Evidence library

The Learning Transfer Constellation

Fifty-one factors that decide whether training changes behaviour on the job. Ten sit inside the training room. Forty-one sit around it.

How to read this. Every factor here is drawn from published transfer research and carries its source. Where a meta-analytic effect size exists it is shown. Blume and colleagues found that when a predictor and transfer are both self-reported by the trainee at once, correlations inflate sharply: for the work environment, from .23 to .54. Where they report that split, the figure shown is the one with the bias removed. Where they do not, it is the pooled estimate, and the honest number is probably lower.

Factors are not independent, and several are subscales of the same instrument or study. Read this as a map of what has been studied, not as fifty-one separate levers.
Drag to rotate. Click any node.
Inside the training room 10Around the training 41Node size = measured effect. Open nodes have no meta-analytic estimate.
Browse all 51 factors as a list

Pick a factor

Click any node on the sphere, or search above. Each factor opens with a definition, what the evidence actually says, its sources, and the other factors it connects to.

The sphere is a map, not a ranking. Blume and colleagues concluded there are no magic bullets for transfer, and that professionals should combine multiple strategies rather than hunt for one.

Principal sources

Baldwin & Ford (1988), Personnel Psychology, 41 · Holton, Bates & Ruona (2000), HRDQ, 11(4) · Burke & Hutchins (2007), HRD Review, 6 · Blume, Ford, Baldwin & Huang (2010), Journal of Management, 36(4) · Grossman & Salas (2011), IJTD, 15(2) · Choi & Ruona (2008), An Update on the Learning Transfer System, AHRD, plus the primary studies cited on individual factors.

Which of these are you already getting right?

Seventeen questions, scored against the factors on this map. You get a Transfer Readiness Score and a plain-English read on what is at risk.