With a little help: improving kindergarten children’s vocabulary by enhancing the home literacy environment
Early linguistic competencies are necessary prerequisites for later reading and writing abilities and thus for a successful school career. Various child and family characteristics have been identified as important predictors of children’s linguistic abilities, such as intelligence or the “home literacy environment” (HLE). Therefore, one way to improve children’s competencies is by enhancing the HLE they live in. Family literacy programs have proven to be successful with this task. However, most interventions used to improve HLE were fairly intensive and costly. In this study a nonintensive intervention procedure was developed to improve both, HLE and linguistic competencies. The sample consisted of 125 German children in their last year of kindergarten (mean child age at the beginning of the study: 5 years, 5 months) and their families who showed an above average socio-economic status. All parents were offered to participate in the intervention, consisting of providing them with relevant information on HLE at one evening meeting and providing an additional individual reading session that introduced them to the concept of dialogic reading. HLE and children’s linguistic competencies were assessed before and after the intervention. Participating and non-participating families did not differ in any of the study variables at the beginning of the study. However, families who participated in the interventions not only improved their HLE, but children in those families also showed greater linguistic competency development when compared with the non-participating group. The results indicate that less intensive interventions can have an impact on home learning environments and children’s linguistic development.
KeywordsHome literacy environment (HLE) Non-intensive intervention Family literacy programs Linguistic competencies Pre-school children
A child’s learning and achievement in school depends on several factors. In addition to individual characteristics of the child and aspects of school and teaching such as school size or quality of teaching, the home environment plays an important role (Hattie, 2013). In particular, the so-called “home literacy environment” (HLE) influences early competencies in reading and spelling and their precursors (e.g. Davidse, de Jong, Bus, Huijbregts, & Swaab, 2011; Niklas & Schneider, 2013). The concept of the HLE comprises all elements of the environment provided by the family that facilitate a child’s acquisition of linguistic skills. Examples of such elements include the reading behaviour of parents, the frequency a child is read to, the number of books in the home, and the number of children’s books in the home.
In accordance with Bronfenbrenner’s ecological theory (e.g. Bronfenbrenner, 1979) research shows that structural characteristics of a family such as the socio-economic status (SES) or migration background influence the HLE, which in turn impacts on early childhood competencies (e.g. Aikens & Barbarin, 2008; Niklas & Schneider, 2013). The HLE thus acts as a mediator. As children’s early linguistic competencies are a good predictor of later reading and spelling (e.g. Näslund & Schneider, 1996; Torppa et al., 2007), the HLE also influences these important abilities. For instance, a more favourable HLE is associated with better linguistic precursors such as vocabulary, phonological awareness or listening comprehension and is thus linked indirectly to later reading performance (Sénéchal & LeFevre, 2002). Therefore, one way to improve children’s literacy development is to extend and improve their literacy experiences at home. This paper investigates whether a non-intensive, short-term intervention is sufficient to support parents in providing more favourable HLEs and thereby to support their children’s linguistic competencies.
First, important specific linguistic precursors are introduced. Thereafter, studies are reported that analyse the influence of the HLE on these precursors. The discussion proceeds with a discussion of the effectiveness of family literacy programs and finally, results of an intervention study conducted in Germany will be presented.
Predictors of reading performance
As the English language has a deeper orthography than the German language, English students often experience more difficulty than German students when learning to read, English dyslexics tend to read more slowly than German dyslexics, and English students make more mistakes when reading (Landerl, Wimmer, & Frith, 1997). However, like English, German has a complex syllable structure with many complex consonant clusters in both onset and coda position (Seymour, Aro, & Erskine, 2003). It is thus not surprising that the same linguistic precursors that have been found to predict later reading performance in English children are also important in German-speaking samples (e.g. Ennemoser, Marx, Weber, & Schneider, 2012; Näslund & Schneider, 1996; Schneider & Näslund, 1999).
One such precursor is vocabulary (e.g. Torppa et al., 2007; cf. Dickinson, Griffith, Michnick, Golinkoff, & Hirsh-Pasek, 2012). Reading comprehension is improved by better vocabulary (Ennemoser et al., 2012; Joshi, 2005). Children will read both faster and more accurately when they know the meaning of all the words they are reading and a below-average vocabulary before school enrolment seems to be a risk factor of being identified as dyslexic later in school (Torgesen, 2002).
Another important predictor for later linguistic competency in school is early letter knowledge (e.g. Torppa, Poikkeus, Laakso, Eklund, & Lyytinen, 2006). Children who know many letters are more likely to succeed in later literacy tasks at school (e.g. Piasta, Petscher, & Justice, 2012). This knowledge helps, in particular, when children start learning to read and need to connect phonemes to graphemes. German children learn letters somewhat later than many English-speaking children as the teaching of letters is regarded as a task for formal schooling, rather than for kindergartens and pre-schools. However, German children seem to benefit similarly from earlier and better letter knowledge: children with better early letter knowledge often outperform other children in reading and spelling competencies later on (Näslund & Schneider, 1996). It is thus important for families to support children’s emerging vocabulary and letter knowledge, particularly in the years before school.
Other important precursors to the development of linguistic competencies include intelligence and rapid naming. As non-specific predictors they predict not only later spelling and reading performance, but also other competencies such as mathematical abilities (Koponen, Aunola, Ahonen, & Nurmi, 2007; Krajewski & Schneider, 2009; Puolakanaho et al., 2008; Schatschneider, Fletcher, Francis, Carlson, & Foorman, 2004; Schneider & Näslund, 1999).
Home literacy environment and linguistic competencies
Early meta-analyses demonstrated that reading to children explained about 8 % of the variance of children’s linguistic competencies (Bus, van IJzendoorn, & Pellegrini, 1995; Scarborough & Dobrich, 1994). This amount of explained variance is not impressive. In addition, intervention studies in which parents were asked explicitly to read to their children were not all successful in improving children’s linguistic abilities (Scarborough & Dobrich, 1994). However, these effects may accumulate over time, and reading to children is only one aspect of the HLE (Dickinson et al., 2012; Dunning, Mason, & Stewart, 1994).
In addition to children benefiting from being read to frequently, the quality of the book reading itself is also important. Here, dialogic reading strategies have been shown to support children’s learning: the (adult) reader interacts with the child, asks questions, encourages remarks given by the child, and expands on the text (cf. Whitehurst & Lonigan, 1998; Whitehurst et al., 1988). Moreover, other aspects of the learning environment such as the numbers of books at home or the reading or TV watching behaviour of the parents seem to be of importance as well (cf. Niklas & Schneider, 2013).
Various studies using different operationalisations of the HLE showed fairly similar results, in that a more favourable HLE is interrelated with better linguistic competencies of children living in those environments (e.g. Davidse et al., 2011; de Jong & Leseman, 2001; Hood, Conlon, & Andrews, 2008; Roberts, Jurgens, & Burchinal, 2005). Small to medium effect sizes were frequently found, and several results indicate that the HLE influences linguistic precursors rather than directly influencing academic achievement in school (e.g. Aikens & Barbarin, 2008; Hood et al., 2008; Niklas, Möllers, & Schneider, 2013; Sénéchal & LeFevre, 2002). However, in some studies the increase of specific competencies, such as phonological awareness, was also directly associated with HLE even when controlling for initial competencies or other important variables (e.g. Burgess, 2002; Niklas & Schneider, 2013).
Unlike in English-speaking countries, few studies have been conducted on the HLE in German-speaking countries and many of these studies are recent (e.g. Kluczniok, Lehrl, Kuger, & Rossbach, 2013; Niklas & Schneider, 2013; Niklas et al., 2013; Wieler, 1997). However, the findings of studies analysing German samples showed the same association of family characteristics such as SES and the HLE (Kluczniok et al., 2013; Wieler, 1997). In addition, the HLE also proved to be a predictor of linguistic precursors in German samples (Niklas et al., 2013). Consequently, the role of the HLE in child development seems to be comparable for both contexts (cf. Niklas & Schneider, 2013).
Effectiveness of family literacy programs
An implication of the association of between the HLE and children’s linguistic competencies is that improving the HLE should result in a better development of these competencies. In an intervention study conducted by Harper, Platt, and Pelletier (2011), children and their families were offered a fairly intensive 9-week family literacy program. Each of the nine sessions lasted about 90 min and included literacy activities, discussion, and singing songs. The gains in English language learners’ letter knowledge and their ability to infer meaning from print was statistically significant for the intervention group when compared with (1) first-language English-speaking children in the intervention and control groups, and (2) English language learners in the control group (medium to large effects).
Other studies have focused on book reading interventions by introducing dialogic reading to parents. Dialogic reading improved the linguistic development of children in comparison with children with no such experiences (e.g. Hargrave & Sénéchal, 2000; Huebner, 2000; Lever & Sénéchal, 2011). Overall, many studies demonstrated effects of family literacy programs that ranged from small to large effects, whilst in other studies no such effects were found (e.g. St. Pierre, Ricciuti, & Rimdzius, 2005).
Meta-analyses on this topic help to obtain a better picture of the overall effect of family literacy programs on children’s linguistic ability. Sénéchal and Young (2008) combined the results of 16 intervention studies on parental involvement in their kindergarten and primary school children’s development of reading and spelling abilities that fulfilled certain criteria and found a mean weighted effect size on children’s reading outcome that was moderately large (Cohen’s d = .65), indicating that parental involvement played a fairly important role. A meta-analysis by Mol, Bus, de Jong, and Smeets (2008) focused solely on dialogic reading and its effect on vocabulary. The authors analysed 16 studies that compared dialogic reading intervention groups and reading-as-usual control groups. Here, a medium mean effect size was found (d = .42), with higher effects for younger children aged 2–3 years old and children without literacy impairments.
By comparison, van Steensel, McElvany, Kurvers, and Herppich (2011) reported a small but statistically significant overall effect on comprehension-related skills and code-related skills with d = .18, based on results of 30 studies that used various intervention approaches. Although each of the three meta-analyses had a different focus, they all showed that literacy interventions in families improve children’s literacy competencies at least to some degree. However, many interventions evaluated in these studies and included in the meta-analyses often took several weeks or even months and thus were fairly extensive (e.g. Harper et al., 2011; Lever & Sénéchal, 2011).
Although the impact of the HLE on precursors of literacy acquisition has been explored in several large-scale studies (e.g. Burgess, Hecht, & Lonigan, 2002; Sénéchal & LeFevre, 2002), and first evaluations of family literacy programs have been conducted (Mol et al., 2008; van Steensel et al., 2011), we still know little about what kind of non-intensive intervention might be sufficient to improve the learning environment provided by the family, and thereby enhance the linguistic competencies of children living in these environments. We thus decided to carry out a non-intensive intervention study in the last year of German kindergarten. The intervention focused on the development of vocabulary, letter knowledge, and rapid naming of pictures.
Between pre- and post-intervention assessments, interested parents were offered general information about measures to improve the HLE. In addition, they were given a suitable children’s book, and during individual sessions, were taught how to implement book-reading strategies that aligned with the dialogic reading approach in order to improve the quality of reading. We assumed that this kind of intervention might be sufficient to increase the quality of the HLE and better support children’s linguistic competencies when compared with families not participating in the intervention. Moreover, in line with the findings of meta-analyses, we hypothesized that the impact on vocabulary gain would be greater than on gains in letter knowledge and in rapid naming (cf. Mol et al., 2008; van Steensel et al., 2011).
Materials and methods
Data assessment was carried out in the context of an intervention study (Project “KLUG”). The assessment started in September 2012, about 12 months prior to school enrolment (t1), when children were about 5 years old. The assessment was repeated in January 2013 (t2), about 4 months after the first assessment and some 8 months prior to school enrolment.
In total, n = 125 children (50.4 % girls and 49.6 % boys) for whom parents had given their consent were tested. At t1, children’s ages ranged from 58 to 78 months, with a mean age of M = 65 months (SD = 7.3 months). Children were distributed across nine kindergartens in and near a large city in southern Germany. About half of the children attended kindergartens in an urban area (51.2 %); the remainder attended kindergartens in rural towns close to the city.
Information regarding children’s migration background was obtained. Only 19 children (15.2 %) had a migration background (at least one parent born outside Germany). This percentage is fairly low for a sample of children this age in Germany. Moreover, in about half of these cases, the children’s parents had been born in countries sharing a similar standard of living to that of Germany such as Switzerland, Canada or the USA. Consequently, migration background was not expected to influence the children’s HLE or the development of literacy competencies. For this reason, children’s migration background was not taken into account in our analyses.
The Wegener scale was used to determine participants’ SES score, based on parental occupation prestige (Wegener, 1988). Such prestige scores are also used in large educational studies such as the Programme for International Student Assessment (PISA), and are closely associated with the education and income of parents. The scale ranged from 20 (unskilled labour) to 186.8 (physician). Occupations of 107 households were obtained. The highest prestige score in a household was on average M = 90.8 (SD = 37.3; median = 81.4), indicating a slightly above average SES in this sample compared with other German samples (e.g. Niklas & Schneider, 2013; Weinert, Schneider, Stefanek, & Weber, 1999).
Testing sessions at both, t1 and t2, were distributed across 2 days. Each individual testing session lasted for about 30 min, and all tests were conducted by trained research assistants who were provided with detailed test instructions. We were able to obtain data from the whole sample at t1 and t2, and thus data of all of the 125 children were included in our analyses.
Test instruments and questionnaires
The middle section of the Revised Vocabulary Test for 3- to 5-year-old Children (“Aktiver Wortschatztest für 3- bis 5-jährige Kinder-Revision”; Kiese-Himmel, 2005) was used to assess children’s active vocabulary. Here, children were required to name 40 pictures. The test includes both nouns and verbs. In our sample, the internal consistency for the vocabulary sum score was acceptable (Cronbachs α = .69 and a retest reliability of rtt = .77).
To assess children’s letter knowledge, small cards with the twelve most common German letters were presented to the children who were invited to name the letters. One year prior to school enrolment, some 7 % of the children were unable to name any letters. Approximately 15 % of the children named all twelve letters. On average, seven letters were known at t1. At t2, 4 months later, only two children were unable to name any letters. On average eight letters were known at t2 (rtt = .88).
Children’s fast access to long-term memory was tested by three rapid naming tasks. All three tasks required children to identify easy pictures (fish, ball, house, dog, tree, ice), colours, or pictures of dice (1–6 dots), in random order, three times. It took children about 10–90 s to name the 18 pictures, with means of 25.5 (pictures), 26.4 (colours), and 25.0 (dice pictures) seconds at t1, and means of 21.6, 21.5, and 20.0 at t2, respectively. As all three tasks were highly correlated, the mean was calculated and used in subsequent analyses (Cronbachs α = .78; rtt = .77).
The Columbia Mental Maturity Scale (CMMS; Burgemeister, Blum, & Lorge, 1972) was used to assess intelligence at t1. Children were required to identify the odd-one-out picture in an array of four or five pictures (e.g. one spoon and three forks). A maximum of 57 items was presented and the number of correctly identified odd-one-out pictures used in the analyses. Reliability and prognostic validity of the CMMS are guaranteed (cf. Esser, 2002; split-half reliability ranging between r = .92 and .96).
A parent questionnaire used at t1 and t2 to obtain information about the HLE. Although the risk of social desirability cannot be excluded, this way of assessment often provides reliable data, and previous studies have shown that there is high concordance with other measures such as knowledge about children’s books or diary entries (Burgess, 2002). We decided to obtain HLE data by asking six questions that focused mainly on the reading behaviour and the value attached to reading in the household (e.g. “How often do you read to your child”, “How much do you agree on the following statement: In our household we like to read”, and “At home we often play word games such as scrabble”). In addition, the number of books and storybooks in the household was considered as these seem to be important elements of the home learning environment in German households (McElvany, Becker, & Lüdtke, 2009; Niklas et al., 2013).
Each item had to be answered on a Likert scale (0–4 points), and data were obtained for a total of 109 families (104 families at t2). Observed values ranged from 17 to 32 with a mean of M = 27.3 (SD = 3.2 at t1; M = 27.8, SD = 3.2 at t2), and an acceptable internal consistency of the sum score (Cronbach’s α = .72; rtt = .79).
Project “KLUG” commenced in August 2012, when several kindergartens were approached by university staff and invited to participate in the intervention study. All parents having children in their final year of kindergarten were then invited to participate in the study and 125 out of 148 families signed consent forms. The first assessment took place in September 2012 and the intervention phase started in the beginning of November 2012.
The intervention consisted of two parts. First, the parents of all children in the sample were invited to a parents’ evening, which lasted for about 40 min. Three such evening meetings were held in three different locations and on 3 different days in 1 week. At least one of these events was held close to the home of each participant. During these evening meetings, a PowerPoint presentation provided parents with general information about the importance of the home learning environment, and about parents’ and families’ important role in contributing to child development from birth to school entry.
In addition, parents were provided with specific suggestions about how to improve the learning environment of their child (e.g. read often to your child, frequently use libraries with your child, read yourself, value reading and talk about reading at home, teach your child grapheme-phoneme correspondence, play word games with your child, listen to audio books with your child, demonstrate the how reading and writing are important in everyday life). After this information was provided, the parents had the opportunity to ask questions and to discuss these suggestions. At the end of each session, parents were provided with an information sheet containing information about all the suggestions to improve the HLE to take home and keep. In addition, they were invited to take part in the second part of the intervention.
The second part of the intervention consisted of a trained research assistant facilitating a 30-min session with one child, with his or her mother and/or father. During these sessions, the parent read the first three pages of a suitable German children’s book to the child and was observed by the assistant. During the reading session, guided by elements of the dialogic reading approach (cf. Whitehurst et al., 1988), the observer noted how the parent read to the child. For example, were questions asked regularly, did the reader expand on the information provided by the text in the book, were connections made between text and pictures and pointed out to the child, was the child encouraged to talk about the illustrations, were comments made by the child repeated and expanded, and did the reader show interest in the reading session?
Afterwards, elements of the parent’s book-reading that aligned with a dialogic reading approach were pointed out as strengths, and recommendations were made about aspects of the book-reading that parents may try to improve, pointing out why these aspects of reading are important. At the end of the session, parents had the opportunity to ask further questions, and each child was presented with the book as a gift.
Fifty-nine parents (47.2 %) attended the parent evening meetings, and 64 parents and their children (51.2 %) took part in the reading intervention. As some of the parents joined in only one of the two steps, only 47 parents (37.6 %) took part in the whole intervention. Thus, our sample consisted of four different groups: (1) no parent evening/no intervention (n = 49), (2) parent evening only (n = 12), (3) intervention only (n = 17), and (4) parent evening and intervention (n = 47). There were no statistically significant differences between these groups with regard to age, sex, SES, migration background, HLE, intelligence, vocabulary, letter knowledge or rapid naming at t1 (in each case p > .05).
Overview of statistical analyses
All analyses were conducted with SPSS 21. Results of correlational analyses and descriptive statistics will be presented first. In a second step, analyses of variance with repeated measurement were used to explore whether the four groups differed in their competency development (controlled for intelligence and sex), and whether the reported HLE changed (repeated-measures ANCOVA and repeated-measures ANOVA). Finally, difference values (values at t2 − values at t1) were calculated and the four groups were compared directly with t tests.
Descriptive statistics and correlational analyses
Means, standard deviations, minima and maxima of all measures and correlational analysis
Age in months t1 (1)
Sex # (2)
HLE t1 (3)
HLE t2 (4)
Vocabulary t1 (6)
Vocabulary t2 (7)
Letter knowledge t1 (8)
Letter knowledge t2 (9)
Rapid naming+ t1 (10)
Rapid naming+ t2 (11)
Effect of the intervention
In a second step, we analysed whether the intervention had some effect on the HLE. First, we checked how often the children’s book was used. Parents who took part in the individual intervention sessions reported having used the children’s book about once a week. Only 14 % of parents reported having used it rarely or never.
As a last step, analyses of variance with repeated measurements for all linguistic measures were conducted with intelligence and sex as control variables (repeated-measures ANCOVAs). In regard to vocabulary, a statistically significant interaction between time and sex was observed (girls learnt more new words than boys). In addition, a statistically significant interaction between time and intervention groups was found; F(3, 119) = 2.82 (p < .05; η2 = .07). Figure 1 presents the gain in vocabulary (vocabulary at t2 − vocabulary at t1) for all four groups and the significance and effect sizes of the differences obtained with t tests.
Turning now to letter knowledge gain, no statistically significant impact of intervention was found, although there was a tendency that the groups who took part in the intervention showed a greater letter knowledge gain; F(3, 119) = 1.35 (p > .05; η2 = .03). Of all groups, the non-participating group had the smallest gain in letter knowledge. For the group who only attended the parent evening the greatest gain was found. Thus, the comparison of these groups resulted in a marginally statistically significant medium effect size; t(59) = 1.73; p < .10; d = .56). In regard to the control variables, only intelligence played a statistically significant role in the development of letter knowledge between t1 and t2; F(1, 119) = 8.34 (p < .01; η2 = .07).
The speed gain in the rapid naming tasks was neither influenced significantly by the control variables nor by the intervention, although for the latter a small effect size was found; F(3, 119) = .80 (p > .05; η2 = .02). Again, the non-participating group had the smallest speed gain of all groups. Small, but statistically non-significant effect sizes (d = .20–.45) were found for all direct comparisons between this group and the groups that participated in at least some part of the intervention.
The significance of the learning environment a family provides for the development of children’s linguistic competencies has been discussed and analysed for a long time (e.g. Dunning et al., 1994; Lonigan, 1994; Scarborough & Dobrich, 1994). Longitudinal studies have shown that the HLE is an important predictor of later linguistic competencies of children, partly through direct connections and mainly by being closely associated with precursors such as vocabulary or letter knowledge (e.g. Aikens & Barbarin, 2008; Niklas & Schneider, 2013). Thus, it is not surprising that family literacy programs are able to improve children’s linguistic abilities (Mol et al., 2008; Sénéchal & Young, 2008).
However, most interventions were of long duration and high-intensity, leading to relatively high costs and effort (e.g. Cronan, Cruz, Arriaga, & Sarkin, 1996; Harper et al., 2011; Lever & Sénéchal, 2011; Saint-Laurent & Giasson, 2005). In our study we used a non-intensive, two-step intervention with a sample of 125 children from families of above average SES. This intervention consisted firstly of a parents’ evening, during which information about how to improve the learning environment was provided. Secondly, a short individual session for each interested parent and child was offered, during which parents were trained in the use of dialogic reading strategies and a children’s book was given as a gift to the child.
Despite of the limitations of this study discussed below, results indicated that this non-intensive intervention may have helped parents to improve not only the home learning environment, but may also have supported children’s linguistic competencies in a fairly short period of time. Thus, even very circumscribed interventions may have an impact on children’s development.
As expected, the effect on vocabulary was greatest as this competence seems to be easily susceptible to interventions such as introducing dialogic reading (Mol et al., 2008). The group who attended the parent evening only showed the highest gain in letter knowledge. This is of interest since the training in grapheme-phoneme correspondence was only mentioned at the parents’ evening. By comparison, the vocabulary gain was greatest in both groups whose parents attended the individual intervention sessions that focused predominantly on dialogic reading.
A statistically significant gain in rapid naming speed due to the intervention was not expected as this ability is fairly hard to train (Marx, 2007). Nonetheless, small but statistically non-significant effect sizes were found when comparing the non-participating group with the intervention groups. This finding indicates that our intervention may, to some degree, have contributed to improving children’s overall linguistic competencies.
Our findings differed from earlier results that found no statistically significant effect of family literacy programs (e.g. St. Pierre et al., 2005). One reason for our results may be the composition of our sample. St. Pierre et al. (2005) concluded that the “Even Start” family literacy program probably showed no effects due to the lack of full participation and to the instruction delivery. Our sample, however, consisted of parents with an-above average mean SES who were very willing to participate. This also led to us being able to assess data from the whole sample at both t1 and t2.
Some limitations mark the present study. Firstly, there was no random assignment and thus no actual control group in our study. Due to limited resources and a small sample size, no waiting-list control group could be used. As a consequence, comparisons were only possible for participating versus non-participating families. This non-random design implies that there might have been a selection effect, and it is possible that only families with a strong interest in educational issues and who were most motivated decided to take part in the intervention. However, we tested for initial differences between the groups and we controlled for intelligence and sex when comparing children’s linguistic development. There were no statistically significant differences between the groups in regard to all study variables at t1, indicating that participating and non-participating families were fairly similar.
In addition, parents were asked how important they deem the education of their children to be and how important it is for them that their children enter the highest education track (i.e. the German “Gymnasium”). Education and school track was more important for non-participating parents than for parents who participated in both phases of intervention (statistically significant, p < .05). Consequently, no differences in regard to parental interest and motivation in favour of participating families were found. Nevertheless, one cannot rule out that the groups differed in regard to other unobserved variables. Therefore, the results must be interpreted with caution.
Furthermore, no follow-up measure was included in our analyses and so we cannot be sure whether the gain in competencies was sustained over a longer period of time. On the other hand, our intervention seemed to have an impact on the HLE. If this change persisted over time, children’s linguistic development should also be influenced more permanently, perhaps cumulative effects may even occur (Dunning et al., 1994).
Unfortunately, no measures of reading and spelling could be included as our sample had not yet commenced formal education. Thus, despite the vocabulary gain and the tendency of improvement in letter knowledge and rapid naming we do not know whether children in the total-intervention group really started school better prepared. However, vocabulary and letter knowledge have proved to be important predictors of reading and spelling skills in school in other studies, and consequently one can expect the competencies gained in these variables to be relevant for reading acquisition in school (cf. Dickinson et al., 2012; Ennemoser et al., 2012; Torppa et al., 2006, 2007).
Finally, a fairly small sample was analysed, resulting in very small group sizes. Consequently, many results remained statistically non-significant, although small to medium effect sizes were found. In addition, many families in our sample had an above-average SES. Therefore, one cannot be sure whether the results would be representative for randomised samples. Given that our intervention was based on the willingness and ability of parents to improve the HLE and support their children, it is possible that this is due largely to the higher mean social status and the values that are typically associated with such a high status in our sample (cf. St. Pierre et al., 2005).
However, repeating the analyses with families with lower SES only did not change the picture. Analysing families with lower SES only reduced group sizes to N = 17 for the non-participating group, N = 8 for the group attending the parent evening, N = 6 for the group attending the second part of the intervention only and N = 23 for the group with complete intervention. Again, the non-participating group showed the smallest gain in HLE and in vocabulary and the smallest speed gain in rapid naming of all groups. For the non-participating group about the same gain in letter knowledge as for the intervention groups was found (about 1 letter on average), whereas the group attending the parent evening only had again the greatest gain in letter knowledge (about 2–3 new letters on average). For families with lower SES in comparison to the whole sample greater vocabulary gains were found in all groups. Whereas children in the non-participating group learnt about 3 new words on average between t1 and t2, children whose parents attended the parent evening only were able to name more than 3.5 new words on average. The greatest gains were again found for the intervention groups with more than 4 new words on average for the complete intervention group and 5 new words on average for the group attending the second part of intervention only. Consequently, the results seem also to be valid for families with lower SES. Nonetheless, replications of this study with samples from different backgrounds and future research on family literacy programs have to test whether our results are valid for other populations and samples as well.
Despite these limitations this study also has several strengths. We were able to show that even non-intensive interventions may help to improve the home learning environment and to support children’s linguistic development. Our results indicate that when parents are provided with relevant information about the learning environment in the family and about dialogic reading, this may contribute to the provision of a more favourable learning environment. Improvements to the HLE may be effected even on a small budget.
Whilst the children did not differ significantly at the beginning of the study, the participating groups showed a somewhat greater gain in linguistic competencies across all measures than the non-participating group and so we argue that the intervention may have at least some effect on overall linguistic competencies (cf. Mol et al., 2008; van Steensel et al., 2011).
Finally, our findings indicate that the HLE and children’s linguistic competencies can be improved even in samples that already provide fairly favourable HLEs. Whereas some studies focused on children at risk or found effects for these children only (cf. Hargrave & Sénéchal, 2000; Harper et al., 2011), we were able to show that a non-intensive intervention may further increase high levels of HLE and thus support the further development of children benefiting from high-level HLEs. The next step would be to use this kind of intervention in other contexts.
The present study shows the effects of a non-intensive intervention on both the HLE and on children’s linguistic competencies. Providing parents with adequate information about how to improve the HLE in the family at one parents’ evening session, and helping parents to use dialogic reading at home in one individual reading session, helped to support the development of the HLE and supported gains in children’s vocabulary in comparison with a non-participating group at least in a sample with above average mean SES. Moreover, statistically non-significant, but nonetheless meaningful, effects were found for the development of letter knowledge and rapid naming, indicating at least to some degree an effect on overall linguistic competencies. Although the findings may be limited to samples with high willingness and an ability to implement the provided information, we were able to show that improvement seems possible with little cost and effort. Therefore, despite budget constraints for interventions for family interventions, inexpensive non-intensive interventions make a difference to children’s learning.
- Burgemeister, B., Blum, L., & Lorge, J. (1972). Columbia Mental Maturity Scale. New York, NY: Harcourt Brace Jovanovich.Google Scholar
- Burgess, S. R. (2002). The influence of speech perception, oral language ability, the home literacy environment, and pre-reading knowledge on the growth of phonological sensitivity: A one-year longitudinal investigation. Reading and Writing: An Interdisciplinary Journal, 15, 709–737.CrossRefGoogle Scholar
- Dickinson, D. K., Griffith, J. A., Michnick Golinkoff, R., & Hirsh-Pasek, K. (2012). How reading books fosters language development around the world. Child Development Research, 1–15. doi:10.1155/2012/602807.
- Ennemoser, M., Marx, P., Weber, J., & Schneider, W. (2012). Spezifische Vorläuferfertigkeiten der Lesegeschwindigkeit, des Leseverständnisses und des Rechtschreibens. Evidenz aus zwei Längsschnittstudien vom Kindergarten bis zur 4. Klasse [Specific precursors of decoding speed, reading comprehension, and spelling: Evidence from two longitudinal studies from kindergarten to Grade 4]. Zeitschrift für Entwicklungspsychologie und Pädagogische Psychologie, 44(2), 53–67.CrossRefGoogle Scholar
- Esser, G. (2002). BUEVA—Basisdiagnostik für umschriebene Entwicklungsstörungen im Vorschulalter [BUEVA—Basis diagnostics for developmental disorders at pre-school age]. Göttingen: Beltz.Google Scholar
- Hattie, J. (2013). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. London: Routledge.Google Scholar
- Kiese-Himmel, C. (2005). AWST-R—Aktiver Wortschatztest für 3- bis 5-jährige Kinder [AWST-R—Active vocabulary test for 3- to 5-year-old children]. Göttingen: Hogrefe.Google Scholar
- Marx, P. (2007). Lese- und Rechtschreiberwerb [Acquisition of reading and spelling competences]. Paderborn: Ferdinand Schöningh.Google Scholar
- McElvany, N., Becker, M., & Lüdtke, O. (2009). Die Bedeutung familiärer Merkmale für Lesekompetenz, Wortschatz, Lesemotivation und Leseverhalten [The role of family variables in reading literacy, vocabulary, reading motivation, and reading behavior]. Zeitschrift für Entwicklungspsychologie und Pädagogische Psychologie, 41(3), 121–131.CrossRefGoogle Scholar
- Niklas, F., Möllers, K., & Schneider, W. (2013). Die frühe familiäre Lernumwelt als Mediator zwischen strukturellen Herkunftsmerkmalen und der basalen Lesefähigkeit am Ende der ersten Klasse [The early learning environment provided by the family as a mediator between structural family background and basic reading abilities at the end of Grade 1]. Psychologie in Erziehung und Unterricht, 60, 94–111.CrossRefGoogle Scholar
- Piasta, S. B., Petscher, Y., & Justice, L. M. (2012). How many letters should pre-schoolers in public programs know? The diagnostic efficiency of various preschool letter-naming benchmarks for predicting first-grade literacy achievement. Journal of Educational Psychology, 104, 945–958.CrossRefGoogle Scholar
- Puolakanaho, A., Ahonen, T., Aro, M., Eklund, K., Leppänen, P. H. T., Poikkeus, A.-M., … Lyytinen, H. (2008). Developmental links of very early phonological and language skills to second grade reading outcomes: Strong to accuracy but only minor to fluency. Journal of Learning Disabilities, 41, 353–370.Google Scholar
- Schneider, W., & Näslund, J. C. (1999). The impact of early phonological processing skills on reading and spelling in school: Evidence from the Munich Longitudinal Study. In F. E. Weinert & W. Schneider (Eds.), Individual development from 3 to 12: Findings from the Munich Longitudinal Study (pp. 126–147). Cambridge: Cambridge University Press.Google Scholar
- Weinert, F. E., Schneider, W., Stefanek, J., & Weber, A. (1999). LOGIC: Introduction and overview. In F. E. Weinert & W. Schneider (Eds.), Individual development from 3 to 12. Findings from the Munich Longitudinal Study (pp. 1–8). Cambridge: Cambridge University Press.Google Scholar
- Whitehurst, G. J., Falco, F. L., Lonigan, C. J., Fischel, J. E., DeBaryshe, B. D., Valdez-Menchaca, M. C., & Caulfield, M. (1988). Accelerating language development through picture book reading. Developmental Psychology, 24, 552–559.Google Scholar
- Wieler, P. (1997). Vorlesen in der Familie. Fallstudien zur literarisch-kulturellen Sozialisation von Vierjährigen [Reading to children in the family. A case-study about literary–cultural socialisation of four-year-olds]. Weinheim: Juventa.Google Scholar