Estimating the replicability of technology education research

Technology education research is a growing field, with the rate of growth increasing over the last 2 decades. As the field grows, it is paramount that credibility is maintained in published findings. To date there is no evidence to suggest a lack trust is warranted, however in the midst of the replication crisis there is need to ensure continued rigour. This article presents a z-curve analysis of the replicability of quantitative research in technology education since 1983 using statcheck for automated data extraction. The results indicate that authors often mis-report p-values, typically due to rounding errors, with a small percentage (1.59%) of inconsistently reported p-values leading to decision errors in terms of statistical inference. With respect to replicability, overall it is estimated that 55.7% of reported quantitative results in technology education would replicate, however since 2020 this estimate appears to be increasing. These results do not indicate specific findings which are likely or unlikely to replicate, but do suggest a need to invest effort in identifying studies which would have a high value in being replicated, particularly in the timeframe of work published from 2010 to 2020.