Is Ability Grouping or Streaming Effective?

We have been approached by a number of teachers over the past few months to discuss the value of streaming students in ability groupings as a way of improving performance. There is also a big push amongst education unions to lower class sizes as a way of again improving performance of students. The logic behind such requests is that, given the increasing variance of competency that is occurring in classes (can be up to 7 years variance between students) , then reducing class sizes or streaming students in to approximate ability groups would enable a teacher to better be able to provide the point of need teaching for the students.

Rather than entering into what can be quite an emotive debate we thought we would extract what John Hattie uncovered in his synthesis of meta-analyses book Visible Learning. We have created a brief summary of the appropriate meta-analyses followed by the conclusion Hattie drew from studying the research. For more detail see his book. The effect size one would look for to decide whether it is worth pursuing is d > 0.4)

1.  Ability Grouping [14 Meta-analyses, 500 studies, Effect -Low (d = 0.12)]

• Fundamentally about whether classes are heterogeneous or homogeneous in ability of achievement
• Tracking
  • At the Upper School level about undertaking different courses
  • At the Middle School level students normally tracked in some subjects (normally English and Maths) and not in others
  • At Earlier Levels typically students take the same subjects as each other but the orientation or pacing of the instruction is intended to match the differing ability levels of the students
• Essence of the research shows that tracking has minimal effects on learning outcomes and profound negative equity effects
• However, there was also qualitative evidence that low track classes (i.e for lower competency students) were more fragmented, less engaging, and taught by fewer well-trained teachers. This points to the quality of teaching and the nature of the student interactions as the key issues not the compositional structure of the classes.
• Hattie points out that there might be some benefits if the lower tracked classes were taught by well-trained teachers and were more stimulating and challenging.

2.       Multi-Grade / Multi-Age Classes [3 Meta-Analyses, 94 studies, Effect -Low (d = 0.04)]

• Multi-age classes include students from more than one year level who are taught in the same classroom by the same teacher. This is common in small schools, developing nations, and also by some schools to allow for “more flexible grouping and learning styles, and having students work cooperatively and collaboratively”
• Research essentially shows that no significant difference to single-grade results in achievement or affective outcomes
• Part of issue is that teacher rarely capitalise on the multi-grade or multi-age arrangement to promote learning to from peers. Instead they tend to teach distinctly different curricula, maintain grade levels, and deliver separate lessons to each grade-level group

3.       Within-Class Grouping [2 Meta-Analyses, 129 studies, Effect – Low (d = 0.16)]

• Defined as “a teacher’s practice of forming groups of students of similar ability within an individual class”
• Research showed for high ability students (d ~ 0.29) compared to the remaining students (d ~ 0.16)
• It is more beneficial for large class sizes (> 35 students gives d = 0.35) than smaller classes (< 26 students gives d = 0.06 – 0.22)
• The beneficial effects are more associated with small group learning and instruction

4.       Small-Group Learning [2 Meta-Analyses, 78 studies, Effect – Medium (d = 0.49)]

• This differs from within-class grouping in that it typically involves assigning a task to a small group and then expecting them to complete this task (only research done is at tertiary level)
• Small group learning had significantly more positive effects than individual learning when
  • Students had group work experience or instruction
  • Cooperative learning strategies were employed
  • Group size was small
  • Small group learning led to greater self-esteem among undergraduate students

John Hattie’s Summary of the effects of grouping (page 95)

• “… that instructional materials and the nature of instruction must be adapted to these specific groups”
• “Simply placing students in small or homogeneous groups is not enough”
• “For grouping to be maximally effective materials and teaching must be varied and made appropriately challenging to accommodate the needs of students at their differing levels of ability”

Final Note

One note that we must add in to this if your school is going down the path of grouping students in some fashion of form. This is a school structural issue that is not addressed in the research. One of the barriers we have found to make such groupings work is the timetables. If you are going to provide the appropriate teacher resources and skills to this model then schools really need to address their fundamental timetabling structures and how they use teacher resources within a school. This is a particularly thorny issue in secondary schools as teachers are quite often teaching across many year levels and timetables are scheduled to suit this need rather than student learning needs. Most secondary schools would need to explore how they are going to manage their staff structures and budget.

An example of a timetable we saw in a New Zealand school (Years 5 – 10) allowed for groupings (and thus allowed for the placing of teacher experience and resources at the points of need) is shown below. Notice that the day is broken into 3 blocks and the actual time assigned to particular KLA would be up to the team of teachers.