This paper examines how contextual (conversational) aspects and socially shared meanings might affect the participants' performance on a standardised memory test using the theoretical framework of social representations. A total of 97 members of centres for older adults located in Rome, Italy participated in a screening using the Montreal Cognitive Assessment test. Prior to testing, a group of volunteers had organised a performance focused on events from the distant past, stimulating intergenerational reminiscence. The participants were randomly assigned to one of two conditions. In the first case, prior to administering the test, a psychotherapist talked to each participant about the performance, focusing on ageing and stressing the neutral aspects of its social representations, such as change and time. In the second case, performance was used to concentrate on positive aspects of the social representations of ageing, namely wisdom and experience. In line with the hypothesis, focusing on positive aspects of social representations of ageing (wisdom and experience) versus their neutral aspects (change and time) has resulted in improved performance on a standardised memory test. Practitioners (psychotherapists – experts in psycho-diagnostics) who administered the tests have been involved in the co-construction of the meaning of ageing, discussing a real-life situation: the common experience of intergenerational activity that involved the participants' memories of their urban environment.