The impact of direct and indirect feedback on the development of Spanish aspect

ABSTRACT

This study explores the impact of direct and indirect feedback on the improvement of Spanish past aspect (imperfect/preterit) in writing activities in a third-semester Spanish course. Ninety learners were divided into three groups: direct feedback, indirect feedback, and control (no feedback). Learners completed a pretest, a posttest, and a delayed posttest. Although it remains unclear which type of feedback is more effective, results indicate that both types improved preterit–imperfect performance. As the delayed posttest showed, however, learners exhibited more lasting improvement (as indicated by the delayed posttest) when working on the production Activity than on the terminology and selection activities.

RESUMEN

Este estudio explora el impacto de la retroalimentación directa e indirecta en la mejora del aspecto de pasado (pretérito/imperfecto) con actividades de escritura en un curso de tercer semestre de español. Se dividió noventa estudiantes en tres grupos: retroalimentación directa, retroalimentación indirecta, y un grupo de control (sin retroalimentación). Completaron un pretest, postest, y postest tardío. Aunque no hubo diferencias significativas entre los grupos experimentales, los resultados indican que los participantes mejoraron en el uso de aspecto verbal. No obstante, fue una mejora más perdurable (postest tardío) en la actividad de producción que en las actividades de terminología y selección.

KEYWORDS:

• Spanish aspect
• preterit
• imperfect
• direct feedback
• indirect feedback
• writing tasks

PALABRAS CLAVE:

• Aspecto en español
• pretérito
• imperfecto
• retroalimentación directa
• retroalimentación indirecta
• actividades escritas

Notes

1. This cutoff was established to reduce the confound of prior knowledge and its accompanying ceiling effect. Participants with high pretest scores were already relatively familiar with past-tense forms and thus would not have been able to show substantial improvement regardless of feedback treatment.

2. This pattern also extends to the trend-level findings on the post hoc tests for the Terminology Activity: Imperfect posttest and the Terminology Activity: Preterit delayed posttest. On the Production Activity: Imperfect delayed posttest, the indirect group outperforms control, but the direct group does not, and no differences are found between feedback groups.

3. The importance of conducting research with large samples is illustrated by the trend-level differences (p ≤.10) found for the Terminology Activity: Imperfect delayed posttest, the Selection Activity: Preterit delayed posttest, and the Production Activity: posttest. Although post hoc tests were not conducted for trend-level findings, an examination of the means show that the treatment groups scored similarly to each other but substantially higher than the control group; this is the predominant pattern in the statistically significant results. With a larger sample, there may have been a significant main effect of group, and post hoc tests may have revealed significant differences between the groups. In all, the findings may have been more robust with a larger sample, especially with regard to longer-term effects at delayed posttest.