Methods
Survey design
We employed a cross-sectional survey to explore the engagement with
research of adult social care staff. The research took place in three
neighbouring local authorities, affiliated as part of an Integrated Care
System11In England, integrated care systems (ICSs) are
partnerships that bring together NHS organisations, local authorities,
and others to take collective responsibility for planning services.
in the South West of England. Ethical approval for the study was given
by the School for Policy Studies Research Ethics Committee, University
of Bristol (UK). The survey built on the survey designed by Gray et al.
(2014) for social work professionals in Australia. We adapted the survey
to apply to all practitioners working in an adult social care
organisation, which included social workers, occupational therapists,
non-registered professionals, service leaders and managers, and other
support staff. The survey covered views about using research; knowledge
and skills in working with research evidence; searching for, finding,
and evaluating evidence; key barriers to using research; and demographic
and employment details.
As well as review from the practice leads and evidence champions for the
ConnectED programme, the draft survey was piloted and reviewed with five
social care practitioners and students. In response to feedback, we
adapted the language of the measurement scales used by Gray et al.
(2014). The phrase ‘research views’ was used instead of ‘research
attitudes’, which was perceived to have moral connotations. Similarly,
reviewer feedback led us to change two sets of scales from measures of
‘ability’ to those of ‘confidence’. A full copy of the final survey is
available as a supplementary file [see Supplementary Materials].
Survey distribution
The final survey was designed as a 49-item, self-complete, and anonymous
questionnaire hosted on REDCap (an online survey platform). It was
distributed as a weblink through email lists collated by administrative
staff in the participating local authorities to all adult social staff.
The survey was promoted by staff working in the ConnectED programme and
through newsletters and publicity at staff team and department meetings,
but completion was voluntary. The survey ran between March 2022 and July
2022, with multiple email-based reminders being sent out. Participants
had the option to apply to win a £50 voucher by giving contact details
separately to the survey (to preserve anonymity). One applicant was
randomly selected for a voucher from each local authority.
Data analysis
The dataset was exported into Microsoft Excel and IBM SPSS version 29
for processing. For measurement scale items where there were up to three
missing data items, we applied imputation (Luengo, García, and Herrera,
2012) to insert the mean for other responses from the respondent (9
cases in total). Given our use of revised questions, we assessed the
internal consistency of the measurement scales where a Cronbach α of .70
or higher was considered acceptable. For the question set on ‘research
views’, the reliability analysis obtained a Cronbach α of .766. For
‘literature search confidence’ measures and ‘critical appraisal
confidence’ measures, the scale reliability analysis was a Cronbach α of
.820 and .866 respectively. Within-group comparisons were made by
Pearson chi-square test for nominal data and the
Mann-Whitney U test for data on ordinal scales, withp <.05 set to determine statistical significance.