Abstract
English learners (ELs) with reading disabilities (RDs) have been among the lowest performers on academic achievement tests that assess vocabulary. To meet academic demands and prepare for college or careers, ELs with RDs clearly need support in terms of vocabulary acquisition; however, relevant research is scarce. This study investigated the effects of the CLUES strategy, a generative vocabulary strategy, on the ability of students to analyze and define unknown science terms relating to biology. The study further evaluated students’ ability to maintain vocabulary gains over time and to generalize the CLUES strategy. The results showed that participants benefited from the use of the CLUES approach to define novel science terms and maintain their ability to use this strategy over time. The participants themselves generally expressed satisfaction with the CLUES strategy and recommended it to their peers.
Students preparing for college or a career must be able to read and write proficiently (Miller, 2009), but in fact more than 8 million secondary students in the United States read below the proficiency levels necessary for positive postsecondary outcomes (Kamil, 2008). Particular groups of students, such as English learners (ELs) with reading disabilities (RDs) have been more vulnerable to reading difficulties. Results from National Assessment of Educational Progress (NAEP, 2017) showed that ELs with RDs have the least favorable academic achievement outcomes compared with ELs without disabilities and their counterparts with learning disabilities (LDs) who speak fluent English; they also are more likely to be further behind in terms of vocabulary acquisition (Carlo et al., 2004; NAEP, 2017).
ELs with RDs have vocabulary deficits for several reasons. To begin with, students with RDs, including ELs with RDs, have core linguistic-related deficits in phonological awareness (e.g., Caravolas et al., 2005; Lonigan et al., 2000; Muter et al., 2004; Wagner et al., 1994) that can compromise word-reading efficiency. Poorly developed word identification skills can further impede vocabulary acquisition in those students who are less likely to attend to unfamiliar words, sound them out accurately, and encode them for future usage (Moats & Tolman, 2009). The vocabulary deficit of ELs with RDs is often further exacerbated by deficits in the working memory in which phonological information is stored (Swanson et al., 2012) that can cause difficulties in learning and specifically in transforming the unfamiliar phonological structures of new words into permanent memories (Baddeley et al.,1998; Chiappe et al., 2002).
Because the extent and depth of vocabulary knowledge is strongly related to the ability to comprehend unfamiliar texts (Garcia, 1991), a limited vocabulary is for ELs, especially those also with RDs, the greatest impediment to academic reading success (Carlo et al., 2004; Proctor et al., 2005). Researchers over the past two decades have accordingly sought to identify effective vocabulary interventions for ELs with or without RDs (e.g., Graves et al., 2012; Swanson et al., 2012). Relatively, little of this body of work, however, has specifically targeted ELs with RDs. In fact, only 11 studies have addressed the vocabulary needs of ELs and only one of which for ELs with RDs (August et al., 2014, 2016; Carlo et al., 2004; Helman et al., 2015; Hwang et al., 2015; Proctor et al., 2007, 2009; Snow et al., 2009; Townsend & Collins, 2009; Vaughn et al., 2009). The vocabulary instruction strategies described in these studies include context clue analysis, morphemic analysis, priming prior knowledge, repeated exposure to targeted words, and the use of external references. These strategies represent one of two broad approaches to vocabulary instruction, namely, non-generative and generative.
Non-generative vocabulary interventions teach students the meaning of individual target words with the aid of some sort of device (Harris et al., 2011). One approach involves having students memorize the definition of each target word through rote repetition (e.g., Barrett & Graves, 1981; Beck et al., 1982; Leong et al., 1990; McKeown et al., 1983, 1985). In another, students relate new words to known words through mnemonic devices (Terrill et al., 2004) or semantic webs (Johnson & Pearson, 1984; Pearson & Johnson, 1978; Pearson & Spiro, 1982; Sinatra et al., 1984). These non-generative approaches have been in common use as vocabulary interventions for students with RDs, and they have been shown to be effective in helping students to learn targeted new words (Bryant et al., 2003; Elleman et al., 2009; Jitendra et al., 2004). There are, however, concerns that non-generative strategies alone are insufficient to bridge the vocabulary gap of ELs with RDs and that this approach does not represent the most efficient use of instructional time. Moreover, at the secondary level, where many novel words in content area readings are complex and polysemous, the non-generative approach is problematic owing to these students’ difficulty memorizing words and generalizing this knowledge (Beck et al., 2002; Graves, 2006; Nagy & Scott, 2000; Nash & Snowling, 2006).
The other main approach to vocabulary acquisition involves generative strategies, which teach students how to use contextual and linguistic cues to derive the meaning of new words (Nagy et al., 2006). These strategies help students become more independent word learners by bolstering their skills at analyzing the linguistic structure of individual words and texts in ways that generalize to novel occurrences. Generative approach usually takes the form of contextual analysis (e.g., August & Shanahan, 2010; Carlo et al., 2004; Proctor et al., 2005), morphemic analysis (e.g., Harris et al., 2011; Katz & Carlisle, 2009; Kieffer & Lesaux, 2008; Tomeste & Arnoutse, 1998), cognate knowledge (e.g., Garcia & Nagy, 1993; Rodriguez, 2001), or combined strategies (Bauman et al., 2002; Baumann et al., 2003; Carlo et al., 2004; Helman et al., 2015).
Contextual analysis involves the use of clues in a text to derive the meanings of a word (e.g., Baumann et al., 2003). This type of analysis can occur incidentally or be taught explicitly, though it has been suggested that students, particularly those with RDs, have better vocabulary outcomes with the latter (Carlo et al., 2004; Ebbers & Denton, 2008; Nagy & Scott, 2000). Through explicit instruction, students learn to use semantic clues in a given context to identify synonyms, antonyms, syntax, and definitional examples that surround an unknown word to infer its meanings (Kuhn & Stahl, 1998). Research has demonstrated that that deriving the meaning of a word from its written context is a very important means of increasing students’ vocabulary (Fukkink et al., 2001). However, the effectiveness of explicit instruction in promoting students’ ability to use contextual analysis remains relatively unexplored (National Reading Panel, 2000).
Another type of generative vocabulary strategy is morphemic analysis, which involves deriving the meaning of a word by combining morphemes such as prefixes, suffixes, and roots (Nagy & Scott, 2000). Specifically, words are broken down into their morphemic parts, meaning is connected with these morphemes, and then the definition of the whole word is deduced (Nation, 1990, 2008; Wei & Nation, 2013). Several literature reviews assessing the effectiveness of morphemic analysis interventions have suggested that this strategy promotes literacy achievement for struggling readers but also that it is more effective in combination with other literacy strategies (Bowers et al., 2010; Goodwin & Ahn, 2010; Reed, 2008).
The third strategy involves cognate knowledge (Klingner et al., 2015). A cognate is “a word in one language which is similar in form and meaning to a word in another language because both languages are related” (Richards & Schmidt, 2002, p. 829). The idea is that, because cognates are semantically related and share phonological and/or orthographic forms, they can facilitate vocabulary building. Research into cognate knowledge has focused on what is referred to as the facilitation effect (e.g., Bravo et al., 2007; Pérez et al., 2010). Although research has suggested that cognates can facilitate the transfer of vocabulary from students’ first languages to English, when ELs’ first language has a shared etymological origin with English, cognate knowledge has remained underutilized as an intervention (Carlo et al., 2004).
Because the generative vocabulary strategies reviewed above each seem to focus on specific aspects of reading (i.e., text level or word level), the strategies can be flexibly combined. Three studies were found to combine the use of both contexture clues and morphemic analysis (Bauman et al., 2002; Baumann et al., 2003; Helman et al., 2015) and one utilized all three strategies (Carlo et al., 2004). All studies explicitly taught students how to find context clues and word-part clues in their interventions, whereas Carlo et al. (2004) also provided explicit instruction in cognate knowledge.
Positive student outcomes showed in studies reviewed above suggested that generative vocabulary strategies and the combined use of them seem to be promising in solidifying and increasing the vocabulary knowledge of students with RDs or ELs. However, relatively little research has targeted ELs with RDs, though these students need particularly intensive support when it comes to improving their literacy skills (August & Shanahan, 2010; Lesaux et al., 2010). In addition, it has been suggested that generative vocabulary strategies can be more efficient than non-generative strategies in terms of their easy application to novel words or text (Nagy et al., 2006). This being the case, there is a need to assess their generalizability. Generalizability of a vocabulary intervention can be critical for secondary-level ELs with RDs because vocabulary constitutes the greatest hindrance to the academic success of ELs with RDs (Proctor et al., 2007), and efficient interventions for them in this regard is imperative.
As such, this study was to extend the literature on generative vocabulary interventions by examining the effectiveness CLUES strategy for secondary ELs with RDs, which is an integrated generative vocabulary strategy, combining contextual analysis, morphemic analysis, and cognate knowledge. In addition, our study focused on science vocabulary because knowledge of science words is an important part of being an educated citizen in the increasing informational and technological world. Nevertheless, science textbooks include words that are challenging for many due to the information-dense and multi-morphemic terminology used to explain new and unfamiliar concepts (Harmon et al, 2005). Considering that students must understand approximately 90% to 95% of the words in a text to adequately comprehend (Nagy & Scott, 2000), difficulty in comprehension of science texts can be attributed, in part, to the high density of unfamiliar vocabulary. Therefore, it is especially important for all secondary students, but especially for ELs with RDs, to acquire strategies to help them in content areas such as science (Lee, 2005). The study addressed two research questions: (a) To what extent did the CLUES strategy promote the accuracy with which ELs with RDs identify the parts of science terms, assign meanings to these parts, and determine the meaning of the entire term? And (b) To what extent did ELs with RDs generalize the CLUES strategy to novel science terms within sentences embedded in scientific texts?
Method
Participants and Setting
Participants in this study were drawn from an urban high school in the eastern United States. The ethnic breakdown of the students in the school was 57% Caucasian, 28% Latino, 11% Black, and 4% Asian; 38% were eligible to receive free or reduced lunch. The school’s English speakers of other languages (ESOL) teacher was asked to refer students who met the following criteria: (a) attending ninth or 10th grade; (b) functioning at Level 3 or 4 language proficiency levels based on the World-Class Instructional Design and Assessment Consortium’s English Language Proficiency Standards (WIDA ELP Standards, 2007); (c) identified as an EL based on the school district’s language assessment; (d) having a specific learning disability in reading; (e) reading at a minimum fourth-grade level based on Read 180 Scholastic Reading Inventory (SRI) Lexile scores (Scholastic Reading Inventory; Davidson & Miller, 2002), to control for basic decoding ability; and (e) scoring at or above the fourth-grade level on all subtests of the Test of Reading Comprehension (TORC-4; Brown et al., 2009). Four students, Tamara, Narcisa, Victor, and Sarita (all pseudonyms) met above criteria and were included in the study; their demographic information is presented in Table 1.
Description of Study Participants and Screening Criteria.
Note. WIDA = World-Class Instructional Design and Assessment; RDs = reading disabilities; SRI = Scholastic Reading Inventory; EL = English Learners; PreLAS = Preschool Language Assessment Scale.
All participants took a biology course in the previous year. They were fully included in general education classrooms and received literacy instruction in the ESOL classroom. The study occurred during the 80-min ESOL literacy instruction periods. The first author, a certified special education teacher, served as the interventionist and provided all of the instruction. All sessions were audiotaped for treatment fidelity.
Science Terms Selection and Sentences Construction
Target ancient Greek and Latin roots and science terms selection
Fifty science terms, each comprised phonemes derived from common Greek and Latin roots, prefix, and/or suffix, were selected from biology and life science curricula (e.g., Campbell et al., 2003; Hasseler, 2005). To be more specific, every selected term included at least one high-frequency Greek or Latin root and a prefix and/or suffix; each science term contained a maximum of three syllables, and each morpheme of each word consisted of a maximum of six letters. The authors selected science terms with Spanish cognates whenever possible to facilitate recognition by the students during CLUES instruction and assessment.
Sentence construction
Sentences included a targeted science term and contextual clues to help define it for the intervention lessons and assessment. The target terms appeared alternately at the beginning, middle, or end of the sentences for variation (Beck et al., 2002). The sentences ranged from 15 to 20 syllables. A panel of experts consisting of three professors with expertise in English acquisition, linguistics, reading instruction, RDs, and science education reviewed the words and the constructed sentences for their appropriateness.
Measures
Dependent measure
Curriculum-based assessment probes, the CLUES probes, aligned with the objectives of the intervention served as the primary dependent measure during the baseline, intervention, and maintenance phases. Each CLUES probe consisted of a sentence including an unknown science term constructed in the manner just described and a CLUES organizer for the students to write out and define the morphemes and terms. In addition, students listed each CLUES step on blank lines provided on the probe. Data were collected for each student on: (a) correctly written morphemes, morpheme meanings, and the meaning of science term (Strategy Use [SU]) and (b) correctly written CLUES steps on the CLUES probes (Strategy Knowledge [SK]). A total SU score was calculated based on points earned for correctly listing and defining the morphemes and definitions of the target words. Each student’s SU score was then converted into a percentage. The number of possible points on the CLUES probe for the SU score ranged from 8 to 11 as follows: 2 to 3 points for each morpheme, depending on how many appeared in the selected term; 2 to 3 points for the correct meanings of the morphemes; 3 to 4 points for the correct meaning of a term; and 1 point for the dictionary definition.
The students’ SK was calculated using a distinct grading rubric on which they were awarded up to 15 points for identifying one or more of the five steps in the strategy, with 3 maximum points for each. Specifically, 3 points could be earned for providing the complete identifier for a given strategy step (e.g., “Connect to the Context”), 2 points for providing at least half of a step but omitting detail (e.g., “Connect Context”), or 1 point for providing less than half (e.g., “Connect”); and incorrect and blank responses were scored as 0. A total SK score was calculated and then was converted into a percentage.
Pre- and posttests
Word Knowledge Test (WKT), Morpheme Test (MT), and Strategy Knowledge Test (SKT) were used as the pre- and posttests in this study. The WKT (adapted from Harris et al., 2011) assessed students’ knowledge of 40 isolated science terms, of which 30 were taught in the intervention and 10 were novel. Three points were earned for each term, with one point each for providing correct morphemes, definitions of morphemes, and definitions of terms in their entirety. A total of 120 points was possible; percentages were thus calculated. The MT used in this study was adapted from Harris et al. (2011) to contain only morphemes (i.e., the roots, prefixes, and suffixes) relevant to science terms. The MT consisted of three sections—one each devoted to prefixes, roots, and suffixes—that contained a total of 57 morphemes, 47 taught in the intervention, and 10 were new to them. One point was awarded for each correct definition for a total possible 57 points. The percentage correct was then calculated. The SKT (adapted from Harris et al., 2011) assessed the rules taught during the training lessons regarding context, morphemes, prefixes, roots, and suffixes on the participants’ strategy knowledge. In its seven items, the students were asked to define context, morpheme, root, prefix, suffix, and the rules associated with each definition and to list the CLUES steps. Fifteen total points were possible based on the number of correct definitions and rules. The percentage correct was calculated. The forms of the SKT administered before and after intervention were distinct but equivalent.
Generalization probes
Two types of generalization probes were administered prior to and post the intervention to measure the extent to which students were able to generalize the strategy use to novel words. The first type, measuring near transfer, consisted of novel words embedded within a single sentence with CLUES graphic organizer provided. Students were asked to use the CLUES strategy to read each sentence, identify a key phrase to help define the science term, identify the morphemes in the term, provide the meaning of each morpheme, and then to define the word. They filled in the CLUES graphic organizer. The near-transfer generalization probes were constructed in like manner as were the CLUES probes. Each item was scored based on the correct identification of steps in the strategy and morphemes and meanings for a total possible 52 points.
The second type of generalization probe, measuring far transfer, contained only the novel words each was embedded in a single sentence but neither visual aids nor prompts for the CLUES steps was provided. Rather, blank spaces were provided for students to list the terms, prefixes, roots, and suffixes in novel scientific terms. The total points possible for each novel word ranged from 6 to 8 depending on the number of morphemes, with 1 point awarded for each correctly identified phoneme, 1 point for correctly defining morphemes, and 2 for defining an entire term. Percentage scores were calculated for the near-and far-transfer generalization probes.
Child intervention rating profile (CIRP)
The CIRP (adapted from Witt & Elliott, 1985) was individually administered to the participants following the intervention to assess their opinions of the CLUES strategy. Questions were adapted slightly to reflect the CLUES procedures, with modifications to the statements that asked participants to rate the fairness of the strategy and offer their opinions of it. The CIRP is an empirically validated tool that uses seven items to assess whether participants consider a given intervention to be acceptable (e.g., fairness, expected effectiveness, and possible negative consequences). For this study, the students rated each of the seven items on a 6-point Likert-type scale ranging from 1 (disagree) to 6 (agree). The total CIRP scores thus ranged from 7 to 42, with the higher scores indicating greater acceptability. The internal consistency reliability was .89. Additional space was provided on the CIRP for students to comment on and offer suggestions for improving the CLUES instruction.
Experimental Design
A multiple probe across participants design (Tawney & Gast, 1984) was employed to evaluate the effects of the CLUES intervention. Data collection followed procedures used for multiple baseline design. The intervention was implemented in staggered fashion to meet the systematic manipulation criterion (Kratochwill et al., 2013). Phase transition decisions from intervention to maintenance were based on achievement of a mastery level of 80%. Tau-U index was used to estimate intervention effect.
Procedures
Pretest and generalization tests
The pretest and generalization tests used in this study included the WKT, MT, SKT, and near-transfer and far-transfer generalization tests. The first author trained three graduate students on test administration. The pretest and generalization tests were administered prior to baseline data collection over a period of 1 week to each participant individually.
Baseline procedures
Either trained graduate students or the interventionist administered the initial five CLUES probes to each student at a frequency of every other day over a period of 2 weeks. Participants were instructed to do their best and were given unlimited time to complete the probes; all did so within 10 min. Visual inspection of graphed intervention probe data guided decisions regarding the point at which the intervention was administered to the various participants, in accordance with procedures of multiple baseline design. The staggered implementation of the intervention commenced when the first student’s baseline probe performance showed a stable trend. A stable trend was defined as a lack of variation in scores such that a line fitting the data in a given phase would not show improvement in performance prior to the intervention (Kratochwill et al., 2013).
Intervention procedures
Intervention lessons consisted of four 30-min training lessons over a 1-week period and 10 40-min CLUES lessons. Intervention lessons administered to each individual student four times per week during the intervention phase.
Training lessons
Four scripted training lessons were provided to ensure that the students had the prerequisite knowledge to benefit from the CLUES lesson. During the training lessons, the key terms, including context, morpheme, prefix, suffix, and root, were explained, and examples were provided; participants were also directed to explain their answers to the interventionist in an interactive manner. At the beginning of each lesson, the instructor and the students reviewed the information learned in previous lessons. Throughout the lessons, the interventionist provided immediate feedback whenever students answered incorrectly. Guided (i.e., teacher-led) practice followed the reviews, with the instructor defining a new term (e.g., context), explaining the relevant rules, and providing examples (e.g., by identifying context and key phrases in a sentence) as well as non-examples to ensure comprehension of the term; the students in the process recorded the information on a guided note sheet. Then, during independent practice, the students completed four examples. At the end of the lessons, the instructor reviewed current and previously learned terms.
CLUES lessons
There were 10 scripted CLUES lessons, which were administered immediately after the training lessons. In these intervention lessons, students learned to use the CLUES steps to identify and define the targeted science terms. These steps are as follows.
Step 1. Connect to the text: The student reads a sentence with an unknown science term aloud twice to initiate understanding of it.
Step 2. Label 2 clues: The student underlines clue words, cognates, or phrases that may help to define the unknown term.
Step 3. Use the clues to define the word parts: The student identifies words or phrases that relate to each part of the target word (e.g., prefix, root, and suffix) and then rereads the sentence to look for additional clues.
Step 4. Explain the word: The student uses the meanings of the parts of the term or to reread the sentence with a definition of the unknown term.
Step 5. See if you are correct: The student reads the definition of the term in a dictionary and reviews it with reference to the answer key organizer. Finally, the student compares the definition of the term with the definitions in the dictionary and on the answer key.
All of the intervention lessons incorporated instructional methods shown to be effective for students with LDs, including teacher modeling with thinking aloud, guided and independent practice opportunities, and encouragement of students’ use of their background knowledge (i.e., cognates), in particular building on their previous science courses. Students were also provided with pictures of the terms whenever possible.
In each lesson, the instructor began with thinking aloud, demonstrating how to navigate the CLUES organizer and use CLUES. Thus, for Steps 1 and 2, the instructor taught students how to re-read each sentence to identify key words, cognates, or phrases that could help to define an unknown scientific term. Next, for Steps 3 and 4, the instructor modeled segmentation of the term to identify morphemes, demonstrating how to use the key phrases to derive the meanings of the morphemes and of the entire term. Finally, at the Step 5, the instructor modeled the use of a dictionary, and students checked their answer keys to monitor their ability to list the CLUES steps correctly and define the morphemes and terms. During guided practice using a different science term, the instructor scaffolded instruction for the students as they applied the various CLUES steps and provided feedback. The instructor directed the students to complete the CLUES organizer as the students wrote their answers on the guided practice sheet. Then, for independent practice, the students on their own used the CLUES steps to define the morphemes and terms. Following the lesson, the instructor reviewed the steps and each of the three graphic organizers that listed the meanings of the newly learned science terms and their morphemes.
Participants received tangible incentives (e.g., candy or a pencil) once weekly, based on their attendance. Students also earned a ticket for the completion of a session, used toward a more expensive incentive of their choice (e.g., a gift certificate) awarded at the end of the study.
Maintenance procedures
Maintenance CLUES probes were administered to each participant at 2 weeks and again at 1 month following intervention. No review took place prior to the administration of these probes. As during the baseline and intervention phases, students had approximately 10 min to complete the probes.
Posttest and generalization tests
Trained graduate students administered the WKT, MT, SKT, and generalization tests to each participant in the week immediately following the intervention. The CIRP was given only at posttest by trained graduate students who were not involved in the intervention.
Treatment Fidelity and Inter-Scorer Agreement
Treatment fidelity was assessed during the implementation of training lessons and the CLUES instruction. The fidelity checklist for implementation of CLUES, created for this study, included 20 of the critical components of instruction. Two trained graduate students independently listened to audiotapes of all of the instructional sessions and assigned one point for each step implemented correctly. The percentage of correct steps was calculated by dividing correct steps by total steps as assessed by the observers. The percentage of correct steps was above 95% in all cases. Interobserver agreement (IOA) data were collected for 30% of the sessions, and a percentage score for IOA was calculated for the instruction by dividing the number of agreements on steps implemented correctly by the combined number of agreements and disagreements and then multiplying by 100. Mean IOA of CLUES intervention lesson implementation was 93% (range = 90%–100%).
The inter-scorer reliability of the assessments was conducted by two graduate students, who independently scored all of the tests using a checklist created by the instructor. The scored tests were compared item by item to tabulate the number of agreements and disagreements. To calculate the percentage of agreement, the number of agreements was divided by the combined number of agreements and disagreements and multiplied by 100. The scorers discussed any disagreements and achieved a consensus before the data were graphed. The overall agreement for the probes, generalization tests, and pre- and posttests was 98%.
Results
CLUES Strategy Probes
Figure 1 shows each participant’s percentage of correct SK and SU on the CLUES probes. During baseline, the mean percentages of SU and of the SK for the four participants were 0. The data were stable for all participants during the baseline phase, showing no variability. After introduction of the CLUES intervention phase, an immediate change in SU was apparent for all participants. An immediate change in level and an accelerating upward trend can be seen in Tamara’s SU data, with a mean of 87.1% (range = 63%–100%) and 100% nonoverlapping data points. For Narcisa, the mean was 82.4% (range = 40%–100%) with 100% nonoverlapping data; her performances showed a slightly decreasing trend during Sessions 4 and 5, but she had an SU score of 100% on Session 6; her score dipped slightly after an absence following Session 7 but increased to 100% mastery during Session 9, with a slight downward trend in the morpheme data during Sessions 6, 7, and 8. Victor’s SU showed a change in level and an upward trend, with a mean of 70.3% (range = 13%–100%) and 100% nonoverlapping data; he was absent during Sessions 7 and 8, but his SU improved to 88% on Sessions 9 and 10. An immediate change in level can also be seen in Sarita’s SU, with a mean of 90% (range = 62%–100%) and 100% nonoverlapping data; she was frequently absent between Sessions 1 through 5, but her morpheme data remained above the baseline level throughout the intervention. The overall Tau-U for the participants on SU was 1.0, indicating a large effect for the intervention package (Parker & Vannest, 2009). The maintenance data for all four students indicated treatment gains for SU, with their performances in every case exceeding the baseline.
Percentage of strategy knowledge, strategy step, and generalization probe.
An immediate change in SK was also apparent for all four participants. This change was accompanied by an upward trend in Tamara’s SK data, with a mean of 88.3% (range = 60%–100%) and 100% nonoverlapping data. Similarly for Narcisa, introduction of the CLUES intervention resulted in an immediate change in level compared with the baseline, and an upward accelerating trend can be seen in her ability to recall the steps of the strategy, with a mean of 83.4% (range = 73%–100%) and 100% nonoverlapping data. Victor’s SK data also showed a change in level and an upward trend with a mean of 88% (range = 47%–100%) and 100% nonoverlapping data; his SK data showed very little variability with the exception of the periods between Sessions 4 and 5 and Sessions 7 and 8. Following the introduction of the CLUES intervention phase, an immediate change in level from the baseline to the intervention was apparent in Sarita’s ability to recall the CLUES steps, with a mean of 72.4% (range = 62%–80%); despite her frequent absences, her SK remained stable. The overall Tau-U on SK for the participants was 1.0, also indicating a large effect for the intervention package (Parker & Vannest, 2009; see Table 2). The maintenance data for all four students exceeded their baseline performances indicated that the treatment gains were persistent.
Pre- and Posttest WKT, MT, and SKT Percentage Scores.
Note. WKT = Word Knowledge Test; MT = Morpheme Test; SKT = Strategy Knowledge Test.
Pretests, Posttests, and Generalization Probes
Tables 2 and 3 display the data from the pretest, posttest, and generalization tests for all of the participants. Each improved in some respects in their performances on the WKT, MT, and SKT tests, indicating gains in knowledge in terms of using the CLUES strategy to identify and define morphemes and to define scientific terms. Each also performed better on both the near-transfer and far-transfer generalization probes.
Percentage Scores for Generalization Tests.
Social Validity
Students’ satisfaction with the intervention was relatively high as measured by the adapted CIRP, the scores for Tamara, Narcisa, Victor, and Sarita being 36, 38, 41, and 42 (out of 42 total point), respectively. All four rated CLUES favorably, including in terms of the effectiveness of the intervention. In the words of one, the intervention “helped me learn words.” All recommended that CLUES be used by their peers.
Discussion
This study was to investigate the effects of a generative vocabulary strategy that combines contextual, morphemic analysis, and cognate knowledge on the acquisition of science vocabulary within a small sample of ELs with RDs. It contributes to the limited literature that has addressed these students’ vocabulary acquisition by demonstrating that the CLUES strategy can be effective as a generative tool for using contextual cues and morphemes to define and learn science terms. All participants in the study made significant gains with the CLUES probes in terms of both SU and SK. Several factors seem to have contributed to the effectiveness of the CLUES strategy. First, this study combined a variety of vocabulary approaches, including the use of cognates, to facilitate the association of known Spanish words with unknown root and a prefix and/or suffix, thereby building upon students’ prior knowledge as they encounter novel science term, as well as the use of external aids such as graphic organizers, pictures, and dictionary consultation. All of these approaches can be thought of as additions to students’ toolkits to be employed when they encounter novel scientific terms. Second, explicit vocabulary instruction, in particular regarding explicit contextual analysis and morphemic analysis, helped students to learn and to apply the CLUES strategy rapidly. These results demonstrate the promise of the CLUES intervention as a generative vocabulary strategy for identifying morphemes and extrapolating the meanings of unknown science terms. This finding is worth noting because it is estimated that nearly half of the words that students encounter in content area texts are unknown to them, so their ability to analyze unknown words while reading content area texts is essential (Harris et al., 2011). Third, the CLUES intervention utilized mnemonic strategy to facilitate the memorization of the steps to follow to figure out the meaning of a newly encountered word. Although several previous studies also used combined generative vocabulary strategies, this study was the first that explicitly broke down the strategies and utilized a mnemonic cue to structure an instruction routine (i.e., CLUES steps) to facility student learning. Because mnemonic instruction has shown to be effective for students with learning problems (Scruggs & Mastropieri, 2000), the positive outcomes of the CLUES intervention may partially attribute to the easy-to-follow routine to tackle new words.
Beyond the immediate intervention effect, all four students maintained their knowledge of the CLUES steps in assessments administered 2 weeks and a month after the intervention, showing greater than 85% accuracy. Nevertheless, despite continuing to identify and define morphemes and scientific terms at levels above the baseline, all four participants demonstrated decreased accuracy in the latter compared with the previous intervention probe. These results suggest that frequent exposure to and regular practice of the generative vocabulary strategy may be necessary for ELs with RDs when using this strategy for acquiring scientific vocabulary.
The pretest and posttest results for WKT and MT showed that, while students did perform better on the latter, the gain was small. Because the WKT and MT presented science terms in isolation without any context and because prefixes, suffixes, and roots tend to be integrated into words and may have slightly different meanings (Fox, 1996), students may experience increased difficulty in identifying and recalling the meaning of science terms or morphemes in the absence of contextual clues, especially given that the participants in this study had practiced morpheme analysis for only 10 lessons. Further analysis of the WKT results showed that, while the students were able to list at least 90% of the instructed roots, they retained few of the prefixes and suffixes. Because there was emphasis on explicit instruction on roots rather than prefixes and suffixes, the results suggest that providing students explicit instruction and multiple exposures of all types of morphemes may lead to retention across morpheme types. On the other hand, the students showed greater gains in SKT, for which they were asked to define the terms context, morpheme, prefix, and suffix. Prior to the intervention, they were unable to define any of these terms, suggesting that they were unaware of what contextual clues were and had very limited knowledge of morphemes. After the intervention, all of the participants in this study were able to recall the definitions of at least one of the terms in its entirety and of at least one other term in part, showing that they were cultivating knowledge of contextual and morpheme analysis, though admittedly all four scored below the mastery criterion of 80% accuracy (Deshler et al., 2001) on the posttest.
After the intervention, the students performed better on the near-transfer than on the far-transfer generalization probes. The implication of this finding is that, because the former were constructed similarly to the CLUES probes, students were able to apply the CLUES steps to define newly encountered scientific terms as they had in the intervention lessons, thereby greatly increasing their accuracy. The students’ somewhat less impressive performances on the far-transfer generalization probes suggest that students still relied on the step-by-step approach to define unknown science words. However, the positive outcome in the far-transfer test still indicates that students were able to use contextual and morpheme analysis to arrive at the meaning of newly encountered science terms even without the scaffolding provided by the CLUES graphic organizer. Given the fact that the generalization of skills is challenging for students with RDs (Jitendra et al., 2004) and the generic nature of the intervention (i.e., the strategy was designed to be applicable to a variety of context and morphemes), this result still shows that the CLUES strategy is promising.
Limitations and Future Research
One limitation of this study is possible experimenter bias, for the investigator administered some of the CLUES baseline and intervention probes, though this limitation is somewhat mitigated by the high fidelity of the procedural integrity data. A second limitation is the relatively brief duration of the intervention; other researchers have suggested that vocabulary instruction be extensive, including 50 or more lessons (Elleman et al., 2009; Jitendra et al., 2004) as compared with the 10 administered in this study. This small dosage may have contributed to the participants’ lack of improvement in WKT and MT and their decreases in accuracy on the second maintenance probe. Future study could accordingly extend the duration of the intervention in an effort to solidify students’ skills in contextual and morpheme analysis or further incorporate certain component of non-generative vocabulary strategy such as mnemonic devices to help retain the meaning of the learned morphemes or science terms. Third, although students had favorable performance in near-transfer test, the far-transfer test results were less impressive. The results suggest that students still relied on the prompts offered by the CLUES steps provided in the CLUES graphic organizer. Students seemed to have difficulty in spontaneously using the CLUES steps without the presence of the scaffold of the CLUES organizer. Future research should also explore a systematic approach to fade the prompts offered by the CLUES graphic organizer after the student master their skills in contextual and morpheme analysis with the support of CLUES graphic organizer. Finally, this study has limited generalizability. Although this design meets the single-subject standard (Kratochwill et al., 2013), the positive effects of the intervention were obtained only for the students participated in the study. A larger scale study can further demonstrate the effectiveness of the intervention.
Instructional Implications
This study has demonstrated the promise of the CLUES approach for facilitating the acquisition of science terms by secondary ELs with RDs. Based on the positive preliminary findings, several recommendations can ensure the effective implementation of this intervention. First, given that practice is critical to vocabulary acquisition (Jitendra et al., 2004), maximized opportunities for students to apply newly learned vocabulary can be essential to improve their fluency in analyzing contextual clues and morphemes. Furthermore, explicit instruction in which teachers think aloud and students respond verbally to generic vocabulary lessons is essential to assist ELs with RDs in detecting the nuances of terms and to deepen their understanding of why and how to employ this approach.
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.