The Correlation between Self-concept and the Students’ Speaking Ability

Abstract

A self-concept is a collection of beliefs about one's own nature, unique qualities, and typical behavior. Your self-concept is your mental picture of yourself. It is a collection of self-perceptions. A self-concept might include such beliefs as 'I am easygoing' or 'I am pretty' or 'I am hardworking. The aspect of self-concept is divided into three; academic self-concept, personal self-concept and social self-concept. The students’ level of self-concept was good. It is shown by the mean score of the written test that was 70.37. Speaking is the unique human ability to communicate information, idea, and emotion to others using oral language. There are five basic types of speaking; they are; imitative, intensive, responsive, interactive and extensive. The students’ speaking skill was under the school’s standard minimum score. It is shown by the mean score of the speaking test which was 64.91, while the school’s standard minimum score was 70. To make the students successful in EFL speaking achievement, the teacher should improve students’ social development and pro social behavior, including increasing liking for students, reducing racial stereotyping and discrimination, increasing self-esteem, self-concept, and self-direction, self-expectations, sense of intellectual competence, and increasing liking for the class. Effective communication in speaking is influenced by superficial cause’s self-concept, motivation to interact with strange, reaction to strangers, social categorization of strangers, situational process, and connection with strangers. According to the interpretation table written by Arikunto (2010, p.245), the value of r takes a position between 0.20 – 0.40. Therefore, the correlation between students’ ability to translate report text and their ability to identify topic sentence is low as it is seen in table 4.8. T-test was calculated 0.9568. The calculation was seen from this real data as a result, ttable: (33) in the standard errors of 5% = 2.750, so, t-observed < t-table (1.8730 < 2.750).