Evidence for Simulation: practising the real thing safely #

Every substantive claim on the Simulation: practising the real thing safely page is checked against current research. Here is each claim, how well today’s evidence supports it, and the sources. The full, de-duplicated source list lives on the references page.

Supported · moderate evidence — Simulator-based practice transfers positively to real-task performance, as shown most clearly for flight training where pooled studies find a reliable transfer-of-training benefit from simulators to actual flying.

Hays, Jacobs, Prince & Salas’s (1992) meta-analysis of flight-simulator training found that, on average, simulator practice combined with aircraft training produced better performance than aircraft training alone, i.e. positive transfer. The general claim that simulation transfers is well supported; the page is careful not to overstate magnitude, and effect sizes in the original were heterogeneous across studies and aircraft types.

Sources: Hays, R. T., Jacobs, J. W., Prince, C., & Salas, E. (1992), Flight simulator training effectiveness: A meta-analysis. Military Psychology, 4(2), 63-74 — https://doi.org/10.1207/s15327876mp0402_1 · full reference ›

Supported · moderate evidence — Adding physical realism to a simulator (e.g. full-motion platforms, ever-richer visuals) tends to add greatly to cost while adding relatively little to the learning of the underlying skills.

The flight-training literature, anchored by Hays et al. (1992) and consistent with later reviews, repeatedly finds that higher physical fidelity (notably platform motion) yields small and inconsistent additional training benefit for many tasks despite large cost increases; the cue/procedure content drives most of the transfer. The direction is well supported, though the precise added value of motion is task-dependent and still debated at the margins.

Sources: Hays, R. T., Jacobs, J. W., Prince, C., & Salas, E. (1992), Flight simulator training effectiveness: A meta-analysis — https://doi.org/10.1207/s15327876mp0402_1 · full reference ›

Supported · moderate evidence — A simpler, cheaper trainer that captures the right task elements can teach the core skill about as well as a far more expensive high-fidelity simulator.

Because most measured transfer in Hays et al. (1992) and related work comes from rehearsing procedures and decisions rather than from physical realism, lower-fidelity devices that reproduce the relevant cues often match high-fidelity ones on core-skill learning. Well supported as a general principle; whether a given low-fidelity rig suffices depends on whether it reproduces the skill-critical cues for that specific task.

Sources: Hays, R. T., Jacobs, J. W., Prince, C., & Salas, E. (1992), Flight simulator training effectiveness: A meta-analysis — https://doi.org/10.1207/s15327876mp0402_1 · full reference ›

Supported · moderate evidence — What drives transfer from practice is functional fidelity — reproducing the cues, decisions, timing and demands of the real task — rather than physical (surface) fidelity.

The physical-versus-functional (psychological) fidelity distinction and the claim that the functional cue/decision content carries transfer is the consensus framing in the human-factors and motor-learning literatures; Schmidt & Bjork (1992) underpin the principle that practice should engage the processes the real task requires. The conceptual point is well established; exact contributions vary by domain.

Sources: Schmidt, R. A., & Bjork, R. A. (1992), New conceptualizations of practice: Common principles in three paradigms suggest new concepts for training. Psychological Science, 3(4), 207-217 — https://doi.org/10.1111/j.1467-9280.1992.tb00029.x · full reference ›

Supported · moderate evidence — Practising under varied, representative conditions — rather than a single idealised version — makes a skill hold up better when the real situation differs from the one trained on.

Variability of practice is a core ‘desirable difficulty’: Schmidt & Bjork (1992) show that variable and interleaved practice can depress performance during training yet improve retention and transfer. Robustly supported in motor and verbal learning, with the caveat that benefits depend on learner expertise and task, and can be smaller or reversed for novices on very complex tasks.

Sources: Schmidt, R. A., & Bjork, R. A. (1992), New conceptualizations of practice — https://doi.org/10.1111/j.1467-9280.1992.tb00029.x · full reference ›

Supported · moderate evidence — A simulation that lets the learner succeed using a cue or shortcut not present in the real task can foster reliance on that crutch and produce negative transfer, so practice conditions must be chosen with care.

Schmidt & Bjork (1992) caution that practice conditions which inflate immediate performance (e.g. by supplying guidance or cues the test will lack) can impair later transfer — the basis for the negative-transfer/over-reliance warning. The mechanism is well supported; the page states it qualitatively rather than claiming a specific effect size.

Sources: Schmidt, R. A., & Bjork, R. A. (1992), New conceptualizations of practice — https://doi.org/10.1111/j.1467-9280.1992.tb00029.x · full reference ›