Spanish Language Development Trajectories
There was a lack of longitudinal studies documenting the trajectory of production of linguistic complexity and accuracy by learners of Spanish. This research project was meant to fill that gap by studying, over one academic year, learners of Spanish in the second year of language classes. I also wanted to investigate the effect of task type on production. Further, this project was designed to capture naturalistic language production in the learners' classroom environment when traditionally, these types of studies took place in a laboratory or experimental setting.
Because I wanted to capture naturalistic data, I got permission from two instructors to recruit participants in their classrooms. One instructor taught at a large, urban university. The other taught at a small, selective liberal arts college. My participants were self-selected based on enrollment in these instructors' classes, and had the option to not participate if they chose not to.
I had three participants at the end of the study due to the natural attrition that occurs in longitudinal research. Each participant wore a microphone and pocket-sized audio recorder for the duration of their Spanish class once a week approximately bi-weekly. Those recordings were then transcribed and the language coded for analysis.
The students also responded to a survey I developed in Google forms. This survey was meant to gather data on their attitudes toward speaking Spanish and people with low, same level, and advanced proficiency in Spanish. It was also meant to gather demographic and background information.
Finally, students participated in semi-structured interviews at various points in the study to determine whether changes in attitudes were occurring over time.
I used five measures of language use: two relating to linguistic accuracy and three related to linguistic complexity. I also categorized all the tasks the students participated in. Once all the data were coded, I ran statistical analyses to determine which measures of linguistic complexity and accuracy changed significantly over time. I also ran analyses meant to determine whether task had an effect on the production of these measures, and then ran further tests to determine whether time interacted with task to affect linguistic complexity and accuracy.
Statistical methods used: ANOVAs, multi-level modeling, and Pearson's Correlations.
Lastly, to visualize the trajectory of production over time, I created visualizations of the dependent variables to show how they changed over time. The top three variables are related to linguistic complexity, and the bottom two are linguistic accuracy.
The analyses showed that students do experience variation in their production over time. It is not a linear process to improve Spanish skills over the course of an academic year. However, students did improve over the course of the academic year, and this improvement was statistically significant.
The analyses also showed that task was significantly correlated with increases in linguistic complexity and decreases in accuracy when the task was an open-ended task. Since students were more likely to produce more language and engage in things like storytelling, hypothesizing, and discussion in open-ended tasks, this result was not a surprise. Rather, it provided further support for allowing students to engage in these types of tasks in the classroom.
This research study was a case study of only three students; therefore, the results are not generalizable, and no broad recommendations can be made with any level of certainty. However, it appears that allowing open-ended tasks will promote development of linguistic complexity, with some gains in linguistic accuracy, though not as dramatic. It also provides support for the idea that language development is not a linear pattern; each student will travel along their own path as they develop their language skills.