Ainsworth devised a controlled observation called the Strange Situation to assess types of attachment in 9–18 month old infants. This involved placing a child and their mother in a novel environment of mild stress, whereby they would be observed and videoed through a one‐way mirror during a series of eight different situations.The purpose of the Strange Situation is to measure four key behaviours, including: exploration behaviours – how the child explores the environment and whether they use the mother as a safe base; separation anxiety – how the child responds to/behaves when the mother leaves the room; stranger anxiety – how the child responds in the presence of a stranger; and reunion behaviours – how the child acts when reunited with their mother.Depending on how the child responds in the Strange Situation would lead to one of three attachment classifications: secure, insecure–avoidant and insecure–resistant.One criticism of Ainsworth’s Strange Situation is the type of observation she used, which was an overt observation. The parents in Ainsworth’s study knew they were being observed through the one‐way mirror and therefore may have displayed demand characteristics. This meant that the mothers may have been overly affectionate towards their children, changing their behaviour to appear excellent parents, as they believed this is the behaviour that the scenario demanded of them. In turn, this could have altered the children’s behaviour and therefore lowers the internal validity of the experiment making the Strange Situation a less valid method of assessing attachment.However, the Strange Situation method of assessing attachment type is said to have high reliability. The observations took place under strict and controlled methods (including video recording) using predetermined behavioural categories. Ainsworth had several observers watching and coding the same infant behaviours independently. Ainsworth et al. (1978) found 94% agreement between observers on attachment classifications. When inter-observer reliability is high, this suggests that the Strange Situation is a reliable technique for assessing types of attachment, as different observers agree on their assessment of child attachment types.