Project-based learning can divide opinion.
Supporters claim that it provides a real-world education and supports so-called non-cognitive skills such as curiosity, time management and motivation.
Detractors argue that such methods do little to teach individuals what they require to pass exams and move up the academic chain.
Quick read: What is project-based learning?
Quick listen: How to spot education myths and read research
Want to know more? Why colleges are embracing neurodiversity tests
Most teachers have anecdotes about when independent learning went well or ended in unmitigated disaster. Like many others, my experiences have been mixed. Coursework was common when I began teaching A-level psychology.
Students would choose a topic of interest, generate a hypothesis, recruit participants, run their experiment, analyse data using statistical tests and present their work as a report that adhered to the requirements of peer-reviewed journals.
This was a period of frantic activity that took place after AS exams and ran until the summer holidays.
It gave students the opportunity to experience being researchers by testing the capacity of short-term memory, the limitations of working memory, the impact of music on learning and even (in one case) geographical memory in guinea pigs.
Most students would keep things pretty simple while others would be overambitious and in need of encouragement to scale back a little.
Some students gained a great deal from the coursework, while others saw it as a laborious chore. Some were motivated enough to consider advancing their psychology studies at university, and a smaller number eventually gained PhDs.
To what extent coursework encouraged this, I’ll never really know because the end result – the grade – was held in higher regard than the process involved.
The latest data
Two recent studies shed light on the effectiveness of project-based learning from different perspectives. The first is a new meta-analysis of 30 research papers published between 1998 and 2017, totalling 12,585 students in 189 schools across nine countries.
The results offer good news for advocates of project-based learning, with a medium-to-large positive effect on students’ academic achievement compared with traditional teacher-led instruction.
These results were seen across all stages of education and size of groups. However, the positive effect was greater in the social sciences than in the natural sciences and mathematics.
It's worth noting, though, that the effectiveness of project-based learning was mediated by the number of hours of instruction, so the findings by no means negate the usefulness of direct teaching methods in preparation for the more hands-on process.
As the authors point out, effect sizes are higher than previous studies, perhaps reflecting how project-based methods have become more effective over the past couple of decades.
Meta-analysis can be a powerful tool when trying to draw the threads of multiple studies together, yet it relies heavily on the validity and reliability of the chosen studies.
The authors of this paper were careful to select studies that employed experimental methods and appropriate data for analyses, yet any deeper methodological problems would obviously be carried over to the meta-analysis so, as with all research, a degree of caution is advisable.
A study conducted by researchers at the University of York took a different approach. Lynda Dunlop and her co-researchers investigated students' experiences of conducting independent research projects using a series of group interviews.
They were particularly interested in what is known as threshold concepts – the ability to differentiate between core learning outcomes that represent seeing things in a new, transformed way, and those that do not.
The overriding intention was to see how research projects allowed students to think about what it would be like to be a researcher, such as developing research questions and design, choices concerning research methods, producing conclusions that deal with concepts rather than facts alone, and making an argument that is supported by the evidence.
In total, 39 students aged 16-19 from two institutions took part in the interviews. Both institutions already had a strong culture of independent research project work that involved collaboration with scientists, universities, industry and non-governmental organisations, so the approach they employed was well established.
Semi-structured group interviews produced interesting responses, with the vast majority of students feeling that they had benefited from the experience both in terms of greater understanding and in potential future aspirations.
Being involved in the process allowed them to discover answers that, as one student put it, “aren’t found at the back of the book”.
Others said that they were motivated to learn new skills, such as learning to code in Python so that they could analyse their data, and had been reading more widely than was required in their A-level studies.
Both studies do appear to support the use of project-based learning methods and may even win over a few sceptics. But what really matters is how schools and students approach such method in the real world.
The second study used institutions with a strong track record or research projects and was already working with external partners with an enhanced understanding of what it is to be a researcher.
This may not be the case in circumstances where such partnerships don't exist or are less established. Such partnerships may therefore provide the answer to effective project-based learning approaches, certainly for older students.