Extending psychometric network analysis: Empirical evidence against g in favor of mutualism?