Bandura's Bobo Doll Experiment: Study Strengths and Weaknesses

Bandura's Study Analysis

Strengths of the Research Method

Structured Observation and Data Collection

One strength of the study is that the research method used was a structured observation where a behavioural checklist was employed to record the aggressive acts of the children. A behavioural checklist allows for standardization as it can be replicated to test for reliability. Furthermore, through time sampling, such as in this study, observers can record the number of acts displayed by the children as per the checklist, which allows for quantitative data to be collected.

Controlled Environment

The study was also a controlled observation as it was conducted in the highly controlled environment of a laboratory. For example, the toys the children were exposed to in each phase were always the same, with the behaviour of the models in each respective condition being standardized. This further allows us to replicate the study to test for reliability.

Internal Validity Enhancement

A second strength of the study is validity. The pre-existing levels of aggression were tested before the participants were allocated to the three groups. This ensures that there would not be individual differences which may interfere with the results of the study. The researcher can therefore be more confident that it is the Independent Variable (IV)—the condition of the model—that is affecting the Dependent Variable (DV)—the aggressive acts displayed by the children—hence improving validity.

Weaknesses of the Research Method

Low Ecological Validity

A weakness, however, is that as it is conducted in a highly controlled and artificial setting of a lab, the study lacks ecological validity. The actions of the model were also not something children experience in their daily lives, as adults do not usually play aggressively around children with children’s toys. This suggests that the study also lacks mundane realism.

Limited Generalisability

A second weakness of the study is that the sample is not representative of a wider population. The sample size is only 72 children of a particular age range (3 to 5 years), all part of the same Stanford nursery, suggesting they may have some similar characteristics. This means the findings cannot be applied to older children or perhaps to children of other nurseries, which limits the generalisability of the study.