ABSTRACT
Writing is one of the four skills, which must be mastered by the student’s who learn English. Writing is a skill, not a gift we are born with and it can be learned. Writing needs practice. However, students usually need help to develop their writing skill appropriate with their readiness and ages or they should be considered as a group of individuals with their own specific nature factors to be taken into account.
Based on this opinion, the writer conducted a study to minimize the students’ errors in building sentences through games. The study that the writer chose was classroom action research.
The subject of this classroom action research was the student of VII B of SMPN 26 Semarang, 40 students were involved, consisting of 21 girls and 19 boys.
The instrument that was used to collect the data was a test. To obtain the expected data for this classroom action research, the procedures used were pre-test and two cycles. Each cycle was analyzed and observed descriptively.
The result obtained on each cycle showed that either the first cycle or the second one had different percentage in minimizing the students’ errors in building sentences. The percentage of error in the pre-test could be reduced in the first cycle, from 53.25% into 27.75% and it could be further reduced in the second cycle into 20%. The mean score of the each cycle also showed the differences. The second cycle had the highest mean score that was 8.
In order to know that game was effective in building sentences, the writer applied the Analysis of Variance (ANOVA) formula. It was used to determine whether there was a significant difference between the means of pre-test, cycle 1, and cycle 2. Eventually, she found that the difference was statistically significant. The writer suggests that the use of various kinds of games is effective to minimize the students’ errors in building sentences.
