Coh-Metrix is among the broadest and most sophisticated automated textual assessment tools available today. Automated Evaluation of Text and Discourse with Coh-Metrix describes this computational tool, as well as the wide range of language and discourse measures it provides. Part I of the book focuses on the theoretical perspectives that led to the development of Coh-Metrix, its measures, and empirical work that has been conducted using this approach. Part II shifts to the practical arena, describing how to use Coh-Metrix and how to analyze, interpret, and describe results. Coh-Metrix opens the door to a new paradigm of research that coordinates studies of language, corpus analysis, computational linguistics, education, and cognitive science. This tool empowers anyone with an interest in text to pursue a wide array of previously unanswerable research questions.

In this study, a corpus of expert-graded essays, based on a standardized scoring rubric, is computationally evaluated so as to distinguish the differences between those essays that were rated as high and those rated as low. The automated tool, Coh-Metrix, is used to examine the degree to which high- and low-proficiency essays can be predicted by linguistic indices of cohesion (i.e., coreference and connectives), syntactic complexity (e.g., number of words before the main verb, sentence structure overlap), the diversity of words used by the writer, and characteristics of words (e.g., frequency, concreteness, imagability). The three most predictive indices of essay quality in this study were syntactic complexity (as measured by number of words before the main verb), lexical diversity (as measured by the Measure of Textual Lexical Diversity), and word frequency (as measured by Celex, logarithm for all words). Using 26 validated indices of cohesion from Coh-Metrix, none showed differences between high- and low-proficiency essays and no indices of cohesion correlated with essay ratings. These results indicate that the textual features that characterize good student writing are not aligned with those features that facilitate reading comprehension. Rather, essays judged to be of higher quality were more likely to contain linguistic features associated with text difficulty and sophisticated language.

No abstract

The opinions of second language learning (L2) theorists and researchers are divided over whether to use authentic or simplified reading texts as the means of input for beginning-and intermediate-level L2 learners. Advocates of both approaches cite the use of linguistic features, syntax, and discourse structures as important elements in support of their arguments, but there has been no conclusive study that measures these differences and their implications for L2 learning. The purpose of this article is to provide an exploratory study that fills this gap. Using the computational tool Coh-Metrix, this study investigates the differences between the linguistic structures of sampled simplified texts and those of authentic reading texts in order to provide a better understanding of the linguistic features that comprise these text types. The findings demonstrate that these texts differ significantly, but not always in the manner supposed by the authors of relevant scholarship. This research is meant to enable material developers, publishers, and classroom teachers to judge more accurately the value of both authentic and simplified texts.

No abstract

A reliable index of lexical diversity (LD) has remained stubbornly elusive for over 60 years. Meanwhile, researchers in fields as varied as stylistics, neuropathology, language acquisition, and even forensics continue to use flawed LD indices — often ignorant that their results are questionable and in some cases potentially dangerous. Recently, an LD measurement instrument known as vocd has become the virtual tool of the LD trade. In this paper, we report both theoretical and empirical evidence that calls into question the rationale for vocd and also indicates that its reliability is not optimal. Although our evidence shows that vocd's output (D) is a relatively robust indicator of the aggregate probabilities of word occurrences in a text, we show that these probabilities — and thus also D — are affected by text length. Malvern, Richards, Chipere and Durán (2004) acknowledge that D (as calculated by vocd's default method) can be affected by text length, but claim that the effects are not significant for the ranges of text lengths with which they are concerned. In this paper, we explain why D is affected by text length, and demonstrate with an extensive empirical analysis that the effects of text length are significant over certain ranges, which we identify.

The words people use and the way they use them can reveal a great deal about their mental states when they attempt to deceive. The challenge for researchers is how to reliably distinguish the linguistic features that characterize these hidden states. In this study, we use a natural language processing tool called Coh-Metrix to evaluate deceptive and truthful conversations that occur within a context of computer-mediated communication. Coh-Metrix is unique in that it tracks linguistic features based on cognitive and social factors that are hypothesized to influence deception. The results from CohMetrix are compared to linguistic features reported in previous independent research, which used a natural language processing tool called Linguistic Inquiry and Word Count. The comparison reveals converging and contrasting alignment for several linguistic features and establishes new insights on deceptive language and its use in conversation.

Two sentences are paraphrases if their meanings are equivalent but their words and syntax are different. Paraphrasing can be used to aid comprehension, stimulate prior knowledge, and assist in writing-skills development. As such, paraphrasing is a feature of fields as diverse as discourse psychology, composition, and computer science. Although automated paraphrase assessment is both commonplace and useful, research has centered solely on artificial, edited paraphrases and has used only binary dimensions (i.e., is or is not a paraphrase). In this study, we use an extensive database (N=1,998) of natural paraphrases generated by high school students that have been assessed along 10 dimensions (e.g., semantic completeness, lexical similarity, syntactical similarity). This study investigates the components of paraphrase quality emerging from these dimensions and examines whether computational approaches can simulate those human evaluations. The results suggest that semantic and syntactic evaluations are the primary components of paraphrase quality, and that computationally light systems such as latent semantic analysis (semantics) and minimal edit distances (syntax) present promising approaches to simulating human evaluations of paraphrases.