Benefits of Repeated Book Readings in Children With SLI

Abstract

In this pilot study, we ask whether repeated storybook reading is also beneficial for word learning in children diagnosed with specific language impairment (SLI). We compared 3-year-old German learning children diagnosed with SLI to typically developing children matched on age and socioeconomic status (SES). One week later, children with SLI retained the target words from the stories just as well as their peers, although they did perform significantly worse on immediate recall.

Keywords

preschool comprehension language learning disorders acquisition/development language/linguistics

Introduction

Joint book reading is highly beneficial for teaching children new vocabulary (e.g., Reese, 2015). Recent research demonstrates that children retain more new words when stories are repeated (Horst, Parsons, & Bryan, 2011; McLeod & McDade, 2011). Repetition is important because word learning extends beyond the initial encounter with a new word, when only some (partial) information about a word and its meaning are encoded (Munro, Baker, McGregor, Docking, & Arciuli, 2012). Critically for joint book reading, each time a story is reread, children have less information to encode because some information was already encoded on a previous reading (Horst, 2015). This, then, allows children to devote more resources to finer details, such as new vocabulary items. The study by Horst and colleagues (2011) provided evidence for this claim by exploring how repetition of joint book readings enables word learning. They taught 3-year-old children novel words across three reading sessions. Children were presented with the novel words either through reading the same story repeatedly (with one new story being repeated 3 times on each session) or through reading three different stories (with three new stories on each session). Using a four-alternative forced-choice picture-pointing task, children’s immediate recall was tested after each session and retention for words from the first two sessions was tested at the end of Session 3. Children who heard the same stories exhibited better word retention than children who heard different stories. Thus, although both groups performed well on immediate recall, only children who experienced the novel words embedded in the same story repeatedly showed a robust long-term learning effect—resulting in better retention of those novel words after a delay of several days.

McLeod and McDade (2011) also examined the effects of repetition in word learning from joint book reading. Children either heard the same story 3 times or one longer story, but all children heard the novel words the same number of times. All children were posttested 1 day after reading. Children learned more words when exposed to the same story repeatedly than when the words were presented within a single story. This suggests that for book reading to be beneficial, it is important to familiarize children with this activity over several readings and to allow the knowledge to consolidate (see also Damhuis, Segers, & Verhoeven, 2015).

In sum, research demonstrates that children learn more new words when stories are repeated (Horst et al., 2011; McLeod & McDade, 2011). The familiarity of this activity seems to facilitate encoding and foster consolidation processes. However, it is not known whether children at risk of language delays also benefit from repeated book readings. Although robust acquisition mechanisms are apparent in children with specific language impairment (SLI; Rice, Oetting, Marquis, Bode, & Pae, 1994), they may require more word repetitions to reach the same level of word learning performance as typically developing age-matched peers (Kan & Windsor, 2010). For example, Niebuhr-Siebert and Ritterfeld (2012) used a pre–post follow-up design to deliver audio books as a complex language input to facilitate language processing in preschool children. When repeated, audio books significantly affected word retrieval—also in children with SLI.

In this pilot study, applying the method of repeated book readings (Horst et al., 2011), we addressed the question of whether a repeated joint book reading activity is as advantageous for children diagnosed with SLI as it was reported for typically developing children. We hypothesized that children with SLI would also benefit from the repetition of the context in which the word learning is embedded. Furthermore, following the findings suggesting that learning effects became prominent within a delayed rather than immediate test, we expected the greatest benefit of repetition in children with SLI to occur during the delayed test (e.g., McGregor, Rohlfing, Bean, & Marschner, 2009).

Method

Sixteen 3-year-old German-speaking children participated: eight children diagnosed with SLI (three girls) and eight typically developing age-matched peers (five girls). Demographic information is presented in Table 1. Children were recruited from a large city in Western Germany. Children with SLI were recruited via language pathologists, day care centers, playgroups, and one otorhinolaryngologist. All the children were monolingual with no further physical and cognitive impairment. We assessed children’s language skills with the SETK 3-5 (Sprachentwicklungstest für drei- bis fünfjährige Kinder; Grimm, Aktas, & Frevert, 2010). This is a diagnostic tool containing these subtests: (a) Comprehension of Sentences, (b) Verbal Descriptions of Pictures, (c) Plural Formation, and (d) Phonological Working Memory.

Table 1.

Demographic Information for the Children Who Participated in the Current Study: Means (SDs in Parentheses).

Demographic information	p	SLI Group	TD Group
Age in months	ns	39.7 (3.5)	41.1 (3.6)
Sentence understanding (SETK 3-5)	***	40.7 (6.4)	59.1 (7.9)
Sentence production (SETK 3-5)	**	42.4 (8.3)	60.3 (8.9)
Morphological Rules (subtest of SETK 3-5)	*	46.1 (12.7)	59.6 (11.4)
Phonological Memory (subtest of SETK 3-5)	***	37.4 (5.9)	60.0 (9.0)
SES	ns	14.8 (4.2)	18.13 (1.6)
Books at home	*	51 (17.0)	77 (28.1)
Book titles	ns	22.6 (9.2)	27.6 (6.3)
Days between Session 1 and 2	ns	2.1 (1.5)	1.5 (1.1)
Days between Session 2 and 3	ns	2.4 (1.3)	2.2 (1.3)
Overall duration for the study (in days)	ns	7.5 (2.4)	6.8 (1.3)

Note. The SETK 3-5 (Grimm, Aktas, & Frevert, 2010) is a German standardized diagnostic tool to assess language capabilities of children between 3 years 0 months and 5 years 11 months of age. SLI = specific language impairment; TD = typically developing children; SES = socioeconomic status.

p < .05. **p < .01. ***p < .001.

Children were visited in their homes 3 times over the course of approximately 1 week. For each reading, children sat next to the experimenter to ensure they could see the illustrations. The experimenter read a different story on each visit, but the story was repeated 3 times during each visit. Children were read German translations of three of the illustrated storybooks created by Horst and colleagues (2011). Each story was appropriate for the German culture and depicted two novel objects; each object was named exactly 4 times during each story. Thus, altogether, children heard each of two target words 12 times per visit (see Horst et al., 2011). In contrast to the original study, the order of the stories was fixed to minimize experimenter errors: On Session 1, children heard the story about a bold dog; on Session 2, they heard a story about the library; and finally, on Session 3, they heard a story about a mysterious aunt. Critically, both testing conditions and book order were the same for the children with SLI and the typically developing children; thus, any effects of book order would be the same for both groups. The experimenter did not emphasize target words or provide definitions or other dialogic cues. Children’s questions and comments were neither encouraged nor discouraged. If children asked questions during the story, the experimenter encouraged them to return attention to the story (e.g., “let’s keep reading, and see!”) and avoided naming any objects (see Horst et al., 2011). Parents completed a demographic questionnaire and a checklist of children’s books and authors to determine familiarity (which can be provided by the authors on request).

Children’s immediate recall and retention was tested via a four-alternative forced-choice pointing task using a test booklet containing four familiar object pictures per practice page and four novel object pictures per test page. On the final visit, after the recall trials, children were also tested on their retention of the words from Session 1 and Session 2 using the same task (eight retention trials, see below). Practice page, trial order, and target locations (e.g., top left) were counterbalanced within and across participants. Children received 12 immediate recall trials (four each in Sessions 1–3) and eight retention trials (four each on the words from Sessions 1–2, tested on Session 3). To facilitate comparison between immediate recall and retention scores, scores were converted into proportion correct out of 12 or 8, respectively.

Results

As can be seen in Table 1, children diagnosed with SLI and typically developing children differed with respect to all measures revealed via the SETK 3-5 test: sentence production, Z = −3.26, p < .001; sentence understanding, Z = −3.00, p < .01; morphological rules, Z = −2.02, p = .05; and phonological memory, Z = −3.22, p < .001, but not age Z = −0.85, df = 16, p = .44.

Both groups demonstrated better immediate recall than expected by chance (25%): for the children with SLI, Z = 1.9, p = .05, d = 1.1, and for the typically developing children, Z = 3.6, p < .001, d = 4.2, see Table 2). A linear mixed-effect model found a significant difference between groups: Typically developing children scored on average 25% better than the children with SLI, F(1, 41) = 8.03, p < .01, η² = .36, immediately after hearing the stories. In addition, we found a group by session interaction, driven by lower performance of the typically developing children after the book about the library (estimate = −.44, p < .05).

Table 2.

Percentage of Correct Recall in the Two Groups.

Child	Immediate Recall		Retention
Child	TD	SLI	TD	SLI
1	0.58	0.25	0.70	0.50
2	0.83	0.42	1.00	0.25
3	0.50	0.42	0.37	0.87
4	0.42	0.33	0.50	0.63
5	0.50	0.17	0.50	0.25
6	0.42	0.25	0.50	0.63
7	0.92	0.42	1.00	0.37
8	0.50	0.50	0.37	0.12
M (SD)	0.58 (0.19)	0.34 (0.11)	0.62 (0.26)	0.45 (0.25)

Note. The chance level was at 25%. SLI = specific language impairment; TD = typically developing children.

Both groups also demonstrated better word retention than expected by chance (25%): for the children with SLI, Z = 1.94, p = .05, d = 1.1, and for the typically developing children, Z = 3.61, p < .001, d = 4.2. A linear mixed-effect model found a significant difference between groups: Retention performance differed significantly for words from Session 1 and Session 2, F(1, 14) = 4.45, p = .052, η² = .23, with children performing better on retention of words from Session 1. Our analysis further revealed that the typically developing children scored on average 17.25% higher than the children with SLI. However, the difference in retention performance between the groups was not statistically significant (p > .19), suggesting that children with SLI benefited from the repetition as did typically developing children.

Correlational analyses (see Table 3) revealed that both sentence understanding and phonological memory (as assessed by SETK 3-5) were associated with the recall performance as we found a significant positive correlation of sentence understanding and recall, r_s = .64, p < .01, and a marginally significant positive correlation of phonological memory and performance in recall, r_s = .46, p = .07, for all children. None of the SETK 3-5 measures correlated with retention performance, which itself did not differ between groups. Correlational analyses further revealed a statistically marginal relationship between the socioeconomic status (SES) of the family and the child’s recall performance, r_s = .47, p = .07. This effect is reported in several studies suggesting that children from families with low SES have an elevated risk of language delays (Adolf, McLeod, & Leftwich, 2014). Furthermore, on the children’s book questionnaire, mothers of the children with SLI reported having significantly fewer books than mothers of typically developing children, Z = −1.96, p = 0.05. Interestingly, how many book titles or authors parents knew and how many books they had at home were not correlated (ps < .6). However, how many book titles parents knew correlated significantly with the numbers of authors parents knew, r_s = .87, p < .001, as well as the book content they knew, r_s = .93, p < .001.

Table 3.

Correlations Between Recall Performances, Language Measures (SETK 3-5), and SES.

Test	Sentence Understanding	Sentence Production	Morphological Rules	Phonological Memory	SES
Recall	.64**	.35	.41	.46^†	.47^†
Retention	.32	.14	.29	.20	.37

Note. SES = socioeconomic status.

†

p < .10. **p < .01.

Discussion

Previous studies demonstrated that children can learn novel words from repeated book readings (e.g., Horst et al., 2011). Our results extend this effect to a group of children diagnosed with SLI. Although children with SLI performed significantly worse than typically developing children in immediate recall, they did retain novel words from the stories at similar rates as typically developing children when tested up to 12 days later. Thus, familiarity with a repeated story appears to have enabled children to devote memory resources to finer details such as new words that may not be the focus on attention when first learning about new characters and plots. We, therefore, conclude that repeating a story with new words is a method that may contribute to improve intervention programs to help children with SLI achieve good long-term learning results. Specific suggestions include repeatedly reading the same books within a short time frame (i.e., within a single session), reading books that introduce few new vocabulary items simultaneously, and, generally, increasing children’s familiarity with the activity of shared book reading.

In addition, the analyses revealed a significant positive correlation between sentence understanding and immediate recall and a marginally significant positive correlation of phonological memory and immediate recall. These findings suggest that deficits in sentence understanding as well as phonological memory might be particularly crucial in the early stages of word learning, but become less relevant during word consolidation. Although relations of phonological and grammatical competencies to short-term memory have already been recognized (see Leonard, 2014, for a summary), the lack of relation between these competencies to long-term learning introduces novel insight to this literature. Finally, our correlational analyses of home reading environment are consistent with studies indicating an elevated risk of language delays in children from homes where fewer books are present (Payne, Whitehurst, & Angell, 1994): Although parents of children with SLI knew as many book titles as parents of typically developing children, fewer books were available to their children at home.

We are aware of the limitations our study bears. First, our method encompasses a small sample, although this is appropriate for exploratory studies (Asendorpf et al., 2013). We consider this a pilot investigation of the effect and nature of context repetition for children with SLI. Clearly, future research is needed to further investigate the effects of different forms of repetition in word learning for children with SLI and at even longer timescales. Second, we tested children’s word retention but not generalization. Thus, it is possible that children in our study only learned to associate the novel word with a specific exemplar and not an entire object category. Furthermore, we tested the acquisition of nouns, whereas children with SLI seem to traverse through a more problematic acquisition process for verbs. For example, in their study with 5-year-olds, Rice and colleagues (1994) found a decrease in verb retention. Thus, further research is needed to confirm the benefits of repeated book readings on the acquisition of words from other word classes. Finally, we did not analyze the interactions during the book readings. It is possible that the groups differ in terms of strategies that aid them to remember novel words. Future research can test whether, for example, children in the SLI group ask more questions than their typically developed peers.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research/work was supported by the Cluster of Excellence Cognitive Interaction Technology (CITEC; EXC 277) at Bielefeld University, which is funded by the German Research Foundation (DFG).

References

Adolf

S. M.

McLeod

A. N.

Leftwich

(2014). Structured narrative retell instruction for young children from low socioeconomic backgrounds: A preliminary study of feasibility. Frontiers of Psychology, 5, 391.

Asendorpf

J. B.

Conner

De Fruyt

De Houwer

Denissen

J. J.

Fiedler

. . . Wicherts

J. M.

(2013). Recommendations for increasing replicability in psychology. European Journal of Personality, 27, 108–119.

Damhuis

C. M.

Segers

Verhoeven

(2015). Stimulating breadth and depth of vocabulary via repeated storybook readings or tests. School Effectiveness and School Improvement, 26, 382–396.

Grimm

Aktas

Frevert

(2010). SETK 3–5. Sprachentwicklungstest für drei- bis fünfjährige Kinder. Diagnose von Sprachverarbeitungsfähigkeiten und auditiven Gedächtnisleistungen [Language development test for 3-5-year-old children. Diagnosis of speech processing abilities and auditory memory proficiency]. Göttingen, Germany: Hogrefe.

Horst

J. S.

(2015). Word learning via shared storybook reading. In Kümmerling-Meibauer

Meibauer

Nachtigäller

Rohlfing

K. J.

(Eds.), Learning from Picturebooks: Perspectives from child development and literacy studies (pp. 194–208). London, England: Routledge.

Horst

J. S.

Parsons

K. L.

Bryan

N. M.

(2011). Get the story straight: Contextual repetition promotes word learning from storybooks. Frontiers in Psychology, 2, 17.

Kan

P. F.

Windsor

(2010). Word learning in children with primary language impairment: A meta-analysis. Journal of Speech, Language, and Hearing Research, 53, 739–756.

Leonard

L. B.

(2014). Children with specific language impairment. Cambridge, MA: MIT Press.

McGregor

K. K.

Rohlfing

K. J.

Bean

Marschner

(2009). Gesture as a support for word learning: The case of under. Journal of Child Language, 36, 807–828.

10.

McLeod

A. N.

McDade

H. L.

(2011). Preschoolers’ incidental learning of novel words during storybook reading. Communication Disorders Quarterly, 32, 256–266.

11.

Munro

Baker

McGregor

Docking

Arciuli

(2012). Why word learning is not fast. Frontiers in Psychology, 3, 41.

12.

Niebuhr-Siebert

Ritterfeld

(2012). Hörspielbasierte Sprachlerneffekte bei Vorschulkindern mit spezifischen Sprachenetwicklungsstörungen [Audiobook language learning effects in preschool children with specific language impairments]. Sprache, Stimme, Gehör, 36, e11–e18.

13.

Payne

A. C.

Whitehurst

G. J.

Angell

A. L.

(1994). The role of home literacy environment in the development of language ability in preschool children from low-income families. Early Childhood Research Quarterly, 9, 427–440.

14.

Reese

(2015). What good is a picturebook? Developing children’s oral language and literacy through shared picturebook reading. In Kümmerling-Meibauer

Meibauer

Nachtigäller

Rohlfing

K. J.

(Eds.), Learning from picturebooks: Perspectives from child development & literacy studies (pp. 194–208). London, England: Routledge.

15.

Rice

M. L.

Oetting

J. B.

Marquis

Bode

Pae

(1994). Frequency of input effects on word comprehension of children with specific language impairment. Journal of Speech and Hearing Research, 37, 106–122.