top of page

A Research Summary of Multi Age Learning

Seann Dikkers & Mandy McRaith

​

Overview

 

Multi-age learning is less prominent than age segregated learning and gets less attention overall in the research community. Multi-age learning has a unique set of benefits and challenges as would be expected for an essentially different model of learning. We will review some of the research and findings around the topic as a summary for the purposes of considering a multi-age format for Portals Learning. 

 

Multi-age learning harkens back to one room school houses as the original Horace Mann “one-room” school model for public education in the United States (Aina, 2001) in the early 1900’s, however as large cities began to industrialize, the overwhelming numbers demanded more cost effective division of labor. Age segregated models were developed to decrease expertise needed in teachers, and increase uniformity in learning for a nation transitioning to a factory workforce (Gatto, 2010). Multi-age learning is still commonly used in lower population environments around the world (Ritland & Eighmy, 2013; Mulcahy, 1992), and is increasingly considered a pedagogical choice under a number of titles (Joan, 1997). 

 

Benefits

The primary benefits of Multi-age education are social development and academic achievement. It is worth mentioning that economic benefits drive much of multi-age learning worldwide (Mason and Dopener, 1998; Thomas, 2012). Individual academic needs can be better attended to in a setting where the teacher can monitor learning for the child over time and develop personalized learning strategies naturally through that relationship. Specifically, since the 1960’s, research on multi-age learning in the USA has been consistent and established. Meta analysis of research (Pratt, 1986) summarizes minimal impact on academic achievement but noticeable increases in student confidence, social networks, and well being. Miller, (1991) reviewed studies again showing comparable academic outcomes and gains in non-academic areas. Some studies, applying different learning theory, show multi-age as the “optimal” setting for learning in Vygatsky’s Zone of Proximal Development (Sivakumar, A., & Thirumoorthy, G., 2018) where students learning at their challenge level see increases in attention, motivation, and retention. 

 

Social learning is a clear and consistent benefit to learners in multi-age learning. So much so that study in the area moved on in the 1990’s, after a series of studies showed fairly conclusive data (Bank, 1997; Psacharopoulos, Rojas & Velez, 1993; Miller 1991). Learning among a community of learners amplifies growth, peer tutoring, and independence (Thomas & Shaw, 1992). Students generally feel more connected, helpful to others, better about themselves, and less negative competition. Knowledge acquisition commonly happens between students where older/advanced students feel comfortable helping learners which reinforces learning through repetition and builds efficacy in the mentors. Conclusively students in multi-age settings have stronger well being, confidence, learning habits, cooperation, pro-social skills, and independent learning skills. All of which have secondary bodies of research showing lasting impact on learning. Some smaller research (i.e. Hoffman 2003) has introduced a line of inquiry around improved student interest and engagement in multi-age settings. While it is reasonable to believe that these would also improve, the research is minimal and the sample size is small. 

 

Where these studies show at worst similar academic achievement, some studies show evidence that multi-age learning and a social construction of knowledge have an almost indisputable positive impact on social learning and positive long term benefits in learning habits and dispositions. However, we were not able to find any studies that followed learners into adulthood as a longitudinal look at impact on lifetime learning habits. 

 

Challenges

Challenges include areas that we have already considered. Notably, parent apprehension has been documented and is largely due to a lack of awareness (Aina, 2001). There are studies that show a perception of lower achievement, not to be confused with actual achievement levels (Quail, A., Smyth, E., 2014). The model is questioned using the concern that multi-age learning may hurt standardized test scores (note these are perception studies, not performance). Perception remains a challenge. Multi-age advocates still have to spend time and effort educating the community and parents of the benefits. 

 

Teacher training is another area of need, as most teachers are trained for age segregated settings. The range of skills and fluidity of practice takes additional training or time to develop as a multi-age teacher (Sag, 2009; Thomas and Shaw, 1992) and the university instruction is simply “inadequate” preparation (Izci, Duran & Tasar, 2010). Taking on our own teacher preparation programming will complement any teacher training they may have already received. As a note, most homeschooling parent-teachers have no teacher training and achieve better learning outcomes on average than government schools. 


 

Aina also identified a need for teacher prep time, materials, and control over the schedule for flexible learning activities. The predominance of learning materials are developed for age segregated settings. Research on multi-age learning universally notes the challenges of curriculum (Cornish, 2009) and testing (Hargreaves, 2001) that are not designed for multi-age learning. As research continues to show differentiated learning models as effective, more and more curriculum is being developed to support it, but not in all subject areas. Teaching in multi-age settings is challenging in itself, and takes support, professional development, and administrative mentoring (Mason & Burns, 1995), and increased prep time (Pratt & Treacy, 1996). Any preparation work that can be done for the teacher and packaged will help to offset this time. If not, we see some research indicating high burnout rates (Richards et al., 2016).

bottom of page