Their teacher can't be an app: Preschoolers' listening comprehension of digital storybooks

Embedded dictionaries/glossaries

Jones and Brown (2011) found that their 3rd grade participants utilized the glossary and audio pronunciation feature of e-books during independent reading time, but did not connect this finding causally to improvements in reading comprehension. In addition, De Jong and Bus (2002) found that preschool-aged students clicked on these features in story listening sessions, wherein the ‘game’ features of the digital storybook media were restricted (when games were enabled, children attended less to the text). Korat and Shamir (2007) also found positive results with dictionaries and audio pronunciations with Kindergarten populations, although their design required that each student utilize the e-book with the dictionary features enabled. Not all e-books are equipped with this feature though (De Jong and Bus, 2003; Korat and Shamir, 2004; Paciga, 2009).

Hotspot animations

There are two kinds of hotspots documented in the research. The first is said to support, or align to, the story. The second is said to disrupt, or not align to, the story. Support for the possibility that interaction with certain kinds hotspots might disrupt word learning specifically and comprehension in general is found in research with adult learners – this body of research has found that the addition of interesting, but irrelevant, information to a passage (i.e. ‘seductive details’) can significantly impede learning (Lehman et al., 2007; Renninger et al., 1992), but similar results have not been consistently confirmed in digital storybook comprehension research with young children. Labbo and Kuhn (2000) found that seductive details interfered with a Kindergarten child's coherence in retelling the story while De Jong and Bus (2004) found children frequently interacted with the animations often embedded in electronic stories, but did not find any evidence to support that interactive animations distracted children from listening to the text, nor that animation interfered with story understanding measured by an emergent reading of the text. Another De Jong and Bus study (2002) provided children with multiple opportunities to interact with enhanced electronic storybooks containing incidental hotspots (i.e. unrelated to the story's plot) and found that, given a choice between interactive animations hidden in the illustrations and text read orally, none of the participants elected to listen to the story multiple times (i.e. without the interactive enhancements), illustrating the attractiveness of incidental hotspots. Most recently, Chiong et al. (2012) concluded that features of the enhanced e-book may have affected children's story recall because both parents and children focused their attention on non-content, more than story-related issues – a finding similar to that of Labbo and Kuhn (2000) when they investigated the story understanding of enhanced e-books. Furthermore, the parents and children surveyed and observed in Chiong and colleagues’ study found print books to be more advantageous for literacy building, whereas the e-books, particularly the enhanced e-book, were more advantageous for engaging children.

Recently, Smeets and Bus (2012) determined that their e-book featuring a digital storybook with embedded extra-textual vocabulary instructions produced significant differences in preschoolers' vocabulary learning. To arrive at this conclusion, Smeets and Bus designed a for-research digital storybook that narrated a page of the story and then had a virtual tutor/teacher ask the students to respond to multiple choice questions to identify an image that illustrated target vocabulary. Their results show that embedding questions about vocabulary was more effective than just exposing children to the definitions or a synonym.

Gestural supports

When texts become animated on screen or when a human reader enacts part of the story, the listener experiences an additional form of semiotic meaning – the gesture. There is some support in the research demonstrating positive effects of gestures on language learning (e.g. Hadar and Pinchas-Zamir, 2004; Sueyoshi and Hardison, 2005) and there are several studies that show positive correlations between animation containing gestures in digital storybooks and measures of story understanding. Ricci and Beal (2002) concluded that exposure to the gestures present in animated digital storybooks (with or without interactivity) yielded better story memory with first grade participants than just hearing the audio track. Later, Verhallen et al. (2006) produced similar results with a preschool population – more story recall after viewing animated interactive stories containing gestural components than just viewing the story with static images. Verhallen and Bus extended their exploration of vocabulary in digital storybook experiences in 2010. In this study, they improved second-language preschoolers' receptive vocabulary learning through experiences with animated digital storybooks.

Story summary

This is a tale of a baby fruit bat that gets separated from her mother when an owl attacks them. Stellaluna falls into a birds' nest and makes friends with the three baby birds that live there. Eventually, the mother bird takes in the stray bat, as long as Stellaluna leaves her bat ways behind and acts like a bird – meaning that she sleeps upright and eats bugs with the other birds. Concomitantly, the mother bat searches for her lost baby. One day Stellaluna and the baby birds go out flying and Stellaluna gets lost in the forest at dusk. The other birds return to their nest without their bat friend. That night, Stellaluna sleeps upright in a mango tree and is approached by several fruit bats (one is Stellaluna's mother) wondering why she is sleeping upside down. After Stellaluna tells her story, the mother realizes that the upside down bat is her lost child. The next day, Stellaluna returns to the birds' nest and invites them to fly at night to the mango tree. The birds agree only to find that they cannot find their way at night. Stellaluna guides the birds to safety and the story's ending conveys the theme – birds and bats are different, and yet alike.

Ensuring preschoolers' listening comprehension of Stellaluna

Comprehension is the understanding and interpretation of what is read. To be able accurately to understand written material, children need to be able to (1) decode what they read (or, for young children not yet reading independently (like the participants of this study) understand the language that is read aloud); (2) make connections between what they read (or listen to) and what they already know; and (3) think deeply about what they have read (or heard). One big part of comprehension is having sufficient vocabulary. Children who have strong comprehension are able to draw conclusions about what they read (or hear) – what is important, what is a fact, what caused an event to happen, which characters are funny or serious, and so on. Thus, comprehension involves combining reading (or listening) with thinking and reasoning.

Given that the role of the text is paramount in the comprehension puzzle (RAND Reading Study Group, 2002), it is important to examine fully the features of the original version of Stellaluna (Canon, 1993); doing so makes clear what it takes for a preschooler to understand this challenging text. First, it is useful to present the elements of the story (i.e. story grammar or dramatic structure) (Black and Wilensky, 1979; Brewer and Lichetnstein, 1981) – characters, setting, problems, events and solution or resolution – and simultaneously to examine the nuances of the language and images employed in the story.

There are a total of eight characters in Stellaluna – Stellaluna, Mother Bat, an owl, Mama Bird, three baby birds (named Pip, Flitter and Flap) and group of other bats (without names). Stellaluna is the main character. She is present in every one of the 13 plot episodes, or events, in the story. The other characters are supporting characters. The owl initiates the problem in the story by attacking Stellaluna's mother. As a result Stellaluna is left on her own. Later, when Stellaluna falls into the birds’ nest, a possible solution to her problem is presented – if she stays with the birds she will no longer be without a mother. This presents another problem, though – because Stellaluna is a bat, she does not know how to behave like a bird. There are a series of episodes wherein Mama Bird and the three baby birds try to teach Stellaluna to ‘act like a bird’. Each of these presents a minor problem and a resolution. Then, at the climax of the story, Stellaluna gets lost again. She subsequently meets up with a group of bats and reunites with her mother. This reunion represents the first steps to the story's resolution. Stellaluna and her mother ‘act like bats’ and Stellaluna wants to share these experiences with her bird family and does so – with an author's twist to exacerbate the differences between bats and birds. The story concludes with an indirect restatement about the differences between bats and birds and an implication that birds and bats can remain friends despite their differences.

A classroom teacher or parent who is adept in supporting their child's or children's listing comprehension in school or home contexts would probably build the child's background knowledge about birds and bats before, during and after the read-aloud. Depending on the depth of the child's existing background knowledge about birds and bats, an adult would provide more or less information that is critical to understanding the story. General scientific knowledge about the habits of bats and birds and understanding of emotions are critical in drawing conclusions about the story's themes – identity and friendship. When Stellaluna is first without her mother, she trembles ‘with cold and fear’. Understanding a baby's need for its mother for warmth would help a young child properly connect a cause to this event. Being able personally to connect with the vocabulary word ‘embarrassing’ would help a young child to understand that Stellaluna really wanted to land on the branch like the birds, but could not do so because her instinct and anatomical structure prevented her from landing on her feet.

General taxonomic (i.e. categorical) knowledge would help the young child connect challenging words like mango or grasshopper to fruit and insect, respectively. These larger taxonomic categories of insects and fruits are less specific terms and therefore are generally acquired first (Gelman et al., 1998). Such connections would be useful in retaining story information in the child's memory. Moreover, properties of specific components of a given taxonomy would accomplish the same goal – retention of story content. For example, bats and birds are both winged creatures. Bats are nocturnal. They sleep hanging upside down. Fruit bats eat fruit. They can sleep anywhere they find space to hang and they use echolocation to ‘see’ where they are going at night. On the other hand, birds are diurnal. They sleep perched upright, build nests and mostly eat insects/crawly things. This information is key in understanding the story's resolution and the theme of the story. Without it a child might think that Stellaluna enjoys ‘acting like a bird’ and, therefore, desires to continue living with and acting like the birds.

In the process of supporting critical background knowledge, the adult would probably present, revisit, explain, highlight or demonstrate some challenging vocabulary presented in Stellaluna to make clear to the young listener the story's content. Among these are words like spied, limp, twig, clutched, trembling, grasshopper, gracefully, embarrassing, clumsy, anxious, ached, peculiar and limb. Because of the nature of children's oral language and vocabulary development, several small misconceptions – primarily related to definitions of words (i.e. superordinates, synonyms, traits etc.) – about each episode's content are possible (see Hoffman et al. (2013) for an explanation of nuanced vocabulary understanding in read-alouds). For example, a child might say that an eagle, rather than an owl, attacked the baby bat. This would be an acceptable misconception because some of the features of an eagle are similar to those of an owl (e.g. winged, claws, larger than a bat).

In addition to building background knowledge and supporting key vocabulary understanding during a storybook reading, an excellent teacher would also make periodic stops throughout the story to summarize events, think aloud about/model his/her own story understanding (e.g. ‘I remember one time I was really embarrassed when …’), rephrase and check for the child's’ understanding. In these ways and others outlined in reputable guides to reading aloud (e.g. Sipe, 2008; Trelease, 2006), teachers continually support and monitor story understanding in a read-aloud event. The researchers believe the four presentation format features of Stellaluna utilized in this study accomplished support of comprehension to varying degrees. The following sections identify the formats of the digital story employed in this study describe and illustrate the presentation features of each.

Presentation format features

There were four commercially available presentation formats of Stellaluna that were given acronyms for this study (all formats were utilized on desktop computers with auxiliary mice). The first and second formats were housed on CD-ROM technology (CD-ROMA and CD-ROMB). The third format is available through http://www.onemorestory.com (OMS), and the fourth is available through http://www.storylineonline.net (SO). The story's plot, theme and general sequence of events were constant across the presentation formats, but there were differences across the four presentation formats in the features associated with the activity structure and the text that might affect story comprehension. We specifically identified filmic features, extra-textual discourse, interactivity through hotspots and the duration of the storybook listening experience as features that differed across some of the four presentation formats of Stellaluna. These differences are summarized in Table 1 and elaborated upon in subsequent paragraphs.

OPEN IN VIEWER

Table 1.

Comparison of main features of Stellaluna presentation formats.

	CD-ROMA	CD-ROMB	OMS	SO
Duration of activity	24:48	Unable to determine	20:56	10:46
Filmic effects	Full animation	Full animation	Static images	Panning, cropping, static image
Extra-textual discourse	Automatic	Upon click	None	None
Hotspots	Absent	Page, text, illustration	Page and text	Absent

Filmic effects

Filmic effects included musical accompaniment, animation, cuts and pans. These features were present in three of the presentation formats for this story. The OMS format does not employ any of the aforementioned filmic features. It presents the participant, rather, with static images that can be found in the paper version of the text. The OMS format does have musical accompaniment between page turns. One limit to this presentation format with respect to filmic features is the lack of visual gestures to help convey the story's meaning. The facial expressions remain constant on each page of the text.

CD-ROMA and CD-ROMB contain full animation with accompanying music. In this way, the CD-ROM formats offer a story-listening experience that replicates many features of television with semiotic components of gestural meaning. The participant in the story-listening activity views the character's gestures and changing facial expressions in the animations. In addition, both CD-ROM formats include many animated images that are ‘gratuitous intrusions into the story’ (Unsworth, 2003: 5) and do not appear in the printed version in any form. For example, when Stellaluna falls into the birds’ nest, there are several hotspots on the page that a child can click on once the text has been narrated. On this particular page (see Figure 1), a ladybug winds around on the tree branches surrounding the nest and music reminiscent of someone tiptoeing plays if a child clicks on the ladybug. This is an example of a ‘gratuitous intrusion’ because the ladybug is neither protagonist nor antagonist nor related to any part of the plot or theme. Intrusions of this type are present on every page of the CD-ROM texts, in both presentation formats, although some have argued that the animations, music and voicing of CD-ROMA are primarily congruent with the plot (Labbo and Kuhn, 2000). OPEN IN VIEWER

The SO format relies primarily on static images in the visual component of the text, but incorporates crops and pans to draw the reader's attention to specific parts of the illustration when they are referred to in the oral rendition. The images are not fully animated, but are not entirely static, either: characters do not flap their wings or talk (as in the CD-ROM formats), but they do appear to move resultant from filmic effects such as panning and cropping. For example, when the bat attacks Stellaluna and her mother, the image is cropped and presented in square segments – first the owl's mouth, then the wing, then the claw, and finally the bats being attacked. Moreover, when the text reads, ‘she clutched the thin branch, trembling with cold and fear …’ the bat shakes back and forth on the branch in the SO format, probably contributing to children's understanding of the word tremble. The participants in this condition also view the narrator's gestures and facial expressions in addition to the static facial expressions of the characters in the static illustrations. This additional source of semiotic meaning may facilitate story comprehension in the SO condition.

Extra-textual discourse

Extra-textual discourse is the explanatory discourse beyond the words contained in the published text. Parents and teachers employ extra-textual discourse when reading with young children – they respond to children's questions, elaborate on the plot, clarify, extent, point out action presented in the illustrations, question to check for understanding, summarize and much more. Some multimedia environments embed this kind of language. Both CD-ROM formats of Stellaluna contain extra-textual discourse. Characters in CD-ROMA (Read to Me mode) engage in some extra-textual comprehension-supportive discourse. For example, when the mother bird brings food to the nest the text reads, ‘Stellaluna was terribly hungry, but not for the crawly things Mama Bird brought’. In the CD-ROMA presentation format, the baby birds question Stellaluna, ‘What's the matter? Aren't you hungry? Don't you want some?’ and probably support children's understanding that Stellaluna does not desire to eat the bugs that Mama Bird offers. The same discourse is embedded in the CD-ROMB version (i.e. the ‘Let Me Play’ mode), but the child has to click on the birds in the illustrations to hear this discourse.

Hotspots

An important subplot of the story of Stellaluna is embedded in the illustrations in the printed text. On every two-page spread, a thumbnail illustration above the printed words shows the mother bat searching for Stellaluna in different parts of the forest. These thumbnail illustrations are not present in two of the four presentation formats associated with this research (OMS and SO). The CD-ROMA and CD-ROMB formats have the thumbnail on the pages of the text. In the CD-ROMB format, this thumbnail is a hotspot. When a child clicks on it, the scene pops out and shows the mother bat looking in a cave, tall grasses etc. and calling, ‘Stellaluna!’ If the participant clicks on these hotspots in CD-ROMB, he or she accesses information about (1) what has happened to the mother bat and (2) whether or not she cares that her child is missing. These pieces of information may contribute to story comprehension. Figure 1 illustrates this hotspot and identifies several other hotspots available only in the CD-ROMB version. The One More Story version of the story has page-turn hotspots only.

Duration of the activity

The first difference in the presentation formats is associated with the story listening activity. Duration is important when examining story comprehension from a cognitive perspective because this perspective recognizes the role of attentional load in comprehension. If the read-aloud activity endures beyond the limits of the child's attention span, he or she is unlikely fully to engage in the activity beyond that point (Paciga et al., 2009), which might affect a child's story comprehension. Anecdotally, teacher-led read-alouds in the classrooms participating in this study ranged from 15 to 35 m.

The duration of an electronic storybook presentation is the amount of time that elapses from the moment the cover of the book is shown and the title is read until the end of the story. CD-ROMA has a duration of 24 m 48 s. It is impossible to determine the duration of the read aloud of CD-ROMB as it depends on the degree to which each child interacts with the hotspots. Pilot data indicate that children engaged in this read-aloud activity for as little as 9 mins (before abandoning the activity) and up to 60 mins in experimental conditions (Paciga, 2010). The OMS format has a duration similar to that of CD-ROMA – 20 m 56 s. The SO format is considerably shorter than the other formats – approximately 11 m.

Story format

Participants were randomly assigned to one of the four story-listening conditions described earlier. The format of the story they experienced was entered as a dummy variable in the regression model to determine the extent to which the story presentation format was related to listening comprehension. In the analyses, the variables z1, z2 and z3 were coded 0 or 1 to indicate having listened to CD-ROMA, CD-ROMB, and OMS formats, respectively. A student who listened to the SO format would be represented by the values for z1, z2, and z3 all being equal 0.

Task completion

In the data-collection stage of the research, participants asked to abandon the listening activity frequently. When this happened, the researcher gave the preschooler three prompts to encourage them to complete the activity. After a fourth request to abandon the activity, the child was permitted to do so, but still responded to all probes for story understanding. The researcher documented whether each child did or did not complete the story-listening activity. This information became a categorical variable that was recoded as a dummy variable with 0 indicating the task was completed and 1 indicating it was not completed.

Assessment of topic knowledge (ATK)

In order to measure prior topic knowledge regarding bats in particular and animals in general, a 14-item, multiple-choice assessment was created. Each item on this assessment had three or four response options, with only one correct choice receiving a score of 1 and three to four distractors scored with a 0. To ascertain internal consistency, the 11 items on the ATK were entered into reliability analysis yielding Kuder–Richardson (KR20) = .71.

Concepts of screen/computer skills (COS) checklist

Because the context of this study was a digital environment, it was necessary to be certain that a child's skill in manipulating the hardware and software on a computer did not interfere with his or her comprehension of digital texts. The COS checklist is a product of Turbill's research (2001a, 2001b) studying computer use in a kindergarten classroom. There are 13 items in this observational checklist that provide information about each child's computer skills: ability to control the mouse, use the hardware and software on the computer, and navigate within a given program. The 13 use items and the one access item allowed for a maximum score of 40 on the COS. Ten per cent of the study's sample was observed by two researchers. Each researcher arrived at COS item-level ratings independent of the other. Comparison scores indicate 91% agreement in observed scores. In a pilot study (Paciga, 2009) data that indicated the items related to mouse control (i.e. use with one hand, clicking with appropriate fingers, scrolling, click and drag, matching hand to mouse movement) accounted for the most variance in COS scores.

Receptive vocabulary

The Peabody Picture Vocabulary Test, 4th Edition (PPVT) (Dunn and Dunn, 2007) is an individually administered, norm-referenced assessment of listening comprehension for spoken words (receptive vocabulary) with standardized administration and is appropriate for use with individuals as young as 2.5 years of age to adulthood. The standard score distribution of the PPVT has a mean of 100 and a standard deviation of 15. Standard scores were normed by age for this study.

Explicit Comprehension Probes (ECP)

Basic story grammar – characters, setting, rising actions and resolutions – addressing the main story elements and sequence of events present in Stellaluna constitutes the ECP. Responses were scored as correct (1 point) or incorrect (0 points) and the sum of the eight items constituted the child's ECP score. All of the information requested in the ECPs was essential to understanding the story of Stellaluna at a literal, basic level. Two researchers observed 10% of the participants responding to the ECP in the data-collection phase of the current investigation. Each researcher arrived at ECP scores independent of the other, with 98% agreement between the two observers in recorded ECP scores.

Implicit Comprehension Probes (ICP)

Six probes were used to assess each child's high-level text comprehension scored as ‘appropriate with response anchored in the text’ (2 points), ‘appropriate without response anchored in the text’ (1 point), ‘inappropriate’ (0 points) or ’unable to code’ (excluded from the analysis). The following responses illustrate the difference between a score of 2 and 1. There is a scene in Stellaluna where she is afraid because she is alone, cold and had recently been attacked by an owl. An ICP question asks participants why Stellaluna is afraid. The child who responded that ‘she was afraid because of a monster’ or ‘she was hurt’ received only 1 point because there was no monster in the text and there are no visual or linguistic cues indicating that the bat is injured (Stellaluna did, however, experience a similar emotion to one that a child might feel when he or she perceives there to be a monster). The child who responded that she was afraid because she had lost her mother received 2 points. The sum of points for this measure served as the ICP score. Two researchers observed 10% of the participants responding to the ICP in the data-collection phase of the current investigation. Each researcher arrived at ICP scores independent of the other, with 86.25% agreement between the two observers on recorded ICP scores.