Implicit Bias in the Recruitment of Racially Diverse Trainee Clinical Psychologists in the UK

This dataset was created as part of a doctoral thesis investigating unconscious bias in Clinical Psychology selection.
The study explored the effect of candidate race and brief unconscious bias training on mock interview ratings given by participants eligible to be DClinPsy selection panellists.

Participants were randomly allocated to one of four conditions (2 × 2 design: Candidate Race × Training) and completed:
- an experience and demographic questionnaire
- pre-interview implicit bias measures (Asian-White IAT and Bodyweight IAT)
- pre-interview self-report questions on unconscious bias awareness

They then observed a mock interview before:
- scoring the candidate across domains (Personal, Academic, Research, Clinical)
- completing post-interview measures (implicit bias and awareness questions repeated, with three additional training evaluation questions)

Data were collected using Qualtrics and the Implicitly online platform.
All personal data (e.g., email addresses) were stored securely and separately from the research dataset. Only anonymised data are included here.

This dataset contains:
- An Excel .xlsx file of anonymised participant data (rather than SPSS, to allow full descriptive labelling and transparency)
- A comprehensive codebook describing each variable, coding scheme, and notes on data processing

Data includes:
- Demographics and experience
- Experimental condition allocation
- IAT scores and labels (pre and post)
- Awareness and knowledge items (pre and post)
- Training evaluation questions (training groups only)
- Interview ratings and placement recommendations

Anonymisation & Processing Notes:
The dataset is fully anonymised. All personally identifying information (e.g., email addresses, participant IDs, timestamps) was removed.
Variables that could contribute to indirect identification (e.g., exact years of panellist experience) were collapsed or excluded.
The data file was created in Excel to allow detailed variable descriptions and overcome SPSS label length constraints.
A full codebook is provided to support interpretation of the dataset.