What the research says: twelve sources on cognition in sport

Twelve studies from the last fifteen years explain why Insporta is built as a sport-specific knowledge test and what such a test can honestly claim. They come from the reference list of Reinhard, Mann and Höner (2025), whose own finding, 34% of playing-level variance explained by a sport-specific test against 13% for a generic composite and 0% for a reactive stress test, is the headline. The twelve sources below supply the theory, the prognostic evidence, the critique and the methodological standard behind that headline.

The map

The literature answers four questions in sequence, and the articles in this section follow the same order.

1. Which kind of cognition separates athletes? Sport-specific decision-making, not generic functions. Kalén et al. (2021).
2. Does it predict the future, not just the present? Moderately, yes. Murr (2018), Ivarsson (2020), Höner (2021, 2023).
3. What does good decision-making look like, and when does it stabilise? A cyclical process that favours quality over speed, stabilising around age 11. Janssen (2023), Heisler (2023), Travassos (2013).
4. What should a test not claim? Executive functions, intelligence, or anything it has not shown it measures. Vestberg (2017), Verburgh (2014), Beavan (2020), Furley (2023).
5. What does a valid study look like? Bergkamp (2019), Williams, Ford and Drust (2020).

Summary table

Source	Main contribution	What it means for a knowledge test
Kalén et al. 2021, meta-analysis, Psychological Bulletin	Splits cognition into three levels; sport-specific decision-making skills are the best discriminator of athlete level	Confirms sport-specific content and topic-level decomposition as the right design
Murr et al. 2018, systematic review, PLoS One	Decision-making predicts future level with d = 0.81; psychomotor measures are strongest overall	Gives a quantitative anchor for what a sport-specific cognitive measure can predict
Ivarsson et al. 2020, meta-analysis, JSAMS	Psychological factors have small but statistically significant prognostic value; perceptual-cognitive factors are among the strongest	Sets realistic expectations: effects are real, not large
Williams, Ford and Drust 2020, two-decade review, Journal of Sports Sciences	Talent identification must be multidisciplinary and longitudinal, not one-off	A test gains value as it accumulates repeated measurements over time
Höner et al. 2021, N = 13,869, Frontiers	Largest prospective talent study in Europe; 9% success rate; multidimensional models beat any single predictor	No single score should stand alone; coach judgment adds information
Höner et al. 2023, 360-degree video, Frontiers	Split-half reliability r = .78 for the decision-making test; prognostic validity for adult level	The reliability standard a sport-specific test should aim for
Heisler et al. 2023, N = 128, mean age 10.69, Psychology of Sport and Exercise	Better players prioritise decision quality over speed; a developmental threshold near age 11	Speed alone is not a measure of expertise; age matters more than difficulty alone
Janssen, Müller and Mann 2023, KNVB framework, Sports Medicine	Decision-making is cyclical; skilled players use a window of opportunity; validity rises with perception-action coupling	Points toward anticipation-style items and more representative response formats
Vestberg et al. 2017, N = 21 + 24, PLoS One	Core executive functions associate with success in young elite players, but age moderates the effect	Any comparison between players must control for age
Verburgh et al. 2014, N = 126, PLoS One	Inhibition and alerting classify talented versus amateur players with 89% accuracy	High classification accuracy is possible, but the base rate must be reported with it
Beavan et al. 2020, N = 343, longitudinal, Psychology of Sport and Exercise	Executive-function trajectories of elite players match the general population	Knowledge, which is trainable, is a different kind of construct from executive function
Furley, Schütz and Wood 2023, critical review, IRSEP	Task impurity and the jingle-jangle fallacy undermine much cognitive talent research	A test must name its construct precisely; Insporta claims knowledge, not intelligence
Bergkamp et al. 2019, position paper, Sports Medicine	Four methodological problems: criterion operationalisation, isolated indicators, restriction of range, base rate	Any validation study of a knowledge test must address all four

Travassos et al. (2013), a meta-analysis of 111 effect sizes on 882 participants, is discussed alongside the decision-making sources: the more the required response resembles the real action, the larger the expertise effect.

Where the sources disagree

The twelve are not a chorus. Two lines of disagreement matter.

Executive functions. Verburgh et al. (2014) and Vestberg et al. (2017) are cross-sectional studies that found executive functions differ between talented and amateur young players. Beavan et al. (2020), a longitudinal study of a professional club’s players over three seasons, found that those functions develop identically in elite players and the general population, which suggests the cross-sectional differences reflect who was selected, not what training produced. Furley, Schütz and Wood (2023) add that most executive-function tasks are impure and inconsistently named. The weight of the longitudinal and critical evidence is why Insporta does not measure executive functions.

How much cognition predicts. Murr et al. (2018) report d = 0.81 for decision-making, a large effect. Ivarsson et al. (2020), pooling more studies, describe psychological factors as small but significant. The difference lies partly in what was pooled and how success was defined, which is exactly the criterion problem Bergkamp et al. (2019) describe. See does sport knowledge predict future level.

What the sources do not say

None of the twelve studied a written multiple-choice knowledge test. The decision-making tests in Höner (2023) and Reinhard (2025) use 360-degree video; the executive-function studies use laboratory tasks. Insporta shares the sport-specific principle and the topic-level decomposition with the decision-making tradition, but its own reliability and validity figures have to be established on its own data, against the standard described by Bergkamp and by Williams, Ford and Drust. The research tells Insporta what to build and how to check it; it does not certify the result in advance.

Reading on

• Sport-specific knowledge, not generic brain training: the headline study in detail.
• Three levels of cognition: the Kalén hierarchy.
• Decision-making frameworks and the age of eleven: Janssen, Heisler, Travassos.
• How Insporta applies the research: the design choices that follow.
• References: the full list.