Index of Productive Syntax (IPSyn) is a method for evaluating and quantifying the grammatical complexity of young children’s spontaneous language samples. IPSyn is based upon the grammatical categories and developmental scheme of Assigning Structural Stage (Miller, 1981). Individual utterances in a sample are scored for the occurrence of 60 different syntactic forms categorized under four subscales: noun phrase, verb phrase, question/negation, and sentence structure forms. Although IPSyn was developed for use with mainstream English speakers, it has been found to be a valid measure for children who speak African American English (Oetting et al., 2010).
IPSyn scores are calculated from children’s spontaneous language samples. After a sample is elicited and recorded, a corpus is formed by transcribing 100 successive, intelligible utterances, excluding imitations, self-repetitions, and routines (stereotyped language that does not represent productive language use). Following guidelines stated in the IPSyn coding manual (Scarborough, 1990), the transcript is then scored for 60 syntactic forms belonging to phrases and sentences that are categorized under four subscales. The first two different occurrences of each form are credited. No additional credit is given for further occurrences in the sample. Certain early-developing structures are automatically credited if there is evidence in the transcript that the child has mastered later-developing structures in the same category. For example, if a child produces a question containing a wh-pronoun and verb (Q4), he also receives credit for intonationally marked question (Q1) and routine do/go or existence/name question (Q2). The points earned for each subscale are summed across all utterances, then added together to yield the total IPSyn score. No examiner qualifications are explicitly stated for this measure; however, the user must be familiar with English grammatical structures.
IPSyn was developed to provide a time-efficient means for measuring the grammatical complexity of children’s utterances from the beginning of oral language through the period of early literacy. Other measures of grammatical complexity, such as Mean Length of Utterance (MLU, Brown, 1973; Miller, 1981) and Developmental Sentence Scoring (DSS, Lee, 1974) had been and continue to be used for that purpose, but these other measures were criticized for their lack of scope and insensitivity to many important developmental language changes (Klee & Fitzgerald, 1985; Klee & Sahlie, 1986; Scarborough, Rescorla, Tager-Flusberg, Fowler, & Sudhalter, 1991). IPSyn derived its grammatical categories and developmental scheme from Assigning Structural Stage (ASS, Miller, 1981), a procedure generally hailed for its comprehensiveness. Furthermore, in contrast to measures such as MLU and DSS, IPSyn does not score every structure of every utterance within a corpus. Instead, it scans for a maximum of two structural types within each of its 60 syntactic forms. This approach results in an analysis that is adequately comprehensive yet not so time consuming as methods that evaluate all the structural forms produced by a child. IPSyn has been used in research on typical language development and disordered language development, including autism (e.g., Tager-Flusberg & Calkins, 1990; Tager-Flusberg et al., 1990). It is especially valuable to researchers using designs that require subject matching by language age (Scarborough et al., 1991).
No psychometric data are provided. IPSyn is not norm referenced.
As a clinical resource, IPSyn can be used to quantify and document linguistic changes that occur over time in successive language samples. A small database (n = 15) exists of typically developing children who were sampled at 24, 30, 36, 42, and 48 months (Scarborough, 1990). Although these numbers fall short of what is needed for standardization, the available data show how IPSyn could be used for peer comparisons, if and when a larger standardization sample is gathered. Two computerized versions of IPSyn have been developed. In the version contained in Computerized Profiling (Long, Fey, & Chapman, 2006), the program performs an automatic analysis that must then be reviewed and edited by the user. Another version, designed to work with CHAT transcripts in the CHILDES system, is completely automatic (Sagae, Lavie, & MacWhinney, 2005).