<![CDATA[This study aims to determine the benefits of using machine images in improving the writing skills of cadets. Quasi-experimental designed was used by the researcher because the subjects were not randomly assigned. The population was the Polytechnic Bumi Akpelni of Marine Mechanical Engineering Technology Cadets in 2022/2023 academic year. In this study, the researcher used purposive sampling technique. The samples consisted of the cadets of Batch 57 semester 3 consisting of two classes. At the beginning of data collection, the researcher prepared the data through a lesson plan, namely the instruction instrument for the writing test and statements for the questionnaire. The teaching phase was carried out to measure whether both were appropriate, pre-test to find cadets' initial abilities and post-the test tests whether the treatment is effective in learning descriptive text. Through data analysis used, namely the scoring system in the form of rubrics, then to see significant differences in scores through studies of normality and homogeneity. From the pre-test result, the average score obtained was 56.66 from the experimental group, while it was 57.77 from the control group. After getting the treatment, the post-test average score of experimental group was 77.77 and control group was 70.10. From the t-test analysis, it was found that t-count was greater than t-table (2.50> 1.55). It indicates that there is a significant difference of the scores from the two groups. Therefore, the researcher concluded that Learned Image Machine is very helpful in improving the ability to write Descriptive Text compared to ordinary teaching without this media.]]>
