Context: The context included UK and international settings within the academic environment (research ecosystem).

Participants: Academic, administrative, and technical staff and students of all levels, grades, disciplines, and professions. This meant including academic and research-enabling staff (e.g., research managers, technicians, administrators, and librarians) to ensure an inclusive approach was taken (and incorporate the principles of ‘team science’ and organisational culture).

Inclusion criteria: Evidence from research institutions only (considering Education, Enterprise and Research, the triple helix approach ^{36 , 37} ) for both academic and grey literature were included. All disciplines within the academic environment were included. All initiatives aimed, partly or wholly, at enhancing or assessing research culture were included.

Exclusion criteria: Anyone undertaking or supporting research outside of a research institution/Higher Education Institutes environment (for example in the health and social care field the National Health Service (NHS) Trusts, hospital settings, primary health care settings, allied health professional settings). Industry and professional service businesses (including consultancies) were not included as they were not considered to have an academic focus. Non-English articles were excluded if no translation was available for the full article.

The database searches and grey literature did not have any limitations on country of origin, apart from news items that were restricted to the UK, Europe, North America, and Australasia.

Several types of contributions were used for the scoping review, which included articles, reports, blogs and web-based articles from both empirical studies and grey literature.

The scoping review considered all types of study designs for inclusion (e.g., randomised controlled trials, non-randomised controlled trials, before and after studies and interrupted time-series studies, analytical observational studies including prospective and retrospective cohort studies, case-control studies, analytical cross-sectional studies, descriptive observational study designs including case series, individual case reports and descriptive cross-sectional studies).

Qualitative studies were also considered that focused on qualitative data including, but not limited to, designs such as phenomenology, grounded theory, ethnography, qualitative description, action research and feminist research. In addition, systematic reviews that met the inclusion criteria were considered, depending on the research question. Editorials and opinion papers were also considered for inclusion in the scoping review.

A range of data were required to be as inclusive as possible due to the diverse nature of how research culture is reported and discussed in the public domain (and its associated parts in Open Access (OA), Equality, Diversity, and Inclusion (EDI) and career paths). Therefore, the review included published material from academic outputs (e.g., journal articles, commentaries, editorials, perspectives, opinion letters) and from grey literature (e.g., reports, blogs, web-based articles, and newsletters including associated webpages of relevance).

The search strategy aimed to locate both published and unpublished references. An initial limited search of Medline and Web of Science (WoS) was undertaken to identify articles on the topic, to develop and pilot the search strategy. The text words contained in the titles and abstracts of relevant articles and reports, and the index terms used to describe the articles were used to develop a full search strategy. ^{15 , 38 – 41} The search strategy, including all identified keywords and index terms, were adapted for each included database and/or information source.

There were no study or language limits applied in the initial retrieval process. The search strategy was limited from 2017 to 2022 but a preliminary search of references during 2015-2022 was initially screened for relevance. Preliminary scoping and piloting of the search terms and strategies suggested that five years was sufficient for literature to be relevant, current, and broad (including relevant references on the reporting of initiatives such as DORA, and any changes due to the COVID-19 pandemic). The review included UK and international literature (including grey literature, although see below for pragmatic restrictions for news items).

Databases: Six databases were searched (Medline, Engineering village, Scopus, JSTOR, ProQuest and WoS) during the period 29 April to 18 May 2022. A range of databases enabled the reviewers to capture several disciplines and to be as inclusive as possible.

Grey literature searches: A pragmatic systematic search was undertaken of the Lexis-Nexis Academic database concentrating on newspapers and news items. The scoping and piloting of the search terms in the database suggested that geographical exclusions were needed due to the scale of results from the searches. As part of discussions with team members as well as an experienced librarian, the results were filtered to only include news outlets and organisations based in UK, Europe, North America, and Australasia. To augment the news searches, purposive sampling of relevant research websites was documented in an Excel spreadsheet to record all platforms and webpages visited. The sampling of websites was drawn from discussions with team members as well as an experienced librarian. Examples of research websites explored include: The Conversation, Nature, Science, UK Research and Innovation (UKRI) and Research on Research Institute (RoRi). Relevant references were identified from Wellcome Trust ^{14 , 15} as well as snowballing. The key references of the included articles and/or reports were screened for additional references to be included as part of the overall screening process (including grey literature).

Following the searches, all identified articles were collated and uploaded into Endnote version 20 ( Clarivate Analytics, PA, USA, https://endnote.com/ ) (a free alternative is Mendeley: https://www.mendeley.com/) and duplicates were removed. Following a pilot test, all titles and abstracts were screened, by two independent reviewers for assessment against the inclusion criteria for the review using Rayyan (rayyan, https://www.rayyan.ai/ freely available for 3 active reviews). Potentially relevant articles were then retrieved for full extraction. At the full citation screening stage, reasons for exclusion were noted independently by both reviewers. Where the independent reviewer was unsure, the article was discussed, and a decision was made by consensus. Screening at both stages (title and abstract and full extraction) was piloted using Rayyan ( https://www.rayyan.ai/) and labels were applied to categorise the focus of the articles based on four areas: •

Security (including career paths, career progression, stability contracts/careers, issues affecting early career researchers etc.)

Both reviewers applied the articles under one of the four areas during extraction, based on the content and context of the article. Some articles were relevant to more than one category, which was reflected in the data extraction table. The reviewers independently checked the included articles to ensure that they were included appropriately across the four areas.

These focus areas were reported in the Wellcome Trust report and formed the basis of the current scoping review, to enable the University of Southampton to build on activity already undertaken, activity underway and enable alignment for future consideration. ^{14 , 15}

The list of included articles for full extraction were then exported to a Microsoft Excel spreadsheet using the labelling of articles from Rayyan (these categories were grouped together under the four focused areas). The results of the search and the study inclusion process are reported in full using the Preferred Reporting Items for Systematic Reviews and Meta-analyses extension for scoping reviews (PRISMA-ScR), including the flow diagram reported in Figure 1.

Both reviewers extracted data from the full text articles using a data extraction tool developed by the reviewers to address the research question. This included the focus of the article, issues and/or problems reported in the article, solutions and/or recommendations provided in the article and details about whether the article related to more than one topic area.

No risk of bias or assessment on quality was conducted due to using a scoping review methodological approach. All the evidence was mapped and categorised into the four areas, which were discussed and agreed between team members at various stages of data extraction and during the write-up of the findings.

From the evidence there was a steady rise in the number of published articles over the last five years, with a notable increase from 2019. Table 1 shows that from the 253 included articles, there were 135 original research articles (this included qualitative and quantitative studies), 20 review articles (using a range of methodological review approaches), 86 perspective articles and 10 conference papers.

The location of the study generation was captured for the included articles (based on location of the research and/or authors location). The included articles covered a global perspective with 71 articles from USA and Canada, 73 from international locations such as Africa (n=13), China (n=7), Australia (n=7) and Pakistan (n=4), 43 from Europe and of these 17 were from the UK.

The grey literature provided 102 additional materials, 40 perspective articles reported in journals, 29 newspaper articles, 17 webpages (including educational webpages such as The Conversation: https://theconversation.com/uk) and including 10 reports. The remaining six were either a podcast, blog or case study. A majority of the grey literature material could not be grouped by location due to the nature of the material (76.5%, 78/102).

The evidence found in the database searches and grey literature was grouped according to the four focused areas, based on the key concepts developed during the full screening of the articles (based on the Wellcome Trust report). ¹⁵ Several included articles were relevant to more than one focus area, which showed the breadth of the topic but also how these areas are overlapping and mutually reinforcing. For example, evidence reported under security was also closely linked to wellbeing and equality of opportunity (especially for early career researchers (ECRs) and Science, Technology, Engineering, Mathematics and Medicine (STEMM)).

The sections below provide a summary of the evidence based on the four focused areas, with particular attention on security, wellbeing, equality of opportunity and teamwork, and research quality and accountability ^{14 , 15} (see Figure 2). The quality or assessment of these initiatives were not explored as part of this scoping review and the key considerations arising from the evidence are not presented in order of priority.

From the evidence it was clear that there is a global drive to expose the challenges and barriers in academic research culture. It was evident that these factors were not exclusive to specific countries, disciplines, or research institutions and the challenge is at a system level. Concerns over job security, career progression and sustainability (e.g., long-term employment, job performance, and organisational commitment) were particularly experienced by PhD students, ECRs and junior researchers from a range of academic environments. ^{42 – 45} The evidence reflects the range of research careers, roles, skills and expertise that are involved in research activity and therefore affected by poor research culture. Research institutions often separate staff into job families such as research, education, technical, clinical, and managerial roles titles reflecting different career pathways. ^{17 , 46 – 48}

Despite the range and diverse roles within academic environments, the evidence suggests there are concerns that research institutions are not managing career progression expectations or providing ways to develop and train staff, across career levels or job families, which can result in inequalities with regard to job security and career pathways. ^{17 , 46 – 48} This can accumulate in feelings of failure by staff, or staff feeling pressured to be successful, which ultimately could promote unhealthy working practices such as excessive workloads and working long hours to meet expectations. ^{49 – 51} These issues affect all research institution staff but the evidence suggests they are particularly acute for several groups such as ECRs, STEMM, people on Fixed Term Contracts (FTC), people with caring responsibilities, and people with disabilities ^{49 – 51} (see Table 2 for a summary of the key considerations associated to security and career progression).

The review suggests that the problem is reinforced by a culture where researchers are incentivized to produce many funding applications and academic publications despite there being high rejection rates. ⁵⁰ The evidence suggests that this can result in a lack of workload oversight, a culture discouraging of appreciation, that in turn can make researchers feel pressured to be successful, often resulting in a significant amount of time in pursuit of success at the cost of their wellbeing. ^{45 , 49 , 50 , 52 – 54}

The concept of research culture and job security was broad and included (but is not limited to) career paths; stability of contracts and careers; and issues affecting ECRs and students. ^{60 , 144 , 148} To improve job security, research institutions and funding organisations need to readdress how research positions are funded, particularly in the early career stage. A wide range of initiatives were explored in the literature covering a broad spectrum of factors (as detailed in Table 2). The evidence showed how, where, and why changes are needed to establish a global cultural change to the research ecosystem to enable fair and transparent progression for all research staff ^{102 , 107 , 164} ; cultivate a culture that fosters diversity across career pathways ^{17 , 58} and; initiate deeper integration of knowledge to ensure institutional stability and innovation. ^{60 , 156}

The evidence suggests that offering potential solutions or supportive actions for research institutions and the research community may enhance (e.g., opportunities for promotion and/or flexible career path changes) and stabilise career paths, particularly those in the early career stage, including those in technical and managerial roles (see Table 2). Although these solutions and supportive actions are by no way exhaustive, they do provide a summary of the range of factors that could go some way in promoting a better research culture. ^{57 , 141 , 157 , 165}

A key consideration from the evidence suggests that there are disparities across the research ecosystem, which are in turn influencing individuals’ wellbeing. ^{42 , 43 , 46 , 73 , 95 , 99 , 102 , 121 , 166 – 169} The impact of these disparities are preventing or slowing down initiatives to seek for a cultural change in the academic environment. Although there has been some progress, the evidence suggests progress is slow and continues to be a challenge, especially for under-represented groups. ^{169 – 172} The disparities highlighted in the literature suggests that under-represented groups are less likely to be promoted or receive funding and have a higher risk of decreased well-being. ^{18 , 67 , 86 , 92 , 173 – 182} However, as Lee (2022) pointed out, although underrepresented groups and junior staff are more likely to experience these challenges anyone can, at some point in their academic career experience some form of microaggression (e.g., bullying, patronage power, exploitation, discrimination, imposter syndrome). ²⁰ Moreover, the way disciplines are taught at university means that curricula focusing on traditional perspectives may not be inclusive to everyone. ^{126 , 179}

The review revealed that the risks associated with a lack of diversity and inclusion often result in individuals leaving academia, low job satisfaction, increased stress, burnout and mental health problems, and decreased productivity. ^{18 , 47 , 77} Moreover, the increasing demands of heavy workloads and the risk of perpetuating a culture of academic rejection can impact an individual’s wellbeing. ^{18 , 49 , 50 , 109 , 183} The evidence suggests that these wellbeing issues can have an impact at an institutional level, resulting in a lack of diversity across leadership roles, ^{76 , 100 , 184} a shortage of role models and peer mentors, ^{166 , 47 , 95 , 185} and driving off talent due to staff leaving academia. ⁸ Table 3 provides a summary of the key considerations associated to wellbeing and equality of opportunity.

The review illustrates that over the last three years, the COVID-19 pandemic has highlighted some important challenges to this already highly pressured working environment, with mixed effects, particularly to those with caring responsibility. ^{18 , 47 , 77} However, there is also emerging evidence showing the potential benefits and opportunities as a result of COVID-19. ^{18 , 47 , 67 , 76 , 77 , 100 , 180 , 197} For example, use of online platforms for training and teaching purposes has opened up opportunities to bring together specific communities and countries. The pandemic led to calls for more ‘kindness in research’ where empathy replaced the usual expectations on life-work balance. ^{76 , 100 , 184} In addition to opportunities emerging from the pandemic, the review also identified clear efforts to improve and raise awareness of the need for academia to embrace the EDI agenda through several initiatives, movements, and policy implementations (as detailed in Table 3). Most prominently, focusing on efforts to improve individuals’ opportunities through networking, collaborations, mentoring and peer-to-peer support, balancing career, and family aspirations can help to guide inclusivity and strengthen infrastructure and local capacity. ^{174 , 180 , 186 , 208}

Collaboration, openness, and transparency were highlighted in the evidence as key indicators of success for driving forward a positive cultural change. However, the emerging evidence suggested that perverse incentives within the research ecosystem, a lack of training, opportunity, support, and infrastructure can undermine ambitions for change, which is further hampered by researcher perceptions of what an academic career entails. ^{4 , 25 , 63 , 157 , 185 , 221 , 222}

The evidence pointed to a range of barriers that have repercussions on the notion of teamwork, such as the ongoing tradition of first and last authors taking most or all of the credit for the work, ²²³ the use of ‘gift authorship’ to enhance research publication of academics with poor research performance, ⁵ and the pressures to have global impact on the scientific community through high-quality scientific writing. ¹⁶⁷ Researchers are incentivized to attain research excellence despite ‘excellence’ being narrowly defined, which can often lead to hyper-competitiveness and unfair working practices. ^{4 , 19 , 25 , 53 , 62 , 91 , 92 , 209} The evidence further suggested that the Research Excellence Framework (REF) in the UK has been criticised as promoting competition between departmental colleagues rather than collaboration due to criteria on who they can contribute with and the increased demand on publications. ^{52 , 92}

Practices that could restrict collaboration as a result of REF can have a detrimental impact for the promotion of research integrity and team science, especially for ECRs. Research careers encompass a range of roles, skills, and expertise but as the evidence suggests this is not universal, with some universities separating research staff from research-enabling colleagues such as research managers, technicians, administrators, and librarians, some of whom have research-level qualifications and experience. ^{44 , 102 , 161} Separation in this way could lead to inequality of how staff are included (or not), trained, mentored and perceived by fellow colleagues (see Table 4 for a summary of the key considerations associated to teamwork and supportive working relationships).

Although there are several challenges related to workplace relationships, any steps to reform these barriers will require accessibility to opportunities and resources that collectively bring research staff together in a unified and cohesive way to promote and create trust (rather than having intensive competitive pressures to achieve based on individual merit). ^{54 , 167 , 250}

The evidence reveals multiple layers of complexity around the notion of ‘teamwork’ and the interrelated social and environmental factors that unfortunately reinforce a status quo. For change to occur there needs to be synergy for collaboration to ensure individuals, Research institutions, funding organisations and society share a unified approach to move beyond solitary, isolated teams to a deeper integration of multi-disciplinary/inter-disciplinary culture. ^{7 , 60} An inclusive, representative, and collaborative research environment contributes to improvement in researchers’ sense of belonging and to positive cultural change. ²⁰⁹

The identified evidence suggests that there is a need to take a holistic and integrated view of the intrinsic (those within disciplines) and extrinsic factors (those outside of disciplines) that affect the research environment to come up with novel ways to tackle the challenges with teamwork and collaboration to ensure openness and a cultural shift in the right direction. ⁷ There is growing evidence that success in research and innovation requires diversity in roles, knowledge, and skills, an inclusive, representative, and collaborative research environment all contribute to improvement in researchers’ sense of belonging and to positive cultural change is required. ²⁰⁹

From the existing evidence it was clear that transparency, open research, and integrity requires collaboration from research institutions, funding organisations, researchers, publishers, and other sectoral organisations such as industry. ^{29 , 263 , 264} A large proportion of the evidence (more than a quarter of articles (133/253) included from the database searches and half of the grey literature (52/102)) suggested a link between lack of open research and reproducibility that could lead to public distrust and inhibit several open research practices. ^{25 , 89 , 219 , 238 , 240 , 260 , 265} Increasing pressures on researchers is causing a ‘publish or perish’ practice, and has meant that researchers are prioritising ‘getting it published’ rather than ‘getting it right’. ^{25 , 89 , 238 , 240 , 257 , 258 , 260}

However, as noted by Munafo (2022), the ‘replication crisis’ could be regarded as an opportunity to promote motivation for improvements. Determining where effort is most needed and what changes are required, provides opportunities for the research ecosystem. Specifically, research institutions and funding organisations can mandate open research practices, and therefore coordinate change at both research integrity and researcher integrity level ²⁶⁴ (see Table 5 for a summary of the key considerations associated to teamwork and supportive working relationships).

The existing evidence demonstrated that open research practices (e.g., research integrity, researcher integrity, open data, open access and transparency) requires a global effort, as well as involvement from all sectors of the research ecosystem (e.g., institutions, researchers, funding organisations, publishers, industry). However, more evidence is needed to demonstrate where and in what circumstances the change is having tangible benefit. ^{46 , 263 , 264 , 285 , 332}

As the evidence suggests, practices should be evaluated to assess whether change has been of value, enhancing the research pathway and align to be evidence informed, therefore avoiding any unintended consequences. ^{5 , 7 , 138 , 264 , 275} Meta research (e.g., research on research, meta science) is one way to evaluate and evidence any innovation taking place, and therefore determine the impact and tangible benefit of these changes to promote and enhance the research ecosystem. ^{7 , 25 , 39 , 250 , 264 , 284 , 332 , 344}

The evidence review found several initiatives such as the UK Reproducibility Network (UKRN), Declaration on Research Assessment (DORA), and the European Network of Research Integrity Offices (ENRIO) to promote, encourage and prioritise the facilitation and creation of open research practices. ^{7 , 22 , 284 , 288 , 328} Adopting such initiatives enhances innovation across all aspects of the research ecosystem but there is variation on how far they have been implemented (including what stage of the development) and the acceptance level from researchers, research institutions and funding organisations. ^{306 , 325 , 328 , 333}

The main strength of the review was using several systematic database literature searches, which was complemented with additional grey literature searches of known online education articles and websites. However, a limitation is that scoping reviews only map the evidence and do not assess the quality of the articles or risk of bias. Most of the database articles had an international focus, with most of them from USA and Canada (28.2%) and other international countries (28.7%). For a large proportion of the grey literature, it was not possible to ascertain the articles countries region. Moreover, on a pragmatic basis, news items from news organisations based outside of UK, Europe, Norther America, and Australia were excluded, which along with the exclusion of non-English language documents means that other initiatives being used to improve research culture may well have been missed. On this basis, there could have been international and regional biases, but it could also be that there is a lack of evidence from these regions rather than missing articles from the systematic searches.

The review found more than 200 articles from the systematic database searches (n=253) and more than 100 from the grey literature searches (n=102) which suggested that there is growing literature around what constitutes a ‘good’ research culture and what this could look like. However, as the literature has shown, progress has been slow and although the evidence provided several examples of established initiatives and networking opportunities, the evidence was more anecdotal, and opinion focused.

Ethical approval and written consent statement were not required.