The Effectiveness of Online and Blended Learning from Schools: A Systematic Review

This systematic analysis examines the research evidence on Online and Blended Learning from schools (rather than in schools). It is particularly relevant at a time when schools have been forced by the covid-19 pandemic to implement some of these, as pupils have been unable to attend school. In addition, the review investigates Educational Games, Computer Supported Cooperative Learning (CSCL) and Computer Assisted Instruction (CAI), which have been largely used in schools but have the potential to be used outside school. Web-based learning in school was included in CAI.

Eight different research databases were searched. Studies not relating to schools, dated before 2000, not in English, without data and duplicates were removed. Remaining were 1355 studies from all over the world: Online studies 134 (7%), Blended 232 (13%), CSCL 129 (7%), Games 488 (26%), and CAI 865 (47%).

Overall, digital technology was found more effective than Traditional Instruction in 85% of studies, while 8% found it the same and only 3% found it worse.

Blended Learning was considerably more effective than Online Learning (83% of studies more effective than Traditional Instruction or other comparator conditions compared to 74%). Of the other conditions, (CAI) was the most effective (91%), with Games and CSCL (both 81%) coming after Blended Learning, but most of these had taken place in school.

Previous reviews of the research evidence numbered 144: Online 19, Blended 20, Games 56, CSCL 14 and CAI 35. Unfortunately, 81% of these covered primary, secondary and higher education together and did not discriminate between sectors. Nonetheless, such indicators of magnitude of effect as were available suggested CAI was the most effective, followed by Blended Learning, Games and CSCL, followed by Online Learning – similar to the findings for all separate studies.

Regarding school sector, Primary and Early Years/Kindergarten both were 87% Better than Traditional Instruction, while Secondary/High came next with 80%, Primary/Secondary next and last was Middle Schools with 73%.

Regarding Subjects, Science and Maths were the most popular, then Thinking, Reading, English as a Foreign Language (EFL), Humanities, Health, STEM, Writing, English, Arts/Music and Foreign Languages in that order.

However, EFL interventions were the most effective, then Writing and STEM, Thinking, Arts/Music, Humanities, Health and Science, Reading and Math, Foreign Languages and English. Thus, the most popular subjects were by no means those where the greatest effectiveness was evident.
Regarding gender, Females were Better in 27 cases (39%), Males and Females were equal in 36 cases (52%), and Males Better in only 6 (9%). Thus, overall, females did better.

Four categories of studies showed high effectiveness (higher than the average for all papers in all studies): HiLo studies (mostly “low ability” children) (total n=76, 92% Better); Socio-Emotional Functioning studies (mostly focused on self-efficacy) (225, 91%); Second Language Learner studies (mainly English as a Foreign Language learners) (69, 90%); Non-English-Speaking Educational Context studies (in a country which did not have English as its first language) (329, 89%).

Two categories showed equal effectiveness (equal to the average for all papers in all studies): Hospitalised or Sick at Home studies (7, 86%) and Rural studies (29, 85%).

Five categories showed lower effectiveness (lower than the average for all papers in all studies). The first was Special Educational Needs (SEN) or Disability (n=89, 80%) (although this was not much lower than the average for all students). Within this, Deaf and Hard of Hearing (100%), Down Syndrome (100%), Writing Difficulty (100%), Attention Deficit and Hyperactivity Disorder (90%), Learning Disability (87%) and Autistic Spectrum Disorder (86%) all scored higher than the average for all students. The others were: Gifted studies (n=10, 80%); Psychological Well-Being (17, 79%); Disadvantaged (46, 79%) (not much lower than the average for all students) and Ethnic Minorities (32, 78%) (not much lower than the average for all students). However, all these categories were effective to different degrees.

Thus, Blended Learning was more effective than Online Learning, although both were effective but to different degrees, and Online Learning may be the only option for students remote from school and those unable to attend school owing to pandemics or other causes. CAI was more effective than either, but much of CAI was in school, and it cannot be assumed that it would be as effective if delivered to homes. The effectiveness of Games and CSCL fell somewhere between Blended and Online Learning, but again much of this was in school, and it cannot be assumed that it would be as effective if delivered elsewhere. The limitations and strengths of the research were discussed and conclusions offered in relation to the research questions. The implications for practitioners, policy-makers and future researchers were outlined.