A Review OF John C. Hattie (2009) Visible Learning: A Synthesis of Over 800 Meta-Analyses Relating to Achievement.
By Donald C. Orlich
Hattie provides a very detailed synthesis of studies describing specific effects on student achievement using meta-analysis. Meta-analysis is a technique of combining several studies related to similar variables. Hattie summarizes this monumental task into sets of variables that he labels as “contributions.” The latter relate to: (1) students (2) homes (3) schools (4) teachers (5) curricula and (6) teaching approaches.
Hattie then converts the statistics of the hundreds of meta-analyses into “effect sizes.” An effect size (d) of 1.0 would show a one standard deviation gain on a normal curve or an increase of 34.13 percentiles. That effect would simultaneously be located at the 85th percentile of achievement if the starting point were at the 50th percentile.
Ninety percent of all 138 computed effect sizes were positive while 10 percent suggested a negative effect on student achievement. Hattie set a benchmark of d=0.40 which would indicate a 16 percent gain. This benchmark argues Hattie is a level at which “real world” differences could be observed in student achievement. The 138 traits and their effect sizes were then organized into “domains.” Of all the variables tested 66 met the d=0.40 benchmark while 72 fell below.
The most effective variables measuring a “d” between 1.44 to 0.80 were:
[ul]student self-reports of grades
Piagetian programs (growth models)
providing formative evaluation
microteaching
acceleration
classroom management.[/ul]
Among the many domains falling in the upper middle range between d=0.77 and 0.60 were:
[ul]teacher clarity
reciprocal teaching
feedback
teacher-student relationships
spaced vs. mass practice
meta-cognitive strategies
prior achievement
vocabulary programs
self-realization
professional development for educators
problem-solving teaching
not labeling students
phonics
teaching strategies.[/ul]
Of the domains that Hattie identified the following were in the [b]bottom[/b] 10 with “d” scores ranging from 0.05 to -0.34. (With negative effects on learning) Included in this embarrassing array were:
[ul]whole language reading
multi-grade-age classes
student control over learning
retention in grade
television
school mobility. [/ul]
Two conclusions may be inferred:
(1) Teacher quality is a key link to student achievement and (2) Most current educational reform efforts have simply been “fads.”