Operant conditioning (Skinner, 1938) is generally proposed as both the initial and primary mechanism for learning with neurofeedback (Hammond, 2007). Operant conditioning uses feedback and reward to modify voluntary behaviours. According to the theory of operant conditioning, there are three critical factors in determining how effective a reward will be in the conditioning feedback loop. The first is the immediacy of the feedback signal: the quicker the feedback, the shorter the learning time and the more rewarding the experience. Conversely, as the time between the behaviour and the reward is increased, known as “reinforcement delay”, learning efficiency is decreased. The second factor is the contingency of the signal; this refers to how accurately or fully the signal represents the activity being trained. And finally saliency refers to how rewarding the reward is to the participant (Skinner, 1950).

Neurofeedback is generally considered to be based on two concepts. Firstly, the activity of the brain in both normal and abnormal functioning is objectively reflected in the EEG (Sterman, 1996; John et al., 1988). Secondly, Neuroplasticity of the brain allows behaviour to be modified by learning (Demarin et al., 2014).

Numerous studies from Pavlov (Pavlov, 1927) and Skinner (Skinner, 1953) have elucidated the mechanism referred to as conditioning and calculated how, when, and for how long, an animal or human can be induced to make a lasting change in behaviour. Both positive and negative conditioning have been codified into a reliable theory that is the basis of much psychology and ethology (Gray, 1970). However in clinical settings usually only positive reward is used in neurofeedback in order to encourage aberrant EEG towards a more normal state (Kluetsch et al., 2014; Birbaumer et al., 2013).