"Most human behaviour is learnt observationally through modelling: from observing others, one forms an idea of how new behaviours are performed, and on a later data this coded information serves as a guide for action."

"Most human behaviour is learnt from the environment through stimulus response mechanisms. Individuals make associations between neutral stimuli in their surroundings with unconditioned responses, until eventually the neutral stimuli becomes the conditioned stimuli, even in the absence of an unconditioned stimuli"

"Most human behaviour is learnt through operant conditioning. According to Thorndike's 'Law of effect' any behaviour which brings about a pleasurable consequence will be repeated. In this way, aggressive behaviour becomes strengthened in one's behavioural systems by the rewards that it produces"

Bandura only used child participants. Firstly, researching with children raises problems with the validity of the study, as children have an increased likelihood of displaying demand characteristics, as by nature children are often eager to please in experiments and get the 'answer right'. As a result, this reduces the internal validity of Bandura's research. For this reason, due to the social learning theory being heavily based around the findings of these experiments, one might argue that the theory does not accurately represent the learning processes associated with aggressive behaviour in humans. What's more, there is concern over the limited applicability of the findings to adults. Therefore, psychologists must be caution when drawing conclusions and applying them to adult aggression as the findings may not be transferable. In this way, the social learning theory could be said to be lacking in universal application, and so can only be used to validly and reliably explain childhood aggression, until further research is carried out with regards to adult participants.

Due to being carried out in a lab setting, and using an artificial doll as the figure to which the aggressive behaviour is directed towards, the study could be criticised for lacking ecological validity. The children's behaviour might be changed somewhat by being in the artificial environment, and so the findings cannot be easily generalised to wider settings. Furthermore, by the aggression being directed at a doll, rather than a human being, there is no significant consequence to the aggressive acts being observed and displayed, which there would be if reproduced in a real-life setting. The study itself supports the idea that motivation and expectancy of reward influences the likelihood of aggressive behaviour being exhibited, and so when there is greater risk of negative consequences e.g. hurting someone, when the aggressive behaviour is imitated in the real world the behaviour seen in the study may be diluted or the patterns may not exist at all. Therefore, psychologists must take this into consideration when drawing conclusions about human aggression from this study, as human aggression does not occur in an artificial, controlled environment, and so the validity and reliability of the behaviours in the study representing those which occur in the natural environment may be poor

The bobo doll study also raises issues of imposed etic. Bandura used Western researchers in a Western country on Western participants, and so the extent to which the findings can be generalised to other cultures is limited. Therefore, the social learning theory may only explain human aggression in terms of that seen in the West, and so have little to no relevance to other individuals from other cultures. As a result, further research is needed, which includes participants and behavioural representations from a variety of different cultures, in order for the social learning theory to be deemed a universally accurate and valid explanation of human aggression.