Abstract
We investigated the efficacy of a morphological awareness intervention on the morphological awareness and reading skills of students from low-socioeconomic-status homes; we also examined whether the intervention was similarly effective for intervention students who differed in their initial morphological awareness abilities. The 8-week intervention was designed to increase awareness of affixes and the relations between base words and their inflected and derived forms for kindergarteners (n = 27) and first- (n = 22) and second-grade (n = 26) students. Students randomly assigned to the small group intervention were provided instruction four times a week, 25 min a day, whereas students assigned to the control group received “business as usual.” Kindergarten and first- and second-grade students receiving the intervention showed statistically significant gains in morphological awareness with large effect sizes on most measures. Students in all three grades who received the intervention demonstrated nonsignificant gains in literacy abilities with null to small effect sizes. Further, students with low morphological awareness abilities at the onset of the study demonstrated similar gains from the intervention as their peers with typical morphological awareness abilities. Our results suggest that explicit morphological awareness instruction may produce gains of practical importance to young elementary students at risk for future literacy difficulties.
Morphological awareness refers to the ability to consciously consider and manipulate morphemes, the smallest units of meaning in language. There is a relatively robust literature base that now demonstrates the significant contributions morphological awareness makes to literacy abilities (e.g., Carlisle & Fleming, 2003; McCutchen, Green, & Abbott, 2008; Singson, Mahoney, & Mann, 2000). There is a smaller yet growing database on the positive benefits of morphological awareness interventions (e.g., Bowers, Kirby, & Deacon, 2010; Carlisle, 2010; Goodwin & Ahn, 2013). However, many of these investigations have been conducted with students in upper primary grades. In this article, we report on a small efficacy study involving kindergarten through second-grade students who received 8 weeks of a morphological awareness intervention. These students were at risk for literacy deficits because they were from high-poverty homes. As such, the intervention also may serve as a model program for students who struggle in the areas of reading and spelling.
Morphological Awareness Intervention
Even for lower elementary students, morphological awareness relates to and uniquely predicts literacy skills when other known contributors to literacy abilities are considered (e.g., Apel, Wilson-Fowler, Brimo, & Perrin, 2012; Carlisle & Stone, 2005; Kirby et al., 2012; Wolter, Wood, & D’zatko, 2009). Because of this strong, influential role of morphological awareness in literacy development, a number of researchers have conducted investigations to determine whether morphological awareness interventions improve students’ morphological awareness and literacy skills. Two recent articles, Bowers et al.’s (2010) systematic review of 22 morphological awareness interventions and Goodwin and Ahn’s (2013) meta-analysis of 30 interventions, provided syntheses of the effect of morphological awareness intervention on literacy outcomes. Although the interventions reviewed varied in many respects (e.g., morphological content, assessment measures, length of intervention, student ages, and student ability levels), in general, morphological awareness interventions lead to moderate improvements in morphological awareness skills (Bowers et al. reported ds ranging from 0.51 to 0.65; Goodwin & Ahn reported a weighted mean effect size of 0.44). Improvements in more distal measures of literacy varied by report. Bowers et al. reported effect size ranges of 0 to 0.41 for improvements in word-level reading and spelling and 0 to 0.28 for improvements in reading comprehension. Goodwin and Ahn reported significant mean effects for decoding (d = 0.59) and spelling (d = 0.30) but nonsignificant effects for reading comprehension (d = 0.09). Bowers et al. noted that effect sizes were greater for interventions that were provided in small groups versus whole classrooms, when the students involved were considered to be “less able,” or when the students were second graders or younger. Similarly, Goodwin and Ahn stated that their findings suggest that instruction is most effective when group size is determined by the instructional needs of the students (e.g., smaller groups for students with literacy challenges). An important finding from both studies is that few investigations included second-grade or younger students (a total of seven studies across the two reviews).
In response to the paucity of information about morphological awareness interventions, particularly with young elementary students, Apel, Brimo, Diehm, and Apel (2013) conducted a feasibility study with primary-grade students considered to be at risk for literacy difficulties: children from low-socioeconomic-status (SES) homes. A total of 19 kindergarteners, 21 first graders, and 21 second graders received an 8-week morphological awareness intervention. The intervention was provided 4 days per week, 30 min per day, in small groups of 3 to 4 students. Instruction focused on increasing the students’ awareness of affixes and the relations between base words and their inflected and derived forms. At the end of the intervention, effect size scores for gains in morphological awareness abilities were medium to very large (ds ranged from 0.66 to 2.26). As expected, effect size scores were smaller for pseudoword and sight word reading measures (ds ranged from 0.21 to 0.81) and reading comprehension (ds ranged from 0.52 to 0.71). Apel, Brimo, et al. suggested the intervention was feasible for the population and showed promise in improving young students’ morphological awareness and reading abilities.
Although Apel, Brimo, et al.’s (2013) results were promising, the lack of a control group prevented drawing any conclusions about the efficacy of the intervention. Thus, we undertook this follow-up efficacy study using a control group. Based on Apel, Brimo, et al., we hypothesized students receiving the intervention would demonstrate statistically significant higher scores on our measures of morphological awareness with accompanying medium to large effect sizes representing practical gains in those skills compared to the control group. We did not anticipate notable morphological awareness gains for the control group, given that minimal morphological instruction occurs in schools (e.g., Bowers, 2012) and teachers have large gaps in their morphological knowledge (e.g., Moats, 2009). Although we were equivocal about whether the groups would be different statistically in their word-level and reading comprehension skills postintervention, we hypothesized that there would be smaller yet identifiable effect size gains in the intervention students’ word-level reading and reading comprehension skills compared to the control group, similar to Apel, Brimo, et al.’s findings and results with older students (e.g., Katz & Carlisle, 2009). This latter hypothesis was based on the fact that morphological awareness aids word-level reading by facilitating pronunciation of multimorphemic words (e.g., understanding disappear is dis + appear vs. di + sappear), which allows a reader to determine the meaning of the word from his or her combined knowledge of the affix and its base. This is true even for pseudowords, as morphological awareness likely aids the reader in recognizing morpheme-like units within nonsense words (e.g., Andrews & Lo, 2013). Further, the reader gains grammatical cues from this analysis of multimorphemic words (e.g., friendly serves as an adverb, teacher serves as a noun). When the meaning of multimorphemic words is accurately determined, then comprehension also improves because (a) the reader has decreased cognitive demands for decoding words, and (b) there is increased semantic and grammatical information on which to develop the overall meaning of the text (Bowers et al., 2010; Goodwin & Ahn, 2013).
Thus, our research question was, will an 8-week, small-group morphological awareness intervention improve the morphological awareness and reading skills of kindergarteners and first- and second-grade students from low-SES homes, compared to a control group of students? Given that children from low-SES homes are known to be at risk for later literacy deficits (e.g., Craig & Washington, 2004; Washington, 2001), successful outcomes for this population could provide potential instructional implications for students who struggle with reading and spelling and for students with identified learning disabilities.
Method
Participants
We recruited 158 students from 6 kindergarten, 5 first-grade, and 5 second-grade classrooms in a public elementary school. At each grade level, the students were randomly assigned to intervention or control groups. A total of 7 students left during the study, 5 because their families moved and 2 because of attendance or behavioral issues, resulting in 151 participating students. See Table 1 for detailed demographic information for all participating students. According to the district records, 74% of the students at the school qualified for free and reduced-price lunch. Of the students, 12 had individualized education plans to receive special services for such conditions as hearing impairment, language impairment, or specific learning disability. English was the primary language for all participants.
Demographic Information.
Measures
All students were administered two morphological awareness tasks, the Relatives and Rehit tasks, and two word-level reading tasks preintervention and postintervention. Two additional morphological awareness tasks, the Affix Identification and Spelling Multimorphemic Words (SMW) tasks, and a reading comprehension task were administered to the first- and second-grade students at the two time points. All measures were administered individually in a quiet area of the school (e.g., library) except three tasks (see below), which were conducted within the students’ classrooms. Seven examiners administered the test battery. Examiners were graduate students attending the local university, former elementary school teachers, or teacher assistants who had participated in specific training on test administration prior to pretesting. Raw scores were used for all analyses.
Morphological awareness
Four experimenter-designed tasks, used previously by Apel, Brimo, et al. (2013) and Apel, Diehm, and Apel (2013), were administered to assess students’ morphological awareness abilities. The base words, or base forms of inflected or derived words, that served as task items on the four measures were at or below the third-grade level of word frequency (SPELL-Links Word List Maker, 2010). All base and multimorphemic words used on all four measures were phonologically transparent in their relationship (e.g., friend/friendly). All multimorphemic words on the morphological awareness tasks contained affixes targeted in the intervention lessons. None of the base words or multimorphemic words used on any of those measures were taught directly within the intervention. The affixes contained on the four morphological awareness tasks and targeted in the lessons were taken from a list of the most common prefixes and suffixes (Berninger & Abbott, 2003).
Relatives task
The Relatives task, based on the work of others (e.g., Carlisle, 2000), was used to assess students’ awareness of the relation of base words and their inflected or derived forms. An examiner provided a student with a base word and then required the student to complete a sentence orally using an inflected or derived form of the base word (e.g., “Cry. The baby bumped his head, so he ______.”) Each correct answer was awarded a point. The task contained 2 practice items, 7 items that required inflected forms, and 19 items that required derived forms (26 test items). The task had adequate internal reliability (α = .87).
Rehit task
The Rehit task was used to determine students’ ability to explicitly combine two morphemes into a novel word, define that word, and then judge its semantic acceptability within the context of a spoken sentence. First, the student was asked to repeat one bound and one free morpheme (e.g., “Say re.” [student responds] “Now, say hit.” [student responds]) and then combine the two morphemes to create a “silly” word (i.e., “rehit”). A student received one point for a correct answer. The correct response was provided if the student did not respond accurately. Second, the student was asked to define the newly created word. Responses were scored via a list of acceptable responses (see Apel, Brimo, et al., 2013; Apel, Diehm, et al., 2013). A correct response was awarded two points. No points were awarded for partially accurate answers. When a student’s response was incorrect, he or she was asked to judge the acceptability of two definitions provided by the examiner (e.g., “Do you think rehit means to hit again or a person who hits?”). A point was awarded for a correct response. Third, the student judged the acceptability of two sentences containing the novel word (e.g., “The rehit is on the bed. Does that sentence make sense?”). A point was awarded for an accurate response to both questions. The total number of points per task item was four. There were 18 items on the task, preceded by one practice item. The task had adequate internal reliability (α = .90).
Affix Identification task
The Affix Identification task measured students’ conscious awareness of printed affixes and the orthographic changes that occur when those affixes are added to base words. The task was group administered to the first- and second-grade students in their classrooms. The students were provided a paper containing a list of pseudowords with real affixes (e.g., famming) and told to circle all the affixes they saw. Four practice items were provided and modeled by the examiner before the start of the task. The students were given 3 min to complete the task. The task consisted of 51 items, 24 inflectional affixes, and 27 derivational affixes. Each correctly circled affix was awarded a point. The task had adequate internal reliability (α = .92).
SMW task
The SMW task also assessed students’ awareness of printed affixes. The first- and second-grade students completed this classroom-administered spelling test of 26 multimorphemic words (e.g., washes, distaste, uneasy). For each item, the examiner said the word, used the word in a sentence, and then repeated the word. The students then wrote the word. Only the affix spellings were scored for accuracy because we were interested in the students’ inflectional and derivational morphological knowledge. The task had adequate internal reliability (α = .89).
Reading
Three tests were administered to measure reading abilities. From the Test of Word Reading Efficiency (TOWRE; Torgesen, Wagner, & Rashotte, 1999), the Sight Word Efficiency subtest was used to assess the students’ ability to read single real words, and the Phonetic Decoding Efficiency subtest was used to determine the students’ ability to read pseudowords. The students had 45 s to complete each task. Alternate form reliability for TOWRE was reported by the authors’ manual at .96. The two TOWRE tasks were administered individually to all students.
The Test of Silent Reading Efficiency and Comprehension (TOSREC; Wagner, Torgesen, Rashotte, & Pearson, 2010) was group administered to assess the first- and second-grade students’ silent sentence reading comprehension. TOSREC required students to read sentences silently and mark Yes or No on an answer sheet to indicate whether each sentence was true or false. Alternate form reliability was reported by the authors’ manual at .91.
Reliability
Interrater score reliability for all tasks was conducted on 10% of the total number of student responses. Each measure was rescored by a second scorer. Interrater agreement ranged from 96% to 100%.
Procedures
The intervention was implemented across 8 weeks between February and April. Preintervention and postintervention testing occurred 2 weeks prior to and 2 weeks following the intervention. The interventionists were the same individuals who had conducted the preintervention and postintervention testing; no interventionist was involved in posttesting for students with whom she intervened. For the intervention, the students were pulled from their classes and seen in groups of 4 or 5 students of the same grade in a quiet room, such as the library. The groups met 4 days per week, 25 min per session. When intervention students were receiving the intervention, the control students were receiving the typical instruction that was occurring in the classroom at the time. This content missed by the intervention students varied by grade and time of day (e.g., computer class, social studies, music).
A total of 11 specific inflectional and derivational affixes were taught across the 8-week intervention, with instruction on each affix occurring across two sessions. After two affix lessons (4 days), a review lesson of the affixes covered in the preceding two lessons was held across 2 days (see Table 2 for the order of lessons and target affixes). These review lessons also included a general discussion of inflectional and derivational forms of a common base word (see description of word relatives below). Awareness of printed and spoken morphology was the focus for first- and second-grade students, whereas kindergarten students’ lessons focused only on awareness of spoken forms. For all students in all activities, the interventionist first modeled and spoke about the focus of an activity and then actively involved the students by requiring them to respond with verbal, physical, and/or written responses. Across all lessons, the students were encouraged to be “word detectives” to help identify affixes, meanings of words, and so forth. All students used magnifying glasses throughout various activities as a tangible means of highlighting their need to investigate words.
Order of Lessons and Targeted Affixes.
Affix lessons
Each affix lesson spanned 2 instructional days. Each day’s lesson contained four main activities selected from a number of possible activities. For Day 1 of an affix lesson, the interventionist provided information about the goal of the lesson (i.e., to learn about a specific affix or “add-on”) and demonstrated the meaning of the word in a sentence (e.g., “This picture shows cutting. The -ing suffix is added to the end of cut to show that the action is happening right now. She is cutting her hair.”). Next, in the listening activity, the interventionist read words with the target affix (e.g., present progressive -ing) and defined the word (e.g., “cutting—to cut right now”). For the second portion of this activity, the students raised their magnifying glasses when they heard a word with the target affix (e.g., “Listen … jumping. Listen … cut”). The first activity of Day 2 was comparable.
For a second activity, the interventionist introduced one of the following three activities: a word-sorting activity (Day 1), “say it another way” activity, or story (both Day 2). In the word-sorting activity, the students sorted picture cards (kindergarten) or word cards (first and second grade) that could be distinguished by the presence or absence of the target affix (e.g., sorting pictures or words that did or did not represent or include the present progressive verb form). The interventionists helped the students identify the differences between pictures or words; for the first- and second-grade students, the interventionists also helped the students identify differences in the spellings based on the target affixes. In the “say it another way activity,” the students verbally produced inflected or derived forms of base words given the interventionist’s definition (e.g., “How can I say … ‘to draw right now’ … another way?”). In the story activity, the interventionists read a story and the students put up their thumbs every time they heard a word containing the target affix; they also were asked to explain what the word meant.
A third activity followed, either the written activity or the add-on activity (Day 1) or the affix book (Day 2). In the written activity, the first- and second-grade students scanned a list of words and circled the words containing the target affix. In the add-on activity, the students either connected or separated blocks as a tangible representation of how to add on or take off affixes from base words. For the affix book activity, the students first defined the target affix, then wrote examples of words (first and second grade), and pasted in pictures that represented words (all grades) that contained those target affixes. Finally, on both days, the lessons ended with a wrap-up activity. During this activity, the interventionist and students reviewed the target affix, its meaning, and brainstormed different words containing the affix.
Review lessons
Four times across the 8-week intervention, the students received a 2-day review lesson. The purpose of the review lessons was twofold. First, the lessons reviewed the specific affixes taught since the prior review. During that portion of the lesson, the interventionist and students talked about the affixes that had been learned, what the affixes meant, and (for first- and second-grade students) how they were spelled. Games incorporating the affixes under review (dice game with affixes on the face of each die) were used to heighten the students’ interest.
The second purpose of the review lessons was to introduce to and discuss the notion of word relatives (Wasowicz, Apel, Masterson, & Whitney, 2004). When word relatives were first introduced, the interventionist began the activity with a picture-sorting activity in which the students separated cards into piles of cats and dogs. The interventionist and students discussed how varieties of cats (and dogs) belonged in the same animal family because even though they did not always look or sound the same, they all meowed or barked. The interventionist also introduced a picture of a hot dog and discussed how it does not belong to the dog family, even though we see and hear the word dog in hot dog, because it does not mean the same as other dogs. Subsequently, the interventionist introduced the notion of word relatives following the same logic: words are related because they share meaning (e.g., eat and eating are relatives because they share meaning, but eat and beat are not). The students brainstormed word relatives of different base words and listened for relatives within a story.
Intervention fidelity
To determine intervention fidelity, approximately 22% of the instructional sessions were observed live. For each affix session observed, the interventionist was expected to complete the four main lesson components (e.g., introduction of target affix, word sorting, writing activity, wrap-up). Likewise, for the review sessions, the interventionist was to cover the three main activities (i.e., review of previously taught affixes, game involving affixes, and word relatives). Fidelity was high; 99% of all lessons contained instruction in all main activities.
Analysis
Our investigation was a pretest/posttest group design treatment with outcome measures of morphological awareness, word-level reading, and reading comprehension. To address our research question, we first examined whether our control and intervention students displayed differences in means on all outcome measures at posttest that were statistically significant. To do so, we conducted a series of 2 × 1 analysis of covariance, with the within-subject variable of posttest performance, the between-subjects variable of condition (intervention vs. control), and the covariate variable of pretest performance to control for any differences between the groups at pretest. Additionally, because we were interested in examining the practical importance of the observed gains from the intervention, we calculated effect size estimates (i.e., Cohen’s d) to interpret the meaning or magnitude of our results (Bothe & Richardson, 2011; Ellis, 2010). To do so, we used the following formula, using adjusted means to control for pretest performance and unadjusted posttest variances:
With small sample sizes, it is possible for statistically nonsignificant results to still have practical importance (Ellis, 2010). Therefore, we calculated effect size estimates for our statistically significant and nonsignificant findings alike, as reliance on significance testing only may sometimes fail to account for the real-world importance of those findings. In the remaining sections, we present and interpret these two constructs in parallel by using the terms statistically significant or significant when discussing the results of null hypothesis statistical tests and using the term practical importance in interpreting effect size estimates.
Results
Descriptive statistics and the results of the analysis of covariance showing statistically significant mean differences between groups at posttest are found in Table 3. Kindergarteners in the intervention group performed significantly higher at posttest than their control peers on the Rehit, F(1, 51) = 23.16, p < .001, partial η2 = .31, and Relatives tasks, F(1, 51) = 13.46, p < .005, partial η2 = .21. There were no significant differences between groups on sight word reading, F(1, 51) = 0.0, p > .05, partial η2 = .00, and pseudoword reading, F(1, 52) = 0.32, p > .05, partial η2 = .00.
Means, Standard Deviations, and Adjusted Means for Pretest-Posttest Measures With Effect Sizes.
Note. K = kindergarten; TOWRE = Test of Word Reading Efficiency.
p < .01. ***p < .001. These identify significant adjusted mean differences between control and intervention groups at posttest.
First-grade students in the intervention group demonstrated significantly higher scores than the control students on the Rehit task, F(1, 40) = 9.73, p < .005, partial η2 = .20; Affix Identification task, F(1, 40) = 73.05, p < .001, partial η2 = .65; and SMW, F(1, 40) = 7.68, p < .01, partial η2 = .16. The two groups did not differ on the Relatives task, F(1, 40) = 3.33, p > .05, partial η2 = .07; sight word reading, F(1, 40) = 0.75, p > .05, partial η2 = .02; pseudoword reading, F(1, 40) = 2.62, p > .05, partial η2 = .06; and reading comprehension, F(1, 40) = 3.11, p > .05, partial η2 = .07.
The second-grade students receiving intervention differed significantly from the control group students on the Rehit task, F(1, 51) = 16.78, p < .001, partial η2 = .25; the Relatives task F(1, 51) = 22.55, p < .001, partial η2 = .31; and the Affix Identification task, F(1, 49) = 38.64, p < .005, partial η2 = .44. There were no statistically significant differences between groups on the SMW task, F(1, 49) = 0.02, p > .05, partial η2 = .00, and the measures of sight word reading, F(1, 51) = 0.71, p > .05, partial η2 = .01; pseudoword reading, F(1, 51) = 2.32, p > .05, partial η2 = .04; and reading comprehension, F(1, 49) = 0.39, p > .05, partial η2 = .01.
We calculated effect size estimates using the formula given above and interpreted the results using Cohen’s (1988) conventions of small (d = 0.20), medium (d = 0.50), and large (d = 0.80). These benchmarks were used to estimate the practical importance of students’ morphological and literacy ability improvements. As can be seen in Table 3, for the morphological awareness tasks, many of the effect sizes were large and in favor of the intervention group. For each grade, effect sizes for morphological awareness tasks were small to large with the exception of performance on the SMW task for second-grade students. We found small, albeit nonsignificant, effects for reading comprehension in first-grade students and pseudoword reading in second-grade students, respectively. Effect size estimates for pseudoword reading in first grade were small and negative, indicating the students in the control group scored higher at posttest than the students in the intervention group. All other effect size estimates for the other reading measures across grades were less than .20, suggesting there was no observable effect of the intervention.
Because we had a range in performance on our morphological awareness measures, we completed a secondary, exploratory analysis of our data using only children involved in the intervention group. Specifically, we were interested in whether students who initially were weaker in their morphological awareness skills demonstrated similar gains as their peers with more typical abilities. To obtain a sense of this, we first calculated gain scores on each measure for every child (i.e., posttest score – pretest score = gain score). Next, we created a standardized morphological awareness composite score for all students by changing the raw score from each morphological awareness measures into a z score and took an average of the z scores across tasks. For kindergarten students, this average was calculated based on performance on two morphological awareness measures (i.e., Rehit and Relatives), whereas first- and second-grade students’ averages were calculated based on performance on all four morphological awareness tasks. We then determined which students fell within the bottom quartile for kindergarten (n = 6), first grade (n = 5), and second grade (n = 6) based on the range of z scores and labeled these children as low-morphological-awareness-ability students at pretest; the remaining students were labeled as having average morphological awareness abilities. Finally, we conducted a series of t tests to determine whether the low-ability versus average-ability groups grew differentially during the course of the intervention (i.e., whether the groups’ mean gain scores differed statistically).
Although the kindergarten control and intervention groups did not differ statistically from each other at posttest on measures of word-level reading (see Table 3), Table 4 shows that the low-ability kindergarten students demonstrated significantly larger mean gains on sight word reading than the students in the average group; on all other measures, the two ability groups made similar amounts of growth, although the low-ability group’s mean gain score neared significance for the Relatives task (p = .051). For first-grade students, the average-ability group achieved a higher gain score than the low-ability group on the Affix Identification task; there were no other statistical differences between the two ability groups for amount of gain made during the intervention. For second-grade students, the low-morphological-awareness-ability students displayed a significantly greater mean gain score than their peers with average morphological awareness abilities on the Affix Identification task. All other comparisons were nonsignificant.
Mean Gains and Standard Deviations for Low-Morphological-Awareness-Ability and High-Morphological-Awareness-Ability Students Within the Intervention Group.
Note. N/A = not applicable—too young for assessment; TOWRE = Test of Word Reading Efficiency.
p < .05.
Discussion
Our main research aim for this study was to determine whether our morphological awareness intervention improved the morphological awareness and reading skills of a group of kindergarteners and first- and second-grade students compared to their counterparts who did not receive the intervention. We hypothesized that we would obtain statistically significant mean differences between groups with accompanying medium to large effect sizes representing gains of practical importance by the intervention group on our measures of morphological awareness; this hypothesis was partially confirmed. Even though we were ambivalent about whether the groups would be different statistically in their word-level and reading comprehension skills postintervention, we hypothesized smaller gains of practical importance on measures of reading, given reading was not directly targeted in the intervention. This hypothesis also was partially confirmed; we obtained null to medium effect sizes although there were no statistically significant differences between groups. Finally, we also conducted a secondary, exploratory analysis that revealed that, in general, students characterized as having low morphological awareness abilities before the intervention benefited as much from the intervention as students with higher morphological awareness abilities. We discuss these findings below.
Effect of the Intervention on Morphological Awareness and Reading Skills
Students who received our morphological awareness intervention made statistically significant and practical gains on 7 of the 10 morphological awareness measures compared to their peers who did not receive the intervention. On the 2 morphological awareness tasks administered to the kindergarten children receiving the intervention, effect sizes for the gain scores were large: 0.82 and 1.25. For the first-grade intervention students, three of the four effect sizes also were medium to large, ranging from 0.67 to 2.54. The fourth effect size, for the Relatives task, was a small effect (d = 0.41), and performance was not statistically different from that of the control group. For the second-grade students participating in the intervention, effect sizes for all but one task (SMW task, d = −0.03; no significant group differences) were large (ds = 0.86 to 1.52). As a whole, these effect sizes suggest that the intervention led to marked improvements in the students’ morphological awareness abilities after 8 weeks of instruction compared to those students who did not receive the intervention.
Our effect sizes for the gains in morphological awareness were in line with those reported by Bowers et al. (2010) for small groups and/or with “less able” students (d = 0.99) and for students in second grade or younger (d = 1.24). Additionally, our findings were similar to, though not as large as, the effect sizes obtained in Apel, Brimo, et al.’s (2013) feasibility study, which used the same morphological awareness intervention, likely because the latter study did not have a control group (i.e., no account for natural maturation). Overall, based on our investigation and the limited number of studies of morphological awareness interventions with young students (Bowers et al., 2010; Goodwin & Ahn, 2013), it appears that a morphological awareness intervention can have a large effect on improving morphological awareness skills when provided in smaller groups and to younger students.
Our tests of statistical significance for the effects of our intervention on reading were nonsignificant. This may have been due to our small sample size. When we conducted our effect size measures, it appeared that our intervention led to small practical gains in reading abilities and only for the first- and second-grade students. The first-grade students demonstrated a small effect size gain (d = 0.26) on reading comprehension. The second-grade students showed a similar effect size for pseudoword reading (d = 0.28). These findings fall generally within the range of previous reports by others for primary-grade students (e.g., Bowers et al., 2010). Although our intervention did not target reading skills directly and did not change them significantly from the control group, the focus on morphological awareness appeared to lead to small practical gains in the reading skills of the intervention students as measured by effect size. It may be that the students’ increased ability to consciously consider base words and their affixes and to understand the meaning of certain affixes aided their decoding of words and extraction of the words’ meaning. Further, reading comprehension may have been positively influenced by decreasing the cognitive demands for decoding words and increasing the semantic and grammatical information available to construct overall text-level understanding (e.g., Bowers et al., 2010; Goodwin & Ahn, 2010). It may be that our effect size estimates provide information about the potential benefit of our intervention even though significance testing did not reveal statistical differences, possibly due to sample size. However, we note that on the majority of the reading measures, there were no marked practical gains.
The kindergarten children receiving our intervention showed no practical gains in word-level reading (ds = 0). We did not focus any instructional attention on print for these students, as our clinical judgment suggested they might not be developmentally ready for learning spoken and written morphological awareness simultaneously. Nunes, Bryant, and Olsson (2003) found that their specific morphology intervention improved the word-level reading abilities of their young students, with or without a focus on written morphology; however, their students averaged 8 years of age. With younger students like ours, embedding written morphology in intervention may be more critical to improving word-level reading.
As a whole, our findings suggest that our morphological awareness intervention led to statistically significant and large practical gains in the kindergarteners’ and first- and second-grade students’ morphological awareness skills and nonsignificant but small practical gains in the first- and second-grade students’ reading comprehension and pseudoword reading skills, respectively. The findings are promising given the relatively short amount of intervention time (i.e., 800 min). It may be that a larger number of students and a longer period of intervention are required to obtain greater practical and significant gains in reading abilities. Although there was no control group involved, Apel, Brimo, et al. (2013) noted statistically significant gains in reading skills with associated large practical gains, as measured by effect size, after their morphological awareness intervention, which involved 900 min of instruction. In the future, investigators will need to determine, among other factors, the impact of intervention duration on outcomes.
Outcomes for Low- Versus Average-Morphological-Awareness-Ability Students
We conducted a secondary analysis to examine whether the intervention led to differences in gains for students we characterized as having low morphological awareness abilities versus average morphological awareness abilities. For the most part, the two groups did not differ in the gains made during intervention. Thus, our findings suggest that students who fall within the lower quartile and students who fall within the higher quartiles on a measure of literacy benefitted equally from the intervention. There were a few noted differences. The kindergarten children in the low-ability group demonstrated greater gains postintervention on the sight word reading task, as did the second-grade students in the low ability group on the Affix Identification task. In contrast, first-grade children in the average-ability group demonstrated significantly greater mean gains on the Affix Identification task than did the low-ability group. It may be that the explicit instruction embedded within the lessons, the relatively large amount of review provided within and across lessons (e.g., reviews at the end of each lesson and review lessons after two affix lessons), and the strict adherence to the protocol (as evidenced by the high intervention fidelity) aided in providing the support needed for the students’ successful learning, regardless of their initial morphological awareness abilities. The intervention, then, may serve as a model instruction for early elementary students, or perhaps even slightly older students, who are known to have learning disabilities or specific struggles with reading and spelling. In the future, researchers should consider further study of differences in responses to intervention by low and typical groups using larger samples and conducting a similar analysis by comparison to a control group.
Limitations and Suggestions for Future Research
We chose our participants, students from a low-SES neighborhood school, because they were considered to be at risk for later literacy difficulties (e.g., Craig & Washington, 2004; Washington, 2001). We reasoned that successful outcomes for this population could provide potential instructional implications for students who struggle with reading and spelling and for students with identified learning disabilities. Nevertheless, our results can be generalized only directly to primary-grade students from low-SES schools. It is unclear at this point whether other student populations would demonstrate similar gains as a result of our morphological awareness intervention. In the future, researchers should determine how students from other backgrounds or with other literacy abilities may benefit from the intervention.
Although we included a control group, that group of students did not receive an alternative intervention; instead, they received the ongoing instruction of the classroom (e.g., computer class, social studies, music) while the intervention group was receiving our intervention. As such, we cannot determine whether some of the gains from our intervention were due to the direct, small-group attention the intervention group received from the interventionists versus the whole-group instruction the control group received within the classroom. In the future, the use of an alternate intervention with a similar delivery schedule would increase confidence in attributing any positive findings to the intervention itself.
After a relatively short amount of time (8 weeks), all the students receiving our intervention made significant and practical gains in their morphological awareness skills and, in the case of the first- and second-grade students, nonsignificant but small practical gains in their pseudoword reading and reading comprehension skills. However, what remains unknown is whether these gains would remain for the long term. We conducted posttesting immediately after the intervention; ideally, follow-up testing several months postintervention should be conducted to determine whether the initial advantages seen for the intervention group remain. Further, other factors that may have affected our outcomes, such as the length of the instructional sessions, the frequency with which they were offered, and the total number of sessions, should be investigated to determine optimal frequency, duration, and time elements. With future knowledge gained, clinical and educational professionals will be well prepared to provide evidence-based instruction that improves students’ literacy and literacy-related skills and, eventually, academic success.
Footnotes
Authors’ Note
At the time of this study, the first author was affiliated with Florida State University.
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the U.S. Department of Education, Institute for Education Sciences Funding, Reading for Understanding Research Initiative Grant Number R305F100027.