Table 1 The steps involved in developing the item pool for a new measure of resilience in dementia

Development of a conceptual model

A scoping review of the literature and qualitative interviews with people with dementia, synthesised into a conceptual model of resilience in dementia. For more information around the development of this conceptual model, see Windle et al. [9].

Question generation

Themes identified in the literature and qualitative interviews used to generate a ‘long list’ of items. Extract quotes from qualitative evidence to inform the development of potential items wherever possible.

Question and response formatting

Discussion among core research team around possible response scales, timeframe and demographics.

Content and face validity discussion, and initial item reduction

Discussion among core research team regarding individual question format and structure. Item list reduced via a process of ranking items and consensus over those deemed most appropriate/least appropriate.

Cognitive interviews for face validity, content validity, and suitability of response format

Interviews with people with dementia (n = 6 people with rare dementias; n = 5 people with more typical forms of dementia). Participants asked to answer each item. Follow-up questions asked by interviewer probing understanding, clarity and usefulness. Participants encouraged to make suggestions for amendment of items where they see fit, and to identify preferred items.

Revising items and final item reduction

Results discussed among core research team. Suggestions for amending questions put forward by interview participants implemented if the core research team agreed that the rewording was clearer and more suitable than the original. Items removed if participants had reported finding difficulty in answering or understanding them. Items with ≤ 1 preference votes also removed.