy and percentage of grammatically correct sentences. The researchers concluded that both groups were similar in language form (syntax/grammatical complexity, phonology) but differed in language content (semantics); the students with dyslexia as a group outperformed the students with LLD on measures of productivity, semantics, and lexical diversity. The findings of Puranik et al. provided initial insight into the differences in expository microstructure performance across two clinical groups and a comparison group of students; however, the study was limited to quantitative measurement of microstructure in one genre. It is equally important to examine the expository macrostructure performance levels and to inspect both microstructure and macrostructure performance levels across genres.
Two research teams have examined the written language skills of students with LLD concurrently in more than one discourse genre, although their analyses were limited to microstructure elements (Blair & Crump, 1984; Scott & Windsor, 2000). Blair and Crump sampled the writing performance of students with LLD from Grades 6, 8, and 10 across descriptive narrative and argumentative genres. The writing samples were analyzed for microstructure by examining grammatical complexity, sentence type, and syntactic patterns. Participants’ argumentative compositions contained more complex syntactic structures than their descriptive narrative compositions. Greater differences in grammatical complexity were found between genres in the compositions produced by students in the later grades (8th and 10th) than in those produced in the 6th grade, suggesting that the younger students with LLD struggled to make distinctions between the different genres. The researchers concluded that greater demands were placed on grammatical complexity in the argumentative genre than in the descriptive narrative genre because an argumentative genre requires a writer to express information at a higher level of abstraction. Writers, therefore, must use more complex structures to meet the demands of the argumentative purpose.
Scott and Windsor (2000) examined a sample of 20 adolescent students with LLD and compared the microstructure of their spoken and written discourse across two genres (narrative and expository) with that of 20 CA-matched and language-age-matched peers. The students were shown two videos that modeled either discourse genre, after which they were asked to provide written and spoken summaries in both narrative and expository forms. The spoken and written language samples were measured for productivity, fluency, lexical diversity, grammatical complexity, and grammatical accuracy. The results indicated that regardless of modality (spoken vs. written), expository summaries were shorter, less fluent, more error prone, and more complex syntactically than narrative summaries. Written summaries were shorter and more error prone than spoken ones. Further inspection of group means for trends indicated that the expository genre was especially challenging for the students with LLD and their language-age-matched peers in the written modality. The researchers concluded that discourse genre and modality clearly impacted language performance and that students with LLD were at a particular disadvantage for written language production due to underlying linguistic deficits.
Taken together, the findings of these studies reflect a need for researchers to further explore the use of micro- and macrostructure elements in both narrative and expository written texts by adolescent students with LLD. To this end, we developed and piloted an experimental measure to capture adolescents’ performance across genres. Specifically, there were two objectives for this study. First, we wanted to determine whether the adolescents’ writing differed across discourse genre (narrative vs. expository) in use of microstructure elements of productivity (number of words, T-units, and different words) and grammatical complexity (number of complex correct sentences [CC sentences], mean length of T-unit [MLT]), and macrostructure elements (genre-specific text structure elements; see Table 2). Second, we wanted to determine whether there is a correlation between use of microstructure elements and use of macrostructure elements in the narrative and expository genres.
We predicted that productivity, as measured by total number of words and T-units, would be lower in the expository genre but the NDW would be higher (Scott & Windsor, 2000). We expected that grammatical complexity (e.g., total number of correct, complex sentences) would be higher in the narrative genre. Past research (Scott & Windsor, 2000) suggests that due to less familiarity with the structure, students produce more syntactic errors in expository genres. However, we considered that a large number of grammatical errors could be present in both genres because grammatical errors often distinguish students with LLD (Gillam & Johnston, 1992; Scott & Windsor, 2000). MLT was expected to be similar in both the narrative and expository genres in the present sample, based on the findings of the cross-genre analyses of Scott and Windsor (2000). We expected that more genre-specific text structure elements would be present in the narrative compositions than those required in the expository compositions because of students’ greater prior experience with the narrative structure (Berman & Verhoeven, 2002). However, we also anticipated that the narrative compositions might reflect an incomplete set of narrative text structure elements because past research indicates that the writing of students with LLD demonstrate underdeveloped use of text structure and discourse schemas in general (Singer & Bashir, 2004). Because students with LLD are thought to experience difficulty with both local microstructure (Mackie & Dockrell, 2004; Puranik et al., 2007; Scott & Windsor, 2000) and global macrostructure levels of text production (Fey et al., 2004; Gillam & Johnston, 1992; McFadden & Gillam, 1996), we anticipated there would be significant correlations between outcomes on the microstructure and macrostructure measures.
Method
Participants
Twelve students in Grades 6 through 12 participated. Participants were previously identified via state education criteria as having a primary exceptionality of language impairment (LI) or specific learning disability (SLD) and were receiving special education services and completed the writing samples as part of an individual evaluation at a university clinic. Research samples of students with LLD typically include students with educational diagnoses such as LI or SLD (Nelson, Roth, & Van Meter, 2009). Participants had no known sensory or intellectual deficits and were monolingual English speakers. The mean age of the participants was 15 years 8 months, with ages ranging from 11 years 6 months to 19 years 5 months. The participants were enrolled in four middle schools and three high schools in a mid-sized city. One high school participant was enrolled in home-based instruction. The school populations varied in the percentage of students receiving free or reduced lunch rates (5%−59%). Participants included six females and six males. The sample included nine Caucasian students and three African American students.
For purposes of this study, LLD is used to denote adolescents with deficits in spoken and/or written language. Evaluations were conducted for the purpose of locating participants for a summer literacy intervention program for adolescents with LLD. To corroborate teacher and parent report of LLD, results of a comprehensive language and literacy evaluation and/or reports of participants’ performance on the state level reading achievement test were used. Each participant received a reading score on the state test below grade-level expectations or a standard score below the 16th percentile on one or more of the following norm- or criterion-referenced academic achievement measures administered as inclusion criteria: Basic Reading Skill cluster of the Woodcock Reading Mastery Test–Revised (WRMT-R; Woodcock, 1998), grade-level appropriate reading passages from the Bader Reading and Language Inventory–Fifth Edition (Bader, 2005), and the Test of Written Spelling–Fourth Edition (TWS-4; Larsen, Hammill, & Moats, 1999). Participants demonstrated average standard scores on the Matrices subtest of the Kaufman Brief Intelligence Test–Second Edition (KBIT-II; Kaufman & Kaufman, 2004). Table 1 shows the mean age and test scores for the participants. Testing occurred during the last month of the school year and took place at a university speech and hearing clinic. Parent/guardian consent was obtained for each participant via an informed consent process approved by the university’s institutional review board.
Mean Scores on Evaluation Tasks.
Note. TWS-4 = Test of Written Spelling-Fourth Edition; WRMT-R = Woodcock Reading Mastery Test-Revised; KBIT-II = Kaufman Brief Intelligence Test-Second Edition. Bader scores are reported in percent correct for comprehension questions.
Procedure
Two writing samples were collected per participant in an individual session at a university clinic via a paper and pencil task with a maximum of 30 minutes per composition. Each participant completed the task in the time allotted. Writing samples were elicited with a counterbalanced genre order. Participants selected from among three prompts per genre. For example, prompt choices included the following for narrative and expository genres, respectively:
“Think about an event in your life that you will always remember. Write a story about what happened.”
“Think about the career you would like to have when you finish school. Think about why you would like to have that career. Write a paper explaining what career you would like to have. Explain why this is what you would like to do.”
Each writing sample was transcribed into a computer database according to the Systematic Analysis of Language Transcripts (SALT, Version 8; J. F. Miller & Chapman, 2005) conventions. The unit of segmentation was the T-unit, as suggested by Nelson et al. (2004) and consistent with previous investigations (Puranik et al., 2007; Scott & Windsor, 2000). Each sample was coded and analyzed for microstructure elements. Specifically, we examined productivity through total words, total T-units, and NDW, values that SALT calculated automatically. Grammatical complexity was examined through sentence type (complex vs. simple, correct vs. incorrect), number of clauses, and grammatical errors. A simple sentence consisted of one main clause and only one verb, whereas a complex sentence included one main clause plus one or more embedded/subordinate clauses, two main clauses, or one main clause and verb phrase joined by a coordinating conjunction. Grammatical errors were defined as errors occurring in verb tense or pronoun usage, agreement or case, omitted or incorrect inflection, omitted or substituted grammatical elements, and violated word order. A sentence without any grammatical errors was considered correct, whereas a sentence with one or more errors was deemed incorrect.
To analyze samples for genre-specific macrostructure, we created a checklist based on the common macrostructure elements in the 6th- through 12th-grade-level curricular standards for language arts. Sufficient information to establish the narrative and expository elements was required to receive a score of “1,” whereas samples without the designated elements were assigned a score of “0.” Operational definitions (see Table 2) regarding the use of those elements were adapted from items on the Informal Writing Inventory (Singer & Bashir, 2002), a measure designed to allow for both qualitative and quantitative writing assessment at any grade level and for use with a variety of genres.
Macrostructure Features Analyzed by Genre.
For measures not calculated by SALT, a subset of four narrative and four expository transcripts, 33% of the sample, was randomly selected for reliability checks performed by a trained graduate research assistant. Interrater agreement (point-by-point agreement) was 88% for segmentation of T-units, 86% for coding of number of clauses, 83% for coding of sentence type (simple/complex, correct/incorrect), and 80% for coding of presence of genre-specific text elements.
Results
Microstructure
A series of five Wilcoxon signed ranks tests were conducted to evaluate whether narrative or expository compositions showed greater total words, total T-units, NDW, number of CC sentences, or MLT. Nonparametric Wilcoxon’s tests were chosen due to nonnormal distribution. Descriptive statistics are displayed in Table 3, whereas the inferential statistics are reported herein. Effect sizes were calculated by converting the z scores into effect size estimates (r) and were interpreted as small (r = .10), medium (r = .30), and large (r = .50) effects.
Descriptive Statistics for Microstructure Measures.
Note. NDW = number of different words; CC sentences = complex correct sentences; MLT = mean length of T-unit.
The results indicated that there was a significant difference between narrative and expository compositions for total number of words with a large effect size, z = −2.55, p < .01, r = −.52. The number of words was significantly higher in the narrative (mean rank = 7.15) than in the expository compositions (mean rank = 3.25). There also was a significant difference between narrative and expository compositions in terms of the NDW used, with a large effect size, z = −2.59, p < .01, r = −.53. Narrative compositions generally included greater lexical diversity (mean rank = 8.00) than the expository compositions (mean rank = 2.00). In addition, a significant difference was found between the numbers of complex correct sentences used across genres with a large effect size, z = −2.40, p < .01, r = −.50. Narrative compositions generally included a greater number of CC sentences (mean rank = 6.95) than the expository compositions did (mean rank = 4.25).
There were no significant differences between narrative and expository compositions in total number of T-units with a medium effect size, z = −1.74, p = .04, r = −.35, or MLT with a small effect size, z = −1.20, p = .12, r = −.24. Further inspection of the results for trends revealed that 9 of the 12 participants had a higher number of T-units in the narrative than in the expository genre. Seven of the 12 participants demonstrated a higher score for MLT in the narrative genre than in the expository genre. The mean rank for narrative compositions with a higher MLT was 7.71, whereas the mean rank for expository compositions with a larger MLT was 4.80.
Macrostructure
We determined the number of macrostructure elements present in each sample across genres. The results indicated the mean number of genre-specific text structure elements evident in the writing samples was 2.5 for both genres. Results of a Wilcoxon signed ranks test confirmed that there was no significant difference between genres for the number of text structure elements used, with a trivial effect size, z = −.14, p = .44, r = −.03. On average, the students included half (i.e., 2.5 of 5) of the required genre-specific text structure elements when writing in either genre. Further analysis was completed to measure the frequency with which individual elements were included by the participants. The narrative macrostructure element most frequently incorporated was sensory detail (9), followed by characters (7), logical sequence (6), plot (5), and context (3). The expository macrostructure element most frequently incorporated was an introduction (11), followed by body (9), conclusion (5), logical sequence (4), and information that adequately addressed the assignment (2). Only two participants produced all expected elements in both genres.
Relations Among Microstructure and Macrostructure Elements
Two-tailed Spearman rank correlations (Table 4) were calculated between the text structure elements used and the five microstructure measures (corrected for Type I error via Bonferroni correction; p < .008). Positive correlations were found between microstructure measures of productivity and grammatical complexity and macrostructure measures of text structure for both genres (correlation coefficients, ρ values, ranged from .73 to .86). Correlations between the number of narrative text structure elements used and two measures of productivity (number of words, NDW) were strong, ρ = .73 and .77, respectively (p < .008). The correlation with a third productivity measure (total T-units) also was strong (ρ = .70, p = .014) but not statistically significant after correction for Type I error. The correlation between the number of narrative text structure elements used and one measure of grammatical complexity (number of CC sentences) was strong and significant (ρ = .75, p < .008); however, the correlation with MLT was weak and not significant (ρ = .22, p = .50).
Spearman (ρ) Correlations for Microstructure and Macrostructure Measures.
Note. NDW = number of different words; CC sentences = complex correct sentences; MLT = mean length of T-unit; TE = text structure elements. Narrative values listed first, then expository.
p < .008.
We found similar results for the expository genre. Correlations between the number of expository text structures used and two measures of productivity (number of words, number of T-units) were strong, ρ = .83 and .86, respectively (p < .008). For a third productivity measure (NDW), the correlation was moderately strong (ρ = .67, p = .02) but not statistically significant at the corrected significance level. The correlation between the number of expository text structure elements used and number of CC sentences (grammatical complexity measure) was significant and strong (ρ = .86, p < .008). However, there was little to no correlation with MLT (ρ = −.02, p = .96).
Discussion
In summary, the writing of adolescents with LLD exhibited significantly greater productivity and grammatical complexity in the narrative genre when compared with the expository genre. Participants, on average, included the same number of text structure elements in both genres. However, their writing samples reflected a range of individual text structure elements. For both genres, two measures of productivity and one measure of grammatical complexity were significantly and positively correlated with use of genre-specific text structure elements. Therefore, three of the five microstructure measures were strongly and significantly related to the macrostructure measures of text structure elements. The results of this study confirm and expand on previous comparative reports of the cross-genre writing performance of adolescents with LLD in a number of ways.
Microstructure
The finding of greater productivity in the narrative genre confirmed our expectations and corroborated results from previous studies citing higher productivity in the narrative genre for students with and without LLD (Berman & Verhoeven, 2002; Scott & Windsor, 2000). Previous narrative experience may influence productivity levels. Children begin to narrate when they are as young as 2 to 3 years of age and continually develop their narrative abilities throughout their school years (Liles, 1993). In contrast, expository discourse is not typically mastered until well into adolescence (Berman & Nir-Sagiv, 2007). Participants’ prior experience was not directly measured. However, it is likely that our participants possessed more prior experience with narrative discourse than expository. Furthermore, narrative and expository discourses vary in purpose and context. Previous researchers have suggested that the organization and structure of expository text is more difficult to master and produce, even within students exhibiting typical language development (Berman & Nir-Sagiv, 2007; Singer & Bashir, 2004). Thus, greater prior experience with narrative discourse, as well as the differing demands of narrative and expository macrostructure, likely account for why narrative productivity was higher in this sample.
There were no differences across genres for MLT, consistent with expectations and the findings of Scott and Windsor (2000) but inconsistent with the results of Blair and Crump (1984). The differences among these studies may be due to the compositions elicited. Like Scott and Windsor, we elicited narrative and expository samples. In contrast, Blair and Crump elicited an argumentative genre structure. If we consider that the text structure used in the narrative genre is more similar to that used in the expository genre than the argumentative genre, then it is possible that MLT is more sensitive to differences between genres that are maximally different in structure. A second possibility is that the argumentative genre requires the writer to assume another’s perspective or opposing point of view. Such a task is cognitively linguistically demanding and can be especially challenging for students with LLD who may experience difficulties with theory of mind (see S. Miller, 2004). There is a need to further investigate the utility of MLT to detect differences in grammatical complexity across genres.
A significantly higher number of CC sentences occurred in the narrative genre, as expected. In addition, the participants demonstrated a lower level of syntactic complexity and a higher grammatical error rate when writing in the expository genre. Similar to results for increased narrative productivity, this finding of greater narrative grammatical complexity was likely due to the influence of prior narrative experiences. That is, the demands for narrative text structure may have been offset due to increased familiarity. Our results for the increased number of complex sentences within the narrative genre were similar to the outcomes of the Scott and Windsor (2000) study but differ from those of Blair and Crump (1984). As noted earlier, increased demands on grammatical complexity may occur in the argumentative genre because it involves a higher level of abstraction. Thus, the differences in findings may be due to the differences in type of expository genres sampled. Future investigations should examine the effect of the expository subtype on the production of specific microstructure and macrostructure elements.
Macrostructure
As anticipated, participants included an incomplete set of macrostructure elements within their narrative writing. However, in contrast to expectations, participants included an equally low number of elements in their narrative and expository compositions. Our results provide preliminary evidence that students with LLD are including only a minimal number of the required text structure elements, a problem which extends beyond the narrative genre into the expository genre. This finding is unique among studies examining cross-genre writing performance and contrasts with developmental expectations that adolescents produce a higher number of narrative elements because of increased experiences with or exposure to narratives (Berman & Verhoeven, 2002; Scott & Windsor, 2000; Singer, 2007). Regardless of levels of productivity and grammatical complexity, a limited text structure, regardless of genre, is particularly problematic as it greatly impacts academic performance.
Relation of Microstructure and Macrostructure
Consistent with expectations, some microstructure elements were significantly and strongly correlated with macrostructure elements for each genre. Positive correlations were found between productivity and grammatical complexity and text structure for both genres (ρ values ranged from .73 to .86). One possible explanation for these findings is that the microstructure measures for both genres were related to composition length (i.e., productivity), and, therefore, longer compositions offered more opportunities for various elements of macrostructure to be incorporated.
Our findings are unique in that we documented specifically the relation between microstructure and macrostructure element use in students with LLD. In the past, some researchers have shown that students with LLD experience difficulty with the microstructure elements of writing (Mackie & Dockrell, 2004; Puranik et al., 2007; Scott & Windsor, 2000), as well as genre-specific macrostructure elements of writing (Englert, Raphael, Anderson, Gregg, & Anthony, 1989; Fey et al., 2004; Gillam & Johnston, 1992; McFadden & Gillam, 1996). By documenting difficulties in both microstructure and macrostructure elements simultaneously, we are better able to capture a broader picture of the pervasive nature of writing deficits in this population and apply the findings directly to intervention.
Methodological Considerations
Several methodological factors warrant attention when interpreting the results. First, this investigation involved a relatively small number of participants who came from different schools and presented with a range of writing skills. We believe our sample is not unlike other clinical samples with a degree of heterogeneity among students in this population. Even so, subsequent investigations with larger samples may be able to detect additional significant differences within and across genres. Nested designs in particular would allow direct comparisons within and across schools. Second, a comparison group was not included. However, given the current knowledge of writing development among adolescents exhibiting typical written language abilities (e.g., Berman & Nir-Sagiv, 2007; Berman & Verhoeven, 2002), it appears that the writing products of the adolescent participants in our study were more deficient of key linguistic components (e.g., microstructure and macrostructure features) than adolescents of typical language skills of comparable ages. Third, given the somewhat broad age range of our participants, differences due to developmental abilities may have been masked by placing them into a single group. However, inspection of the data suggests that some middle school students outperformed some of the high school students. Nevertheless, future research should include narrower age ranges when examining the writing skills of students with LLD. Fourth, we elicited one writing sample per genre. Multiple samples would help establish a more consistent pattern of performance. Sampling in a variety of contexts and expository tasks will assist in better identifying relative strengths and weaknesses, lead to increased generalizations about how students with LLD may perform in different situations (Singer, 2007), and help document the range of writing abilities that may exist in individual participants.
Conclusion
Writing assessment among adolescents with LLD can be a daunting task. Based on the current findings, writing assessment should include measures of microstructure as well as macrostructure elicited in multiple discourse contexts (e.g., narrative, expository). The assessment methods used in this study can be utilized by educational teams to obtain a more detailed measure of a student’s writing proficiency than with a single holistic writing score, and can be administered for purposes of evaluation, assessment, preintervention baseline, or to monitor progress over time. To provide effective writing interventions and instruction, professionals must capture a thorough understanding of students’ written language deficits and problems that underlie them, such as difficulty incorporating a variety of linguistic microstructure and macrostructure elements.
There is a pressing need for additional writing research with students with LLD. This pilot investigation represents one attempt to highlight and compare the cross-genre writing abilities of adolescents with LLD using an experimental measure. We plan to extend this work to develop more fine-grained macrostructure assessment across genres and monitoring tools sensitive to changes occurring over the course of an instructional and/or intervention period. With a better understanding of the contributions of the microstructure and macrostructure elements used in these students’ compositions, it may be that future interventions designed specifically with those components in mind will be successful in increasing the written language abilities of this population.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.