We are often bombarded with research claims in education. The reality is that many of these claims are shaky, unsubstantiated, or even false. However, some researchers have attempted to make some sense of the often contradictory evidence. See how your intuition or experience align with the research base below.
Below is a list of 30 influences on student learning. Rate your estimate of their impacts based on the research base. (Answers at the end of the post).
Influence Impact Impact
Ability grouping/tracking/streaming High Medium Low
Acceleration (for example skipping a year) High Medium Low
Comprehension programs High Medium Low
Concept mapping High Medium Low
Cooperative vs. individualistic learning High Medium Low
Direct instruction High Medium Low
Feedback High Medium Low
Gender (male vs female achievement) High Medium Low
Home environment High Medium Low
Individualizing instruction High Medium Low
Influence of peers High Medium Low
Matching teaching with student learning style High Medium Low
Meta-cognitive strategy programs High Medium Low
Phonics instruction High Medium Low
Professional development on student achievement High Medium Low
Providing formative evaluation to teachers High Medium Low
Providing worked examples High Medium Low
Reciprocal teaching High Medium Low
Reducing class size High Medium Low
Retention (holding back a year) High Medium Low
Student control over learning High Medium Low
Student expectations High Medium Low
Teacher credibility in eyes of student High Medium Low
Teacher expectations High Medium Low
Teacher subject matter knowledge High Medium Low
Teacher-student relationships High Medium Low
Using simulations and gaming High Medium Low
Vocabulary programs High Medium Low
Whole language programs High Medium Low
Within-class grouping High Medium Low
I have been reading a, recently published, book by John Hattie, Visible Learning for Teachers, Maximizing Impact on Learning (2012). This is a followup to Hattie’s Visible Learning, the largest synthesis and collection of evidence-based research on what actually works in schools to improve learning. Visible Learning for Teachers was written to bring this synthesis to teachers. It has proven to be more interesting and nuanced than Marzano’s Classroom Instruction that Works, maybe because it doesn’t feel like Hattie is selling a product or system of products. Editorials aside, I would strongly recommend this to teachers wishing to investigate what the research says on teaching.
Much of his scholarship deals with the synthesis of meta-analyses of research. This work involved more than 800 meta-analyses of 50,000 research studies involving 240 million students. Out of these Hattie and others have settled on studies that show an effect size of 0.4 as being worthy of replication (an effect size is described below for those wanting to get all statistical). An effect size of 0.4 is considered significant as it is the mean effect size of all 50,000 studies and also turns out to be the average effect of a year of schooling on a student’s learning. In fact students in younger grades show an effect size slightly above this and students in higher grades slightly below this for a year of schooling.
The results for the influences above are based on Hattie’s research, influences that were classified as low impacts had effects sizes below 0.3, medium between 0.3-0.6, and high above 0.6. See how your intuition and the research literature align here.
What is an effect size?
An effect size quantifies the difference between two groups. It allows researchers and consumers of research to move beyond asking questions such as “does it work” to questions such as “how well does it work”? Thus a large effect size indicates not only that the effect is significant, but the importance with which we should view the treatment. This measure takes into account the amount of variability within the two compared groups. The more variability, the greater the difference between the averages for both groups must be for the effect size to be large.
A significant medium effect size is considered to be between 0.3-0.6. One important aspect of using effect sizes is that it allows researchers like Hattie to compare different studies, using different measures, to one another in a fair manner.
Looking for more go here.