Los desafíos de realizar la investigación en el aula: el caso de un estudio réplica sobre la ansiedad en el aprendizaje del inglés

Abstract

Many authors in the field of education (Cronbach, 1980; Cronbach; Shapiro, 1982; Shadish; Cook; Campbell, 2002; Schneider, 2004) have recommended that investigations be replicated so that more faith can be placed in their results. Schneider affirmed that replication is “essential for being able to generalize to more people and settings than are represented in a single study” (2004: 1472). The same author favours replications in order to reap the benefits of “conducting an investigation repeatedly with comparable subjects and conditions so as to achieve what would be expected to be similar results” (2004: 1472). In addition to supporting or extending the findings of other researches, replications can lead to their reassessment or to their rejection.

This article reports on our replication of Phillips’s (1992) study, in which she investigated links between foreign language anxiety and performance on oral exams in university students. The respective academic situations of Phillips and ourselves were quite similar. For example, her students, like ours, were not foreign language majors, and the level of the foreign language subject was intermediate. Due to administrative constraints imposed by our respective universities, both Phillips and ourselves were obliged to work with an intact group of students, and were therefore unable to design a true experiment with random allocation of participants to more than one group.  As regards differences between the two studies, our participants were nearly all Spanish students of English, while Phillips’s were all North American students of French. Our language subject was elective, while Phillips’s was compulsory.

Our academic setting, therefore, was sufficiently like Phillips’s to warrant a replication which would allow us to further verify her conclusions or not. Using the same instrument to measure foreign language anxiety, the Foreign Language Classroom Anxiety Scale, FLCAS (Horwitz; Horwitz; Cope, 1986), the same criteria for grading the exams, and the same statistical analyses as Phillips did, we obtained results which were largely similar to hers. For instance, the Pearson correlations found between anxiety and oral exam scores in both investigations were observed to negative and statistically significant. On the other hand, Phillips’s analyses of variance did not show any statistically significant differences among three anxiety groups (low, moderate, and high), while in our investigation statistically significant differences were revealed. In this way, our study buttresses many of Phillips’s findings and offers new and valuable insights into anxiety in the learning of foreign languages.