Purpose The Research Autism Cygnet Mentoring project was a two-year pilot study, completed in 2016, which aimed to develop, trial and evaluate a mentoring scheme designed with input from autistic people, their families and supporters. The paper aims to discuss these issues. Design/methodology/approach The mentoring scheme involved 12 matched pairs (mentor/mentee) meeting once per week for one hour, over a six-month period. All mentors attended a training day, led by the principles of personal construct theory and an emancipatory research ethos. The project and training involved significant involvement of autistic people in both its design and delivery. Findings Participants on the autism spectrum found their mentoring experience very helpful in enabling them to progress towards self-identified goals, and mentees felt empowered by the person-centred ethos and the methods employed on the project. However, a number of aspects of the mentoring project have been identified that require further investigation, including: caution over offering mentoring without formal structures, boundary setting, supervision, flexibility and the matching of mentees with mentors. Originality/value The project has highlighted the potential benefits of time-limited goal-orientated mentoring and the negligible evidence base underpinning current mentoring practice with adults on the autism spectrum. In order for the project to realise its emancipatory aim, there is a need for a large-scale quantitative study and a health-economics analysis to provide the necessary evidence base for mentoring to be recommended as a cost-effective intervention with clear benefits for individual wellbeing.

Expert or skilled behaviors (for example, face recognition or sporting performance) are typically performed automatically and with little conscious awareness. Previous studies, in various domains of performance, have shown that activities immediately prior to a task demanding a learnt skill can affect performance. In sport, describing the to-be-performed action is detrimental; whereas in face recognition, describing a face or reading local Navon letters is detrimental. Two golf-putting experiments are presented that compare the effects that these three tasks have on experienced and novice golfers. Experiment 1 found a Navon effect on golf performance for experienced players. Experiment 2 found, for experienced players only, that performance was impaired following the three tasks described above when compared to reading or global Navon tasks. It is suggested that the three tasks affect skilled performance by provoking a shift from automatic behavior to a more analytic style.By demonstrating similarities between effects in face recognition and sporting behavior, it is hoped to better understand concepts in both fields.Navon effects in golf performance 3