디지털 리터러시는 디지털 대전환 시대를 살아가는 학생들에게 필수적인 역량으로, OECD와 유럽연합(EU) 등 국제기구를 중심으로 그 중요성이 지속적으로 강조되고 있다. 디지털 리터러시는 Gilster의 개념화 이후 단순한 디지털 기기 활용 능력을 넘어, 정보 창출, 협업, 문제 해결, 윤리적 의사결정까지 포함하는 종합적 역량으로 확장되어 왔다.

우리나라는 2022 개정 교육과정을 통해 디지털 리터러시를 ‘디지털 지식과 기술에 대한 이해를 바탕으로 정보를 수집·분석·비판적으로 이해하고 새로운 정보를 창출·활용하는 능력’으로 정의하고 있으며, 학습자가 디지털 환경에서 능동적으로 가치를 창출하는 주체로 성장하는 것을 교육 목표로 제시하고 있다.

한편, 한국교육학술정보원(KERIS)은 2007년부터 초·중학생의 디지털 리터러시 수준을 지속적으로 측정·분석해 왔으며, 2022년에는 수행형 검사 체제로의 전환을 위한 기초연구를 통해 디지털 리터러시 개념을 디지털 시민성과 문제 해결 역량을 포함하는 방향으로 재정의하였다. 이어 2023년에는 학습자가 실제 디지털 환경과 유사한 맥락에서 역량을 발휘할 수 있도록 수행형 검사 체제로 전환하였다. 수행형 검사는 실생활 시나리오를 기반으로 학습자의 문제 해결 과정을 측정할 수 있다는 장점이 있으나, 온라인 검사 시스템의 고도화와 기존 선다형 검사와는 다른 분석 방법의 적용이 요구된다.

본 연구는 이러한 배경을 바탕으로 국가 수준에서 초·중학생의 디지털 리터러시 수준을 측정하고, 수행형 디지털 리터러시 검사 도구의 개선 방향과 디지털 리터러시 교육의 방향성을 제시하는 것을 목적으로 한다. 이를 위해 초등학교 4학년부터 중학교 3학년까지의 학생을 대상으로 디지털 리터러시 수준을 측정하고, 향후 교육 정책 수립을 위한 실증적 자료를 제공하고자 한다.

본 연구의 세부 내용은 다음과 같다. 첫째, 디지털 리터러시 관련 문헌 분석 및 전문가 협의회 등을 통해 수행형 디지털 리터러시 검사 도구의 신뢰도와 타당도를 제고한다. 둘째, 전국 단위 층화 표집을 통해 초·중학생의 디지털 리터러시 수준을 체계적으로 측정·분석하고, 배경 변인과 로그 데이터를 활용한 심층 분석을 실시한다. 셋째, 수행형 검사의 특성을 반영하여 로그 데이터를 활용한 다각적인 수준 해석 방안을 모색한다. 넷째, 분석 결과를 토대로 디지털 리터러시 역량 강화를 위한 교육 정책 수립에 활용 가능한 기초 자료를 제공한다.

이를 위해 본 연구는 전국 17개 시도별로 지역 규모, 성별, 학급당 학생 수 등을 고려하여 전국 초등학교 4·5·6학년 및 중학교 1·2·3학년 학생 수의 약 1.6%에 해당하는 학생들에 대해 유층 무선 표집을 실시하였다. 최종적으로 초등학교는 총 266개교 17,718명, 중학교는 총 255개교 22,687명의 데이터를 결과 분석에 활용하였다.

본 연구의 주요 결과는 다음과 같다. 첫째, DigComp 2.2, OECD PISA 2025 LDW, ICILS, UNESCO AI 역량 프레임워크 등 국내외 주요 프레임워크를 분석하여 현행 디지털 리터러시 검사 체계의 개선 방향을 검토하였다. 분석 결과, 본 검사 도구는 정보·데이터 활용, 디지털 의사소통과 협력, 디지털 자원 생산, 디지털 안전과 건강, 실생활 문제 해결 등 국제적으로 강조되는 핵심 역량을 포괄하고 있는 것으로 확인되었다. 다만 최근 국제 프레임워크가 AI 활용 역량, 생성형 AI의 이해와 비판적 활용, 인간–AI 협업과 윤리적 판단을 점차 핵심 요소로 포함하고 있다는 점에서, 향후 국가 수준 디지털 리터러시 측정에서도 이러한 요소를 보다 체계적으로 반영할 필요성이 제기된다. 본 연구에서 일부 문항에 생성형 AI 관련 내용을 포함한 것은 이러한 흐름에 대한 초기적 대응으로 볼 수 있다.

둘째, 초등학교 디지털 리터러시 검사 총점은 100점 환산 점수 기준 평균 60.46점으로, 전반적으로 부적 편포를 보였다. 중학교는 평균 59.98점으로 나타났으며, 점수 분포 역시 부적 편포를 보였다. 초등학교와 중학교 모두에서 하위 영역별 성취 특성과 배경 변인에 따른 차이가 확인되었으며, 그 양상은 학교급에 따라 다르게 나타났다. 두 학교급 모두 디지털 도구 활용과 디지털 정보·데이터 탐색 영역은 비교적 안정적인 성취를 보인 반면, 디지털 자원 생산 영역은 상대적으로 낮은 성취 수준을 보였다. 초등학교에서는 디지털 자원 생산 영역 전반에서 낮은 성취가 관찰된 반면, 중학교에서는 해당 영역 중 알고리즘 저작형 문항에서만 상대적으로 낮은 성취가 확인되었다. 배경 변인 분석 결과, 학년이 높아질수록 성취 수준이 높아지는 경향이 나타났으며, 여학생의 성취 수준이 남학생보다 높게 나타났다. 지역 규모별로는 특별·광역시 및 중소도시 학생들의 평균 점수가 읍면 지역보다 높았으나, 중학교에서는 도서벽지 지역 학생들의 평균 점수가 상대적으로 높게 나타나 표본 수를 고려한 신중한 해석이 요구된다.

셋째, 다층 잠재 프로파일 분석 결과, 초등학교와 중학교 모두에서 디지털 리터러시 성취 조합에 따라 서로 다른 4개의 학습자 유형이 도출되었다. 두 학교급 모두에서 계층 1은 모든 평가 영역에서 가장 높은 성취를 보였으며, 디지털 자원 생산 영역에서도 상대적으로 높은 수준을 유지하였다. 반면 중간 및 하위 계층에서는 디지털 자원 생산과 디지털 정보·데이터 영역의 취약성이 함께 나타나는 유형이 확인되었다. 이는 디지털 자원 생산 영역의 낮은 성취가 모든 학생에게 보편적으로 나타나는 현상이 아니라, 집단에 따라 뚜렷하게 분화된 특성임을 보여준다.

넷째, 잠재 프로파일 유형에 대한 영향 요인 분석 결과, 초등학교와 중학교에서 상위 성취 집단에 도달하는 경로는 질적으로 상이한 것으로 나타났다. 초등학교에서는 학생 개인의 디지털 활용 경험과 효능감, 흥미와 동기가 상위 계층 진입에 중요한 요인으로 작용한 반면, 중학교에서는 이러한 개인 요인과 함께 교사의 교수·학습 실천과 인식이 보다 직접적인 영향을 미쳤다. 특히 중학교에서는 비판적 정보 판단을 강조하는 수업일수록 상위 계층 소속 가능성이 높게 나타났다.

다섯째, 로그 데이터 분석 결과, 문항 유형과 검사 설계 방식에 따라 학생들의 문제 해결 과정과 참여 양상이 뚜렷하게 구분되었다. 선다형 문항은 안정적인 응답 패턴을 보인 반면, 기능형·저작형·알고리즘형 문항에서는 로그 발생량과 체류 시간이 크게 증가하였다. 검사 후반부로 갈수록 무응답과 문항 건너뛰기가 증가하는 경향이 나타났으며, 일부 문항에서는 조작 미완료로 인한 오답 사례도 다수 확인되었다. 또한 로그 데이터를 통해 ‘시도하지 않은 경우’와 ‘시도했으나 해결하지 못한 경우’를 구분할 수 있는 가능성이 확인되었다.

이러한 결과를 토대로 디지털 리터러시 교육 및 정책을 위한 시사점은 다음과 같다. 첫째, 우리나라 학생들의 디지털 리터러시 향상을 위해 디지털 활용 능력 중심의 접근을 넘어, 디지털 자원 생산과 문제 해결 중심의 학습 경험을 체계적으로 강화할 필요가 있다. 본 연구 결과 디지털 자원 생산 영역은 전체적으로 취약한 영역으로 나타났으나, 상위 성취 집단에서는 안정적인 성취가 확인되어 적절한 학습 경험과 교육 환경이 제공될 경우 충분히 형성 가능한 역량임을 보여준다. 이에 따라 향후 디지털 리터러시 교육은 평균 수준 향상과 더불어 상위 집단의 효과적인 학습 경험을 확산하고, 중간 및 하위 집단의 영역별 취약성을 정밀하게 지원하는 방향으로 설계될 필요가 있다.

둘째, 수행형 검사를 고려한 평가 모델 및 수행 체계를 갖출 필요가 있다. 로그 데이터 분석 결과는 기존 결과 중심 평가의 한계를 보완할 가능성을 보여주며, 문제 해결을 위한 다양한 시도, 탐색, 중도 이탈 양상 등은 학생들의 문제 해결 전략과 인지적 부담을 반영하는 중요한 정보로 활용될 수 있다. 향후에는 로그 데이터를 활용한 단계별 수행 분석과 부분 성취 반영 등 보다 정교한 평가 모델 도입이 요구된다.

셋째, 국제적으로 강조되고 있는 AI 관련 역량을 고려할 때, 디지털 리터러시 측정과 교육 체계 역시 AI의 이해와 비판적 활용, 인간–AI 협업, 윤리적 판단 역량을 점진적으로 포함하도록 확장될 필요가 있다. 본 연구에서 일부 생성형 AI 관련 문항을 시범적으로 포함한 것은 이러한 변화에 대한 초기적 대응으로 의미를 가지며, 향후에는 AI 리터러시 요소를 디지털 리터러시의 하위 요소로 체계화하고 학교급별 발달 수준에 맞게 단계적으로 반영하는 정책적 논의가 요구된다.

종합하면, 향후 우리나라의 디지털 리터러시 교육 정책은 교육과정, 교수·학습, 평가, 데이터 기반 분석을 유기적으로 연계하는 방향으로 발전할 필요가 있으며, 특히 국가 수준 검사 결과와 로그 데이터를 교육 정책과 학교 현장에 환류할 수 있는 구조를 마련하는 것이 중요하다. 이를 통해 디지털 리터러시 측정은 단순한 성취도 보고를 넘어, 학생 개개인의 학습 특성과 성장 경로를 지원하는 정책적 도구로 기능할 수 있을 것으로 기대된다.

요 약······························································································································ i
Ⅰ. 서론························································································································· 1
1. 연구 필요성 및 목적························································································1
가. 연구 필요성·································································································1
나. 연구 목적·····································································································3
2. 연구 내용 및 범위····························································································4
3. 연구 방법··········································································································· 5
가. 문헌 분석·····································································································5
나. 전문가 자문 및검토 ················································································5
다. 통계 분석·····································································································6
라. 로그 분석·····································································································7
Ⅱ. 디지털 리터러시 관련 선행 연구 분석···························································8
1. 디지털 리터러시 개념 및 정의·······································································8
2. 국내·외 디지털 리터러시 관련 연구······························································9
가. 디지털 리터러시 프레임워크 관련 연구 ··············································9
나. 디지털 리터러시 수준 측정관련 연구 ················································24
3. AI 리터러시 관련 연구··················································································44
가. AI 리터러시 개념 및 프레임워크···························································44
Ⅲ. 디지털 리터러시 검사 프레임워크 및 도구 개발·······································58
1. 2025년 디지털 리터러시 검사 프레임워크··················································58
가. 디지털 리터러시 검사 프레임워크 구성 요소······································58
나. 디지털 리터러시 성취 수행예시···························································59
Ⅵ. 디지털 리터러시 로그 데이터 분석 결과···················································152
1. 로그 데이터 개요 및 분석 방법·································································153
가. 로그 데이터 일반 특성분석개요·······················································153
2. 주요 분석 결과·····························································································154
가. 초등학교 분석 결과················································································154
나. 중학교 분석 결과····················································································161
Ⅶ. 결 론··················································································································· 168
1. 연구결과 논의·······························································································168
2. 제언················································································································ 172

✅ 연구책임자: 김한성(고려사이버대학교)
✅ 공동연구자: 이현숙(건국대학교), 유수진(한성대학교), 박주연(덕성여자대학교), 서정희(한국교육학술정보원), 임영수(코댑트)
✅ 연구보조원: 노혜림(건국대학교), 배경림(건국대학교), 남상국(건국대학교), 김보아(고려사이버대학교)
✅ 과제책임자: 최미애(한국교육학술정보원)