TBLT implementation and evaluation: A meta-analysis

52 studies

Language programmes implementing task-based language teaching across institution types and world regions

Synthesises PROGRAMME IMPLEMENTATION studies rather than controlled comparisons of task-based against another method, which is why it is graded D despite its size. Many included studies are programme evaluations conducted by people implementing the programme; the pooled d = 0.93 is largely a pre-post and programme-satisfaction literature, and "positive stakeholder perceptions" is reported as a finding alongside the effect size. Recorded so the strongest-sounding number in favour of communicative, task-based teaching is on file WITH its design.

Overall positive and strong effect (d = 0.93) for task-based language teaching implementation across a variety of learning outcomes, moderated by programme region, institution type, whether a needs analysis was conducted, and cycles of implementation. Positive stakeholder perceptions across both quantitative and qualitative data.

Not the main threat. The main threat is design: programme evaluations without adequate counterfactuals, conducted by implementers.

No independent replication with adequate counterfactuals. This archive has a standing finding that practitioner satisfaction carries no information about effect - it was over 90% in a randomised grammar trial that returned -0.02 - so the stakeholder-perception result should not be read as support.