Evidence-based education

Evidence-based education is an approach to all aspects of education—from policy-making to classroom practice—where the methods used are based on significant and reliable evidence derived from experiments.

It shares with evidence-based medicine the aim: to apply the best available evidence, gained from the scientific method, to educational decision making. "Evidence-based teaching" refers to the teaching aspects.

Sources of evidence

Meta-studies of classroom-based experiments

As with the testing of new drugs, evidence-based teaching methods are derived from controlled trials. When several of these studies are compared, and their conclusions combined, we get a meta-study or meta-analysis. This is significantly more reliable than the results of individual studies due to the difficulty in controlling variables and individual bias.

Two sources of meta-analyses in education include: Visible Learning from a team in New Zealand under John Hattie and Classroom Instruction that Works from a Colorado, USA team under Robert Marzano.

According to the Marzano study, there are ten classroom methods which have been shown to work significantly better than many others:

• using analogies and similes
  • identifying similarities and differences;
• note making and summarising;
• developing a growth mindset;
• repetition and practice;
• graphical organisers and methods;
• cooperative learning;
• setting goals in advance
  • providing feedback (formative assessment);
• hypothesis testing;
• activating prior knowledge;
• advance organisers.

Although Hattie's work does not exactly mirror this list, the main reason is that the New Zealand study looks at everything related to education, including family effects and changes to the curriculum, while the Colorado study looked only at classroom methods. There are, however, no incompatibilities and most of Marzano's top-ten appear high on Hattie's list.

Hattie points out that there is no shortage of effective methods - almost anything you try in education seems to have a small beneficial effect. He therefore uses a scale of effect size which measures by how much the learning is improved. As an effect-size of 0.4 is the average for all interventions (and also the effect of a hard working, well organised and enthusiastic teacher), he suggests that methods with an effect size above 0.4 should be used as a priority. This ties with Marzano, whose list starts at an effect size of 0.59 for Advance Organisers and increases up the list.

What Works Clearinghouse

Another source for evidence based education includes "What Works Clearinghouse", that evaluates educational programs by evidence and effectiveness and is operated by the federal National Center for Education Evaluation and Regional Assistance (NCEE). The 10 interventions below are currently the interventions with both the highest evidence rating grade of "medium or large" and the highest effectiveness rating grade of "positive effects".

Program	Intervention Description	Targeted Students	Effects
Sound Partners	Phonetics based one-to-one tutoring four times a week designed to require minimal expertise from the tutor. The program consists of scripted lessons that use oral reading to practice phonics skills.	K1-3, especially high risk and low performing students	Improvement Index Effect size Alphabetics 21% Reading comprehension 21% Reading fluency 19%
Success For All	A school wide reading program to grades K2 with emphasis on early detection and prevention of reading problems. Key program elements include daily 90-minute reading classes where students are grouped by performance regardless of their age. Daily one-on-one tutoring is provided to students with learning difficulties. First year includes curriculum that focuses on language development and phonemic awareness. The second year emphasizes learning activities that happen in teams or pairs.	K2	Improvement Index Effect size Alphabetics 13% Reading comprehension 8% Reading achievement 10%
Positive action	Teaches children positive and constructive way of thinking about themselves and acting towards others using methods such as discussions, role-playing and games. The program uses factsheets, booklets and songs as teaching material.	K1-K12 education	Improvement Index Effect size Behavior 19% General academic achievement 14%
Coping power	The program consists of a child and a parent component. The child component consists of thirty-four 50 minute group sessions and periodic individual sessions over the course of 15–18 months. The parent component consists of 16 group sessions and periodic invidula meetings. The child component emphasizes goal setting, problem solving, anger management and peer relationships and consists, while the parent lessons emphasize setting expectations, praise, discipline, managing stress, communication and child study skills.	K0-K12 students with emotional disturbances	Improvement Index Effect size External Behavior 8% Social outcomes 6%
Too Good for Drugs and Violence	The program promotes prosocial behavior and norms, and consists of 14 core lessons with additional 12 lessons that include roleplaying and co-operative learning. Pupils are encouraged to apply the skills taught, for example by infusing the lessons into subjects such as English, science or social studies	K5, with additional K8 program	Improvement Index Effect size Knowledge, attitudes & values 16%
I CAN Learn Algebra	The program is mastery based and uses self-paced educational software.	K6-K8	Improvement Index Effect size Mathematics achievement 5%
Pre-K Mathematics	The program uses small group activities (4 - 6 children) with concrete manipulatives and includes take home picture strips and activities that are designed to help parents support their learning, as well as a software with activities to reinforce the lessons.	Kindergarten children	Improvement Index Effect size Mathematics achievement 19%
Literacy Express	includes lessons on oral language, emergent literacy, basic math, science, general knowledge, socioemotional development. It offers the staff with recommendations for room arrangement, daily schedules, classroom management and activities, and provides them with teaching materials.	Kindergarten students, especially those with special needs	Improvement Index Effect size Print knowledge 15% Oral language 12% Psychological processing 12%
Fast for word	The program uses a computer software designed to train cognitive skills that support reading and learning and to adapt the content to the individual students responses. The program consists of two language and literacy series that are used in 30–100 minutes lessons 5 days a week for 4–16 weeks.	K1-K3	Improvement Index Effect size Alphabetics 6% Reading comprehension 6% Reading fluency 7%
Accelerated Middle Schools	The intervention gives additional teaching and attempts to cover an additional year of curriculum during its 1 or 2-year duration. Classes link multiple subjects and are designed to have a 'hands on' practical approach.	High risk and low performing students in grades who are behind their grade levels	Improvement Index Effect size Progressing in school 35% Staying in school 18%

The Coalition For Evidence-Based Policy Congressional Top Tier Programs

The Coalition For Evidence-Based Policy, which is a non-partisan, non-profit organization advocating the use of well conducted randomized controlled trials in policy, has recognized three educational programs as "Top Tier". Programs classified as "Top Tier" must have been proven to be effective in randomized controlled trials that were well designed and implemented.

Program	Intervention Description	Sample size	Sample characteristics	Outcomes	Costs
Career Academies	The intervention happens in learning communities consisting of 150 - 200 students. It combines academic and technical curricula in partnership with local employeers, typically with a specific career theme.	1 764 students in 9 high schools	86% of the students in the study sample were Hispanic or African-American and the high schools were in high poverty urban areas.	11% increase in average annual earnings ($2,460 per year), sustained over the eight years after scheduled high school graduation. The effect was concentrated among men (who experienced a 17% earnings increase), and was not statistically significant for women.	Approximately $2,300 per student for a three-year Career Academy, or $3,000 per student for a four-year Career Academy.
Success For All	A school wide reading program to grades K2 with emphasis on early detection and prevention of reading problems. Key program elements include daily 90-mnute reading classes where students are grouped by performance, not age and in addition daily one-on-one tutoring to students with learning problems. First year includes curriculum that focuses on language development and phonemic awareness and second year learning activities that happen in teams or pairs.	41 schools with a total student population of 2,694	56% African-American, 10% Hispanic of whom 72% of students were eligible for federally subsidized lunches.	Schools that were allocated to the program increased their average reading achievement by 25-30% of a grade-level three years after random assignment.	Approximately $220,000 per school (K5) or $510 per student, over the full three-year period.*
H&R Block College Financial Aid Application Assistance	Offering personal assistance in completing a college financial aid application.	Approximately 1045 students	57% female, 55% white, 39% African American who had family income was below average (23,000 $) and average age was 17.7.	A sizable increase in college attendance and persistence over the 3½-4 years following the intervention, 29% greater likelihood of attending college for two consecutive years.	Approximately $90 per person to deliver the intervention, in 2012 dollars. $375 per person in federal need-based (“Pell”) grants for college.

Educational neuroscience

Educational neuroscience uses the methods and insights from the study of the brain to investigate educationally inspired questions.

The evidence derived from neuroscience now provides broad guidelines for various learning theories since, while the details have yet to be revealed, we now know some of the mechanisms underlying the learning process. The brain is now understood to perform a large number of separate processes, several of which need to work together for any given task (e.g. reading). Neuroplasticity refers to the brain's ability to adapt and change, and learning difficulties can reflect significantly weak development in one or more area so that learning is seriously impaired. A wide range of books are now available which attempt to translate the sometimes complex ideas of neuroscience into teacher-friendly language. David Sousa has a series headed by 'How the brain learns'. 'Learning and the Brain' is an excellent introduction to the subject as the authors have included only material relevant to teachers and done so in a jargon-free way using diagrams. Links to other relevant books can be found on the Evidence-Based Teachers Network website.

Myths and low effect-size methods

Myths

Neuroscience has identified a number of common beliefs (or neuromyths) which are not supported by evidence and include:

• the belief that students have Learning styles (commonly visual, auditory or kinaesthetic);
• that they may be left or right-brain dominant;
• that there are critical periods during school-years when certain learning needs to take place.

Other myths include

• the belief that students need water available at their desk to maintain hydration,
• that special diets or brain foods (rather than a balanced diet) can improve learning
• that Neuro-Linguistic Programming can help learning.
• that the start of the school day should be delayed since teenagers go to sleep and wake-up late.

Low effect size interventions

John Hattie shows that many of the interventions favoured by government in many countries have low effect-sizes, but often high cost:

• Setting or grouping by intelligence
• Retention – keeping a low achieving pupil down to retake the year
• Reducing class size (the effect is not great until the class size gets below 12)
• Charter schools and Academies – schools freed from local authority control, funded by government
• School finances
• New buildings
• Teacher subject knowledge (specialists do not get better results on average)
• E-learning, including interactive white boards (IWB), voting systems and computer suites
• Passive teaching assistants (who sit with the pupil and help them in lessons)

While all these things can show a positive effect, this is at an effect-size of around 0.2 - about the same improvement which can be achieved by partial use of some of the top-ten methods such as giving feedback, or using an advance organiser.

Effective professional development

For students' results to reflect these high effect-sizes, teachers need to develop the skills of their use. According to several studies, the time taken to do this lies somewhere between the learning of new facts and the development of a musical or sporting skill. While facts can be learned with a few repetitions, skills may need several hundred hours to develop. The evidence is that teachers start to become skilled with a particular method after about 10 repetitions with improvement plateauing after 6 months to 2 years of use. Continuing professional development (CPD) needs to reflect these findings. Teaching staff need the opportunity to learn about and then practice these skills. The role of CPD managers is to ensure that the time is available and the process takes place, not to instruct the teachers to follow directions. Where staff self-select their training either from external providers or from a range of sessions on a training day, they do not have the chance to develop their skills. Training, development and discussion of a smaller list of high-effect-size methods will be more effective. This process is sometimes referred to as supported experimentation or peer mentoring.

Implications for teachers

Teachers have more effect on the outcomes for their students than anyone else. The difference in outcomes for 2 teachers in the same college is significantly greater than the average of teachers in a 'good' rather than a 'weak' school. The main reason why some schools do better is that they have a higher percentage of teachers who use high effect-size methods. While individual teachers can improve their students' results using these methods in isolation, it is far more effective if they are adopted department or college-wide so that the discussions, observations and sharing-of-practice can take place easily.

External links

• The Evidence Based Teachers Network (EBTN)
• Institute for Effective Education (IEE)
• Best Evidence Encyclopedia