This workshop will provide attendees with a framework for improving the representativeness of person centred care research. Person centred care research frequently involves the collection of patient and public views. This may be quantitative (eg patient experience surveys) or qualitative (eg ethnography). Typically the focus is on understanding the needs, preferences, and experiences of people who use health and care. There is a growing focus on issues of equality, diversity, and inclusion. Despite this, person centred research often addresses only limited barriers to representativeness, which can lead to a superficial interpretation of representation, undermining efforts to understand equality and equity. We propose a framework of non-representation in research that identifies five causes for users to address: Non-access - groups are excluded because of their inability or unwillingness to access care services. Non-invitation - groups are excluded in selection, often for practical reasons. Non-response - groups are systematically less likely to respond to requests to participate. Non-identification - research fails to identify people from different communities. Non-review - data is not analysed according to available diversity markers, or issues important to certain communities are not considered. Workshop content: CG to explain the framework, with clear examples and discussion of the impact of underrepresentation - 20 minutes. CG, JK and AT to facilitate group discussions of two case study projects - one quantitative and one qualitative - to analyse barriers to representation and identify improvements - 30 minutes. We will use paper templates, pens, and post-its to record the discussion. Reconvene to one group providing feedback in a plenary discussion led by CG - 10 minutes. For inclusivity and to support participation, plain English will be used throughout. Following the event we will - with consent - e-mail participants the workshop content including presentation, case studies and analysis conducted in the groups.