이 연구의 목적은 청소년의 다문화 행동의도와 관련변인인 다문화 인식, 다문화 자기 효능감, 다문화 태도 간의 인과적 관계의 구조를 분석하는 데 있었으며, 그 구체적인 목표는 다음과 같다. 첫째, 청소년의 다문화 행동의도와 다문화 인식, 다문화 자기 효능감 및 다문화 태도의 구조적 가설모형을 설정하고 설정한 모형이 변인 간의 구조적 관계를 적합하게 예측하는지 검증한다. 둘째, 청소년의 다문화 행동의도에 대한 다문화 인식, 다문화 자기 효능감 및 다문화 태도의 영향 관계를 밝힌다. 셋째, 청소년의 다문화 행동의도와 관련 변인 간의 관계에서 다문화 자기 효능감과 다문화 태도가 나타내는 매개효과를 밝힌다. 이 연구의 모집단은 한국 청소년 전체이나 연구 주제의 적합성과 용이성을 고려하여 2015년 기준으로 고등학교 재학생을 목표모집단으로 설정하였다. 표집은 비확률적 표집 중 유의표본 추출법을 이용하여 총 840명을 표집하였다. 조사도구는 다문화 인식, 다문화 자기 효능감, 다문화 태도, 다문화 행동의도 및 인구통계학적 특성 조사문항으로 구성되었다. 조사도구는 다음과 같은 과정을 거쳐 개발되었다. 첫째, 다문화 인식, 다문화 자기 효능감, 다문화 태도, 다문화 행동의도 측정도구는 연구자가 선행연구 고찰을 토대로 우리나라 학교에 재학 중인 청소년에 적합한 내용을 추출하여 직접 문항을 개발하였다. 둘째, 문항의 내용타당도를 확보하기 위하여 전문가의 내용타당도를 검증하였다. 셋째, 1차 예비조사를 통하여 문항의 내용타당도를 점검하였다. 마지막으로, 2차 예비조사를 통하여 문항의 분포 및 신뢰도와 타당도를 검증하였다. 자료 분석은 측정도구의 내용타당도 검증을 위해 내용타당도 지표(Content Validity Index: CVI)를 활용하였다. 신뢰도와 타당도 등을 검증하기 위해 SPSS 22.0를 사용하여 기술통계, 왜도, 첨도를 산출하고 정규성 분포와 다중공선성을 확인하였으며, 상관관계와 신뢰도를 검증하였다. 그리고 각 잠재변인의 요인구조를 확인하기 위하여 탐색적 요인분석을 실시하였다. 또한 AMOS 22.0을 활용하여 문항의 타당성 확보를 위해 확인적 요인분석을 실행하였다. 추리통계 결과에 대한 통계적 유의성은 .05를 기준으로 판단하였으며, 모형 적합도의 판단 기준은 절대적합도 지수의 χ²값, Q값, SRMR, RMSEA, GFI, 증분적합 지수의 NFI, TLI, CFI, 간명적합도 지수의 PRATIO, 총 9개를 선정하여 구조모형의 적합도를 검증하였다. 이 연구의 주요 결과는 다음과 같다. 첫째, 청소년의 다문화 행동의도와 다문화 인식, 다문화 자기 효능감 및 다문화 태도에 관한 가설적 인과모형의 적합도가 전체적으로 양호한 결과로 나타나 다문화 행동의도와 관련 변인 간의 인과관계를 예측하는 데 적합한 모형이라고 판단되었다. 둘째, 청소년의 다문화 인식은 다문화 행동의도(β=-.058)에 직접적인 영향이 나타나지 않았으나, 다문화 자기 효능감(β=.783)과 다문화 태도(β=.223)에는 직접적 정적 영향을 미쳤다. 그리고 다문화 자기 효능감은 다문화 태도(β=.711)와 다문화 행동의도(β=.215)에 직접적인 영향을 미치며, 다문화 태도는 다문화 행동의도(β=.806)에 직접적인 영향을 미치는 것으로 나타났다. 셋째, 청소년의 다문화 인식과 다문화 행동의도 간의 관계에서 다문화 자기 효능감(=.231)과 다문화 태도(=.170)는 매개효과가 유의한 것으로 나타났다. 그리고 다문화 인식과 다문화 자기 효능감 간(=.667), 다문화 인식과 다문화 태도 간(=.299)의 직접적 효과가 있는 것으로 나타났다. 또한 다문화 인식은 다문화 행동의도에 직접효과(=.018)가 유의하지 않았으나 간접효과(=.624)가 유의하여 완전매개 효과가 있는 것으로 확인되었다. 한편 청소년의 다문화 자기 효능감이 다문화 태도(=.588)에 직접적 영향을 미치는 것으로 나타났다. 그리고 다문화 자기 효능감과 다문화 행동 의도와의 관계에서 다문화 태도(=.335)의 매개효과가 있었으며 다문화 태도와 다문화 행동의도 간의 직접효과(=.570)가 있는 것으로 확인되었다. 이 연구의 결과를 토대로 후속 연구를 위하여 다음과 같이 제언한다. 첫째, 이 연구에서 개발된 청소년 다문화 행동의도 모형의 일반화를 위해 다른 연령 집단이나 다른 지역 청소년 집단에게 적용하는 후속 연구가 이루어질 필요가 있다. 둘째, 이 연구에서 설정한 잠재 변인 이외에 청소년의 다문화 행동의도에 영향을 미친다고 실증된 변인(예: 다문화 교육 경험, 다문화 감수성, 다문화 민감성 등)을 포함한 다문화 행동 모형을 개발하여 검증하는 연구가 필요하다. 셋째, 청소년을 둘러싼 맥락(환경) 효과까지 추정할 수 있도록 위계선형모형(Hierarchical Linear Models: HLM) 분석을 활용하는 연구가 필요하다. 넷째, 이 연구에서 청소년의 다문화 행동의도에 영향을 미친다고 밝혀진 다문화 인식, 다문화 효능감, 다문화 태도를 청소년에게 함양하기 위해서는 중재프로그램을 개발하는 후속 연구가 필요하다. 더 나아가 중재프로그램을 지속적으로 적용한 후에 잠재 성장모형(Latent Growth Models) 분석을 통하여 중재 프로그램의 종단적 효과를 밝히는 연구도 필요하다.
The purpose of this study was to identify a structural relations among multicultural behavior intention of adolescents and its related variables, which are multicultural cognition, multicultural self efficacy and multicultural attitude. The specific goals set for the research are as follows; First, to set the structural relation model for multicultural behavior intention, multicultural cognition, multicultural self efficacy and multicultural attitude of adolescents and look deep into whether this model is appropriate for predicting the cause-and-effect relations among each variable. Second, to reveal the effect relation of multicultural cognition, multicultural self efficacy and multicultural attitude regarding adolescent’s multicultural behavior intention. Third, to find out the mediating effect produced by multicultural self efficacy and multicultural attitude amid the relation of adolescent’s multicultural cognition and multicultural behavior intention. Fourth, to find out the mediating effect produced by multicultural attitude amid the relation of adolescent’s multicultural self efficacy and multicultural behavior intention. The population of the research was all adolescents in Korea. However, considering the suitability with the research and accessibility, a target population of 1,788,266 students currently attending high school (Korea Education Development Institute, 2015) in the year 2015, was set. Sampling method used was purposive sampling, one of non-probability sampling, and 840 students were sampled. The research tool was a set of survey questions of multicultural cognition, multicultural self efficacy, multicultural attitude, multicultural behavior intention and demographic characteristics. The measuring tool for multicultural cognition, multicultural self efficacy, multicultural attitude and multicultural behavior intention was self-extracted survey questions which are appropriate for adolescents currently attending school in Korea. They were made through considerations after a pilot study, also carried out by the researcher herself. The research tool was made through several procedures. To mention the steps, there was a first preliminary research to comprehend the understanding of the questions, verification of content validity to secure the questions content validity and a second preliminary research to secure the validity and reliability of the distribution of questions. Concerning data collection, the preliminary research had surveys from 425 students in 3 schools in capital areas of the country and for the actual research, 825 students from 2 schools in capital areas and 1 school in the provincial area, a total of 3 schools, answered the survey. Exempting 92 surveys, which had unanswered or insincerely answered questions, the valid analysis ratio was 88.8% and the total number of survey papers used for analysis was 733. Data analysis used the Content Validity Index(CVI) to verify the content validity of research tools. To secure the validity and reliability of the result, SPSS 22.0 was used to verify frequency, ratio, average, standard deviation, descriptive statistics such as skewness value and kurtosis, Item total correlation and credibility. Moreover, Exploratory Factor Analysis(EFA), with the application of maximum likelihood and oblimin, was practiced. In the next stage, Confirmatory Factor Analysis(CFA) was conducted in order to secure the validity of the questions by using AMOS 22.0. At this stage, in order to understand the compatibility of the model and path analysis among variables, the number of trials of the Bootstrap test was set to 500 to conduct structure analysis. The statistical significance of the inferential statistics result was judged with 0.5 as the standard point and the criteria of the model suitability was verified with 9 of the selected; χ², Q, SRMR, RMSEA, GFI of absolute fit index, NFI, TLI, CFI of incremental fit index and PRATIO of parsimonious fit index. The result of the experiment can be summarized as below. First, the validity of the hypothetical cause-and-effect model, regarding adolescents’ multicultural behavior intention, multicultural cognition, multicultural self efficacy and multicultural attitude, had an overall satisfying result (SRMR=.0383, GFI=.894, NFI=.930, TLI=.906, CFI=.935, PRATIO=.691) and was concluded to be an appropriate model to predict the cause-and-effect relation of the variables. Second, adolescents’ multicultural cognition did not have direct influence (β=-.058) on multicultural behavior intention. However, on multicultural self efficacy (β=.783) and multicultural attitude (β=.223), it had direct static influence with the significance level of .001. Also, multicultural self efficacy had direct influence on multicultural attitude (β=.711) and multicultural behavior intention (β=.215) and multicultural attitude had direct influence on multicultural behavior intention (β=.806). Thirdly, in the relation between multicultural cognition and multicultural behavior intention of adolescents, multicultural self efficacy (=.231) and multicultural attitude (=.170) there was significance of indirect mediating effect. Also, a significance of direct influence was revealed between multicultural cognition and multicultural attitude (=.299), and multicultural cognition and multicultural self efficacy (=.667). It was verified that there is direct effect between multicultural attitude and multicultural behavior intention (=.570) and despite the fact that there was no sign of direct effect between multicultural cognition and multicultural behavior intention (=.018), indirect effect (=.624) had the significance level of .001. Fourth, in the relation between adolescents’ multicultural self efficacy and multicultural behavior intention, the mediating effect of multicultural attitude was significant (=.335) and it showed that multicultural self efficacy had direct influence on multicultural attitude (=.588). Based on the results of the study, some recommendations for future research were suggested as follows; First, in the task of analyzing data, it is required to distinguish the regions, the school and the class where the adolescents are involved in, and also the individual level. Secondly, there should be consideration on latent variables such as multicultural education experience, multicultural sensitivity and multicultural impressionability which were not dealt in this project. Thirdly, it is necessary to take advantage of Hierarchical Linear Models analysis in order to estimate the context effects on the youth in the further research. Fourthly, in order to foster multicultural awareness, multicultural efficacy and multicultural attitude that found to influence on the multicultural behavior intention of youth, it is necessary to develop an intervention program in the further research. Furthermore, the research on the longitudinal effect of the intervention is needed by using Latent Growth Models analysis after consistently applying the intervention program.
목 차국문초록Ⅰ. 서론1. 문제의 제기 ···································································· 12. 연구의 목적 ···································································· 53. 용어의 정의 ···································································· 64. 연구의 범위 ···································································· 8Ⅱ. 이론적 배경1. 다문화 행동의도 ································································· 92. 다문화 행동의도 관련 변인 ················································· 263. 다문화 행동의도와 관련 변인 간의 관계 ·································· 734. 연구 모형 및 가설····························································· 86Ⅲ. 연구 방법1. 연구 대상 및 표본 ····························································· 892. 조사 도구 ······································································· 903. 자료 수집 ······································································ 1294. 자료 분석 ······································································ 131Ⅳ. 연구 결과1. 응답자의 일반적 특성 ························································· 1392. 관찰변인의 기술통계 및 정규성 검증 ········································ 1393. 측정모형 분석 결과 ···························································· 1434. 구조모형 분석 및 모형 수정 ·················································· 1465. 최종 구조모형 분석 및 가설 검증 ············································ 153Ⅴ. 논의 및 결론1. 논의 ············································································· 1612. 결론 ············································································· 171참고문헌 ·········································································· 174Abstract ·········································································· 216부록 ··············································································· 221