Table 3 Common themes, subthemes and quotes identified in response to the question “Do you think that PPI is relevant to statistical methodology work?” from “Conclusions” section of the questionnaire

Positive

Aids in communication of work

“Most of the work that we do is funded by taxpayers, so I think that they should always have a say in our research. It also helps the public know more about what we do (i.e. no more scary statistics!) and improves our own communication skills and how we interact with the public.” (P13)

Provides motivation for the research

“For statistical methodology research, it is easy to get caught up in the methodology. PPI provides an opportunity for the researcher to get first hand account on how important the topic is to the people that will be affected by the research. They become the motivation for your research.” (P35)

Shapes the research question and determines prioritisation of work/funding

“Can confirm importance of question and also ensure the approach reflects real-world experiences” (P104)

“PPI is very important at informing what questions we should ask and where research needs to be directed.” (P42)

“Many methods projects are not developed in a way that allows the work to be adopted widely and applied for patient and public benefit.” (P103)

Public confidence/understanding of data

“Any study using patient and public data is inherently linked to the interests of patients and the public. Involvement allows them to have a say in how their data is used.” (P65)

“If the general public had better understanding of the importance of data sharing and safe guards in place to protect them I hope in the future they are more likely to share there personal data.” (P120)

Patients/public are often underestimated and can contribute

“Yes, we are using new statistical methods or more advanced methods. It can be tricky but using common language and giving participants an active role in selecting variables, or allowing them to expand on their feelings of certain approaches can offer valuable thinking on how the results are interpreted by the groups who they affect.” (P68)

Relevance of work to the public

“The methods we use will impact the information they are provided with. Therefore, we need to make sure we are developing methodology which provides answers they are interested in.” (P46)

Neutral

The PPI group should be statisticians

“The target for this work is statisticians, not patients. So eliciting the thoughts of target end users (methodologists) might be helpful.” (P38)

“The findings are relevant to those with knowledge of the area who are likely to use them, not the layperson.” (P114)

Dependent on the research

“Sometimes it can be difficult to see where public can meaningfully contribute (or have much interest) if the work is especially theoretical and patient/public implications for quite far down the line. Other methodology work has more obvious implications for public/patients. However, I’d hope public would help identify where they can input meaningfully rather than me dictate that.” (P86)

“There are some aspects which are very relevant (e.g., choosing appropriate outcome measures and how to present them) and others where it would be difficult for PPI members to contribute meaningfully without specialist training—e.g. comparing different modelling techniques.” (P116)

“It depends on the project. There is a place for it in methods being developed in a specific clinical area, but I can’t see how the public would be able to comment on a purely methodological statistics project without expert knowledge themselves.” (P79)

“A lot of statistical methodology work is driven by the need to obtain unbiased answers to questions. This does not necessarily require PPI input

There are some areas of research where PPI input can help frame the methodological questions addressed, but not all—a blanket requirement for PPI in methodological work risks being a check box exercise.” (P47)

Not sure on the relevance

“PPI is a confusing world. Should patients be involved? Yes they should. Are the ideas for how to involve them sensible or a good use of money? No they’re not.” (P66)

“I’ve seen how it can be both helpful and unhelpful, so my thoughts on its relevance are unclear.” (P71)

Negative

The research is too technical for PPI

“Methodological research must be driven by scientific priority. Most methodological research is completely beyond the understanding of a lay person.” (P93)

“I don’t think it is possible to include patients/public when carrying out high level methodological work without forcing the patient/public to become an ‘expert’ in the area, and defeating the point of PPI.” (P80)

“It was challenging to convey complex ethical and design concepts to patient partners but the experience was nevertheless useful and rewarding. As much as I would like to, I cannot imagine involving community partners in the more technical statistical projects I have been involved in. These typically require statistical simulation and theoretical development. I guess partners could help refine research questions/objectives but I'm having a difficult time imagining how to engage them in technical aspects.” (P27)

Patients/public cannot have a meaningful impact

“The statistical methods shouldn’t alter depending on what patients think…” (P119)

“It doesn’t really inform/change your ideas of what you are planning to do.” (P30)

“I often struggle to think how inviting patients to a session would help inform the design of the work.” (P96)

Tick box exercise

“For statistical methodology it feels like a square peg in a round hole…if there is no meaningful way a non-statistician can contribute then this becomes a tick box exercise that can be detrimental to forming the positive relationships required to make PPI effective.” (P40)

Low priority for patients to know about statistical methodology

“Deriving a standard error isn’t probably on the top 100 things patients care about…” (P32)