Input Subject Diversity Accelerates the Growth of Tense and Agreement: Indirect Benefits From a Parent-Implemented Intervention Purpose This follow-up study examined whether a parent intervention that increased the diversity of lexical noun phrase subjects in parent input and accelerated children's sentence diversity (Hadley et al., 2017) had indirect benefits on tense/agreement (T/A) morphemes in parent input and children's spontaneous speech. Method Differences in input ... Research Article
Free
Research Article  |   September 18, 2017
Input Subject Diversity Accelerates the Growth of Tense and Agreement: Indirect Benefits From a Parent-Implemented Intervention
 
Author Affiliations & Notes
  • Pamela A. Hadley
    Department of Speech and Hearing Science, University of Illinois at Urbana–Champaign
  • Matthew Rispoli
    Department of Speech and Hearing Science, University of Illinois at Urbana–Champaign
  • Janet K. Holt
    Illinois Educational Research Council, Southern Illinois University, Edwardsville
  • Disclosure: The authors have declared that no competing interests existed at the time of publication.
    Disclosure: The authors have declared that no competing interests existed at the time of publication. ×
  • Correspondence to Pamela A. Hadley: phadley@illinois.edu
  • Editor-in-Chief: Sean Redmond
    Editor-in-Chief: Sean Redmond×
  • Editor: Jan de Jong
    Editor: Jan de Jong×
Article Information
Development / Research Issues, Methods & Evidence-Based Practice / Normal Language Processing / Language Disorders / Specific Language Impairment / Attention, Memory & Executive Functions / Speech, Voice & Prosody / Language / Research Articles
Research Article   |   September 18, 2017
Input Subject Diversity Accelerates the Growth of Tense and Agreement: Indirect Benefits From a Parent-Implemented Intervention
Journal of Speech, Language, and Hearing Research, September 2017, Vol. 60, 2619-2635. doi:10.1044/2017_JSLHR-L-17-0008
History: Received January 6, 2017 , Revised March 12, 2017 , Accepted April 26, 2017
 
Journal of Speech, Language, and Hearing Research, September 2017, Vol. 60, 2619-2635. doi:10.1044/2017_JSLHR-L-17-0008
History: Received January 6, 2017; Revised March 12, 2017; Accepted April 26, 2017

Purpose This follow-up study examined whether a parent intervention that increased the diversity of lexical noun phrase subjects in parent input and accelerated children's sentence diversity (Hadley et al., 2017) had indirect benefits on tense/agreement (T/A) morphemes in parent input and children's spontaneous speech.

Method Differences in input variables related to T/A marking were compared for parents who received toy talk instruction and a quasi-control group: input informativeness and full is declaratives. Language growth on tense agreement productivity (TAP) was modeled for 38 children from language samples obtained at 21, 24, 27, and 30 months. Parent input properties following instruction and children's growth in lexical diversity and sentence diversity were examined as predictors of TAP growth.

Results Instruction increased parent use of full is declaratives (ηp 2 ≥ .25) but not input informativeness. Children's sentence diversity was also a significant time-varying predictor of TAP growth. Two input variables, lexical noun phrase subject diversity and full is declaratives, were also significant predictors, even after controlling for children's sentence diversity.

Conclusions These findings establish a link between children's sentence diversity and the development of T/A morphemes and provide evidence about characteristics of input that facilitate growth in this grammatical system.

How do children use language input to acquire the grammar of their native language? Research traditions in psychology and linguistics have tackled this fundamental question by focusing on different aspects of the problem space (Chomsky, 1995; Ingram, 1989; Lidz & Gagliardi, 2015; Tomasello, 2003; Valian, 1999; Yang, 2004). Psychologists have contributed descriptions of the language input directed to children, insights about the social and cognitive underpinnings of learning, and evidence of domain-general, statistical learning mechanisms that children use to detect patterns in language input. Generative linguists have characterized children's grammatical representations, allowing hypotheses to be tested about the way abstract grammatical structures are related to one another in mental grammar and in development. Unfortunately, research programs about natural learning environments (e.g., Huttenlocher, Vasilyeva, Cymerman, & Levine, 2002; Huttenlocher, Vasilyeva, Waterfall, Vevea, & Hedges, 2007; Rowe, 2012; Vasilyeva, Waterfall, & Huttenlocher, 2008), learning mechanisms (e.g., Gómez, 2002; Thompson & Newport, 2007), and the nature of grammatical representations (e.g., Yang, 2002; Lidz & Gagliardi, 2015) often proceed in isolation. However, these enterprises need not remain independent. Moreover, the integration of insights from different disciplinary traditions is crucial for developing evidence-based grammatical interventions for young children.
By manipulating properties of language input, language interventions can be used to test hypotheses about learning mechanisms while addressing the practical need for effective and more efficient language interventions. For children with specific language impairment (SLI), efforts to evaluate the efficacy of grammatical interventions are particularly important, given the pronounced difficulty they have with the acquisition of grammar (Leonard, 2014; Oetting & Hadley, 2017). In light of this need, many clinical investigators have evaluated the efficacy of interventions for tense/agreement (T/A) morphemes (Fey, Cleave, & Long, 1997; Fey, Cleave, Long, & Hughes, 1993; Fey & Finestack, 2009; Leonard, Camarata, Brown, & Camarata, 2004; Leonard, Camarata, Pawłowska, Brown, & Camarata, 2006; 2008; Proctor-Williams, 2009; Tyler, Lewis, Haskill, & Tolbert, 2002, 2003). More recently, investigators have moved beyond comparisons of treatment agents, procedures, dosage, and goal attack strategies to explore how differences in input properties promote learning (Leonard, Fey, Deevy, & Bredin-Oja, 2015; Plante et al., 2014) and how differences in children's grammatical knowledge allow them to make better use of the therapeutic input provided (Leonard et al., 2004, 2006; Pawłowska, Leonard, Camarata, Brown, & Camarata, 2008).
For example, Plante et al. (2014)  manipulated statistical properties of input known to promote learning in artificial miniature learning studies in a grammatical intervention for preschoolers with SLI. The input manipulation was combined with conversational recasting, an empirically supported technique (Cleave, Becker, Curran, Van Horne, & Fey, 2015). Recasts are adult responses to a child's prior platform utterance (e.g., pizza cooking) that maintain the semantic content of the child's utterance while modeling the treatment target in a well-formed adult version of the utterance (e.g., The pizza is cooking). Recasts are inherently responsive turns, building upon the child's attentional focus and intention to communicate. Children received 30-min individual intervention sessions daily for approximately 5 weeks. Treatment targets (e.g., auxiliary is, past –ed) were individualized. Clinicians structured the play activities and discourse to increase children's attempts at treatment targets and responded to these attempts with recasts. Children were exposed to 24 recasts per session; however, the number of verbs co-occurring with treatment targets varied by condition. Children in the low-variability condition heard 12 different verbs recast twice, whereas children in the high-variability condition heard 24 different verbs recast only once. Plante et al. (2014)  observed significantly greater treatment effects for children in the high-variability condition. They concluded that high verb variability within a single session shifted children's attention toward their treatment targets—the more stable and frequent structures in the input (Gómez, 2002).
In another study, Pawłowska et al. (2008)  explored the relationship between pretreatment grammatical knowledge and the ability of preschoolers with SLI to learn agreement morphemes from focused input (Leonard et al., 2006). In this focused stimulation intervention, children received concentrated exposure to either third person singular –s or auxiliary is/are/was in each session, specifically 12 models in scripted stories and 12 conversational recasts in play. Children participated in two sessions per day, two times per week over 6 months for a total of 96 sessions. The authors hypothesized that children's developmental readiness to acquire agreement morphemes would depend upon their pretreatment ability to mark number with plural –s and produce subject–verb constructions, two basic skills underlying the production of sentence subjects marked for number. In evaluating children's treatment progress, Pawłowska et al. (2008)  found that posttreatment accuracy on the target morpheme was predicted by a combination of age, pretreatment accuracy with plural -s, and the percentage of utterances with lexical verbs co-occurring with a subject. In contrast, treatment progress was not predicted by mean length of utterance (MLU) or by grammatical structures unrelated to agreement (i.e., progressive –ing, nonthematic of) that develop in the same age range as plural –s. These findings demonstrate that children's grammatical knowledge has an influence on their ability to learn from focused input.
In a third study, Hadley et al. (2017)  used a parent-implemented intervention to manipulate the diversity of sentence subjects in parent input to toddlers with typical development to facilitate sentence development, specifically unique subject–verb combinations. Hadley et al. (2017)  hypothesized that lexical noun phrase (NP) subjects in parent input would alter cues that signal the boundary between subject and predicate constituents, and this would, in turn, strengthen children's representation of clause structure. Parents received instruction on responsive interaction and language modeling strategies during a 1.5-hr parent group education session and two 1-hr individualized coaching sessions (see the Method section and Hadley et al., 2017, for more details). Two novel toy talk strategies were used to alter parents' lexical NP subjects in declarative sentences: (a) Talk about the toys increased third person sentence subjects referring to concrete objects in the play environment and (b) give the object its name increased nouns in subject position. Following toy talk instruction, parents in the treatment group increased their frequency and diversity of lexical NP subjects in toy talk sentences (e.g., Your castle is wobbling) compared with parents in a quasi-control group. Although treatment condition (i.e., treatment vs. control) was only a marginal predictor of children's growth in sentence diversity after controlling for growth in children's number of different words (NDW), parent input subject diversity, operationalized as the number of different lexical NP subjects each parent produced, was a significant predictor. The significant input effects revealed at the level of the individual parent–toddler dyads make sense in the context of the parent-implemented intervention. That is, parents varied in how well they learned and delivered the intervention's active ingredient (i.e., the number of different lexical NP subjects in toy talk sentences).
Hadley et al. (2017)  proposed that multiple cues to clause structure were altered by parents' lexical NP subjects in declarative sentences. First, input samples with more low-frequency lexical NP subjects would have lower transitional probabilities or less predictability between the subject and the following verb (e.g., The oven → is hot) than input samples with primarily high-frequency pronominal subjects (e.g., This → is hot). Lower transitional probabilities between the nouns in the subject NPs and the verbs that follow should make the subject–predicate constituent boundary more prominent to the learner (Thompson & Newport, 2007). Second, declarative input sentences with lexical NP subjects might better align prosodic cues, such as syllable lengthening, pausing, and pitch changes with the subject–predicate constituent boundary (Fisher & Tokura, 1996), because declarative sentences with lexical NP subjects are more likely to be followed by uncontracted copula and auxiliary forms (e.g., The pig is hiding) than those with pronominal subjects (e.g., He' s hiding; Frank & Jaeger, 2008). In addition, lexical NP subjects could increase cross-sentential cues to constituent structure, specifically moved phrases and pronominalization (Hoff-Ginsberg, 1985; Morgan, Meier, & Newport, 1989), as parents moved lexical NPs from object position to subject position (e.g., I can't find the pig ; the pig is hiding) and varied pronouns with lexical NPs in the subject position of their input sentences. In sum, by increasing the frequency and diversity of lexical NP subjects in declarative input sentences, toy talk instruction shifted transitional probabilities, prosodic cues and cross-sentential cues to highlight the subject constituent, which, in turn, helped strengthen the child's representation of basic clause structure. The current study expands the investigation of Hadley et al. (2017)  by using the same study participants to determine whether toy talk instruction also changed properties of T/A morphemes in parent input without any specific instruction on these morphemes and whether these properties of parent input, in concert with children's growth in sentence diversity, promoted growth in T/A morphemes.
Toy talk instruction was expected to alter parent input for T/A morphemes in two ways, leading to additional benefits for children's T/A growth. First, we predicted increases in parents' input informativeness (i.e., the percentage of verb forms providing unambiguous evidence for tense marking; Hadley, Rispoli, Fitzgerald, & Bahnsen, 2011; Legate & Yang, 2007). Legate and Yang (2007)  revealed cross-linguistic differences in the percentage of verb forms unambiguously marked for tense. They found 80.1% and 69.8% of verb forms were unambiguous in child-directed Spanish and French, compared with only 52.9% in English. These percentages, importantly, aligned with empirical differences in the age of acquisition of tense across these languages. Applying this measure to individual parents, Hadley et al. (2011)  documented variation among 15 English-speaking parents, with the percentage of overtly marked verb forms ranging from 33.1% to 69.8% (M = 50.6%). Parents' input informativeness was also a significant predictor of toddlers' tense/agreement productivity (TAP) scores at 30 months, accounting for 23.0% of the variance between children. Given these findings, it was expected that toy talk instruction would increase parents' use of declarative statements about objects in the play environment (i.e., third person subjects) and therefore the third person singular cell of the English verb paradigm. Third person is most consistently and distinctively marked for agreement, so an increase in sentences with third person subjects was expected to increase input informativeness (Fitzgerald, Hadley, & Rispoli, 2013; Hadley & Walsh, 2014; Rispoli & Hadley, 2011; Rispoli, Hadley, & Holt, 2012). More informative input, in turn, was predicted to facilitate children's acquisition of the T/A system.
As treatment parents produced more lexical NP subjects in Hadley et al. (2017), we predicted they would use more full is declaratives as well, because full is forms in declaratives are associated with lower frequency NP subjects, whereas contracted 's forms are associated with higher frequency pronominal subjects (Frank & Jaeger, 2008). We hypothesized that parents' full is declaratives would facilitate the child's entry into the T/A system for several reasons. First, the derivational simplicity of declarative sentences may support a child's ability to parse the syntactic structure of input sentences (Rispoli, Papastratakos, Stern, & Hadley, 2015). Second, we hypothesized that high input subject diversity with is in declarative sentences would facilitate its analysis, whereas low input subject diversity would encourage rote learning of copula and auxiliary is and hinder morphemic analysis (Rispoli & Hadley, 2014). Four pronoun–'s combinations found in declarative contexts (i.e., that's, it's, he's, there's) are listed among the 100 most frequently produced words in parent input (Li & Shirai, 2000; MacWhinney, 2000). Note that this kind of rote learning is dependent on high transitional probabilities between the pronominal subject and the verb that follows. Third, the phonetic substance of syllabic is may facilitate its identification as a distinct word (Pinker, 1984). Thus, both lexical NP subjects and the associated tendency to use full is forms with these subjects in declarative input sentences could promote children's analysis of is. In turn, the identification of copula and auxiliary is in the input was predicted to facilitate children's acquisition of the T/A system through cross-morpheme facilitation (Leonard et al., 2004, 2006; Rispoli, 2016; Rispoli et al., 2012; Rispoli & Hadley, 2011, 2014). In cross-morpheme facilitation, the learning of one morpheme facilitates the learning of other morphemes that share grammatical features. The effects of cross-morpheme facilitation have been demonstrated from copula is to verb -s (i.e., third person singular present tense) for young children with typical development (Rispoli, 2016) and from verb -s to auxiliary is/are/was and from auxiliary is/are/was to verb -s for preschool children with SLI (Leonard et al., 2004, 2006).
Also, we expected children with stronger clausal representations to have an advantage analyzing the surface forms that encode T/A features in the input over children with weaker clausal representations. To detect and interpret T/A morphemes in the input, knowledge of clause structure is crucial because T/A features are properties of clauses, not words. As previously noted (Hadley et al., 2011), tense in English has scope over the clause, and agreement is the result of checking features of the subject with features of the verb phrase (Chomsky, 1995). In other words, we predicted children's ability to produce diverse sentences would promote acquisition of the T/A system because children use their developing grammar to parse input sentences and detect new grammatical structures (see Lidz & Gagliardi, 2015, for discussion). To test this hypothesis, we examined whether children's growth in sentence diversity, an indicator of the strengthening of clausal representation (Rispoli et al., 2012; Rispoli & Hadley, 2011), was a significant predictor of growth in T/A marking.
In summary, the purpose of this study was to expand our original investigation to determine whether toy talk, a parent-implemented, language modeling strategy previously shown to accelerate children's sentence diversity, also promoted children's acquisition of the T/A system. Our first research question examined whether toy talk instruction had indirect benefits on parents' use of T/A morphemes, specifically input informativeness and full is declaratives. The second research question examined whether developmental changes in children's sentence diversity and parent input properties facilitated growth of T/A morphemes in children's spontaneous speech.
Method
Design
The original study (Hadley et al., 2017) evaluated the efficacy of toy talk instruction on children's growth of sentence diversity by using a quasi-experimental design. Control families were drawn from an earlier longitudinal study of language development (Hadley, Rispoli, Holt, Fitzgerald, & Bahnsen, 2014); therefore, families were not randomly assigned to the treatment and control groups. The use of a quasi-experimental design is considered a cost-effective design element for scaling up language intervention research (Fey & Finestack, 2009). Although collected at different points in time, all treatment and control families were recruited from English-only speaking households in Champaign County, Illinois, and surrounding counties, following the same strategies. Information was distributed to parents through newspapers, community facilities, and listservs. Control families did not receive any instruction on language development or language facilitation strategies. All spontaneous language sampling procedures were the same for the treatment and control families, and all language samples from the control families were retranscribed to eliminate the potential for investigator bias in the current study.
Participants
Nineteen families and their toddlers with typical development participated in the intervention. All children passed the communication section of the Ages and Stages Questionnaire–Third Edition (Squires & Bricker, 2009) administered during a phone interview at 20 months of age and were reported to produce at least 25 different words. Children were excluded if parents reported neurological or sensory impairments, delayed onset of walking or talking, or regular exposure to a language other than English. Children were also excluded if parents reported any use of four-word combinations, an indicator that expressive abilities were too advanced for the intervention. The children in the treatment group (11 girls and 8 boys) were matched to 19 children from an archival database (9 girls and 10 boys) on parent-reported expressive vocabulary at 21 months from the MacArthur–Bates Communicative Development Inventories (CDI; Fenson et al., 2007) and on sex and parent level of education, when possible.
All treatment families were White, non-Hispanic (n = 19). The age of parent participants (17 women and 2 men) was approximately 35 years (M = 34.94; SD = 5.19), and their highest educational levels included associate's degree or some college (n = 1), bachelor's degree (n = 7), and advanced degree (n = 11). The mean CDI total at 21 months was 120.94 (SD = 63.00). Nine children were first born, nine were later born, and one was a twin. Seven children were in child care 5 hr or less per week, four were in child care 6 to 29 hr per week, and eight were in child care 30 hr or more per week.
The majority of families in the control group were also White, non-Hispanic (n = 15). One family was White Hispanic, and three families were Black. The age of the parent participants (18 women and 1 man) was approximately 30 years (M = 30.05; SD = 4.38), and their highest educational levels were high school (n = 1), associate's degree or some college (n = 3), bachelor's degree (n = 11), and advanced degree (n = 4). The mean CDI total at 21 months was 120.31 (SD = 56.65). Ten children in the control group were first born, eight were later born, and one was a twin. Eight children were in child care 5 hr or less per week, five were in child care 6 to 29 hr per week, and six were in child care 30 hr or more per week.
Procedures
The procedures of the original investigation are sumamarized here (for full details, see Hadley et al., 2017). The parent instructional component involved three education sessions between the 21- and 24-month measurement sessions, one 1.5-hr group session and two 1-hr individualized coaching sessions. The group session was divided into three segments. In the first segment (approximately 20 min), information on language development between 18 and 30 months was presented, including characteristics of single-word users, word combiners, and childlike sentence users. In the second segment (approximately 40 min), information on responsive interaction strategies and general language modeling strategies were presented by using the You Make the Difference parent resource (Manolson, Ward, & Dodington, 2007) and instructional videos. During this segment, examples of expanding child utterances referred to as adding language in the parent resource were highlighted. However, alternative ways to expand child utterances using sentences were also presented.

{Direct parents' attention to example in the book} The book suggests imitating pretty and then adding flowerpretty flower. You can also add flower in a short, complete sentence and use the expression in your voice to STRESS the two content words. The FLOWER is PRETTY.

In the final segment (approximately 25 min), the two toy talk strategies (i.e., talk about the toys; give the item its name) were formally introduced by using handouts (Hadley & Rispoli, 2015) and demonstrations. Two investigators, with one playing the role of the child, demonstrated responsive interaction strategies and toy talk during play with various toys (see below). Then, parents practiced the strategies through role play with the investigators and received feedback.

C {holding up a puzzle piece} Cow.

A You found a cow.

 = expansion, describes C's action

C {vocalizing, tries to put piece in}

A Help? {waits} Do you need help?

 = interprets C’s vocalization

C {puts cow in puzzle} Cow.

A The cow fits!

 = expansion with toy talk (both strategies)

A {puts another piece in}

The pig goes in too.

C {feeds baby a bottle}

 = A observes C

A You're feeding the baby.

 = describes C's actions, gives toy its name

A The baby is drinking.

 = toy talk (both strategies)

C Thirsty {points to teddybear}.

A Teddy looks thirsty.

 = expansion with toy talk (both strategies)

A He wants a drink.

 = talks about toy

A treatment fidelity checklist was used to document the delivery of 20 key instructional components for the 10 group education sessions. Two research assistants (RAs) who had not conducted the sessions completed the checklist. Point-by-point interobserver agreement for delivery of the 20 key components was 95.0% (SD = 5.27%).
Coaching sessions began with a review of responsive interaction and toy talk strategies. Parents described their use of the two types of strategies at home, then engaged in 20 min of parent–toddler free-play. The first 10 min of free-play was video-recorded. An investigator then played the video recording for the parent on a laptop computer, pausing approximately once each minute to identify positive instances of the parent's strategy use and to offer constructive feedback on child-centered and interaction-promoting strategies, general language modeling strategies, and/or toy talk strategies. The investigator provided two to three specific suggestions for practice at home, and a written summary was sent by e-mail after the session. Feedback to parents on responding with short, well-formed sentences and using toy talk included 33.3% and 24.2% of the video feedback, respectively.
The measures of parent input and child outcomes were obtained from conversational language samples collected in two contexts. The first context included 30 min of parent–child free-play with age-appropriate toys. Parents were instructed to play with their children “as they would at home.” For the second context, an investigator joined the parent–child dyad and used semistructured play scenarios to create opportunities for diverse verbs, sentence subjects, and T/A morphemes (see Hadley, 2014; Hadley et al., 2014, 2017; Oetting & Hadley, 2017). Transcription was completed by an unbiased team of undergraduate RAs who were unaware of the children's ages, treatment condition, toy talk strategies, and research questions. Language samples for the quasi-control group were also retranscribed alongside the language samples collected from the treatment group to eliminate investigator bias. All sessions were transcribed by using Systematic Analysis of Language Transcripts software (SALT; Miller & Iglesias, 2012). During a subsequent coding pass, RAs used computer-assisted searches to standardize spelling for NDW counts and added standard SALT conventions for all bound morphemes (e.g., jump/ed) and contractions (e.g., he/'s) for parents and children. They also examined child utterances to exclude imitations, self-repetitions, and routine expressions (e.g., counting, lyrics) from further analyses.
Input Measures
Input measures were computed at 21 and 24 months from the child-directed, spontaneous, complete, and intelligible parent utterances produced in the first 30 min of the parent–child sampling context. Toy talk sentences with lexical NP subjects previously coded for Hadley et al. (2017)  were also used in this study. Toy talk sentences met the following requirements. For syntax, clauses were required to be finite with canonical subject–verb–(object) word order as in declarative statements (e.g., The bubbles made a mess; I believe the piece goes right there) or discourse questions with no structural movement (e.g., The egg is hot?). Grammatical subjects could be either common nouns (e.g., The baby needs a bath) or proper nouns (e.g., Nina likes juice). In addition, the referent of the subject was required to be present in the playroom or part of the pretend play (e.g., {mm} This soup tastes good). An explicit predicate was also required. Predicates could describe the states, actions, properties, locations, or possessions of the referential subject. From all sentences meeting these criteria, the number of different lexical NP subjects in toy talk sentences was determined, reflecting the diversity of NP subjects in parent input.
Parent input informativeness was computed following the procedures of Legate and Yang (2007)  and Hadley et al. (2011) . Verb forms marked unambiguously (i.e., overtly) for T/A morphemes were coded as +Tense [+T], and ambiguous forms were coded −Tense [−T]. The overtly marked verb forms included all regular and most irregular past tense (e.g., The bubble pop/ed ; You did it), present tense third person singular (e.g., Danny love/3sgrapes), and finite forms of copula BE and auxiliaries BE, DO, and HAVE. Auxiliary–main verb combinations received only one code (e.g., What do you think?). Ambiguous verb forms included irregular past tense verbs that do not change their surface form (e.g., You put it away), zero-marked present tense verbs (e.g., I get the green one), and modal auxiliaries (e.g., can, will, should). Bare verb forms, such as imperatives (e.g., put your shoes on), serial verbs (e.g., go get your shoes), and bare infinitives (e.g., let's put them on) were also coded −Tense. Infinitives overtly marked with to were not coded (e.g., to blow) nor were small clauses (e.g., I see him playing ) because these forms are distinctly nonfinite rather than ambiguous. Frequency counts of all +Tense and −Tense verb forms were summed for each parent, and input informativeness was computed as the percentage of overt +Tense forms out of all coded verb forms.
All parent uses of copula and auxiliary is and 's forms in spontaneous, complete, and intelligible utterances were extracted from the language transcripts, then coded for syntactic structure by a trained undergraduate volunteer. Five syntactic contexts were possible: declarative, wh-question, yes/no question, tag question, or ellipsis. Syntactic context coding was reviewed with the second author on a weekly basis, and all disagreements were resolved and corrected in the transcripts. Only full is declaratives were used for this study (e.g., The baby is crying; this is a baby cow). Declarative contexts were identified in main clauses and embedded clauses and included lexical NP subjects (e.g., the baby) and pronoun subjects (e.g., this, there).
Child Measures
Child measures were computed at the following sessions: 21, 24, 27, and 30 months. Children's NDW was computed automatically by SALT software for the 30-min parent–child sample. Additional coding and scoring procedures were required to compute the measures of sentence diversity and TAP. For these measures, 60 min of conversational interaction were used by combining the parent–child and the examiner–child–(parent) samples.
Children's declarative statements and structural questions with explicit subjects and lexical verb predicates in a finite clause were coded for sentence diversity (Hadley, 2006, 2014; Ingram, 1989; McKenna, 2013). The subject–verb combinations were further coded for person and number of the subject. Only sentences with third person singular or plural subjects and lexical verbs (e.g., baby drinking; they fall down) were used in this study because they allow lexical flexibility of both subjects and verbs (Hadley et al., 2017). Third person subjects can also appear as either pronouns or expanded NPs. Routine questions (e.g., Where NP go/going? or What NP do/doing?) were excluded because these utterances have the potential to be formulaic (Miller & Chapman, 1981). Sentences with a conversational partner in the initial position (e.g., Mommy) were excluded because it is often challenging to determine whether a partner name was used as an addressee term or as a sentence subject. Children were credited with unique combinations if the same grammatical subject was used with different verb stems (e.g., baby need pillow; baby 's go ing to you), the same verb was used with different grammatical subjects (e.g., fence fall over vs. they fall down), or if subjects appeared in both singular and plural forms with the same verb (e.g., baby need pillow vs. baby/s need a snack).
Children's sentences were also scored for productive uses of T/A morphemes. All child sentences with copula BE, auxiliary BE, auxiliary DO, third person singular present tense /3s, and past tense regular /ed were extracted from the transcripts to compute a noncumulative TAP score at each measurement point. The TAP score was designed to protect against the undue influence of rotes and limited scope formulae by filtering out morphemes potentially produced by rote (Hadley et al., 2014; Hadley & Short, 2005; Rispoli, Hadley, & Holt, 2009). Uses of copula and auxiliary forms contracted to pronouns are excluded (e.g., he's, it's, what's, where's, here's) because these combinations are high frequency in the input and are packaged together as a single syllable (Li & Shirai, 2000; MacWhinney, 2000); however, forms contracted to noun subjects are allowed (e.g., baby's going to you). In addition, irregular forms (e.g., went, has) and negative contractions (e.g., don't) are excluded. Children received 1 point for each unique combination, up to a maximum of 5 points in each morpheme class. For copula and auxiliary forms, unique uses were based on different combinations of grammatical subjects and T/A morphemes, including the same subject with different T/A morphemes (e.g., Where did it go? Does it fit?) or different subjects with the same T/A morpheme (e.g., This is mine; farm is right there). Verb affixes /3s and /ed were counted as unique uses if they appeared with different verb stems, including overregularizations on irregular verb stems (e.g., broke/ed, have/3s).
Reliability
Approximately 16% of each sample (i.e., 5 min of each 30-min parent sample for adult utterances; 10 min of each 1-hr sample for child utterances) was randomly selected and transcribed independently by a second transcriber. If independent reliability was unacceptable (< 80% child, < 90% adult), a consensus pass was completed with digital video files. Average independent transcription reliability was 95.0% (SD = 3.2%) for adults and 82.1% (SD = 11.1%) for children. Adult transcription reliability fell below the 90% criterion only five times when a parent spoke very quietly or very rapidly. Child transcription reliability improved with age, as children talked more and general intelligibility improved. At 21 and 24 months of age, transcripts for 20 and 17 children required a consensus pass, respectively, whereas nine transcripts required a consensus pass at 27 and 30 months of age.
Twelve transcripts (six treatment and six control), were randomly selected and coded independently for adult toy talk and input informativeness by a second RA. For toy talk, coders were required to make decisions about whether an utterance was coded as toy talk with a NP subject, pronominal subject, labeling, or received no code. Cohen's kappas ranged from .88 to .99, with a mean of .95. These kappas exceeded .80, the levels of agreement conventionally considered to be acceptable (Sprent & Smeeton, 2001). For input informativeness, the percentage agreement was computed between the codes from the original and reliability transcripts. Mean agreement was 96.0% (range = 92% to 99%). Given the high levels of agreement, data analysis for input informativeness was based on the original coder's transcript.
Given the low frequency with which full is declaratives were produced, syntactic context coding for all full is forms was reviewed by the first author. Eight coding errors out of 1,972 is uses (0.4%) were identified and corrected.
Independent reliability was reported for children's sentence diversity in Hadley et al. (2017) . Sentence diversity measures were independently computed by a second RA for six randomly selected child participants at 21, 24, 27, and 30 months. Agreement was based on the coding of sentences with all subject–verb codes. A total of 618 child subject–verb codes were compared, resulting in a Cohen's κ of .94. For this study, TAP scores were verified through consensus procedures. All original scoring decisions for every transcript were checked by a second coder. The second coder was in complete agreement with the original coder for 142 (88.8) of the 160 transcripts. For the 18 transcripts with disagreements, 17 had only one scoring decision difference. All scoring disagreements were resolved and corrected prior to data analysis.
Data Analysis
To address the first research question, we examined change in parents' use of input informativeness and the use of full copula and auxiliary is forms in declaratives as a function of Condition (treatment vs. control) before and after instruction. Repeated measures analyses of variance (ANOVAs) were used with Condition as a between-participants factor and Time as a within factor. An interaction effect was expected with treatment parents providing more informative input for tense and producing more full is declaratives than control parents after instruction.
For the second research question, hierarchical linear modeling was used to analyze developmental change in children's TAP scores and to identify predictors of growth from 21 to 30 months (Holt, 2008 : Raudenbush & Bryk, 2002; Raudenbush, Bryk, & Congdon, 2007). When applied to the study of longitudinal change, hierarchical linear modeling involves two levels of analysis: (a) an individual growth model that represents changes in each child's score over time and (b) a between-child model that represents differences in children's growth trajectories. Because we expected the instruction to exert its effect on sentence diversity by approximately 27 months, and subsequently on TAP, equations for TAP growth were age centered at 30 months. Once the best-fit growth model was determined, a random slopes model was fit to allow TAP growth trajectories between 21 and 30 months to vary randomly across participants. This allowed for estimating the growth trajectory for the full group of 38 participants, as well as the individual differences from the overall growth trajectory. In addition, we explored developmental and input predictors of children's growth rates. Developmental effects would be evident if children's NDW and/or sentence diversity were significant predictors of children's ability to produce sentences marked for tense and agreement. Input effects would be evident if parent input variables were significant predictors of children's TAP growth rates. Recall that Hadley et al. (2017)  found a significant input effect for individual parents' use of lexical NP subjects on their children's growth in sentence diversity but only a marginal treatment effect between the treatment and control groups. Therefore, the analyses of input variables in this study were also conducted at the level of the individual parent–toddler dyads collapsing across the two groups. In other words, if the intervention altered T/A properties of parent input as expected, variability in these input properties would expand in the full sample. If so, the analyses conducted on the full sample would allow us to determine which specific input properties were significant predictors, or active ingredients, of children's grammatical growth.
Results
Treatment Effects on Parent Input Informativeness and Full is Declaratives
Our first question focused on the indirect benefits of toy talk instruction on parents' T/A marking, in general, and their use of full copula and auxiliary is forms, in particular. Descriptive statistics for lexical NP subject diversity, +Tense and −Tense verb forms, input informativeness, and full is declaratives are reported in Table 1 for parents in the treatment and control conditions prior to instruction at 21 months and following instruction at 24 months. At 21 months, there were no group differences for any parent input measure all ps ≥ .581.
Table 1. Means and standard deviations for parent input measures by time and condition.
Means and standard deviations for parent input measures by time and condition.×
Measure Control
Treatment
Full sample
M SD M SD M SD
Lexical NP subject diversity
 21 months 4.68 3.04 5.58 2.84 5.13 2.93
 24 months 6.00 2.96 18.05 9.99 12.03 9.49
+Tense verb forms
 21 months 148.05 57.25 157.42 51.12 152.74 53.74
 24 months 167.79 59.76 161.37 54.48 164.58 56.50
−Tense verb forms
 21 months 136.21 53.73 136.11 54.31 136.16 53.29
 24 months 136.95 59.10 126.42 36.26 131.68 48.65
Input informativeness
 21 months 52.42 10.69 54.13 8.08 53.27 9.38
 24 months 55.66 9.14 55.60 6.63 55.63 7.88
Full is declaratives
 21 months 3.32 3.92 2.74 2.56 3.03 3.28
 24 months 5.53 5.37 13.74 10.38 9.63 9.15
Note. Control n = 19; treatment n = 19; full sample N = 38. NP = noun phrase.
Note. Control n = 19; treatment n = 19; full sample N = 38. NP = noun phrase.×
Table 1. Means and standard deviations for parent input measures by time and condition.
Means and standard deviations for parent input measures by time and condition.×
Measure Control
Treatment
Full sample
M SD M SD M SD
Lexical NP subject diversity
 21 months 4.68 3.04 5.58 2.84 5.13 2.93
 24 months 6.00 2.96 18.05 9.99 12.03 9.49
+Tense verb forms
 21 months 148.05 57.25 157.42 51.12 152.74 53.74
 24 months 167.79 59.76 161.37 54.48 164.58 56.50
−Tense verb forms
 21 months 136.21 53.73 136.11 54.31 136.16 53.29
 24 months 136.95 59.10 126.42 36.26 131.68 48.65
Input informativeness
 21 months 52.42 10.69 54.13 8.08 53.27 9.38
 24 months 55.66 9.14 55.60 6.63 55.63 7.88
Full is declaratives
 21 months 3.32 3.92 2.74 2.56 3.03 3.28
 24 months 5.53 5.37 13.74 10.38 9.63 9.15
Note. Control n = 19; treatment n = 19; full sample N = 38. NP = noun phrase.
Note. Control n = 19; treatment n = 19; full sample N = 38. NP = noun phrase.×
×
Given the increased frequency of toy talk sentences with third person subjects following instruction reported in Hadley et al. (2017), we predicted a related increase in input informativeness. However, toy talk instruction did not alter treatment parents' percentage of informative input sentences. The treatment and control groups produced the same average percentage of input sentences with overt T/A marking (treatment, M = 55.6%; control, M = 55.7%). A repeated measures ANOVA revealed no significant main effects for Condition, F(1, 36) = 0.11, p = .743, ηp2 = .003, or for the Time × Condition interaction, F(1, 36) = 0.44, p = .513, ηp2 = .01, although the main effect for Time approached significance, F(1, 36) = 3.08, p = .09. ηp2 = .08.
Toy talk instruction was also predicted to alter full copula and auxiliary is forms in declarative contexts because full is declaratives are more likely with lower frequency grammatical subjects (Frank & Jaeger, 2008), such as those found in toy talk input sentences. This prediction was supported. Instruction magnified the treatment group's production of full is declaratives, increasing their average production from 2.74 to 13.74 instances in 30 min of parent–child interaction (see Table 1). A repeated measures ANOVA for full is declaratives revealed significant main effects for Condition, F(1, 36) = 5.81, p = .02, ηp2 = .14, for Time, F(1, 36) = 26.23, p < .001, ηp2 = .42, and for the Time × Condition interaction, F(1, 36) = 11.61, p = .002, ηp2= .24.
For the final adult analysis, we examined associations between lexical NP subject diversity, the input measure altered as a direct result of toy talk instruction (Hadley et al., 2017), with the input variables predicted to change as indirect benefits of instruction (see Table 2). Correlational analyses for the full sample revealed that lexical NP subject diversity was positively associated with +Tense verb forms, r = .45, p = .003, and with full is in declarative contexts, r = .78, p < .001, but not with −Tense verb forms nor with input informativeness, r = .22, p = .09. Also, only full is declaratives were related to children's 30-month TAP scores, r = .29, p = .041.
Table 2. Correlations for 24-month parent input measures and 30-month child tense agreement productivity score.
Correlations for 24-month parent input measures and 30-month child tense agreement productivity score.×
Adult input measures 2 3 4 5 Child TAP score
1. Lexical NP subject diversity .45** .19 .20 .78*** .22
2. +Tense verb forms .52*** .41** .33* .19
3. −Tense verb forms −.53*** .07 .06
4. Input informativeness .22 .04
5. Full is declarative 29*
Note. N = 38. TAP = tense agreement productivity; NP = noun phrase.
Note. N = 38. TAP = tense agreement productivity; NP = noun phrase.×
* p < .05 (one-tailed).
p < .05 (one-tailed).×
** p < .01 (one-tailed).
p < .01 (one-tailed).×
*** p < .001 (one-tailed).
p < .001 (one-tailed).×
Table 2. Correlations for 24-month parent input measures and 30-month child tense agreement productivity score.
Correlations for 24-month parent input measures and 30-month child tense agreement productivity score.×
Adult input measures 2 3 4 5 Child TAP score
1. Lexical NP subject diversity .45** .19 .20 .78*** .22
2. +Tense verb forms .52*** .41** .33* .19
3. −Tense verb forms −.53*** .07 .06
4. Input informativeness .22 .04
5. Full is declarative 29*
Note. N = 38. TAP = tense agreement productivity; NP = noun phrase.
Note. N = 38. TAP = tense agreement productivity; NP = noun phrase.×
* p < .05 (one-tailed).
p < .05 (one-tailed).×
** p < .01 (one-tailed).
p < .01 (one-tailed).×
*** p < .001 (one-tailed).
p < .001 (one-tailed).×
×
Growth in Children's T/A Marking
Descriptive statistics for children's MLU, NDW, third person sentence diversity, and TAP scores by measurement point are reported in Table 3 by condition and for the full sample, because the growth analyses were conducted on the full sample. Recall that child participants were required to be typically developing but not too advanced to participate in the intervention. At 21 months, child participants in both conditions had MLUs in the below average to average range, relative to the expectations of Miller and Chapman (1981; i.e., M = 1.62, SD = 0.39). Note that there were no significant group differences for any child variables at 21 months, reducing the likelihood that the parent changes observed were attributable to initial child characteristics. However, because the differences in NDW approached significance, t(36) = −1.77, p = .09, this variable was used as a time-varying covariate in the growth analyses.
Table 3. Descriptive statistics for child measures by condition and full sample by measurement point.
Descriptive statistics for child measures by condition and full sample by measurement point.×
Parameter Control
Treatment
Full
M SD M SD M SD
MLU
 21 months 1.20 0.22 1.23 0.22 1.22 0.21
 24 months 1.55 0.46 1.52 0.48 1.53 0.46
 27 months 1.93 0.48 1.95 0.47 1.94 0.47
 30 months 2.26 0.48 2.34 0.66 2.30 0.57
NDW
 21 months 18.53 12.38 26.11 14.02 22.32 13.56
 24 months 47.11 25.91 52.26 18.95 49.68 22.54
 27 months 75.11 31.64 79.37 25.03 77.24 28.22
 30 months 91.79 25.36 102.00 36.35 96.89 31.36
Sentence diversity (in third person sentences)
 21 months 0.16 0.69 0.37 0.68 0.26 0.68
 24 months 2.11 3.13 2.47 2.50 2.39 2.89
 27 months 6.00 4.87 5.11 4.31 6.05 4.90
 30 months 9.58 6.23 12.79 8.00 12.21 7.91
Tense and agreement productivity score
 21 months 0.11 0.46 0.32 0.67 0.21 0.57
 24 months 1.63 2.36 1.05 1.75 1.34 2.10
 27 months 4.11 3.80 2.74 2.83 3.47 3.45
 30 months 6.53 4.97 8.16 5.28 7.42 5.12
Note. Control group n = 19; treatment group n = 19; full sample N = 38. MLU = mean length of utterance; NDW = number of different words.
Note. Control group n = 19; treatment group n = 19; full sample N = 38. MLU = mean length of utterance; NDW = number of different words.×
Table 3. Descriptive statistics for child measures by condition and full sample by measurement point.
Descriptive statistics for child measures by condition and full sample by measurement point.×
Parameter Control
Treatment
Full
M SD M SD M SD
MLU
 21 months 1.20 0.22 1.23 0.22 1.22 0.21
 24 months 1.55 0.46 1.52 0.48 1.53 0.46
 27 months 1.93 0.48 1.95 0.47 1.94 0.47
 30 months 2.26 0.48 2.34 0.66 2.30 0.57
NDW
 21 months 18.53 12.38 26.11 14.02 22.32 13.56
 24 months 47.11 25.91 52.26 18.95 49.68 22.54
 27 months 75.11 31.64 79.37 25.03 77.24 28.22
 30 months 91.79 25.36 102.00 36.35 96.89 31.36
Sentence diversity (in third person sentences)
 21 months 0.16 0.69 0.37 0.68 0.26 0.68
 24 months 2.11 3.13 2.47 2.50 2.39 2.89
 27 months 6.00 4.87 5.11 4.31 6.05 4.90
 30 months 9.58 6.23 12.79 8.00 12.21 7.91
Tense and agreement productivity score
 21 months 0.11 0.46 0.32 0.67 0.21 0.57
 24 months 1.63 2.36 1.05 1.75 1.34 2.10
 27 months 4.11 3.80 2.74 2.83 3.47 3.45
 30 months 6.53 4.97 8.16 5.28 7.42 5.12
Note. Control group n = 19; treatment group n = 19; full sample N = 38. MLU = mean length of utterance; NDW = number of different words.
Note. Control group n = 19; treatment group n = 19; full sample N = 38. MLU = mean length of utterance; NDW = number of different words.×
×
Developmental change in the TAP score was modeled according to Equation 1. In this equation, TAP score ti is the observed score for child i at t months, and e ti is the deviation of child i from his or her growth trajectory at time t. The e ti are assumed to be normally distributed with mean 0 and variance σ2. The parameter, π0i , represents the status of child i at 30 months (the centering point), and π1i is the linear rate of change for individual i at the centering point of 30 months, interpreted as the instantaneous rate of linear change at 30 months. The quadratic growth parameter, π2i , reflects the curvature or acceleration or deceleration in each child's overall growth across time. Larger positive values of π2i imply more rapid growth in the TAP score. Successive models were compared with one another to determine the best fit to the data. Using likelihood ratio tests, the deviance of each successive growth model was compared with the more restricted model. These tests were conducted by using the full information maximum likelihood estimation. Display Formula
T A P t i = π 0 i + π 1 i age t i 30 + π 2 i age t i 30 2 + e t i π 0 i = b 00 + r 0 i π 1 i = b 10 + r 1 i π 1 i = b 20 + r 2 i π 2 i = b 20 + r 2 i
(1)
A linear growth trajectory was examined and tested for its difference from the unconditional means model. The linear model was a statistically significant better fit than the unconditional means model, χ2(3) = 153.00, p < .001. The estimates indicated that the fixed linear coefficient and the between-child variability in linear growth were both statistically significant (see Model 1, Table 4). That is, the linear increase in the TAP score was significantly different from 0, b10= 0.78, p < .001, 95% confidence interval (CI) [0.60, 0.96], indicating growth of .78 T/A morphemes per month for the group with significant variability in linear growth among children (i.e., VAR(r1) = 0.22). The intercept was also significantly different from 0, b00= 6.45, p < .001, 95% CI [4.99, 7.90], with significant variability in the intercept remaining among children (i.e., VAR(r0) = 17.03).
Table 4. Children's tense agreement productivity score growth models.
Children's tense agreement productivity score growth models.×
Effects Model 1 linear Model 2 quadratic Model 3 quadratic with NDW Model 4 quadratic with NDW, sentence diversity
Fixed effects
 Intercept, b 00 6.45** 7.00** 5.71** 5.06**
 Linear growth, b 10 0.78** 1.44** 1.19** 0.86**
 Quadratic growth, b 20 0.08** 0.08** 0.06*
 NDW, b 30 0.04** 0.02
 Sentence diversity, b 40 0.20**
Random effects
 Intercept, VAR(r 0) 17.03** 21.12** 18.99** 14.61**
 Linear growth, VAR(r 1) 0.22** 1.62** 1.44** 0.94**
 Quadratic growth, VAR(r 2) 0.01** 0.011** 0.008**
 Level 1, VAR(e) 3.87 1.52 1.37 1.35
Note.N = 38. Models age centered at 30 months. NDW = number of different words.
Note.N = 38. Models age centered at 30 months. NDW = number of different words.×
* p < .01 (two-tailed).
p < .01 (two-tailed).×
** p < .001 (two-tailed).
p < .001 (two-tailed).×
Table 4. Children's tense agreement productivity score growth models.
Children's tense agreement productivity score growth models.×
Effects Model 1 linear Model 2 quadratic Model 3 quadratic with NDW Model 4 quadratic with NDW, sentence diversity
Fixed effects
 Intercept, b 00 6.45** 7.00** 5.71** 5.06**
 Linear growth, b 10 0.78** 1.44** 1.19** 0.86**
 Quadratic growth, b 20 0.08** 0.08** 0.06*
 NDW, b 30 0.04** 0.02
 Sentence diversity, b 40 0.20**
Random effects
 Intercept, VAR(r 0) 17.03** 21.12** 18.99** 14.61**
 Linear growth, VAR(r 1) 0.22** 1.62** 1.44** 0.94**
 Quadratic growth, VAR(r 2) 0.01** 0.011** 0.008**
 Level 1, VAR(e) 3.87 1.52 1.37 1.35
Note.N = 38. Models age centered at 30 months. NDW = number of different words.
Note.N = 38. Models age centered at 30 months. NDW = number of different words.×
* p < .01 (two-tailed).
p < .01 (two-tailed).×
** p < .001 (two-tailed).
p < .001 (two-tailed).×
×
Next, a quadratic model was fit to the data to model any change in linear growth occurring over time. A likelihood ratio test indicated that the quadratic model was a significantly better fit to the data than the linear model, χ 2 (4) = 53.05, p < .001. In addition, the test of the homogeneity of variance in the residuals was no longer significant in the quadratic model χ 2 (35) = 10.18, p > .5. These findings indicated that a quadratic model was the optimal model to characterize growth in children's TAP scores over time.
The quadratic model had a statistically significant intercept, indicating that the TAP score was significantly different from 0 at 30 months, b00= 7.00, p < .001, 95% CI [5.44, 8.55], a statistically significant linear term, indicating that the linear growth rate at 30 months was significantly different from 0, b10= 1.44, p < .001, 95% CI [0.98, 1.91], and the quadratic trajectory from 21 to 30 months was also statistically significant, b20= 0.08, p < .001, 95% CI [0.03, 0.12] (see Model 2, Table 4). Model 2 indicated that the average child in this sample produced seven unique T/A morphemes with significant linear growth at 30 months and overall acceleration. There was also significant variability among children for all three growth components: for intercept, VAR(r0) = 21.12, p < .001, linear growth, VAR(r1) = 1.62, p < .001, and quadratic growth, VAR(r2) = 0.01, p < .001, respectively.
Developmental Predictors of Children's Growth in TAP
The next two models examined whether measures of children's growth in NDW and sentence diversity would improve the within-child estimation of TAP growth. Given that TAP scoring criteria require T/A morphemes to be used with different subjects and verbs, we expected the diversity measures to improve the estimation of TAP growth. However, we expected sentence diversity, requiring lexical flexibility with basic clause structure, to be the primary predictor of children's ability to mark sentences with tense and agreement because tense and agreement are properties of clauses, not words.
Children's NDW and sentence diversity from 21 to 30 months were group mean centered and included as time-varying covariates (see Equations 2 and 3, respectively) to determine if spontaneous use of diverse words and subject–verb combinations were significant predictors of TAP growth. To reduce the complexity of the model relative to the sample size, NDW and sentence diversity were estimated as fixed coefficients, although all the growth parameters were allowed to randomly vary across individuals. Display Formula
T A P t i = π 0 i + π 1 i ag e t i 30 + π 2 i ag e t i 30 2 + π 3 i N D W t i + e t i
(2)
Display Formula
T A P t i = π 0 i + π 1 i ag e t i 30 + π 2 i ag e t i 30 2 + π 3 i N D W t i + π 4 i SenDiv t i + e t i
(3)
Children's NDW was a statistically significant predictor of TAP trajectories over time, b30= 0.04, p < .001, 95% CI [0.02, 0.05]. This indicates that TAP scores increased by .04 points with each different word produced or by 1 point with every 25 different words. The addition of NDW in Model 3 accounted for an additional 10.4% of the within-child error variance in TAP growth (see Table 4). As reported in Model 4, sentence diversity was also a statistically significant predictor of TAP growth over time, b40= 0.20, p < .001, 95% CI [0.11, 0.29] (see Table 4). This indicates that TAP scores increased by .20 points with each unique combination of third person subjects and verbs, or by 1 point with every five different sentences. Furthermore, once sentence diversity was added, NDW was no longer a significant predictor of TAP growth, b30= 0.02, p = .21, 95% CI [−0.002, 0.04].
Input Predictors of Children's Growth in TAP
To identify the contribution of parent input to children's TAP growth, the three input variables from the 24-month measurement point were grand mean centered and added as Level 2 predictors in three parallel models (see Equation 4). Model 5a examined the effects of input informativeness (i.e., percentage of verb forms marked overtly for tense and agreement); Model 5b examined the effects of input subject diversity (i.e., number of different lexical NP subjects); and Model 5c examined the effects of full is declaratives. Display Formula
T A P t i = π 0 i + π 1 i ag e t i 30 + π 2 i ag e t i 30 2 + π 3 i N D W t i + π 4 i SenDiv t i + e t i π 0 i = b 00 + b 01 Input Variable 24 i + r 0 i π 1 i = b 10 + b 11 Input Variable 24 i + r 1 i π 2 i = b 20 + b 21 Input Variable 24 i + r 2 i π 3 i = b 30 + b 31 Input Variable 24 i π 4 i = b 40 + b 41 Input Variable 24 i
(4)
Parents' input informativeness was not a significant predictor of any component of children's TAP growth trajectories, all p > .28 (see Table 5, Model 5a). However, parents' lexical NP subject diversity and the frequency of their full is declaratives were both significant predictors of TAP growth, albeit for different growth components. Model 5b identified lexical NP subject diversity as a significant predictor of linear growth in TAP scores at 30 months, b11= 0.05, p = .032, 95% CI [0.005, 0.10], and quadratic growth, b21= 0.004, p = .037, 95% CI [0.0003, 0.008] (see Table 5). That is, parents' input subject diversity accounted for 10.4% of the between-child variance in linear growth and 12.8% of the between-child variance in quadratic growth. The linear coefficient of .05 indicates that for every 20 different lexical NP subjects in parent input at 24 months, children's TAP scores at 30 months increased by one morpheme per month. The quadratic coefficient of .004 indicates that the number of different lexical NP subjects was associated with acceleration in TAP growth between 21 and 30 months. A median split was used to illustrate the effect of high versus low parent input subject diversity on children's TAP growth (see Figure 1). The full sample was divided at the median of 8.50, with the vast majority of treatment parents falling above the median (i.e., 16 of 19). Model 5c revealed that parents' frequency of full is declaratives was also a significant predictor of TAP intercept at 30 months, b01= 0.17, p = .033, 95% CI [0.00, 0.34], as well as linear growth in TAP scores at 30 months, b11= 0.05, p = .031, 95% CI [0.005, 0.10] (see Table 5). Full is declaratives accounted for 8.2%, and 14.0% of the between-child variance in estimates of the intercept and linear growth, respectively. As a significant predictor of TAP status at 30 months, approximately every six parent uses of full is declaratives at the 24-month measurement point increased a child's 30-month TAP score by approximately 1 point. The linear coefficient of .05 indicates that for every 20 full is declaratives, children's TAP scores increased by one morpheme per month. Figure 2 illustrates the effects of parent differences in full is declaratives on children's TAP growth for parents above and below the full sample median of 6.50. On this input variable, 68.4% of the treatment parents were above the median (i.e., 13 of 19). In both Models 5b and 5c, the fixed effects for intercept, linear growth, and quadratic growth remained statistically significant after controlling for the developmental and input predictors (see Table 5). Likewise, significant random variation remained in all growth parameters for Models 5b and 5c, with significant between-child variation.
Table 5. Input predictors of children's tense agreement productivity growth.
Input predictors of children's tense agreement productivity growth.×
Effects Model 5a input informativeness Model 5b lexical NP subject diversity Model 5c full is declarative
Fixed effects
 Intercept 5.14*** 5.15 *** 5.06 ***
  Input informativeness 5.83
  Lexical NP subject diversity 0.12
  Full is declarative 0.17 *
 Linear growth 0.90*** 0.90 *** 0.87 ***
  Input informativeness 2.89
  Lexical NP subject diversity 0.05 *
  Full is declarative 0.05 *
 Quadratic growth 0.06** 0.06 ** 0.06 **
  Input informativeness 0.24
  Lexical NP subject diversity 0.004 *
  Full is declarative 0.003
 NDW 0.02 0.02 0.02 *
  Input informativeness 0.04
  Lexical NP subject diversity −0.00009
  Full is declarative −0.001
 Sentence diversity 0.19*** 0.19 *** 0.19 ***
  Input informativeness −0.71
  Lexical NP subject diversity -0.004
  Full is declarative −0.0004
Random effects
 Intercept, VAR(r 0) 15.32*** 14.28* 13.41***
 Linear growth, VAR(r 1) 0.98*** 0.84*** 0.81***
 Quadratic growth, VAR(r 2) 0.008*** 0.007*** 0.007***
 Level 1 error, VAR(e) 1.30 1.30 1.32
Note. N = 38. Model age centered at 30 months. NP = noun phrase.
Note. N = 38. Model age centered at 30 months. NP = noun phrase.×
* p < .05 (two-tailed).
p < .05 (two-tailed).×
** p < .01 (two-tailed).
p < .01 (two-tailed).×
*** p < .001 (two-tailed).
p < .001 (two-tailed).×
p < .10 (two-tailed).
p < .10 (two-tailed).×
Table 5. Input predictors of children's tense agreement productivity growth.
Input predictors of children's tense agreement productivity growth.×
Effects Model 5a input informativeness Model 5b lexical NP subject diversity Model 5c full is declarative
Fixed effects
 Intercept 5.14*** 5.15 *** 5.06 ***
  Input informativeness 5.83
  Lexical NP subject diversity 0.12
  Full is declarative 0.17 *
 Linear growth 0.90*** 0.90 *** 0.87 ***
  Input informativeness 2.89
  Lexical NP subject diversity 0.05 *
  Full is declarative 0.05 *
 Quadratic growth 0.06** 0.06 ** 0.06 **
  Input informativeness 0.24
  Lexical NP subject diversity 0.004 *
  Full is declarative 0.003
 NDW 0.02 0.02 0.02 *
  Input informativeness 0.04
  Lexical NP subject diversity −0.00009
  Full is declarative −0.001
 Sentence diversity 0.19*** 0.19 *** 0.19 ***
  Input informativeness −0.71
  Lexical NP subject diversity -0.004
  Full is declarative −0.0004
Random effects
 Intercept, VAR(r 0) 15.32*** 14.28* 13.41***
 Linear growth, VAR(r 1) 0.98*** 0.84*** 0.81***
 Quadratic growth, VAR(r 2) 0.008*** 0.007*** 0.007***
 Level 1 error, VAR(e) 1.30 1.30 1.32
Note. N = 38. Model age centered at 30 months. NP = noun phrase.
Note. N = 38. Model age centered at 30 months. NP = noun phrase.×
* p < .05 (two-tailed).
p < .05 (two-tailed).×
** p < .01 (two-tailed).
p < .01 (two-tailed).×
*** p < .001 (two-tailed).
p < .001 (two-tailed).×
p < .10 (two-tailed).
p < .10 (two-tailed).×
×
Figure 1.

Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on number of different lexical noun phrase subjects.

 Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on number of different lexical noun phrase subjects.
Figure 1.

Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on number of different lexical noun phrase subjects.

×
Figure 2.

Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on full is declaratives.

 Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on full is declaratives.
Figure 2.

Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on full is declaratives.

×
In summary, toy talk instruction modified parent input in both direct and indirect ways. In the original study, Hadley et al. (2017)  documented that treatment parents, relative to control parents, significantly increased their use of lexical NP subjects as a direct result of instruction. The current study documents indirect benefits as well. Treatment parents used more acoustically salient, full is declaratives than control parents, without any explicit instruction on T/A morphemes. Two input variables, lexical NP subject diversity and full is declaratives, were also significant predictors of between-child differences in TAP growth, together with developmental change in sentence diversity from 21 to 30 months.
Discussion
This study provides the first evidence that modifying lexical NP subject diversity in parents' declarative input sentences promotes children's growth in T/A morphemes, along with growth in sentence diversity. In this discussion, we address the significance of the findings, the limitations of the study, and future directions for clinical research and practice.
Toy Talk Instruction and T/A Morphemes in Parent Input
Recall that toy talk instruction altered parent use of lexical NP subjects in declarative sentences by teaching parents two simple strategies, talk about the toys and give the object its name (Hadley et al., 2017). This study explored whether toy talk instruction also had indirect benefits on parents' use of T/A morphemes. The lack of change in input informativeness in this study differed from an earlier feasibility study (Hadley & Walsh, 2014) in which both lexical NP subjects and input informativeness were altered. Methodological differences may explain the different results. In the feasibility study, 2.5-min silent video clips of parent–child play were used to assess change in language input, with the adult participants talking for the parent in the video before and immediately after 15 min of toy talk instruction. This method allowed participants to think about their input without the added challenge of engaging a child in play. In contrast, the parents in this study received instruction on more strategies over a longer period. Parent instruction focused on both responsive interaction strategies and language modeling strategies, and the measures of language input were based on 30 min of parent–child free-play obtained several weeks after the last instructional session. The additional task demands of engaging children in play while using toy talk strategies may explain the lack of change in input informativeness. That is, utterances that direct young children's attention and behavior typically take the form of imperatives with ambiguous, −Tense verb forms (e.g., Look at this; drink your juice; let's eat). Thus, modest increases in object-focused, descriptive talk were insufficient to offset frequent use of imperative verb forms (Fitzgerald et al., 2013).
On the other hand, toy talk instruction did change the way T/A morphemes were packaged in parent input. Following instruction, parents in the treatment group produced more full is declaratives than parents in the control group. Correlational analyses also revealed a positive, moderately strong association between lexical NP diversity and the frequency of full is declaratives. Collapsing across parents in each group, treatment parents produced 261 full is declaratives, with 65.1 of these forms (170 of 261) appearing with lexical NP subjects. In comparison, control parents produced 104 full is declaratives, with only 20.2% (21 of 104) appearing with lexical NP subjects. Control parents produced the majority (48 of 104) of full is declaratives in phonologically uncontractible contexts, primarily with the subject pronoun this (e.g., This is a cow). In addition, when producing lexical NP subjects in contractible contexts, treatment parents were twice as likely as control parents to produce full is declaratives (61.5% vs. 31.0%) as opposed to contracted 's forms. Although the change observed in parents' production of full is declaratives was consistent with our a priori prediction, parents were also encouraged to slow down their rate of speech and to stress content words. This may have influenced their use of full is declaratives as well. Future research is needed to distinguish between these two possibilities.
Developmental and Input Predictors of Growth in Tense and Agreement
Our second research question explored how children's developmental abilities and parents' input worked together to facilitate the growth of tense and agreement. Given the mounting evidence for stability in language development (Bornstein, Hahn, & Putnick, 2016; Bornstein, Hahn, Putnick, & Suwalsky, 2014) and the potential for children's developmental levels to influence the input they receive (e.g., Huttenlocher, Waterfall, Vasilyeva, Vevea, & Hedges, 2010), we included both children's NDW and sentence diversity as time-varying covariates in the model-building process before exploring the contribution of parent input to morphosyntactic growth. With both types of diversity in the growth model, NDW was not a significant predictor of TAP growth trajectories, but growth in sentence diversity was. For every five different subject–verb combinations, TAP scores increased by 1 point. This finding demonstrates that children's knowledge of clause structure, as predicted, contributes to the acquisition of the T/A system.
The empirical link established between children's growth in sentence diversity and TAP is consistent with the predictions of gradual morphosyntactic learning (Rispoli & Hadley, 2011; Rispoli et al., 2012), an account that emphasizes the role of sentence diversity in morphosyntactic development. Gradual morphosyntactic learning posits that children's production of increasingly diverse sentences with lower frequency subjects and verbs is an important indicator of their ability to generate sentences with an internal representation of abstract grammatical relations and clause structure. In addition, diverse subjects, distributed across the grammatical features of person and number, provide opportunities to encode subject–verb agreement with different surface forms (e.g., I am vs. the penguin is; the blocks are), and diverse verbs provide opportunities to encode the interaction between aspect and tense (e.g., is eating vs. tastes;Hadley, Rispoli, & Hsu, 2016; Rispoli et al., 2012).
The current findings also provide new insights about stability in language growth over time. Children's ability to learn words and grammatical patterns from language input is clearly influenced by what they already know. For example, Fernald and Marchman (2011)  have demonstrated that between-child differences in vocabulary knowledge at 18 months influence how efficiently children comprehend words in language input and that differences in the processing efficiency of familiar words confer an advantage on subsequent expressive vocabulary growth. Vocabulary knowledge also provides a foundation for learning and producing basic syntactic structure (Connor, Fisher, & Roth, 2013; Hadley et al., 2017). This study demonstrates that the strength of children's clausal representations, as indicated by the lexical flexibility of subject–verb combinations, contributes to the acquisition of the T/A system.
Empirical support for the hypothesized input predictors was mixed. Although parents exhibited variation in input informativeness, ranging from 36.3% to 68.5%, this input property was not a significant predictor of children's TAP growth. The failure to replicate the previous findings of Hadley et al. (2011)  may have been related, in part, to the inclusion of the two developmental time-varying covariates. This nonsignificant finding also points to the possibility that quality may outweigh quantity in the competition between overt and ambiguous signals of tense marking. In particular, not all overt markings of tense may be equally helpful. On the other hand, this study replicated the average input informativeness previously reported for English child-directed speech. In the full sample, verb forms were marked overtly 55.6% of the time, comparable to values previously reported by Legate and Yang (2007)  and Hadley et al. (2011) . The substantial ambiguity of English verb forms in naturally occurring conversational input (i.e., 44.4%) underscores why this grammatical system is challenging to acquire.
On the other hand, two rare properties of parent input at 24 months, lexical NP subject diversity and full is forms in declarative sentences, were significant predictors of children's TAP growth. Lexical NP subjects in declarative sentences accounted for 10.4% of input sentences for treatment parents and less than 3% of input sentences for the control parents. Full is declaratives were even less common, occurring in 3.6% of input sentences for treatment parents and less than 1% of input sentences for control parents. However, both rare input properties predicted linear growth in TAP scores at 30 months. In addition, lexical NP subject diversity explained additional variance in overall acceleration, and full is declaratives explained additional variance in the intercept at 30 months. We expected parents' use of full is declaratives to increase children's ability to identify the word is in the input by increasing its phonetic substance (Pinker, 1984), especially because copula is is the most frequent T/A morpheme appearing in parent input (Fitzgerald et al., 2013; Rispoli & Hadley, 2014). In turn, we expected children's acquisition of copula is to promote the growth of other T/A morphemes through cross-morpheme facilitation (Rispoli, 2016). On the other hand, we expected parents' lexical NP subjects to exert their effect primarily through the strengthening of children's clause structure. However, parents' lexical NP subject diversity also influenced TAP growth, even after controlling for children's sentence diversity. The additional effect of lexical NP subject diversity may be related to its positive correlation with full is declaratives. Parents who used additional different lexical NPs as subjects also produced more full copula and auxiliary is forms in declaratives.
Why might toy talk input sentences with both lexical NP subjects and full is forms be particularly facilitative? It is possible that these declarative input sentences combine multiple cues to clause structure in a powerful way. As outlined in the introduction, declarative sentences with lexical NP subjects align low transitional probabilities between the noun and the following verb with prosodic cues to a key constituent boundary: the separation of subject from predicate. As the representation of clause structure is strengthened, children should be better able to detect T/A morphemes in input sentences. When lexical NP subjects are combined with full is forms, morpheme identification may be enhanced in two ways, by the phonetic substance of is as a word (cf. It's hot → The egg is hot; He's running → The dog is running) and by situating the morpheme prosodically in the predicate. For example, in the input sentences The egg is hot and The dog is running, the unstressed is falls in the phonological phrase with hot and running (Gerken & McGregor, 1998). In contrast, prosodic structure and subject–predicate constituent structure are not as well aligned in sentences such as This is an egg and He's running. In these sentences, the T/A morpheme is situated prosodically with the subject, disguising the subject–predicate boundary. From this perspective, input sentences with noun subjects improve the reliability of prosodic cues available to children for detecting morphemes and constituent structure. In addition, optimal alignment of constituent structure and prosodic structure occurs when the input sentence is derivationally simple, with declarative sentence structure. Structural questions with subject–auxiliary inversion lack this alignment, insofar as T/A morphemes are moved out of the predicate and toward the periphery of the sentence.
Given that parents' lexical NP subjects and full is declaratives were both coded only in declarative sentences, it is important to point out that the coding schemes had several other distinctive elements. That is, toy talk sentences with lexical NP subjects could occur with lexical verbs, modal auxiliaries, or any copula or auxiliary BE form. Likewise, full is declaratives could occur with any noun or pronoun subject, regardless of whether the subject referred to a concrete referent in the playroom, was used in existential sentences (e.g., Here is your cup), or was used to label objects (e.g., This is broccoli). Thus, the coding schemes used in this study considered the two input properties separately. Future investigations could explore the contribution of these two rare input properties operating in tandem.
Design Limitations
Despite the intriguing new input effects revealed, several design limitations must be recognized. First, a quasi-control group was used to evaluate changes in parent input properties. This design element was deemed a cost-effective approach for an initial investigation of a new input variable. Second, parent educational levels and racial or ethnic diversity were not well balanced between the treatment and quasi-control groups. The demographic differences raise concerns regarding the generalizability of the instructional procedures to more culturally, linguistically, and socioeconomically diverse families (see Hadley et al., 2017, for discussion) and warrant further evaluation of the feasibility and efficacy of the low-intensity instructional procedures with more diverse families. At the same time, the demographic differences between groups are less likely to have influenced the primary findings of this study because input effects were examined at the level of individual parent–toddler dyads rather than as a function of treatment condition. In addition, no direct instruction was provided to either group regarding use of T/A morphemes. Third, it is difficult to disentangle the contribution of parents' lexical NP subject diversity and full is forms in declarative contexts because these input properties were correlated. Given that these input properties occurred at such low frequencies, it will be necessary to increase input subject diversity beyond the rates produced in this study to decouple the contribution of these distinct input properties. Additional factors should also be examined in future studies. For example, children's abilities to parse and comprehend language input in real time plays a substantial role in the extent to which children can make use of the grammatical properties of input (e.g., Lew-Williams & Fernald, 2007). Additional properties of responsive interaction and language input strategies, such as expansions and recasting, should be examined to characterize the nature of optimal input sentences for facilitating grammatical development, in general, and the onset of tense and agreement, in particular.
Clinical Implications
Although this study was conducted with toddlers with slow typical development, the findings have implications for clinical interventions for preschool children with SLI. The most important empirical finding is that children's sentence diversity was a significant predictor of children's acquisition of tense and agreement. If the same relationship holds for children with SLI, intervention activities designed to promote more flexible use of known words in diverse subject–verb combinations could be conceptualized as a preventative approach for addressing the morphosyntactic weaknesses commonly observed in children with SLI. The plausibility of this proposal is strengthened by the converging evidence linking subject–verb combinations and morphosyntactic growth in this study and in Pawłowska et al. (2008), despite methodological differences in the computations of subject–verb combinations. Pawłowska et al. (2008)  used the percentage of clauses with lexical verbs and subjects from a 100-utterance pretreatment language sample. Subjects could be first, second, or third person, and repetitions were allowed. Imperatives were included in the total number of clauses with lexical verbs, even though they did not obligate the use of a subject. In contrast, our measure focused on unique third person subject–verb combinations assessed from 60-min language samples collected at 3-month intervals over a 9-month period. Third person sentences were of primary interest because T/A morphemes commonly appear first in this sentence context (Hadley, 2006; Hadley & Short, 2005; Rispoli et al., 2012) and because input subject diversity can only be manipulated in this sentence context. Together, the converging evidence suggests that lexical flexibility with basic sentence structure is important to establish prior to initiating intervention focused on the emergence of T/A morphemes. Future clinical research studies are needed to determine whether efforts to increase sentence diversity prior to targeting the emergence of T/A morphemes would improve the efficiency of morphosyntactic interventions.
The current study also contributes to a growing body of work documenting the beneficial effects of input diversity on language development (Hsu, Hadley, & Rispoli, 2017; Plante et al., 2014; Rowe, 2012). In Plante et al. (2014), the investigators manipulated verb input diversity as part of a clinical intervention for children with SLI. They demonstrated that recasting grammatical morphemes by using twice as many different verbs facilitated the acquisition of grammatical morphemes, including both auxiliaries that appear before the verb and suffixes that follow it. In this study, parents' input subject diversity promoted children's growth of sentence diversity and T/A morphemes. Together, these studies demonstrate that input diversity for open-class items supports children's analysis of closed-class items, such as T/A morphemes.
These findings raise important clinical issues about how to balance the presentation of diverse nouns in subject position in language input, while also providing platform utterances that children will comprehend easily and spontaneously imitate more readily. However, increasing input subject diversity does not require exposing children to unfamiliar, low-frequency words. Rather, nouns from a child's expressive vocabulary can be presented in subject position. The following examples from Parents 5 and 6 illustrate how children's single words were expanded into toy talk sentences. The example from Parent 9 begins with an ordinary expansion in which the noun glasses occupies the direct object position of the verb phrase, followed by additional description about the glasses. In Parent 9's next turn, the noun glasses is promoted to subject position. The example from Parent 18 illustrates the natural tendency to produce pronouns in subject position. We explained this to parents as the way language works and encouraged them to “give the item its name” in a following utterance if they responded first with a pronoun. In this case, Parent 18 first used the pronoun he and later the bug. Note that the child incorporated the parent's noun as the subject for his next utterance, bug alright, and then the parent recast the childlike utterance into an adult sentence, the bug's alright.
Parent 5  C Stuck.
 C Stuck.
 M Stuck, yeah.
 M These plates are stuck together.
Parent 6  C Penguin.
 M The penguin fell down.
Parent 9  C Glasses on
 M Put his glasses on?
 C xxx.
 M These glasses are green.
Parent 18  M {woah} He crashed.
 M I think he needs help.
 C xx xx.
 M The bug needs help.
 C Bug alright.
 M The bug's alright.
Conclusions
The current study identified two new properties of language input that facilitate children's acquisition of grammar. In particular, the diversity of parents' lexical NP subjects and their use of full is forms in declarative input sentences promoted the growth of tense and agreement. These properties appear to increase the availability and alignment of cues to constituent structure, facilitating children's acquisition of basic clause structure. As the representation of clause structure is strengthened, children are also better able to detect T/A morphemes in parent input. The use of toy talk instruction warrants further investigation to determine if these input–acquisition links can be replicated in independent samples with more diverse family demographics and if these input properties are sufficiently robust to accelerate grammatical development for children with SLI.
Acknowledgments
Funding for this study was supported by National Institute of Child Health & Human Development Grant R21 HD071316 (awarded to Pamela Hadley). Data collection for the quasi-control group was supported by National Science Foundation Grant BCS-08-22513 (awarded to Matthew Rispoli). Portions of this article were previously presented at the 2015 Callier Prize Conference on Children with Specific Language Impairment (SLI): Structuring Language Input to Improve Language Learning, Callier Center, Dallas, TX, and the 2015 Boston University Conference on Language Development. We extend sincere appreciation to the parents, children, and research assistants who made the work possible, as well as to Mary Kubalanza and Megan McKenna for their contributions to parent education and coaching and to Theodora Papastratakos, Ning Hsu, and Colleen Stern for their contributions to data coding.
References
Bornstein, M. H., Hahn, C. S., & Putnick, D. L. (2016). Stability of core language skill across the first decade of life in children at biological and social risk. Journal of Child Psychology and Psychiatry, 57, 1434–1443. [Article] [PubMed]
Bornstein, M. H., Hahn, C. S., & Putnick, D. L. (2016). Stability of core language skill across the first decade of life in children at biological and social risk. Journal of Child Psychology and Psychiatry, 57, 1434–1443. [Article] [PubMed]×
Bornstein, M. H., Hahn, C. S., Putnick, D. L., & Suwalsky, J. T. (2014). Stability of core language skill from early childhood to adolescence: A latent variable approach. Child Development, 85, 1346–1356. [Article] [PubMed]
Bornstein, M. H., Hahn, C. S., Putnick, D. L., & Suwalsky, J. T. (2014). Stability of core language skill from early childhood to adolescence: A latent variable approach. Child Development, 85, 1346–1356. [Article] [PubMed]×
Chomsky, N. (1995). The minimalist program. Cambridge, MA: MIT Press.
Chomsky, N. (1995). The minimalist program. Cambridge, MA: MIT Press.×
Cleave, P., Becker, S., Curran, M., Van Horne, A., & Fey, M. (2015). The efficacy of recasts in language intervention: A systematic review and meta-analysis. American Journal of Speech-Language Pathology, 24, 237–255. [Article] [PubMed]
Cleave, P., Becker, S., Curran, M., Van Horne, A., & Fey, M. (2015). The efficacy of recasts in language intervention: A systematic review and meta-analysis. American Journal of Speech-Language Pathology, 24, 237–255. [Article] [PubMed]×
Connor, M., Fisher, C., & Roth, D. (2013). Starting from scratch in semantic role labeling: Early indirect supervision. In Villavicencio, A., Poibeau, T., Korhonen, A., & Alishahi, A. (Eds.), Cognitive aspects of computational language acquisition (pp. 257–296). Berlin, Germany: Springer-Verlag. [Article]
Connor, M., Fisher, C., & Roth, D. (2013). Starting from scratch in semantic role labeling: Early indirect supervision. In Villavicencio, A., Poibeau, T., Korhonen, A., & Alishahi, A. (Eds.), Cognitive aspects of computational language acquisition (pp. 257–296). Berlin, Germany: Springer-Verlag. [Article] ×
Fenson, L., Marchman, V., Thal, D., Dale, P., Reznick, J., & Bates, E. (2007). MacArthur–Bates Communicative Development Inventories: User's Guide and Technical Manual (2nd ed.). Baltimore, MD: Brookes.
Fenson, L., Marchman, V., Thal, D., Dale, P., Reznick, J., & Bates, E. (2007). MacArthur–Bates Communicative Development Inventories: User's Guide and Technical Manual (2nd ed.). Baltimore, MD: Brookes.×
Fernald, A., & Marchman, V. A. (2011). Individual differences in lexical processing at 18 months predict vocabulary growth in typically developing and late‐talking toddlers. Child Development, 83, 203–222. [Article] [PubMed]
Fernald, A., & Marchman, V. A. (2011). Individual differences in lexical processing at 18 months predict vocabulary growth in typically developing and late‐talking toddlers. Child Development, 83, 203–222. [Article] [PubMed]×
Fey, M., Cleave, P., & Long, S. (1997). Two models of grammar facilitation in children with language impairments: Phase 2. Journal of Speech and Hearing Research, 40, 5–19. [Article]
Fey, M., Cleave, P., & Long, S. (1997). Two models of grammar facilitation in children with language impairments: Phase 2. Journal of Speech and Hearing Research, 40, 5–19. [Article] ×
Fey, M., Cleave, P., Long, S., & Hughes, D. (1993). Two approaches to the facilitation of grammar in children with language impairment: An experimental evaluation. Journal of Speech and Hearing Research, 36, 141–157. [Article] [PubMed]
Fey, M., Cleave, P., Long, S., & Hughes, D. (1993). Two approaches to the facilitation of grammar in children with language impairment: An experimental evaluation. Journal of Speech and Hearing Research, 36, 141–157. [Article] [PubMed]×
Fey, M., & Finestack, L. (2009). Research and development in children's language intervention: A five-phase model. In Schwartz, R. G. (Ed.), Handbook of child language disorders (pp. 513–531). New York, NY: Psychology Press.
Fey, M., & Finestack, L. (2009). Research and development in children's language intervention: A five-phase model. In Schwartz, R. G. (Ed.), Handbook of child language disorders (pp. 513–531). New York, NY: Psychology Press.×
Fisher, C., & Tokura, H. (1996). Acoustic cues to grammatical structure in infant‐directed speech: Cross‐linguistic evidence. Child Development, 67, 3192–3218. [Article] [PubMed]
Fisher, C., & Tokura, H. (1996). Acoustic cues to grammatical structure in infant‐directed speech: Cross‐linguistic evidence. Child Development, 67, 3192–3218. [Article] [PubMed]×
Fitzgerald, C., Hadley, P., & Rispoli, M. (2013). Are some parents' interaction styles associated with richer grammatical input? American Journal of Speech-Language Pathology, 22, 476–488. [Article] [PubMed]
Fitzgerald, C., Hadley, P., & Rispoli, M. (2013). Are some parents' interaction styles associated with richer grammatical input? American Journal of Speech-Language Pathology, 22, 476–488. [Article] [PubMed]×
Frank, A., & Jaeger, T. F. (2008). Speaking rationally: Uniform information density as an optimal strategy for language production. In Love, B. C., McRae, K. & Sloutsky, V. M. (Eds.), Proceedings of the 30th Annual Meeting of the Cognitive Science Society (pp. 933–938). Washington, DC: Cognitive Science Society.
Frank, A., & Jaeger, T. F. (2008). Speaking rationally: Uniform information density as an optimal strategy for language production. In Love, B. C., McRae, K. & Sloutsky, V. M. (Eds.), Proceedings of the 30th Annual Meeting of the Cognitive Science Society (pp. 933–938). Washington, DC: Cognitive Science Society.×
Gerken, L., & McGregor, K. (1998). An overview of prosody and its role in normal and disordered child language. American Journal of Speech-Language Pathology, 7, 38–48. [Article]
Gerken, L., & McGregor, K. (1998). An overview of prosody and its role in normal and disordered child language. American Journal of Speech-Language Pathology, 7, 38–48. [Article] ×
Gómez, R. L. (2002). Variability and detection of invariant structure. Psychological Science, 13, 431–436. [Article] [PubMed]
Gómez, R. L. (2002). Variability and detection of invariant structure. Psychological Science, 13, 431–436. [Article] [PubMed]×
Hadley, P. (2006). Assessing the emergence of grammar in toddlers at risk for specific language impairment. Seminars in Speech and Language, 27, 173–186. [Article] [PubMed]
Hadley, P. (2006). Assessing the emergence of grammar in toddlers at risk for specific language impairment. Seminars in Speech and Language, 27, 173–186. [Article] [PubMed]×
Hadley, P. (2014). Approaching early grammatical intervention from a sentence-focused framework. Language, Speech, and Hearing Services in Schools, 45, 110–116. [Article] [PubMed]
Hadley, P. (2014). Approaching early grammatical intervention from a sentence-focused framework. Language, Speech, and Hearing Services in Schools, 45, 110–116. [Article] [PubMed]×
Hadley, P., & Rispoli, M. (2015). Toy talk strategies: An instructional resource. Retrieved from http://hdl.handle.net/2142/78010
Hadley, P., & Rispoli, M. (2015). Toy talk strategies: An instructional resource. Retrieved from http://hdl.handle.net/2142/78010 ×
Hadley, P., Rispoli, M., Holt, J., Fitzgerald, C., & Bahnsen, A. (2014). The growth of finiteness in the third year of life: Replication and predictive validity. Journal of Speech, Language, and Hearing Research, 57, 887–900. [Article]
Hadley, P., Rispoli, M., Holt, J., Fitzgerald, C., & Bahnsen, A. (2014). The growth of finiteness in the third year of life: Replication and predictive validity. Journal of Speech, Language, and Hearing Research, 57, 887–900. [Article] ×
Hadley, P., Rispoli, M., Holt, J., Papastratakos, T., Hsu, N., Kubalanza, M., & McKenna, M. (2017). Input subject diversity enhances early grammatical growth: Evidence from a parent-implemented intervention. Language Learning and Development, 13, 54–79. [Article] [PubMed]
Hadley, P., Rispoli, M., Holt, J., Papastratakos, T., Hsu, N., Kubalanza, M., & McKenna, M. (2017). Input subject diversity enhances early grammatical growth: Evidence from a parent-implemented intervention. Language Learning and Development, 13, 54–79. [Article] [PubMed]×
Hadley, P., Rispoli, M., Fitzgerald, C., & Bahnsen, A. (2011). Predictors of morphosyntactic growth in typically developing toddlers: Contributions of parent input and child sex. Journal of Speech, Language, and Hearing Research, 54, 549–566. [Article]
Hadley, P., Rispoli, M., Fitzgerald, C., & Bahnsen, A. (2011). Predictors of morphosyntactic growth in typically developing toddlers: Contributions of parent input and child sex. Journal of Speech, Language, and Hearing Research, 54, 549–566. [Article] ×
Hadley, P., Rispoli, M., & Hsu, N. (2016). Toddlers' verb lexicon diversity and grammatical outcomes. Language, Speech, and Hearing Services in Schools, 47, 44–58. [Article] [PubMed]
Hadley, P., Rispoli, M., & Hsu, N. (2016). Toddlers' verb lexicon diversity and grammatical outcomes. Language, Speech, and Hearing Services in Schools, 47, 44–58. [Article] [PubMed]×
Hadley, P., & Short, H. (2005). The onset of tense marking in children at-risk for SLI. Journal of Speech, Language, and Hearing Research, 48, 1344–1362. [Article]
Hadley, P., & Short, H. (2005). The onset of tense marking in children at-risk for SLI. Journal of Speech, Language, and Hearing Research, 48, 1344–1362. [Article] ×
Hadley, P., & Walsh, K. (2014). Toy talk: Simple strategies to create richer grammatical input. Language, Speech, and Hearing Services in Schools, 45, 159–172. [Article] [PubMed]
Hadley, P., & Walsh, K. (2014). Toy talk: Simple strategies to create richer grammatical input. Language, Speech, and Hearing Services in Schools, 45, 159–172. [Article] [PubMed]×
Hoff-Ginsberg, E. (1985). Some contributions of mothers' speech to their children's syntactic growth. Journal of Child Language, 12, 367–385. [Article] [PubMed]
Hoff-Ginsberg, E. (1985). Some contributions of mothers' speech to their children's syntactic growth. Journal of Child Language, 12, 367–385. [Article] [PubMed]×
Holt, J. K. (2008). Modeling growth using multilevel and alternative approaches. In O'Connell, A. A. & McCoach, D. B. (Eds.), Multilevel analysis of educational data (pp. 111–159). Charlotte, NC: Information Age.
Holt, J. K. (2008). Modeling growth using multilevel and alternative approaches. In O'Connell, A. A. & McCoach, D. B. (Eds.), Multilevel analysis of educational data (pp. 111–159). Charlotte, NC: Information Age.×
Hsu, N., Hadley, P., & Rispoli, M. (2017). Diversity matters: Parent input predicts toddler verb production. Journal of Child Language, 44, 63–86. [Article] [PubMed]
Hsu, N., Hadley, P., & Rispoli, M. (2017). Diversity matters: Parent input predicts toddler verb production. Journal of Child Language, 44, 63–86. [Article] [PubMed]×
Huttenlocher, J., Vasilyeva, M., Cymerman, E., & Levine, S. (2002). Language input and child syntax. Cognitive Psychology, 45, 337–374. [Article] [PubMed]
Huttenlocher, J., Vasilyeva, M., Cymerman, E., & Levine, S. (2002). Language input and child syntax. Cognitive Psychology, 45, 337–374. [Article] [PubMed]×
Huttenlocher, J., Vasilyeva, M., Waterfall, H., Vevea, J., & Hedges, L. (2007). The varieties of speech to young children. Developmental Psychology, 43, 1062–1083. [Article] [PubMed]
Huttenlocher, J., Vasilyeva, M., Waterfall, H., Vevea, J., & Hedges, L. (2007). The varieties of speech to young children. Developmental Psychology, 43, 1062–1083. [Article] [PubMed]×
Huttenlocher, J., Waterfall, H. R., Vasilyeva, M., Vevea, J. L., & Hedges, L. V. (2010). Sources of variability in children's language growth. Cognitive Psychology, 4, 343–365. [Article]
Huttenlocher, J., Waterfall, H. R., Vasilyeva, M., Vevea, J. L., & Hedges, L. V. (2010). Sources of variability in children's language growth. Cognitive Psychology, 4, 343–365. [Article] ×
Ingram, D. (1989). First language acquisition: Method, description, and explanation. Cambridge, England: Cambridge University Press.
Ingram, D. (1989). First language acquisition: Method, description, and explanation. Cambridge, England: Cambridge University Press.×
Legate, J., & Yang, C. (2007). Morphosyntactic learning and the development of tense. Language Acquisition, 14, 315–344. [Article]
Legate, J., & Yang, C. (2007). Morphosyntactic learning and the development of tense. Language Acquisition, 14, 315–344. [Article] ×
Leonard, L. (2014). Children with specific language impairment (2nd ed.). Cambridge, MA: MIT Press.
Leonard, L. (2014). Children with specific language impairment (2nd ed.). Cambridge, MA: MIT Press.×
Leonard, L., Camarata, S., Brown, B., & Camarata, M. (2004). Tense and agreement in the speech of children with specific language impairment: Patterns of generalization through intervention. Journal of Speech, Language, and Hearing Research, 47, 1363–1379. [Article]
Leonard, L., Camarata, S., Brown, B., & Camarata, M. (2004). Tense and agreement in the speech of children with specific language impairment: Patterns of generalization through intervention. Journal of Speech, Language, and Hearing Research, 47, 1363–1379. [Article] ×
Leonard, L., Camarata, S., Pawłowska, M., Brown, B., & Camarata, M. (2006). Tense and agreement morphemes in the speech of children with specific language impairment: Phase 2. Journal of Speech, Language, and Hearing Research, 49, 749–770. [Article]
Leonard, L., Camarata, S., Pawłowska, M., Brown, B., & Camarata, M. (2006). Tense and agreement morphemes in the speech of children with specific language impairment: Phase 2. Journal of Speech, Language, and Hearing Research, 49, 749–770. [Article] ×
Leonard, L., Camarata, S., Pawłowska, M., Brown, B., & Camarata, M. (2008). The acquisition of tense and agreement morphemes by children with specific language impairment during intervention. Phase 3. Journal of Speech, Language, and Hearing Research, 51, 120–125. [Article]
Leonard, L., Camarata, S., Pawłowska, M., Brown, B., & Camarata, M. (2008). The acquisition of tense and agreement morphemes by children with specific language impairment during intervention. Phase 3. Journal of Speech, Language, and Hearing Research, 51, 120–125. [Article] ×
Leonard, L., Fey, M., Deevy, P., & Bredin-Oja, S. (2015). Input sources of third person singular -s inconsistency in children with and without specific language impairment. Journal of Child Language, 42, 786–820. [Article] [PubMed]
Leonard, L., Fey, M., Deevy, P., & Bredin-Oja, S. (2015). Input sources of third person singular -s inconsistency in children with and without specific language impairment. Journal of Child Language, 42, 786–820. [Article] [PubMed]×
Lew-Williams, C., & Fernald, A. (2007). Young children learning Spanish make rapid use of grammatical gender in spoken word recognition. Psychological Science, 18, 193–198. [Article] [PubMed]
Lew-Williams, C., & Fernald, A. (2007). Young children learning Spanish make rapid use of grammatical gender in spoken word recognition. Psychological Science, 18, 193–198. [Article] [PubMed]×
Li, P., & Shirai, Y. (2000). The acquisition of lexical and grammatical aspect. Berlin, Germany: Mouton de Gruyter. [Article]
Li, P., & Shirai, Y. (2000). The acquisition of lexical and grammatical aspect. Berlin, Germany: Mouton de Gruyter. [Article] ×
Lidz, J., & Gagliardi, A. (2015). How nature meets nurture: Universal Grammar and statistical learning. Annual Review of Linguistics, 1(1), 333–353. [Article]
Lidz, J., & Gagliardi, A. (2015). How nature meets nurture: Universal Grammar and statistical learning. Annual Review of Linguistics, 1(1), 333–353. [Article] ×
MacWhinney, B. (2000). The CHILDES Project: Tools for analyzing talk (3rd ed.). Mahwah, NJ: Erlbaum.
MacWhinney, B. (2000). The CHILDES Project: Tools for analyzing talk (3rd ed.). Mahwah, NJ: Erlbaum.×
Manolson, A., Ward, B., & Dodington, N. (2007). You make the difference: In helping your child learn. Toronto, Ontario, Canada: The Hanen Centre.
Manolson, A., Ward, B., & Dodington, N. (2007). You make the difference: In helping your child learn. Toronto, Ontario, Canada: The Hanen Centre.×
McKenna, M. (2013). Developmental expectations for child-like sentences (Unpublished master's thesis) . University of Illinois, Urbana–Champaign.
McKenna, M. (2013). Developmental expectations for child-like sentences (Unpublished master's thesis) . University of Illinois, Urbana–Champaign.×
Miller, J., & Chapman, R. (1981). The relation between age and mean length of utterance in morphemes. Journal of Speech, Language, and Hearing Research, 24, 154–161. [Article]
Miller, J., & Chapman, R. (1981). The relation between age and mean length of utterance in morphemes. Journal of Speech, Language, and Hearing Research, 24, 154–161. [Article] ×
Miller, J., & Iglesias, A. (2012). Systematic Analysis of Language Transcripts (SALT) (Research Version 2012) [Computer software] . Middleton, WI: SALT Software.
Miller, J., & Iglesias, A. (2012). Systematic Analysis of Language Transcripts (SALT) (Research Version 2012) [Computer software] . Middleton, WI: SALT Software.×
Morgan, J., Meier, R., & Newport, E. (1989). Facilitating the acquisition of syntax with cross-sentential cues to phrase structure. Journal of Memory and Language, 28, 360–374. [Article]
Morgan, J., Meier, R., & Newport, E. (1989). Facilitating the acquisition of syntax with cross-sentential cues to phrase structure. Journal of Memory and Language, 28, 360–374. [Article] ×
Oetting, J. B., & Hadley, P. A. (2017). Morphosyntax in child language disorders. In Schwartz, R. G. (Ed.), The handbook of child language disorders (2nd ed., pp. 365–392). New York, NY: Psychological Press.
Oetting, J. B., & Hadley, P. A. (2017). Morphosyntax in child language disorders. In Schwartz, R. G. (Ed.), The handbook of child language disorders (2nd ed., pp. 365–392). New York, NY: Psychological Press.×
Pawłowska, M., Leonard, L. B., Camarata, S. M., Brown, B., & Camarata, M. N. (2008). Factors accounting for the ability of children with SLI to learn agreement morphemes in intervention. Journal of Child Language, 35, 25–53. [Article] [PubMed]
Pawłowska, M., Leonard, L. B., Camarata, S. M., Brown, B., & Camarata, M. N. (2008). Factors accounting for the ability of children with SLI to learn agreement morphemes in intervention. Journal of Child Language, 35, 25–53. [Article] [PubMed]×
Pinker, S. (1984). Language learnability and language development. Cambridge, MA: Harvard University Press.
Pinker, S. (1984). Language learnability and language development. Cambridge, MA: Harvard University Press.×
Plante, E., Ogilvie, T., Vance, R., Aguilar, J. M., Dailey, N. S., Meyers, C., … Burton, R. (2014). Variability in the language input to children enhances learning in a treatment context. American Journal of Speech-Language Pathology, 23, 530–545. [Article] [PubMed]
Plante, E., Ogilvie, T., Vance, R., Aguilar, J. M., Dailey, N. S., Meyers, C., … Burton, R. (2014). Variability in the language input to children enhances learning in a treatment context. American Journal of Speech-Language Pathology, 23, 530–545. [Article] [PubMed]×
Proctor-Williams, K. (2009). Dosage and distribution in morphosyntax intervention: Current evidence and future needs. Topics in Language Disorders, 29, 294–311. [Article]
Proctor-Williams, K. (2009). Dosage and distribution in morphosyntax intervention: Current evidence and future needs. Topics in Language Disorders, 29, 294–311. [Article] ×
Raudenbush, S., & Bryk, A. (2002). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks, CA: Sage.
Raudenbush, S., & Bryk, A. (2002). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks, CA: Sage.×
Raudenbush, S., Bryk, A., & Congdon, R. (2007). HLM for Windows (Version 6.04) [Computer software] . Lincolnwood, IL: Scientific Software International.
Raudenbush, S., Bryk, A., & Congdon, R. (2007). HLM for Windows (Version 6.04) [Computer software] . Lincolnwood, IL: Scientific Software International.×
Rispoli, M. (2016). Cross-morpheme facilitation: The systematic emergence of agreement in 2-year-olds. Language Acquisition, 23, 293–306. [Article]
Rispoli, M. (2016). Cross-morpheme facilitation: The systematic emergence of agreement in 2-year-olds. Language Acquisition, 23, 293–306. [Article] ×
Rispoli, M., & Hadley, P. (2011). Toward a theory of gradual morphosyntactic learning. In Arnon, I. & Clark, E. (Eds.), Experience, variation and generalization: Learning a first language (pp. 15–33). Amsterdam, the Netherlands: John Benjamins. [Article]
Rispoli, M., & Hadley, P. (2011). Toward a theory of gradual morphosyntactic learning. In Arnon, I. & Clark, E. (Eds.), Experience, variation and generalization: Learning a first language (pp. 15–33). Amsterdam, the Netherlands: John Benjamins. [Article] ×
Rispoli, M., & Hadley, P. (2014). Input effects on the acquisition of finiteness. In Chu, C., Coughlin, C., Prego, B. L., Minai, U., & Tremblay, A. (Eds.), Proceedings of the Fifth Conference on Generative Approaches to Language Acquisition North America (pp. 121–127). Somerville, MA: Cascadilla Press.
Rispoli, M., & Hadley, P. (2014). Input effects on the acquisition of finiteness. In Chu, C., Coughlin, C., Prego, B. L., Minai, U., & Tremblay, A. (Eds.), Proceedings of the Fifth Conference on Generative Approaches to Language Acquisition North America (pp. 121–127). Somerville, MA: Cascadilla Press.×
Rispoli, M., Hadley, P., & Holt, J. (2009). The growth of tense productivity. Journal of Speech, Language, and Hearing Research, 52, 930–944. [Article]
Rispoli, M., Hadley, P., & Holt, J. (2009). The growth of tense productivity. Journal of Speech, Language, and Hearing Research, 52, 930–944. [Article] ×
Rispoli, M., Hadley, P., & Holt, J. (2012). Sequence and system in the acquisition of tense and agreement. Journal of Speech, Language, and Hearing Research, 55, 1007–1021. [Article]
Rispoli, M., Hadley, P., & Holt, J. (2012). Sequence and system in the acquisition of tense and agreement. Journal of Speech, Language, and Hearing Research, 55, 1007–1021. [Article] ×
Rispoli, M., Papastratakos, T., Stern, C., & Hadley, P. (2015, June). Input packaging and the acquisition of copula is. Paper presented at the Symposium for Research in Child Language Disorders, Madison, WI.
Rispoli, M., Papastratakos, T., Stern, C., & Hadley, P. (2015, June). Input packaging and the acquisition of copula is. Paper presented at the Symposium for Research in Child Language Disorders, Madison, WI.×
Rowe, M. L. (2012). A longitudinal investigation of the role of quantity and quality of child-directed speech in vocabulary development. Child Development, 83, 1762–1774. [Article] [PubMed]
Rowe, M. L. (2012). A longitudinal investigation of the role of quantity and quality of child-directed speech in vocabulary development. Child Development, 83, 1762–1774. [Article] [PubMed]×
Sprent, P., & Smeeton, N. (2001). Applied nonparametric statistical methods. Boca Raton, FL: CRC Press.
Sprent, P., & Smeeton, N. (2001). Applied nonparametric statistical methods. Boca Raton, FL: CRC Press.×
Squires, J., & Bricker, D. (2009). Ages and Stages Questionairres–Third Edition. Baltimore, MD: Brookes.
Squires, J., & Bricker, D. (2009). Ages and Stages Questionairres–Third Edition. Baltimore, MD: Brookes.×
Thompson, S., & Newport, E. (2007). Statistical learning of syntax: The role of transitional probability. Language Learning and Development, 3, 1–42. [Article]
Thompson, S., & Newport, E. (2007). Statistical learning of syntax: The role of transitional probability. Language Learning and Development, 3, 1–42. [Article] ×
Tomasello, M. (2003). Constructing a language: A usage-based theory of language acquisition. Cambridge, MA: Harvard University Press.
Tomasello, M. (2003). Constructing a language: A usage-based theory of language acquisition. Cambridge, MA: Harvard University Press.×
Tyler, A., Lewis, K., Haskill, A., & Tolbert, L. (2002). Efficacy and cross-domain effects of a morphosyntax and a phonology intervention. Language, Speech, and Hearing Services in Schools, 33, 52–66. [Article] [PubMed]
Tyler, A., Lewis, K., Haskill, A., & Tolbert, L. (2002). Efficacy and cross-domain effects of a morphosyntax and a phonology intervention. Language, Speech, and Hearing Services in Schools, 33, 52–66. [Article] [PubMed]×
Tyler, A., Lewis, K., Haskill, A., & Tolbert, L. (2003). Outcomes of different speech and language goal attack strategies. Journal of Speech, Language, and Hearing Research, 46, 1077–1094. [Article]
Tyler, A., Lewis, K., Haskill, A., & Tolbert, L. (2003). Outcomes of different speech and language goal attack strategies. Journal of Speech, Language, and Hearing Research, 46, 1077–1094. [Article] ×
Valian, V. (1999). Input and language variation. In Ritchie, W. & Bhatia, T. (Eds.), Handbook of child language acquisition (pp. 497–530). San Diego, CA: Academic Press.
Valian, V. (1999). Input and language variation. In Ritchie, W. & Bhatia, T. (Eds.), Handbook of child language acquisition (pp. 497–530). San Diego, CA: Academic Press.×
Vasilyeva, M., Waterfall, H., & Huttenlocher, J. (2008). Emergence of syntax: Commonalities and differences across children. Developmental Science, 11, 84–97. [Article] [PubMed]
Vasilyeva, M., Waterfall, H., & Huttenlocher, J. (2008). Emergence of syntax: Commonalities and differences across children. Developmental Science, 11, 84–97. [Article] [PubMed]×
Yang, C. (2002). Knowledge and learning in natural language. Oxford, England: Oxford University Press.
Yang, C. (2002). Knowledge and learning in natural language. Oxford, England: Oxford University Press.×
Yang, C. (2004). Universal grammar, statistics, or both? Trends in Cognitive Sciences, 8, 451–456. [Article] [PubMed]
Yang, C. (2004). Universal grammar, statistics, or both? Trends in Cognitive Sciences, 8, 451–456. [Article] [PubMed]×
Figure 1.

Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on number of different lexical noun phrase subjects.

 Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on number of different lexical noun phrase subjects.
Figure 1.

Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on number of different lexical noun phrase subjects.

×
Figure 2.

Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on full is declaratives.

 Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on full is declaratives.
Figure 2.

Differences in tense agreement productivity growth trajectories from 21 to 30 months, controlling for children's number of different words and sentence diversity, with median split on full is declaratives.

×
Table 1. Means and standard deviations for parent input measures by time and condition.
Means and standard deviations for parent input measures by time and condition.×
Measure Control
Treatment
Full sample
M SD M SD M SD
Lexical NP subject diversity
 21 months 4.68 3.04 5.58 2.84 5.13 2.93
 24 months 6.00 2.96 18.05 9.99 12.03 9.49
+Tense verb forms
 21 months 148.05 57.25 157.42 51.12 152.74 53.74
 24 months 167.79 59.76 161.37 54.48 164.58 56.50
−Tense verb forms
 21 months 136.21 53.73 136.11 54.31 136.16 53.29
 24 months 136.95 59.10 126.42 36.26 131.68 48.65
Input informativeness
 21 months 52.42 10.69 54.13 8.08 53.27 9.38
 24 months 55.66 9.14 55.60 6.63 55.63 7.88
Full is declaratives
 21 months 3.32 3.92 2.74 2.56 3.03 3.28
 24 months 5.53 5.37 13.74 10.38 9.63 9.15
Note. Control n = 19; treatment n = 19; full sample N = 38. NP = noun phrase.
Note. Control n = 19; treatment n = 19; full sample N = 38. NP = noun phrase.×
Table 1. Means and standard deviations for parent input measures by time and condition.
Means and standard deviations for parent input measures by time and condition.×
Measure Control
Treatment
Full sample
M SD M SD M SD
Lexical NP subject diversity
 21 months 4.68 3.04 5.58 2.84 5.13 2.93
 24 months 6.00 2.96 18.05 9.99 12.03 9.49
+Tense verb forms
 21 months 148.05 57.25 157.42 51.12 152.74 53.74
 24 months 167.79 59.76 161.37 54.48 164.58 56.50
−Tense verb forms
 21 months 136.21 53.73 136.11 54.31 136.16 53.29
 24 months 136.95 59.10 126.42 36.26 131.68 48.65
Input informativeness
 21 months 52.42 10.69 54.13 8.08 53.27 9.38
 24 months 55.66 9.14 55.60 6.63 55.63 7.88
Full is declaratives
 21 months 3.32 3.92 2.74 2.56 3.03 3.28
 24 months 5.53 5.37 13.74 10.38 9.63 9.15
Note. Control n = 19; treatment n = 19; full sample N = 38. NP = noun phrase.
Note. Control n = 19; treatment n = 19; full sample N = 38. NP = noun phrase.×
×
Table 2. Correlations for 24-month parent input measures and 30-month child tense agreement productivity score.
Correlations for 24-month parent input measures and 30-month child tense agreement productivity score.×
Adult input measures 2 3 4 5 Child TAP score
1. Lexical NP subject diversity .45** .19 .20 .78*** .22
2. +Tense verb forms .52*** .41** .33* .19
3. −Tense verb forms −.53*** .07 .06
4. Input informativeness .22 .04
5. Full is declarative 29*
Note. N = 38. TAP = tense agreement productivity; NP = noun phrase.
Note. N = 38. TAP = tense agreement productivity; NP = noun phrase.×
* p < .05 (one-tailed).
p < .05 (one-tailed).×
** p < .01 (one-tailed).
p < .01 (one-tailed).×
*** p < .001 (one-tailed).
p < .001 (one-tailed).×
Table 2. Correlations for 24-month parent input measures and 30-month child tense agreement productivity score.
Correlations for 24-month parent input measures and 30-month child tense agreement productivity score.×
Adult input measures 2 3 4 5 Child TAP score
1. Lexical NP subject diversity .45** .19 .20 .78*** .22
2. +Tense verb forms .52*** .41** .33* .19
3. −Tense verb forms −.53*** .07 .06
4. Input informativeness .22 .04
5. Full is declarative 29*
Note. N = 38. TAP = tense agreement productivity; NP = noun phrase.
Note. N = 38. TAP = tense agreement productivity; NP = noun phrase.×
* p < .05 (one-tailed).
p < .05 (one-tailed).×
** p < .01 (one-tailed).
p < .01 (one-tailed).×
*** p < .001 (one-tailed).
p < .001 (one-tailed).×
×
Table 3. Descriptive statistics for child measures by condition and full sample by measurement point.
Descriptive statistics for child measures by condition and full sample by measurement point.×
Parameter Control
Treatment
Full
M SD M SD M SD
MLU
 21 months 1.20 0.22 1.23 0.22 1.22 0.21
 24 months 1.55 0.46 1.52 0.48 1.53 0.46
 27 months 1.93 0.48 1.95 0.47 1.94 0.47
 30 months 2.26 0.48 2.34 0.66 2.30 0.57
NDW
 21 months 18.53 12.38 26.11 14.02 22.32 13.56
 24 months 47.11 25.91 52.26 18.95 49.68 22.54
 27 months 75.11 31.64 79.37 25.03 77.24 28.22
 30 months 91.79 25.36 102.00 36.35 96.89 31.36
Sentence diversity (in third person sentences)
 21 months 0.16 0.69 0.37 0.68 0.26 0.68
 24 months 2.11 3.13 2.47 2.50 2.39 2.89
 27 months 6.00 4.87 5.11 4.31 6.05 4.90
 30 months 9.58 6.23 12.79 8.00 12.21 7.91
Tense and agreement productivity score
 21 months 0.11 0.46 0.32 0.67 0.21 0.57
 24 months 1.63 2.36 1.05 1.75 1.34 2.10
 27 months 4.11 3.80 2.74 2.83 3.47 3.45
 30 months 6.53 4.97 8.16 5.28 7.42 5.12
Note. Control group n = 19; treatment group n = 19; full sample N = 38. MLU = mean length of utterance; NDW = number of different words.
Note. Control group n = 19; treatment group n = 19; full sample N = 38. MLU = mean length of utterance; NDW = number of different words.×
Table 3. Descriptive statistics for child measures by condition and full sample by measurement point.
Descriptive statistics for child measures by condition and full sample by measurement point.×
Parameter Control
Treatment
Full
M SD M SD M SD
MLU
 21 months 1.20 0.22 1.23 0.22 1.22 0.21
 24 months 1.55 0.46 1.52 0.48 1.53 0.46
 27 months 1.93 0.48 1.95 0.47 1.94 0.47
 30 months 2.26 0.48 2.34 0.66 2.30 0.57
NDW
 21 months 18.53 12.38 26.11 14.02 22.32 13.56
 24 months 47.11 25.91 52.26 18.95 49.68 22.54
 27 months 75.11 31.64 79.37 25.03 77.24 28.22
 30 months 91.79 25.36 102.00 36.35 96.89 31.36
Sentence diversity (in third person sentences)
 21 months 0.16 0.69 0.37 0.68 0.26 0.68
 24 months 2.11 3.13 2.47 2.50 2.39 2.89
 27 months 6.00 4.87 5.11 4.31 6.05 4.90
 30 months 9.58 6.23 12.79 8.00 12.21 7.91
Tense and agreement productivity score
 21 months 0.11 0.46 0.32 0.67 0.21 0.57
 24 months 1.63 2.36 1.05 1.75 1.34 2.10
 27 months 4.11 3.80 2.74 2.83 3.47 3.45
 30 months 6.53 4.97 8.16 5.28 7.42 5.12
Note. Control group n = 19; treatment group n = 19; full sample N = 38. MLU = mean length of utterance; NDW = number of different words.
Note. Control group n = 19; treatment group n = 19; full sample N = 38. MLU = mean length of utterance; NDW = number of different words.×
×
Table 4. Children's tense agreement productivity score growth models.
Children's tense agreement productivity score growth models.×
Effects Model 1 linear Model 2 quadratic Model 3 quadratic with NDW Model 4 quadratic with NDW, sentence diversity
Fixed effects
 Intercept, b 00 6.45** 7.00** 5.71** 5.06**
 Linear growth, b 10 0.78** 1.44** 1.19** 0.86**
 Quadratic growth, b 20 0.08** 0.08** 0.06*
 NDW, b 30 0.04** 0.02
 Sentence diversity, b 40 0.20**
Random effects
 Intercept, VAR(r 0) 17.03** 21.12** 18.99** 14.61**
 Linear growth, VAR(r 1) 0.22** 1.62** 1.44** 0.94**
 Quadratic growth, VAR(r 2) 0.01** 0.011** 0.008**
 Level 1, VAR(e) 3.87 1.52 1.37 1.35
Note.N = 38. Models age centered at 30 months. NDW = number of different words.
Note.N = 38. Models age centered at 30 months. NDW = number of different words.×
* p < .01 (two-tailed).
p < .01 (two-tailed).×
** p < .001 (two-tailed).
p < .001 (two-tailed).×
Table 4. Children's tense agreement productivity score growth models.
Children's tense agreement productivity score growth models.×
Effects Model 1 linear Model 2 quadratic Model 3 quadratic with NDW Model 4 quadratic with NDW, sentence diversity
Fixed effects
 Intercept, b 00 6.45** 7.00** 5.71** 5.06**
 Linear growth, b 10 0.78** 1.44** 1.19** 0.86**
 Quadratic growth, b 20 0.08** 0.08** 0.06*
 NDW, b 30 0.04** 0.02
 Sentence diversity, b 40 0.20**
Random effects
 Intercept, VAR(r 0) 17.03** 21.12** 18.99** 14.61**
 Linear growth, VAR(r 1) 0.22** 1.62** 1.44** 0.94**
 Quadratic growth, VAR(r 2) 0.01** 0.011** 0.008**
 Level 1, VAR(e) 3.87 1.52 1.37 1.35
Note.N = 38. Models age centered at 30 months. NDW = number of different words.
Note.N = 38. Models age centered at 30 months. NDW = number of different words.×
* p < .01 (two-tailed).
p < .01 (two-tailed).×
** p < .001 (two-tailed).
p < .001 (two-tailed).×
×
Table 5. Input predictors of children's tense agreement productivity growth.
Input predictors of children's tense agreement productivity growth.×
Effects Model 5a input informativeness Model 5b lexical NP subject diversity Model 5c full is declarative
Fixed effects
 Intercept 5.14*** 5.15 *** 5.06 ***
  Input informativeness 5.83
  Lexical NP subject diversity 0.12
  Full is declarative 0.17 *
 Linear growth 0.90*** 0.90 *** 0.87 ***
  Input informativeness 2.89
  Lexical NP subject diversity 0.05 *
  Full is declarative 0.05 *
 Quadratic growth 0.06** 0.06 ** 0.06 **
  Input informativeness 0.24
  Lexical NP subject diversity 0.004 *
  Full is declarative 0.003
 NDW 0.02 0.02 0.02 *
  Input informativeness 0.04
  Lexical NP subject diversity −0.00009
  Full is declarative −0.001
 Sentence diversity 0.19*** 0.19 *** 0.19 ***
  Input informativeness −0.71
  Lexical NP subject diversity -0.004
  Full is declarative −0.0004
Random effects
 Intercept, VAR(r 0) 15.32*** 14.28* 13.41***
 Linear growth, VAR(r 1) 0.98*** 0.84*** 0.81***
 Quadratic growth, VAR(r 2) 0.008*** 0.007*** 0.007***
 Level 1 error, VAR(e) 1.30 1.30 1.32
Note. N = 38. Model age centered at 30 months. NP = noun phrase.
Note. N = 38. Model age centered at 30 months. NP = noun phrase.×
* p < .05 (two-tailed).
p < .05 (two-tailed).×
** p < .01 (two-tailed).
p < .01 (two-tailed).×
*** p < .001 (two-tailed).
p < .001 (two-tailed).×
p < .10 (two-tailed).
p < .10 (two-tailed).×
Table 5. Input predictors of children's tense agreement productivity growth.
Input predictors of children's tense agreement productivity growth.×
Effects Model 5a input informativeness Model 5b lexical NP subject diversity Model 5c full is declarative
Fixed effects
 Intercept 5.14*** 5.15 *** 5.06 ***
  Input informativeness 5.83
  Lexical NP subject diversity 0.12
  Full is declarative 0.17 *
 Linear growth 0.90*** 0.90 *** 0.87 ***
  Input informativeness 2.89
  Lexical NP subject diversity 0.05 *
  Full is declarative 0.05 *
 Quadratic growth 0.06** 0.06 ** 0.06 **
  Input informativeness 0.24
  Lexical NP subject diversity 0.004 *
  Full is declarative 0.003
 NDW 0.02 0.02 0.02 *
  Input informativeness 0.04
  Lexical NP subject diversity −0.00009
  Full is declarative −0.001
 Sentence diversity 0.19*** 0.19 *** 0.19 ***
  Input informativeness −0.71
  Lexical NP subject diversity -0.004
  Full is declarative −0.0004
Random effects
 Intercept, VAR(r 0) 15.32*** 14.28* 13.41***
 Linear growth, VAR(r 1) 0.98*** 0.84*** 0.81***
 Quadratic growth, VAR(r 2) 0.008*** 0.007*** 0.007***
 Level 1 error, VAR(e) 1.30 1.30 1.32
Note. N = 38. Model age centered at 30 months. NP = noun phrase.
Note. N = 38. Model age centered at 30 months. NP = noun phrase.×
* p < .05 (two-tailed).
p < .05 (two-tailed).×
** p < .01 (two-tailed).
p < .01 (two-tailed).×
*** p < .001 (two-tailed).
p < .001 (two-tailed).×
p < .10 (two-tailed).
p < .10 (two-tailed).×
×