Two experiments examined students' ability to understand definitions of novel verbs. In both experiments, upper elementary students were given definitions of pseudowords paired with sentences using those words, and asked to evaluate whether the use of the word in the sentence was consistent with the meaning given in the definition. The sentence was either appropriate for the definition, consistent with a fragment of the definition but not the definition as a whole (fragment selection error), or completely inconsistent with the definition. Fragment selection error sentences were intended to represent a frequently made error observed in previous research in which children had been asked to generate sentences based on definitions. In the first experiment 60 fourth‐grade participants were able to correctly judge appropriate or completely inconsistent sentences accurately, but incorrectly judged sentences containing fragment selection error sentences as consistent with the definitions about 50% of the time, indicating a chance level of performance on such items. In the second experiment 55 fourth‐grade and 45 sixth‐grade students were administered the same task with three types of definitions: traditional definitions, traditional definitions with an example sentence, and innovative definitions written in an informal and presumably more user‐friendly style. Provision of example sentences produced a statistically significant but small improvement in performance for fragment selection error sentences, but for both grade levels, all levels of verbal ability, and all three definition types, students' judgments about these sentences were not substantially different from chance. Results indicate that students' difficulties with definitions observed in previous research using a writing task are not limited to this task; nor can they be attributed to unfamiliarity with the format of traditional definitions. Rather, students appear to experience fundamental difficulties using information provided in definitions about general syntactic or semantic categories of new words.

EN DOS experimentos se examinó la habilidad de los estudiantes para comprender definiciones de verbos desconocidos. En ambos experimentos, a estudiantes del segundo ciclo de la escuela primaria se les dieron definiciones de pseudo‐palabras apareadas con oraciones en las que se usaban esas palabras y se les pidió que evaluaran si el uso de la palabra en la oración era consistente con el significado dado en la definición. La oración era apropiada para la definición, o consistente con un fragmento de la definición pero no con la definición como un todo (error de selección de fragmento), o completamente inconsistente con la definición. Las oraciones del tipo error de selección de fragmento intentaban representar un error frecuente observado en una investigación previa, en la que a los niños se les había pedido que generaran oraciones basadas en definiciones. En el primer experimento, 60 participantes de cuarto grado fueron capaces de juzgar correctamente oraciones apropiadas o completamente inconsistentes, pero juzgaron incorrectamente oraciones que contenían error de selección de fragmento como si fueran oraciones consistentes con las definiciones, en un 50% de las veces. Esto indica un desempeño en el nivel del azar en estos ítemes. En el segundo experimento, a 55 estudiantes de cuarto grado y a 45 de sexto grado se les adminstró la misma tarea con tres tipos de definiciones: definiciones tradicionales, definiciones tradicionales con una oración de ejemplo y definiciones innovadoras escritas en un estilo más informal. La inclusión del ejemplo produjo un progreso …