Muh. Fadil. 2026. Improving Students’ English Speaking Skills through
Lingopanda: A Deep Learning-Based Speaking Practice Application at SMAN 4
Palopo. Thesis. English Language Education Study Program, Faculty of Teacher
Training and Education, Cokroaminoto Palopo University. Supervised by Andi
Mangnguntungi S. and Reski Yusrini Islamiah Yunus.
This research aimed to Improve students’ English-speaking skills using the
Lingopanda Application at SMAN 4 Palopo. The research employed a quantitative
approach using a pre-experimental method with a one-group pretest-posttest design.
The population of this research consisted of all Grade XI students of SMAN 4
Palopo in the academic year 2025/2026. The sample consisted of 16 students
selected through purposive sampling. The research instrument was an oral speaking
performance test that assessed two aspects of speaking skills, namely pronunciation
and fluency. The research procedures consisted of a pretest, four treatment meetings
using the Lingopanda application, and a posttest. The collected data were analyzed
using descriptive statistics and inferential statistics through the Shapiro-Wilk
normality test and the Paired Sample t-test in SPSS version 26. The findings showed
that the students' mean score increased from 40.00 in the pretest to 54.38 in the
posttest. The Paired Sample t-test revealed that the significance value (2-tailed) was
0.000, which was lower than the significance level of 0.05 (0.000 < 0.05).
Therefore, the alternative hypothesis (H₁) was accepted and the null hypothesis (H₀)
was rejected. The findings indicate that the use of the Lingopanda application
significantly improved students’ English speaking skills, particularly in the aspects
of pronunciation and fluency. Although the average achievement remained within
the Fairly Poor classification, students demonstrated measurable progress in
speaking performance after the implementation of the treatment.