Genuine association or spurious responses? An experimental investigation of the association between cryptocurrency trading and Problem Gambling Severity Index scores
Abstract
Background and aims: Cryptocurrency trading has become a popular activity worldwide, despite being largely unregulated and posing high risks. Previous studies have correlated cryptocurrency trading with harmful gambling as measured by the Problem Gambling Severity Index (PGSI), but no experimental research has explored whether these represent genuine associations or are spuriously inflated by methodological weaknesses. Methods: In a large-scale pre-registered 2x2 online experiment (N = 4,781), we hypothesised that if observed correlations were driven by spurious responding, then they would be significantly weakened by screening out non-gamblers (compared to the unscreened control), and by explicitly instructing participants to not consider cryptocurrency trading as a form of gambling (compared to a no-instruction control). Cryptocurrency trading engagement was assessed via self-reported trading frequency, price monitoring frequency, and the number of coins owned over the past 12 months. Results: Results While screening significantly lowered reported PGSI scores, the association between trading frequency and the PGSI remained robust and statistically significant across all conditions, and was not moderated by screening or providing explicit instructions. Exploratory analyses found the same patterns with price monitoring frequency and the number of coins owned, and in a combined model with all predictors, trading frequency explained the most variance in the PGSI outcomes. Conclusions: Together, these findings provide strong support for a genuine association between cryptocurrency trading and harmful gambling. Financial and gambling regulators should consider this evidence when designing consumer protection frameworks for cryptocurrencies.
Related
- Experimental investigation on human thermal responses to static and dynamic ramp temperatures in simulated hot-humid underground environments
- Nutritional profiling and pharmacological investigation of Baccaurea motleyana peel and seed through experimental and computational approach
- Experimental investigation of pile diameter and group configuration effects on soil pile interaction and lateral group efficiency in sand
- Experimental Investigation of Tear Strength, Stress Relaxation, and Anisotropy of Polypropylene Surgical Meshes
- Experimental and theoretical investigation of parameter-driven stability and chaos in a non-linear neuron circuit model
BibTeX
@misc{andrade2026genuine,
title = {{Genuine association or spurious responses? An experimental investigation of the association between cryptocurrency trading and Problem Gambling Severity Index scores}},
author = {Maira Andrade and Leonardo Weiss-Cohen and Richard James and Leon Y. Xiao and Philip Newall},
howpublished = {PsyArXiv},
year = {2026},
month = oct,
doi = {10.31234/osf.io/ckdne_v3},
url = {https://osf.io/preprints/psyarxiv/ckdne_v3/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Genuine association or spurious responses? An experimental investigation of the association between cryptocurrency trading and Problem Gambling Severity Index scores
AU - Maira Andrade
AU - Leonardo Weiss-Cohen
AU - Richard James
AU - Leon Y. Xiao
AU - Philip Newall
PY - 2026
DA - 2026/10/20
PB - PsyArXiv
DO - 10.31234/osf.io/ckdne_v3
UR - https://osf.io/preprints/psyarxiv/ckdne_v3/
AB - Background and aims: Cryptocurrency trading has become a popular activity worldwide, despite being largely unregulated and posing high risks. Previous studies have correlated cryptocurrency trading with harmful gambling as measured by the Problem Gambling Severity Index (PGSI), but no experimental research has explored whether these represent genuine associations or are spuriously inflated by methodological weaknesses. Methods: In a large-scale pre-registered 2x2 online experiment (N = 4,781), we hypothesised that if observed correlations were driven by spurious responding, then they would be significantly weakened by screening out non-gamblers (compared to the unscreened control), and by explicitly instructing participants to not consider cryptocurrency trading as a form of gambling (compared to a no-instruction control). Cryptocurrency trading engagement was assessed via self-reported trading frequency, price monitoring frequency, and the number of coins owned over the past 12 months. Results: Results While screening significantly lowered reported PGSI scores, the association between trading frequency and the PGSI remained robust and statistically significant across all conditions, and was not moderated by screening or providing explicit instructions. Exploratory analyses found the same patterns with price monitoring frequency and the number of coins owned, and in a combined model with all predictors, trading frequency explained the most variance in the PGSI outcomes. Conclusions: Together, these findings provide strong support for a genuine association between cryptocurrency trading and harmful gambling. Financial and gambling regulators should consider this evidence when designing consumer protection frameworks for cryptocurrencies.
ER -