A survey of the training experiences and needs on Wellcome Trust PhD programmes

Background Training for PhD researchers was previously identified by the Wellcome Trust funded Emerging Research Cultures project as an area for further investigation to ensure an inclusive culture which enables PhD students to become well-rounded researchers. Methods The Taskforce on Training conducted a survey of 35 Wellcome Trust funded PhD students and 10 programme administrators to evaluate the provision of training in eight key areas. This survey examined a number of issues, such as availability and knowledge of training, potential gaps in training, and the perceived usefulness of training. Results PhD students reported that training was generally useful and viewed as important; with technical training being particularly highly valued. However, the survey identified that students desired additional training in project management and personal development. Surveying programme administrators highlighted the wide variety in training availability for students across different Wellcome Trust programmes currently running in the UK. Conclusion In response to these findings, several recommendations were suggested. Examples include; promotion of peer mentoring for PhD students, and alternative methods for delivery of wellbeing training. However, this report only explores the views of a small number of Wellcome Trust funded PhD students and would benefit from further research into the experiences of PhD students, programme administrators, and PhD supervisors.


Methods
. In contrast to these other focuses, the ERC project was launched to identify both positive actions and challenges faced in different work environments by researchers, to allow for the critical evaluation of current practices and the promotion of a healthy research culture.
Shortly after the project launch, group discussions were held with Wellcome Trust funded PhD students and staff from 18 of the 23 UK-based Wellcome Trust funded PhD programmes at 15 different institutions.The discussions considered several key concerns including; the ongoing negative impacts of the COVID-19 pandemic on funding and career prospects, good supervisory practices, the role of flexible working and mental health for researchers.Upon the conclusion of these group discussions in 2021, a report was published (Carusi et al., 2022a) highlighting four areas of interest in which further investigation was needed.
One of the key areas identified by Carusi et al. (2022a)  Training (hereafter referred to as the taskforce) was formed consisting of first-and second-year PhD students across current Wellcome Trust funded PhD programmes.The research undertaken by the taskforce is therefore reflective not only of the experiences of study participants but is also driven by the experiences of the taskforce who developed the questionnaire and performed the subsequent analysis and interpretation.The taskforce specifically focused on what and how training is delivered to students, and how training opportunities are promoted in their institutions.Moreover, it also assessed students' levels of satisfaction about training and where students felt improvements were needed.More information about the Emerging Research Cultures Project and its associated taskforces is available here: https://interchangeresearch-my.sharepoint.com/:p:/g/personal/erc_inter-chang-eresearch_com/eOu5MswTKLI3tVnQBGTBBxRt5aJCGOOR_WAkHyQ?e=wtCV6S The following report summarises the findings and recommendations of the training taskforce from the ERC project.
It can be used to reflect on current training practices across programmes, their successes and concerns, and can be used to standardise or implement training to ensure every Wellcome funded PhD student has the required skills to become a well-rounded researcher.Although this report specifically investigates the experiences of students on Wellcome Trust funded PhD programmes, the findings may be beneficial to other PhD programmes.

This research in context
Over the last 20 years, there have been a significant increase in the number of postgraduate research (PGR) students worldwide (Barry et al., 2018).As the number of PGR students has grown, a need to standardise and capture the PGR experience and training has been of academic interest.These findings are consistent with recent scholarly work concerning the training needs of PhD students.A previous study of postgraduate researchers from a wide range of disciplines demonstrated that a majority of PhD students would like more training to be integrated into their programme of study (Jones, 2017).This accords with an increasing appreciation of the fact that PhD students, upon graduation, will not necessarily enter academia: they will follow a wide range of career paths (Gould, 2015;Jones, 2017 andWoolston, 2017) and so need diverse, transferable skills (Doonan et al., 2018;Jones, 2017).Personal development training has been specifically highlighted in previous work as an area where students felt they had not received sufficient support (Jones, 2017).Training was also identified as a key suggestion for improving the PGR experience by Williams (2019) with 23% of suggestions from 10,303 free text answers regarding programme design, including training.
Despite an interest in identifying the key skills developed by PhD students, the history and development of the PhD concept has resulted in significant programme differences between countries, often making them difficult to compare (Barnett et al., 2017).Consequently, guidelines such as ORPHEUS were developed to define best practice in research training programmes in Europe, allowing for comparisons between countries to become more comprehensive over the last 20 years.Despite this, in the UK, doctoral candidates are considered students and so the undergoing of training is likely expected; whilst the international doctoral candidate may instead be linked with the role of an early career researcher, which can further complicate comparisons in PhD programmes across countries.
The introduction of frameworks such as the Salzburg Principles (2005) and more recent developments have focused attention on doctoral training experiences.More recent concerns around doctoral training are whether the use of taught elements of doctoral training programmes such as training are leading to a loss in the independence expected of PGR students (EUA-CDE, 2010).Additionally, the rise of globalisation and digitalisation has resulted in fresh concerns on how to implement open and ethical research (EUA-CDE, 2016).Also highlighted in more recent years is the rise in the number of PGR students seeking non-academic employment following the completion of their PhD.The EUA-CDE (2016) report highlights the importance of PGR students exploring non-academic settings such as industry through work placements, whilst Hancock (2023) demonstrates that two thirds of PGR students are in nonacademic careers following the conclusion of their PhD.Consequently, many countries and institutions are seeking to reform programmes and implement training for PGR students around transferable skills and professional development, allowing students to seek a wide range of career opportunities (Hancock, 2023).
In France, Durette et al. (2016) conducted a survey of 2,794 individuals who had completed their PhDs to identify the core competencies which had been developed through their PhD training programmes.No competencies were suggested in this survey, to ensure that individuals' responses were not restricted by closed categories defined by the researchers.This survey indicated that transferable skills, including both formalised and non-formalised skills, in addition to technical skills were highly prioritised by former PhD students (Durette et al., 2016).Furthermore, recent evidence has indicated that current training for PhD students needs to be updated to align with the more recent shifts in career outcomes and the post-PhD landscape (Reeves et al., 2022).Although these changes may be required, more research is needed to identify what changes would most support students and how best to implement these into existing frameworks (Reeves et al., 2022).
Although this survey of Wellcome Trust PhD students drew from a relatively small cohort, it is clear the training experiences and needs expressed by participants are reflective of a broader perception of training in research degree programmes.The Wellcome Trust studentships which inspired this research are therefore representative of the practices which constitute doctoral training programmes more widely.

Aims and objectives
The taskforce began by evaluating the training needs of PhD students.This process included identifying key areas of interest, such as gaining professional skills which play an important part of students' developing academic portfolio.The taskforce further laid down a research framework on how to approach research on this topic with consideration of the expected research outputs, culminating in the production of this report.
The first concern of this taskforce was to identify students' current thoughts about training they had received, which may differ between cohorts due to differing years of study and due to the effects of the COVID-19 pandemic (Carusi et al., 2022b).Upon completion of these early discussions, the initial aims for the taskforce were: • To explore the training needs and experiences of staff and students across 23 Wellcome Trust funded PhD programmes.

•
To create a report describing the findings of those training needs and experiences which could be used as guidance to standardise training across all programmes and provide insight into the student experience.

Preliminary research methods
The objectives and initial intentions of the taskforce with regard to data collection were presented at the Wellcome Trust PhD students' meeting in London for first-and second-year students in mid-July 2022.These students were current firstand second-year PhD students from 23 Wellcome Trust funded PhD programmes across the UK.In addition to describing the work of the taskforce to date, the delivery of this presentation also provided a valuable opportunity to engage a wider audience of students in Wellcome Trust programmes.This allowed the taskforce to gain feedback in the form of a Mentimeter 'word cloud' and additionally to recruit further members interested in pursuing these objectives.
During the presentation, the taskforce requested instinctual responses to the preliminary research question, 'how would you describe the training that you have received so far as a postgraduate researcher?'This was intentionally phrased in an abstract manner to attract a range of answers, thereby providing a gauge as to the attitudes of students in each year group to training.Using free access to the Mentimeter website, students were able to virtually submit short answers which were then compiled in a 'cloud' format and shown to the audience during the presentation.All answers were anonymous and thus able to be retained for further reflection.
To achieve the study aims, surveys (see Extended data, (Fawcett et al., 2023b)

Ethics
The research for this publication was conducted under the auspices of the Emerging Research Cultures Project Ethics Approval, and falls under its ethics protocol, as approved by an ad hoc ethics panel on 9th December 2020.The panel was comprised of three independent ethicists as the project did not fall under the remit of any institutional review boards.Written consent was obtained from participants prior to inclusion in the ERC community of practice, with an agreement that collected data could be used for future research up to 5 years following the conclusion of the Emerging Research Cultures project (currently funded until August 2025).The student survey was conducted after obtaining informed consent from participants and under ethics approval for the ERC project.However, as no personal data was collected from participants in the administrator survey, informed consent was not required.
Prior to the release of the survey, all questions were approved by the ERC team.

Surveys
Following the initial students' response on training experiences at the Wellcome Trust event in July 2022, the taskforce developed a novel student survey to further understand; 1) which areas of training students had received, 2) how training was delivered, recorded, and shared; and 3) areas in which students felt they may benefit from further training (Fawcett et al., 2023b).Additionally, the authors of this report conducted a survey of Wellcome No exclusion criteria were used, but one fourth-year student was later excluded from analysis as they had only completed two of the survey questions.No incentives were provided to encourage participation.All contacted students began their PhD studies in September 2021 or earlier.As the survey was disseminated in September and October 2022, all contacted students had experienced at least one year of PhD study and so were able to comment on the training they had received so far.Students who were contacted had prior contact with the ERC Project, which was established to create a community of practice across the 23 Wellcome Trust-funded PhD programmes in 2019, and so were attending an in-person event with the ERC Project (254 total students across both events).
The survey was sent to the mailing list of students and administrators held by the ERC Project, alongside other ERC surveys.Both surveys are unvalidated, but may serve as a potential pilot of this questionnaire.
For ease of recording, the administrator survey was divided into eight distinct areas of training.These eight areas were: 1) Professional development (including presenting or writing workshops); 2) Wellbeing and support (including mental health support and signposting); 3) Equality, diversity, and inclusion training; For each of these categories, administrators were asked to consider the source of the training provided; whether this represented a one-off session; how the training was delivered; the length of the training session; how the training was advertised; and when in the course of study this training was provided.Since the study was designed to allow comparison of the administrators' survey with those of the student survey, the student survey was developed using the same eight training categories.
To adhere to General Data Protection Regulations (GDPR) (European Parliament and Council of the European Union, 2016), the taskforce did not request that participants identify the Wellcome Trust programme on which they were enrolled during the surveys.However, it was necessary to ascertain the year group of students to analyse the appropriateness of the timeframe in which training occurred.Additionally, to comply with GDPR, information such as gender and age were not collected by the taskforce, meaning that genderspecific analyses cannot be performed.The administrator survey also requested that the specific programme be identified for purposes of comparison, although this did not involve any personally identifiable information.For the survey questions, please see Extended data (Fawcett et al., 2023b).Where information about the programme name was gathered, this was pseudonymised.This data was processed and stored according to the ethical approval of the ERC Project, on secure servers subject to UK law.Additional information was not collected as this did not fall under the prior ethical approval sought by the ERC Project.
The student survey investigated the availability, effectiveness, and knowledge of training opportunities, amongst other points of interest.Students were asked to rate the usefulness of types of training and the appropriateness of the format, whether it be through a workshop, lecture, or online session.It was hoped this would identify any training considered extraneous and highlight potential means of improving engagement with training.The student survey was distributed via an email which included a link to a Microsoft form, where the questions were listed (Fawcett et al., 2023b).
Questions were presented as a five-point sliding scale or a 'yes' or 'no' format, to ensure for time efficient completion of the questionnaire and efficient data analysis.Additional free text questions included were: "What training modules did you find most useful?","Are there any aspects of the PhD project or life as a PhD student which you do not feel equipped to manage?", and "What training would you like to be offered?"These questions sought to better understand the student experience and engage with student opinions as how to potentially enhance training and so free text answers allowed students to fully answer these questions.
Potential gaps in the training provision for PhD students were considered a significant investigative aim of the survey.Preliminary findings were obtained through interrogation of the Mentimeter word clouds, resulting in specific enquiries into the provision of training and its usefulness using multiple-choice questions included in the student survey.To hear students' own voices, some free text answers were included so that the survey might benefit from a wide range of opinions on training needs.Gaps were identified through questioning firstly which specific types of training students had found useful, allowing students to indicate where they had not received training in a particular area (Fawcett et al., 2023a).This was then supplemented with a follow-up question asking students to rank types of training according to usefulness, which demonstrated the priorities accorded by students to different training types.The survey was divided into a series of sections relating to training: awareness, scheduling, content, format, usefulness, recording, and overall satisfaction.

Analysis
After survey completion, the data was anonymised and analysed.Analysis was conducted by the data analysis action group using R (version 4.1.0)and RStudio (Build 554).In the student survey, quantitative questions with multiple choice responses were analysed by tallying the number of responses for each outcome and plotting these as bar charts, also producing by year group bar charts.Bar charts were generated using the R package 'ggplot2' (version 3.3.6)and were used to identify data trends.Open questions were analysed by identifying and extracting the key themes, categorising these themes, and plotting them as bar charts.This allowed for answers to open questions to be explored via shared themes to identify the key concerns and interests of students.
The administrator survey included only quantitative data and was analysed by tallying the number of responses and producing bar charts to demonstrate trends across different programmes (Fawcett et al., 2023a).This allowed for the taskforce to report on areas of training delivered by the programmes, in addition to how and when this training is delivered.

Preliminary research results
The number of students who responded to the preliminary research question 'how would you describe the training that you have received so far as a postgraduate researcher?' differed by year group, with 60 first-year students recording their responses, whilst 43 second-year students had their response recorded on the Mentimeter word cloud.The number of individuals at the second-year event was 75, including Wellcome staff, whilst there were 105 attendees at the first-year event including Wellcome staff.A similar rate of non-response was shown amongst both year groups, with a response rate of 57%.The reasons for non-response are unknown as information about non-respondents was not collected.
Responses, across both year groups, were varied and contained approximately equal numbers of phrases with connotations which were positive, negative, and neutral (Extended Data, Fawcett et al., 2023b).First-year students submitted 23 positive terms, 24 negative terms, and 24 neutral terms.Second-year students submitted 17 positive terms, 18 negative terms, and 25 neutral terms.Note that this does not include terms which were repeated in submissions, indicated by an enlarged font in the word cloud.Some examples of positive terms included: 'brilliant', 'thorough', 'supportive', or 'helpful'.Other responses were negative, such as those which mentioned a sentiment that training was 'not rigorous', 'unstructured', or 'sometimes irrelevant'.A great many responses also included neutral phrasing such as: 'fine', 'interdisciplinary', or 'sufficient'.These responses indicated that whilst there are areas of training students felt happy about, there were areas of training which could be improved and these could be identified through a more comprehensive survey of student experiences regarding training.Concerns raised about the advertising of training activities and access to training information raised an issue not identified in the original discussion groups (see Figure 1).This issue therefore requires further investigation as it may suggest that students are not aware of training opportunities available to them at their institutions.
The software used for this poll, Mentimeter, changes the size of words in the word cloud to represent the number of times that word was submitted by the audience.Figure 1 shows that more than one first-year student felt that their training was 'varied', 'good', 'challenging', 'broad', and 'relevant'.However, this does not suggest that the overall experience of first-year students was positive, as there remains an approximately equal incidence of positive, negative, and neutral word choice.Similarly, Figure 2 shows that more than one second-year student submitted the words 'unstructured', 'inadequate', or 'good', but this does not mean that second-year students reported a generally more negative training experience.Across both cohorts, the Mentimeter word clouds revealed that students had very varied responses to the training experienced.
Immediately following the presentations at the event, the taskforce engaged in smaller group discussions with students interested in contributing to the proposed research.There was a high level of interest from attendants at the Wellcome events, with several students attending a drop-in session to find out more about the taskforce and its goal.This allowed for input from non-taskforce members about their concerns and queries around training and provided the taskforce with some new areas of interest when conducting their research.It also showed the importance that students felt this topic had.
Use of the word cloud format ensured a broad range of opinions could be reflected in the exploration of training  experiences as the taskforce engaged in further research.It aided in refining the questions included in the student experience survey to investigate this topic in further depth.Furthermore, the Wellcome Trust events demonstrated the engagement of students with this topic and its importance to their future as researchers, in addition to highlighting differences between the training experiences of year groups, which were of later interest to the taskforce.

Student survey
Responses were obtained from 35 Wellcome Trust funded PhD students in different years of study at institutions across the UK, in addition to 10 administrator surveys (Table 1).As the institution of the participant was not recorded for GDPR reasons, it is unknown how many of the 23 UK based Wellcome Trust programmes were captured.Although the same survey was sent to all participants, some chose to answer the survey only partially.During analysis, missing data was not considered and instead each question was analysed separately, with all students who answered a specific question being included at each stage of analysis.Second-year PhD students represented the largest number (15/35) of the respondents of the survey conducted.However, it is important to note that the survey was distributed in September and October 2022, during the start of the new academic year: as such, querying students about their year of study could yield different results depending on when within that period they completed the survey.Additionally, there was a high level of non-response, even excluding the members of the taskforce, as a total of 180 first-and second-year Wellcome Trust students had attended the Wellcome Trust events in July 2022 and at least 254 students had been sent the survey.However, the reason for this cannot be determined as information about non-respondents was not collected.
The following section will discuss the results of the student survey (see Underlying data) by training section.

Awareness of training
One of the key areas of interest for the taskforce was student awareness of training opportunities.Although most participants agreed that they were informed about the upcoming training events, 18% did not know where to find information on upcoming training opportunities (Figure 3) (Fawcett et al., 2023a).Fewer participants in Year 3 of their studies knew where to find training than in other year groups.
When asked whether their Wellcome Trust PhD programme made them aware of training opportunities, only 54.3% (19/35) agreed with this statement.For first-and second-year PhD students, there was an approximately 50% split in responses, whilst 72.7% (8/11) third-year students reported that their programme had made them aware of available training opportunities.Overall, the data shows that most of the students participating in the survey were informed of the training opportunities available and knew where the find the relevant information.

Scheduling of training
A range of responses were obtained from students when questioned on whether their training was delivered at an appropriate time for use in the PhD project: responses were broadly positive and consistent across year groups, with the most popular response being 'agree' (Figure 4).Second-year students had a marginally higher number of negative answers when compared with other year groups.
Although most students felt that the timing of training within the programme was appropriate, 57.6% (19/33) of survey participants stated that training was not repeated for students who have scheduling conflicts.This experience was generally consistent across the different year groups.
Overall, the findings show that most people agreed that training was provided at an appropriate stage in their studies but raised the need for regularly repeated training sessions to account for scheduling conflicts.

Content of training
Students were asked to reflect on the usefulness of training in the eight previously identified categories individually, with a later question asking students to compare and rank the usefulness of different types of training.When considered independently, technical and cohort-specific training were both almost universally considered to have been useful across all year groups.More specifically in regards to technical training, the vast majority of surveyed students found the training to be useful with only a single respondent finding it to be 'extremely not useful'.Similarly, most students noted that cohort-specific training had been useful.
In contrast to an apparently high uptake of these forms of training, comparatively few students reported having received training on equality, diversity, and inclusion reflected by a 'not applicable' response to the question of usefulness.Of 33 respondents, half had not experienced this type of training.However, the students who had received equality, diversity and inclusion training generally found it useful when asked to discuss this form of training alone, (page 18, Taskforce on Training -Student Questionnaire Analysis, Fawcett et al., 2023a).
In addition to being asked about the individual usefulness of the eight categories of training, students were also asked to rank them in order of helpfulness (Figure 5).The overall results show that the most helpful category of training was technical, with the second most commonly chosen category being professional development.The least helpful training was commonly reported to be wellbeing training, and equality, diversity, and inclusion training.Third-year students displayed a more diverse range of the most popular training category than other year groups, but still indicated that technical training was the most useful.It is notable that this question did not allow students to select 'not applicable' for any  area in which they had received no training, as was possible with regard to the previous question on usefulness wherein equality, diversity, and inclusion training was more positively regarded.
Other forms of training saw similarly limited provision.Across the remaining five categories, a high proportion of respondents selected 'not applicable' answers regarding the usefulness of this type of training: 28.1% (9/32) for engagement and outreach; 31.3%(10/32) for both professional development and personal development; 37.5% (12/32) for wellbeing; and as many as 53.3% (16/30) for career development.Where students had received training in these categories, their experiences were generally positive, with very few responses indicating that the training had not been helpful.In addition, 3 of 32 respondents found that personal development training had been 'somewhat not useful', 2 of 30 respondents stated the same for career development training, and one single response suggested that wellbeing training had been 'somewhat not useful'.Overall, results suggest that some students may not have received training in these areas, but the majority of those who had, found that training was helpful.
When asked to answer an open-ended question relating to the most useful training modules experienced across all year groups, subject specific training was overwhelmingly held to be the most useful, with 53.3% of respondents indicating that this was the case.Other modules mentioned included: coding, ethics, project management, presenting, public engagement, science communication, writing, or simply 'all' available training (Figure 6).A majority of students (62.5%) reported that they had used skills learned in personal development training during the PhD, but a substantial remainder felt that they had not yet done so.
Students were given an opportunity to provide open answers to the question 'are there any aspects of the PhD project or life as a PhD student you do not feel equipped to manage?'.20 of the 33 respondents chose not to answer this question, while 3 stated that there were no concerns of this nature.Of the remaining 13 who answered 'yes', 11 gave detailed answers (Fawcett et al., 2023a).These responses raised a range of different issues.3 students noted challenges relating to their wellbeing, which included struggles with 'loneliness' and 'self-doubt'.

Format of training
Training was reported to be generally compulsory, with 23 of 33 respondents stating this is the case.Additionally, training was considered to be a good use of time and resources by 75.8% (25/33) of students participating in the survey; a further 15.2% responded that this was not applicable, and 9.1% disagreed with this sentiment.
Students were asked to indicate their preferred mode of training and responses varied across the available options, which included: group discussions and activities; hands-on tutorials; lectures; mentoring; online training; required reading; and workshops (Figure 7).The most highly favoured delivery methods were workshops, hands-on tutorials and group discussions and activities, whilst the least popular modes of training were required reading and mentoring.Most students felt that training was delivered in an appropriate format, with 22 of 33 respondents selecting the answer 'agree' in response to the statement 'I feel the format of the training is generally appropriate to the subject matter'.Only two respondents disagreed with the statement.
Overall, the findings for this section were that most students felt that the delivery method used for training was appropriate for students and training was mostly reported to be compulsory.Students also reported that training was considered a good use of time and resources.

Overall satisfaction
Students were finally queried as to their overall satisfaction with their training during their PhD programme.Firstly, they were asked whether the training provided to them prepared them to become a good researcher, with 87.5% (28/32) respondents agreeing or strongly agreeing with this statement.This statement was only disagreed with by 4 students, all of whom were in second-year.Students were then asked whether their PhD training had prepared them for life outside of academia, with 75% (24/32) of students agreeing or strongly agreeing to this statement.Again, second-year students made up the majority disagreeing with this statement.
Students were asked as to their main objectives when engaging with training in an open-ended question.Most students reported 'learning relevant skills' to be their main objective, although other themes identified were: becoming a better researcher, open-mindedness, and personal development.Finally, students were given an opportunity to provide any final comments on training.Only five students responded, with three suggesting that their training had been good but incomplete, one reporting that they would prefer more in-person training, and another student suggested that sharing training course information across Wellcome Trust-funded PhD programmes would be beneficial.
In summary, students reported that they were satisfied with their training and felt it will prepare them for life in research.
Students also reported wanting to learn new and PhD-relevant skills when undergoing training.However, second-year PhD students reported that they did not believe that the skills they were learning would be transferable outside of academia.

Administrator survey
As part of this project, a survey presented in a Microsoft Form was sent to PhD programme administrators across 23 Wellcome Trust-funded PhD programmes across 15 UK-based institutions.Ten responses were received and will be described in this section (Fawcett et al., 2023a).
Administrators Lastly, administrators were queried about how training was advertised to students, which included being part of the programme syllabus, being included in emails, newsletters, or the programme calendar (Figure 8).Information about training across these categories was delivered using a range of methods, with dissemination through email being the most common response, followed by inclusion in a newsletter.
In summary, the output from the administrator survey was highly variable by programme, demonstrating a lack of consistency between programmes in delivering training to students.Equality, diversity, and inclusion training were often reported to vary between programmes, more so than other training categories.

Discussion
Interpreting the survey findings Firstly, it should be noted that whilst this study identified some areas in which training provisions may need to be reconsidered to meet PhD student needs, it was agreed by most participants that training was a good use of time and resources.When reporting on training previously received, respondents appeared to recognise that provision had been generous and beneficial.This  or belonging to ethnic minorities who are at higher risk of experiencing poor wellbeing (Creaton, 2021).
Furthermore Additionally, despite known evidence that non-academic careers are increasingly attracting PGR students, career training was ranked similarly to equality, diversity, and inclusion training and wellbeing training (Figure 5).Interestingly, it was perhaps lowest ranked by third-year students, who may be in most need of considering their future options as they approach the end of their PhDs.However, similarly to equality, diversity, and inclusion training, career insights may be better suited to an alternative metric of measurement e.g., through placements, workshops, or wider communication as this may allow PGR students to identify previously unknown career paths.who identified that peers can help build a 'researcher' identity through the sharing of experiences and perspectives including through the use of online platforms.Mentoring has also been highlighted as important in the development of a scientific identity, which has been linked to a researchers' career trajectory, although this mentoring mostly focuses on the student-supervisory mentoring relationship (Pfeifer et al., 2024).
Training scheduling was generally reported to not be repeated, although there was no provision for students to say whether they had access to recordings of missed training at a later time.
Although methods such as recording training can allow students to still access the material, it can affect the level of engagement students have with the content and may also affect the behaviour of the recorded students.Furthermore, if training is not repeated to account for scheduling conflicts, this may adversely affect specific groups of students due to the timing of training (e.g., all day events).This includes PhD students who may be carers, parents, suffer from chronic illness and others working more flexibly.Care must therefore be taken when arranging training for consideration of these students and ensuring that they have an opportunity to attend live training sessions.However, no individual circumstances were assessed during the survey to establish whether there had been specific difficulties in accessing training.An important consideration here is the increased use of virtual training methods during the peak of the COVID-19 pandemic (Bashir et al., 2021) which may have impacted year groups differently.For example, students beginning their PhDs in 2020 are likely to have received the majority of their first-year training online, whilst cohorts which began earlier or later may have had more opportunities for in-person training.COVID-19 also had known effects on the mental health of PGR researchers, contributing to the mental health crisis amongst researchers (Creaton, 2021) and may have been mediated by additional wellbeing support for PGR students both at the time and more recently.A survey of early career researchers in 2020 also demonstrated the impact of COVID-19 and the lockdown, with 80% of 5,000 surveyed individuals showing symptoms of mental distress (Byrom, 2020).No specific questions regarding the impact of the pandemic on PhD training were included within the survey, though future research should consider how this may account for the differing experiences between year groups.
Students also described their experiences of keeping a record of their training, with most students stating that they had access to a university-maintained training record.However, there are year group differences in whether students feel they can access those logs, with first-year students more commonly reporting that they can access their logs and lower rates for second-and third-year students.This may be due to programmes changing how this information is conveyed to new students or may be due to over-confidence on the part of students who have not yet actually attempted to access their training log.
When students were queried about their overall experiences of training, most identified their main goal of undergoing training was to develop relevant skills, although other reasons were identified.This links to the preference for technical training, as this would be classed as skills relevant to their PhD project.Additionally, the survey found year group differences when asked whether this training prepared them for life inside or outside academia.Dissatisfaction with these statements was highest in second-year students, which may be due to considering these skills but not yet having received training to support them e.g., careers training, which was not reported in the administrator survey to take place specifically during first year.This may affect students' perception of their readiness for the future beyond their

Other considerations
The results from this survey show a range of experiences from Wellcome funded PhD students, identifying areas where students feel well supported but also areas where they feel more training is required.However, it is important to note that only a total of 35 students responded to the survey, from a total of at least 254 students who were sent the survey and 10 programme administrators across the 23 programmes.This shows a significant level of non-response to the survey, which needs to be considered and addressed in future work.It may also be the case that the sample was biased as survey respondents may have had a pre-existing interest in PhD training.Moreover, the small sample size can make it difficult to draw robust conclusions by year group, although it is possible to report observed trends.The low sample size could have been due to survey fatigue as the Taskforce on Training survey was one of several surveys disseminated at the same time by the ERC Project, which may have caused fewer students to complete all surveys.Due to a tight timeline, the study was limited in addressing uptake and future work should consider allocating more resource and a longer timeline to explore this area.Furthermore, as the surveys do not report which institution the student was attending, this may also impact the results as the administrator survey indicates variability in how different programmes deliver training.Consequently, without data regarding which programme a student is affiliated with, it is difficult to link specific administrator outputs with student outputs.
In addition, the timing of survey dissemination further complicates this issue as it was administered across September and October, during a time when many PhD students are transitioning between year groups.This means that students in the same cohort, depending on when they completed the survey, may have reported to be in different year groups.
It was intended that Year 1 students would have started their research in September 2021, Year 2 students would have started their research in September 2020 and Year 3 students would have started their research in September 2019.However, students who completed the survey in September 2022 may have reported themselves to be Year 1 students, whilst students who completed the survey in October who had started their programme at the same time may have identified themselves as Year 2 students, which may add confusion to interpretations around specific year groups.In future surveys it would be more useful to ask students to select the start year of their PhD, rather than their current year of study.This would also provide additional clarification as some students may have had periods of absence during their PhD and so their year of study and start date may differ.Furthermore, as surveys can make it difficult to obtain more detailed information regarding the experiences of PhD students, future work into this area could also include the use of interviews with participants to obtain a more comprehensive understanding of the PhD training experience alongside these survey results.
Furthermore, although a survey was used, the taskforce recognises the limitations of such methods, particularly in determining detailed experiences of PhD students.Surveys may have high levels of non-response, which appeared to be the case in this report, with only 35 student responses.This was a novel questionnaire developed by the taskforce but was not validated prior to use.However, this study could be used as a pilot to validate and improve the use of this questionnaire for future research into training provisions.This survey was also only targeting students on a UK-based PhD programme, where doctoral candidates are considered students, so the findings may not be applicable to other countries where the role of a doctoral candidate is broader and more similar to that of an early career researcher.
Although the survey findings of Wellcome Trust students and programme administrators have been recorded here, staff who are supervising these students were not surveyed.To further tie together information about the training received by Wellcome Trust funded individuals, a survey of supervisors should be conducted to determine what training supervisors have received to support students with their areas of concern and any support that supervisors themselves might require.Additionally, supervisors may be able to report on gaps they had identified through supervisory meetings with students e.g., students who report a lack of understanding or training about a specific area during their supervisory meetings.However, it could be difficult to determine this as supervisors may have multiple students in different year groups which can make untangling the relationship between year group and training provisions difficult.Furthermore, it would depend on students self-reporting problems to their supervisors and this may depend on the student-supervisor relationship and the confidence of the student to report such issues.Despite this, a survey of supervisor experiences may further enhance this study because previous research has suggested that supervisors can play a key role in supporting a student's mental health during their PhD and so should be provided with additional training for this (Mackie & Bates, 2019).The role of supervisors was also highlighted by Bégin and Gérard (2013), who reported that supervisors play an important role in helping students to develop the skills needed to become a research specialist.This may particularly apply to technical skills, but supervisors can also play a coaching role in the development of wider transferable skills.
Consequently, future work into this area may include validation of this questionnaire, surveying supervisors to obtain their views on training, use of interviews to obtain more detailed information, and a need to promote response rates from the PhD students surveyed.

Recommendations
Despite the limits of our report, the taskforce has identified the following recommendations for the Wellcome Trust regarding the training of PhD students:

Data availability
Underlying data Figshare: ERC Taskforce 3 Underlying Data https://doi.org/10.6084/m9.figshare.c.6765435.v2(Fawcett et al., 2023a) This project contains the following underlying data: • Data file Taskforce on Training -Student Questionnaire Analysis (R analysis of student survey output) • Data file Taskforce on Training -Administrator Questionnaire Analysis (R analysis of administrator survey output) • Data file Taskforce on Training -Raw Data (Excel file containing the raw data from the student survey used for later analysis) • Data file Taskforce on Training -Raw Data from Administrators (Excel file containing the raw data from the administrator survey used for later analysis)

Extended data
Figshare: ERC Taskforce 3 Extended Data https://doi.org/10.6084/m9.figshare.c.6765441.v6(Fawcett et al., 2023b) This project contains the following extended data: •  There are a few minor issues the authors still need to address.I detail these issues below.I suggest the authors revise the manuscript and check for typos and other minor edits.They must ensure there are no missing or repeated words.

1.
Methods.The author(s) do not report the total number of staff administrators participating in the preliminary study.They state reaching out to the main point of contact for the 23 Wellcome Trust PhD-funded programmes.These programmes are hosted in 15 UK-based institutions.They also state both ERC events were attended by staff members who participated in the study, but they did not disclose how many.Furthermore, there is no information on how many staff administrators were contacted and asked to participate and/or collaborate in the study (either in its preliminary stage or the survey).

2.
Results.I suggest the authors use horizontal bar charts for some of the figures and see if this option helps increase their readability.It is sometimes difficult to be sure to which bar the legend concerns.In some other graphs, the series numbers do not show.

3.
Results.When reporting the results from the Administrators' survey, the authors often use the expression "the programmes reported/reporting".Programme administrators and not the programmes answered the questions in the survey.

4.
Discussion.The authors tend to focus on the specific topic they are discussing (e.g.content of the training, scheduling of the training).The discussion would gain depth if the authors cross-referred the findings from the different types of questions from the administrators' and the doctoral researchers' surveys.Illustrative are the apparent contradictions in the doctoral researchers' answers to the training on career and personal development, wellbeing and equality, diversity and inclusion.When reflecting on these answers (as the authors did) several explanations can be found.A better understanding of their meaning can be achieved when adding the authors' discussion about the administrators' answers.In combination, this shows different priorities, understandings, needs and concerns on the part of the doctoral researchers and the administrators.It also allows us to raise the question -as the authors did -if training is always the best option to address the specific needs of doctoral researchers, namely regarding wellbeing or career development.

5.
Discussion.Based on the results reported, doctoral researchers attending different years of the PhD programme expressed also different views on the usefulness of the training they received to date.This is an interesting finding that could merit some additional considerations on the part of the authors.I understand the need to be cautious based on what they discuss in the section on Other Considerations (Limitations?).However, I would have liked to have read more on their take on the finding.

6.
Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Research careers, early career researchers advocacy and policy, career guidance and development I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
Version 2 1. Introduction 1.1.The present study has at its background the notion of research culture(s).In the introductory paragraph of the manuscript, the authors provide an overview of the concept that does not adequately reflect its depth and complexity.Over the years, several organizations or movements, such as the San Francisco Declaration on Research Assessment (DORA), the Wellcome Trust or the Coallition for Advancing Research Assessment (CoARA), have addressed the topic in statement positions, reports, or other policy-oriented documents.Some of the leading members of these organizations or movements also published journal articles focusing on the topic of the contemporary (hypercompetitive) research culture and the need for change.I suggest the authors look into these publications as they will help them better frame their study.1.2.It struck me as odd that at no point in the introductory section of the manuscript, the authors attempted to discuss the notion of doctoral training: What are its scope and aims?How has it changed over time?What are the main challenges doctoral training currently faces?For some time now, these questions have been framing discussions around doctoral training and important documents such as the Salzburg Principles, which have been updated several times since its publication in 2005, have provided insights into discussions and directions to consider when reflecting on the issue.1.3.The section on Aims and objectives includes information that is not strictly about the study aims or objectives.The authors included information that is about the procedure and that should be in the section on the Method.Also, how the section is structured obfuscates the aims of the study, which should be the first thing to be presented to the reader.These aims should be clear and provide a background to all the steps the authors took when designing and implementing the study.
1.4.I suggest the authors rethink the structure of the manuscript.This will help remove some redundancy from the text.It will also help clarify the study aims and objectives as well as the study design and procedure.Specifically, what the authors have called "Early findings" appears to be an exploratory or pilot study whose findings were used to support the design of a following study.My suggestion is that the structure of the article (i.e., how the different sections are presented and the information included in each section) better reflects this two-step approach to the study.

Method
2.1.The authors opted to use a narrative approach to their description of the study method that does not consider the typical subsections in a section on the Method (e.g., participants, instrument(s), data collection or analysis procedures).I suggest the authors reconsider the current structure of the section on the Method and better organize the information to increase clarity on what they have done, how and why.Important information about the different components of the method, for example, who the study participants were, is scattered across the sections, which makes the reporting hard to follow.What data the authors collect and how?How did the procedure for collecting and analysing the data address the study aim(s) and/or research questions?What instruments did they use to collect the data and which target population?2.2.The procedure the authors used to collect the data from the doctoral researchers differed from the procedure they used to collect the data from the Ph.D. programme administrators not just in terms of the type of data collected but on the instruments the authors used -at least, this is what I understood from their reporting.Regarding the type of data collected, the authors claim GDPR compliance restricted their ability to collect certain information from doctoral researchers.Wouldn't anonymization of the data (which they did with the data they collected from the administrators) address the issue?No clear explanations are provided about why different instruments were used to collect data from the administrators and the doctoral researchers.
Considering one of the aims was to compare doctoral researchers' and programme administrators' results, it makes no sense to me why paired data and other sociodemographic data (e.g., gender, Ph.D. programme attended, discipline or field of research) were not collected from both groups of participants.The information the authors chose to not collect from the surveyed doctoral researchers would also have been important to understand partial responses and missing data patterns -as the authors themselves admit.This is an important limitation of the study that requires additional consideration and reflection on the part of the authors.Furthermore, their decisions must be better explained, including the decision to send the survey at the start of the academic year and to include first-years doctoral researchers.

Results
3.1.Overall, the section on the Results is clear and well-structured.The authors reported differently the results from the doctoral researchers' and the administrators' surveys.Perhaps this has to do with the fact that there were fewer answers, and these answers were scattered, as the authors highlight.However, no section comparing the doctoral researchers with the administrators' answers is included.The authors stated this as one of the study's aims.What were the main similarities and differences observed in the answers the doctoral researchers and the administrators gave to the survey?To what extent are these answers/the results comparable?3.2.Other minor issues: -When reporting doctoral researchers' perceptions on the usefulness of the training they received, the authors discuss their results by relating them to the results of another study.I suggest all discussions are left to the section on the Discussion and in the Results section only the study results are reported.
-Sometimes it is hard to read to which category each bar in the graph corresponds.Perhaps using a smaller size in the font would help increase readability.

○
There was no mention of this, but I wondered if there was any incentive provided to the survey respondents.
○ I would love to see a table that summarizes the response rates in addition to being embedded in the text.This will make it easier for the reader to get a quick snapshot of the total population of students and administrators surveyed.

○
Please clarify in the text how you are defining 'mentoring'.When I think of mentoring, my first thought is faculty or advisor mentoring and in this manuscript, it seems that the focus is on peer mentoring.I would invite the authors to define this more precisely.

○
This work is in part motivated by the desire to understand the ongoing negative impacts of the COVID-19 pandemic, yet there was no real discussion about it except in relation to virtual training.I agree that future research should ask students to identify the year that they started their Ph.D. but I also wonder if it would be possible to define the years of each group (Years 1-4).This might indicate if these students were pre-or post-pandemic or both.

○
One thing that I continued to question while reading the results and discussion was the inclusion of Year 1 students in this research.To what extent are students being oriented and exposed to training opportunities prior to the start of their first year?Would Year 1 students be able to see the full range of training that will be available to them during their programme?I ask because there is an expectation being made in this study that these Year 1 PhD students are going to be able to respond to a survey about their needs and training experiences.It is indicated on page 7 that this survey was distributed in September and October 2022, so I, as a reader, have to assume that Year 1 students did in fact start this same year.While these new students would be able to speak to their needs, I am not sure how much information they have regarding the opportunities that will be made available to them.
○ I would agree with the first reviewer to be cautious about your findings regarding EDI training.It seems that the survey did not allow students to rank the training that they had engaged in.As a result, the findings still seem to suggest to me, even after the first revision, that students do not value EDI training.I do not feel that you have the evidence to make that claim and would recommend stating that additional research is needed.
○ I hope that these comments will be useful and invite any questions if any of the above is unclear.Overall, the manuscript is well written and interesting, and it discusses a topic that may be relevant beyond Wellcome Trust programmes in improving doctoral education.Yet in reading the manuscript I noticed elements that raised important concerns which would need to be addressed by the authors.I believe that the manuscript should be revised substantially to address or clarify such concerns.I will go through these concerns in the order that they appear in the manuscript.

Is the work clearly and accurately presented and does
First, in the introduction, the authors describe the findings from Mentimeter word clouds they captured during a taskforce meeting.As a first, yet minor, thought on this point, it would be useful if the colour scheme used in the word clouds could reflect the positivity/negativity/neutrality of the terms, although I understand that this may not be possible to do with fixed Mentimeter results.But my main concern is in the way the word clouds were interpreted by the authors, and I believe that the way they are described needs to be revised to adequately reflect the data embedded in the word clouds rather than a superficial interpretation of 'bigger words'.In fact, in describing the word clouds, the authors focus on the few 'bigger words', disregarding the smaller words inputted in the word clouds.Having used Mentimeter in the past, I know that 'bigger' words are sometimes bigger only from 2 or 3 responses using the same term.To be able to judge the positivity/negativity of the students' input, the full data (i.e., all words and frequencies) should be used.For example, at first glance, I counted approximately 31 negative terms and 26 positive terms in Figure 1.Adapting for the few bigger words and assuming these embedded 2-3 answers, it still means that the authors' interpretation of Figure 1 as an obvious positive impression may not be justified.I therefore suggest that, if the authors wish to use the word cloud to introduce their work, they should use the meta-data received with these word clouds to interpret their meaning.
In the methods section, it would be very useful if a copy of the survey was included.I could not find the survey in the online material and making the survey available would without doubt help clarify (although not solve entirely) a few of the points I raise below.I also noted the several use of the adjective 'novel' to describe the questionnaire, and wondered whether a less value-laden term could be substituted for the overly-used and badly-defined term 'novel'.
Later in the manuscript, the authors explain that "Potential gaps in the training provision for PhD students were considered the most significant investigative aim of the survey…" (p.6 in the PDF version), yet these gaps do not seem to be clearly addressed nor covered by the survey questions.
The authors add that the word clouds and text answers helped to identify these gaps, but it was not clear to me where responses really helped identify gaps vs. where they identified impressions and feedback.I believe that this information is in the manuscript, but I would encourage authors to make it more clear where these gaps were identified in the responses of the survey throughout the manuscript.
Still on p.6 of the PDF manuscript, when introducing the respondents, the authors mention that "Responses were obtained from 36 Trust funded PhD students in different years of study at institutions across the UK", yet it is not until the 'Other considerations' section on p. 14 that the reader is aware of the total cohort and, by consequence, the response rate of the survey.I believe that this response rate should be mentioned already on page 6.
On page 9, figure 6 is identical to figure 8 on page 11.Both figure contain an irrelevant title at the bottom "Most to least preferred mode of training" since both refer to topics rather than to teaching modes.This should be rectified and the missing figure should be added.
In interpreting Figure 6, the authors also introduce conflicting information about equality and diversity training, stating that "Of 33 respondents, exactly half had not experienced this type of training.However, the students who had received equality and diversity training generally found it useful."Later in the same section, the authors then state that "The least useful training was commonly reported to be wellbeing training, and equality and diversity training."Only to mention on page 10 that "Overall, the data shows that students were lacking in training on equality, diversity, and inclusion, although those who had received this training found it useful."These contradicting findings are very confusing and it would be important for the authors to clarify this, especially in the section 'Content of Training'.Specifically, it would help if the authors could clarify where the information about the question in which "students who had received equality and diversity training generally found it useful" is captured.Is this from the survey?Is this in an open text question or is it from somewhere else?It seems important that the authors clarify this information and make sure it is reflected throughout the manuscript (including abstract, discussion, and recommendation).If contradictory information was obtained this contradiction should also be explored further and more explicitly.Considering the points above, the summary paragraph before the section 'Administrator survey' and sections of the abstract and discussion may also need to be adapted to correspond to the clarifications.
On page 11 of the PDF version, the authors announce Figure 9 in a section that does not correspond to the elements displayed in figure 9.I believe that figure 9 should be referenced in the penultimate paragraph before the Discussion section.
In the Discussion section, several conclusions should be revised to better reflect the findings more accurately.For instance, the authors state that "it was nearly universally agreed that training was a good use of time and resources" where this appears to have been agreed by 75.8% of respondents -if I am not mistaken -a high proportion but not one I would qualify as 'nearly universally'.A bit later, the authors reinstate that "wellbeing training was not prioritised when students were asked to rank the helpfulness of various types of training: it was consistently rated as the least useful" whereas according to Figure 6 EDI training was also rated as the least useful by as many respondents as wellbeing training (see point on EDI above).A few other instances of the discussion section appear to paint a positive outlook that is not entirely reflective of, or not easy to tie to, the results presented.I would thus recommend that the discussion should be revised to clarify the elements mentioned and to transparently reflect the possible drawbacks and possible conflicts in the responses.On a more general note, I would also have liked to see further comparisons with past literature on the training needs of PhD students, and more discussion of the meaning of this work in a broader sense.The manuscript only refers to two references in the discussion section, both seemingly about the impact of COVID in teaching, but I am sure that several studies could provide useful information to complement and enrich the findings from work.
Finally, the section about 'other considerations' is very helpful, but I had the impression that the authors focused this section mostly on the reasons for the small response rate rather than on what the small response rate may mean for the data.To ensure that the readers understand the strength and limitations of the work, it would be important to expand further on what the data may provide and on where used of these findings should be mindful.For instance, it seems valuable to bring back the issue of the survey not being able to identify the institutions from which the respondents came and to explain whether this impacts the conclusions given the diversity in training and training delivery between institutions.In addition, the authors mention the lack of validation for the questionnaire, suggesting that the study may be used as a pilot for the questionnaire; this point is not reflected anywhere else in the manuscript but may benefit by being added in other sections, especially in the methods section.
I hope that my comments will be useful to the authors in revising their manuscript.I would welcome them reaching out via the comment section if any of my concern is unclear.

Kathryn Sandilands
Thank you to the reviewer for your time and comments.We have found this very helpful and have made amendments to the article to address the minor improvements suggested in the peer review.The amendments addressed in turn below: Unfortunately, we are unable to alter the colour scheme of the Mentimeter word clouds to clarify the designation of 'positive', 'negative', or 'neutral' submissions.As regards the interpretation of the relative size of the words represented in the word cloud, we have included some additional explanation indicating that an enlarged word within the cloud indicates repeated incidence in submissions and have contextualised this within the overall findings regarding the nature of the phrases submitted by students.We have also quantified the 'positive', 'negative', or 'neutral' responses for each word cloud within the report, as well as adding these quantifications to the Figshare collection.An additional citation has been included to enable ease of locating the student and administrator surveys in the Extended data files (Fawcett et al., 2023b).Where appropriate, the word 'novel' has been omitted or altered to better reflect that the survey had not previously been piloted.This is further discussed within the article on page 14 under 'Other Considerations'.We have clarified that the use of the Mentimeter word clouds were intended to provide overall impressions of student perspectives, while multiple-choice questions permitted more targeted enquiries and free-text responses were encouraged to allow students to elaborate on these issues.The ordering of questions and their different purposes have been added to the methods section to explain how gaps in training were identified from student responses.The response rate to the survey has been more prominently highlighted in the methods section of the article so as to provide context to the reader when approaching the results which follow.The total number of students contacted is now noted at this earlier stage.This is discussed in more detail in the section entitled 'Other Considerations' at the end of the article.Corrections have been made to Figure 6 and Figure 8.The axis of Figure 6 has been edited with a more appropriate title.Figure 8 has been updated to include the correct data and legend as described in the text.
Both have been updated within the Extended Data files on Figshare, with the collection amended to reflect these changes also.The reviewer noted that there appeared to be a contradiction in the results regarding EDI training.This has been clarified within the article: two questions within the survey asked respondents to reflect on the usefulness of EDI training.In the first, respondents were asked about EDI training independently and half of those respondents submitted a 'non applicable' response, indicating that they had not undergone EDI training, while the remainder answered either 'extremely useful', 'somewhat useful', or 'neutral'.In the second question, respondents were asked to compare different forms of training and rank them.No option of 'non applicable' was offered, which meant that EDI training was ranked as one of the least useful forms of training, though this may not represent the experiences of students who had actually undertaken EDI training.A citation for Figure 9 has been moved within the article to a more appropriate context where its results are specifically discussed.The phrase 'nearly universally' has been altered to better reflect the results.More references to the literature have been included, including Williams in the introduction, Meulenders in the results, and Wilson in the discussion.This should enhance the contextualisation of the findings within the article.A review of the discussion section better reflects the drawbacks and potential conflicts of responses.Comments regarding potential future research have been added to the section entitled Other Considerations to better reflect areas which might be improved through further investigation.As well as this, we have sought to contextualise the response rate more clearly in this section.We have also included an additional paragraph to highlight the potential differences between institutions and how this links administrator and student outputs, which we are unable to determine.A sentence was also added to the methods section to suggest that this research could act to pilot this questionnaire for future use.

The
Taskforce on Training conducted a survey of 35 Wellcome Trust funded PhD students and 10 programme administrators to evaluate the provision of training in eight key areas.This survey examined a number of issues, such as availability and knowledge of training, potential gaps in training, and the perceived usefulness of training.
training (including training in a specific software or laboratory skills).

Figure 1 .
Figure 1.Mentimeter word cloud made up of responses from first-year students at the PhD Wellcome Trust Meeting.Size of the word represents the number of times that idea was submitted by students during the online poll.

Figure 2 .
Figure 2. Mentimeter word cloud made up of responses from second-year students at the PhD Wellcome Trust Meeting.Size of the word represents the number of times that idea was submitted by students during the online poll.

Figure 3 .
Figure 3. Bar chart showing responses regarding knowledge of where to find information about training offered.Overall and breakdown by year responses are shown.

Figure 4 .
Figure 4. Bar chart showing responses regarding the appropriateness of timing of PhD training.Overall and breakdown by year responses are shown.

Figure 5 .
Figure 5. Bar charts showing the helpfulness ranking counts for each category of training.Respondents ranked each training category by its helpfulness, with 1 indicating the most helpful category of training and 8 indicating the least helpful category of training.

Figure 6 .
Figure 6.Bar chart showing responses regarding the most useful training modules.Overall and breakdown by year responses are shown.

Figure 7 .
Figure 7. Bar charts showing the enjoyment ranking counts for seven different modes of delivering training.Respondents ranked seven different modes of delivering training by their enjoyment, with 1 indicating the most enjoyed delivery method and 8 indicating the least enjoyed delivery method.
reflection emphasises the significance of the taskforce's work to assess the training needs and experiences of students engaged in Wellcome Trust-funded PhD programmes.Training is evidently considered to be generally useful and important by the respondents to this study.Tailoring this training to account for student experiences to date and identified gaps in training needs could therefore aid in the further development of these and other students engaged in Wellcome Trust PhD programmes.In contrast to previous research byDurette, Fournier and Lafon (2016)   this survey used pre-assigned competency categories as there was interest in comparing students' perceptions and thoughts on different categories of training, not in initially identifying these categories.Students were asked to reflect on the usefulness of the training they had received, and technical training was particularly identified as being widely provided and was acknowledged to have been especially useful to students.Questions pertaining to the usefulness of training also highlighted perspectives on supporting wellbeing, as responses were somewhat contradictory.A large percentage of respondents indicated that they had not received wellbeing training.Those who had, reported that it had generally been useful.When presented with an opportunity to provide free text comments regarding aspects of the PhD project or life as a PhD student which they did not feel equipped to manage; three respondents indicated wellbeing concerns such as struggles with 'loneliness' and 'self-doubt'.This supports previous work by Sverdlik, Hall and McAlpine (2020) who identified that imposter syndrome was predictive of psychiatric distress and recommended increased provisions for such wellbeing training to doctoral research students.When asked what type of training they would like to be offered, two respondents suggested wellbeing training.Despite this, wellbeing training was not prioritised when students were asked to rank the usefulness of various types of training: it was consistently ranked as the least useful.This lack of prioritisation may suggest that while students seek some greater provision of wellbeing support, this should not necessarily take the form of training.Additionally, this could be due to the conflict between the number of students who reported that wellbeing support was not applicable, thereby suggesting they had not undertaken this form of training and the inability to reflect this when ranking.The importance of PGR wellbeing has been highlighted in numerous recent publications, particularly in light of the COVID-19 pandemic which was shown to worsen the pre-existing mental health crisis among researchers(Creaton, 2021).A recent survey of research staff, including PGR students(Moran et al., 2020), found that 44% found workplace wellbeing support only adequate, whilst only 28% found it appropriate to their needs.This demonstrates the need to tailor wellbeing support to the specific circumstances of PGR students and the needs of these individual students, particularly those who are international

Figure 8 .
Figure 8. Bar chart showing how different programmes advertise upcoming training by category.
Data file Taskforce on Training -Mentimeter Wordclouds (Tables used to interpret the Mentimeter Wordclouds) • Data file Taskforce on Training -Figures, Graphs for publication (Figures included in the final report) • Data file Taskforce on Training -Administrator Questionnaire (Novel questionnaire developed for administrators) • Data file Taskforce on Training -Student Questionnaire (Novel questionnaire developed for students) Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).

○
I do not see the value and believe you should exclude the one student who responded from year 4.They are only represented in Figure1, and I could not find further mention of findings from students in their fourth year.○ I really struggled with Figure 6 on page 9 and Figure 8 on page 11.I think a separate stacked bar chart for each of the groupings would make this much easier to interpret.
the work clearly and accurately presented and does it cite the current literature?Partly Is the study design appropriate and is the work technically sound?Partly Are sufficient details of methods and analysis provided to allow replication by others?Partly If applicable, is the statistical analysis and its interpretation appropriate?Not applicable Are all the source data underlying the results available to ensure full reproducibility?Yes Are the conclusions drawn adequately supported by the results?Partly Competing Interests: No competing interests were disclosed.Reviewer Expertise: Research assessment; research culture; research careers; research integrity I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
IntroductionBackgroundIn 2020, the Emerging Research Cultures (ERC) project was launched to share the practices and experiences of research culture across Wellcome Trust funded PhD programmes which have started since 2019.Research culture generally includes ideas such as equality and diversity, mental health, bullying and harassment and other aspects such as encouraging good scientific values which can, in turn, help to build a successful and engaging research community.
A number of more recent initiatives e.g., the San Francisco Declaration on Research Assessment (DORA) (Cagan, 2013) and the Wellcome Trust or the Coallition for Advancing Research Assessment (CoARA) (Coalition for Advancing Research Assessment, 2022) have focused on identifying and defining factors relating to research culture and considering how a positive research culture can be encouraged by both individuals and institutions.DORA has focused on the importance of developing metrics for use in evaluating a range of scientific outputs openly and transparently (Cagan, 2013), whilst CoARA has focused predominantly on the need to consider how researchers are assessed (Coalition for Advancing Research Assessment University of Dundee, University of East Anglia, University of Edinburgh, University of Glasgow, University of Leicester, University of Nottingham, University of Oxford, and University of Sheffield.Invitations to participate were issued through emails from the ERC project which were circulated by programme co-ordinators and administrators.The survey was accessible by a link leading to a Microsoft Office form, which students could complete by selecting responses on a sliding scale and offering free text comments to some questions.35 student responses were analysed from the survey, although not all questions were mandatory and some students elected not to respond to all questions. ) were conducted of students and administrators to identify the training currently available, training students found most useful, how training was delivered, and areas of training students felt were lacking.Students on Wellcome Trust Doctoral Training Programmes across the UK were invited to take part in the study in August and September 2022, drawing from the following institutions: King's College London, Queen Mary University of London, University College London, University of Bristol, University of Cambridge, Recruited students were current students on one of 23 Wellcome Trust funded UK PhD programmes, across 15 UK-based institutions.Access was provided to these students and administrators through the ERC community of practice(Emerging Research Cultures, 2023).
including programme administrators and programme leads were contacted for the 23 different programmes for the administrator survey.Administrators were recruited by the ERC project, as part their community of practice(Emerging Research Cultures, 2023).No personal information about administrators was collected; all answers pertained to the Wellcome Trust programme with which they were affiliated.

Table 1 . Table showing the total number of respondents to
the student and administrator survey.Year breakdowns are also included, NA indicates this is not applicable.Total Number of RespondentsYear 1 Year 2 Year 3 , including struggles with networking, planning for next steps after the PhD, and managing finances.
There were 6 responses indicating issues related to academia or the PhD project, including concerns about selecting a supervisor, project management, and grant applications.A further 4 responses mentioned concerns regarding a career beyond research skillsOverall, the data suggests students were lacking in training on equality, diversity, and inclusion, with some indication that those who had received this training found it useful.Other results suggest that such training is deemed less useful than training in other areas.Students also felt that their training in professional, personal, wellbeing, outreach & engagement, and career development was helpful, but some students reported that they had not had training in some of these areas.The most important training reported was technical and cohort-or subject-specific training and that was consistent across all year groups.
cohort-specific training was least commonly delivered through the university and was one of the training categories most commonly delivered through the PhD programme.Only three of the categories reported using external providers to supplement training, in professional development, technical training and wellbeing.
were asked about training providers for the eight categories of training segregated in the student survey, with training being offered through the university, PhD programme, department or externally.Equality, diversity, and inclusion training was most commonly delivered through the university, whilstThe timing of training during the academic year and PhD student progression was also reported on.Although some types of training were reported to take place at the beginning of the academic year e.g., public engagement, equality, diversity and inclusion, technical and wellbeing training, training was generally reported to take place throughout the year and to vary.Administrators also reported that training for all categories except for careers development training was delivered in the first year of study, although training was generally reported to take place across all four years of study.Equality and diversity training had the most mixed reporting, with most programme administrators reporting it happened in first year, others reporting it happened after first year or in all years of study and two programmes reporting that the question was not applicable.The mode of training delivery was also considered, including seminars, workshops, and self-taught modules.Other forms of delivery were also mentioned, such as that careers development training could be delivered through industrial training partnerships.The results found that for all eight categories of training, the mode of delivery was diverse, with most training being delivered through varying methods.
, our results on the usefulness and provisions of outreach training are comparable to recent work which identified that only 40% of 30 German universities surveyed offered public outreach training to their PGR students(Meuleners  et al., 2023), showing that a lack of provision in some areas of training may be applicable internationally.Survey results regarding equality, diversity, and inclusion training suggest that further research in this area is required.Few students reported receiving this form of training, though those who did indicated that it had been useful.However, when asked to rank the helpfulness of different forms of training; equality, diversity, and inclusion training was frequently ranked as less helpful than training in other areas.This might be accounted for by the lack of a 'not applicable' option, which led students who had not experienced equality, diversity, and inclusion training to rank it below other forms of training because they were not able to exclude it from their rankings.The results on this are therefore somewhat inconclusive.Future studies should consider the provision of equality, diversity, and inclusion training and investigate whether, when provided, it is considered helpful to students.Alternatively, it might be suggested that such training is not best measured through 'usefulness' and 'helpfulness' and another metric could be better suited for further exploration of this issue.
This was highlighted in EUA-CDE (2016) but also inMulvany (2013)andMcGagh et al. (2016)who suggest alternative forms of training outside the scope of this report.Both suggest that exposure to industry and relevant employment sectors, in the form of placements, will allow PhD students to develop and practice their skills in a non-academic setting.As the Wellcome Trust offers opportunities for students to take part in placements during their PhD, this could be considered in future examinations of student training.Of those who responded to the open-ended question regarding aspects of the PhD which they do not feel prepared to manage, a number of students indicated areas of personal development such as project or financial management or future planning.However, when asked whether they have used any personal development training to date, a third of respondents said that they had not done so.While this could appear to suggest that personal development training has not been useful, the lack of a 'not applicable' response for this question means that respondents may instead have been suggesting that they haven't received such training.Alternatively, if in the earlier years of their PhD, students may have not yet had an opportunity to use these skills but it is unknown how they may utilise the skills and knowledge provided to them through training at a later stage of their study.
ing, which was commonly ranked near to the bottom, despite other forms of interactive learning being much more highly ranked.This may suggest that students instead have no or limited experience with this kind of training, resulting in it falling to the bottom of the ranking by default.Providing an option which allowed participants to make clear that they have not received a particular form of training may have clarified this, particularly in light of recent work byWilson et al. (2023) PhD studies and how their training is relevant to this.It is possible that restrictions imposed due to the pandemic may have impacted the training experiences of these students, as discussed inCarusi et al. (2022a).) students reported that their programme did not make them aware of training opportunities.This could be because emails about training come from alternative avenues e.g., those organised and supported by their institution or self-organised training.Interestingly, personal development training is most reported to only take place in the first-year by programme administrators, whilst this was reported by students in their free text answers to be an area of concern.This could have been due to differences in what the administrators and students considered to be personal development training or a belief that personal development related training is less required and students will develop these skills in the course of their studies e.g., project or financial management.However, it may highlight a need to further investigate personal development training and more clearly understand both perspectives on this.

The role of supervisors in light of the experience of doctoral students.
Policy Futures in Education.2013; 11(3): 267-276.

EU) 2016.679 of the European Parliament and of the Council
. [Of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/ EC (General Data Protection Regulation)].Retrieved 1 st March 2023, 2016.
2.3.Was 254 the total number of doctoral researchers attending Wellcome Trust PhD programmes that have started since 2019?If not, what was the total number of doctoral researchers attending a Wellcome Trust Ph.D. programme and what criteria presided to the selection of the 254 participants.Regardless, the response rate is quite low overall.Although acknowledging this as a limitation in the section on the Other considerations, no reflection on the reasons for such a low response rate is presented.
and Recommendations 4.1.When reflecting on the results, the authors make almost no references to the literature on the topic.It is unclear the extent to which the study results overlap, complement, or contradict previous research on the topic.There are several studies examining the training needs and priorities of doctoral researchers, including reports produced by the Wellcome Trust, that would have helped give additional depth to the results of the present study.Also, recently, Nature published several interesting articles addressing the wellbeing and work-life balance of early career researchers.I suggest the authors consider using these and other relevant publications to frame the discussion of their results.Similar considerations can be made to almost all topics in the discussion of the results (e.g., mentoring/peer mentoring, the impact of COVID-19 on training, professional and/or career development).4.2.Although mentioning career development as a category of training, no additional information is provided about its content or its specific results.No consideration of the motives making doctoral researchers choose to train at the Ph.D. level is included.Universities or research institutes' ability to attract and retain talent, including at the level of Ph.D. training, has recently started to permeate discussions around the need for reforms, including the reform of doctoral training and its assessment.It is unclear the extent to which the present study considered these discussions in its framing or used them to reflect and interpret findings.4.3.No reflection on the similarities/differences between the results from the survey answered by the doctoral researchers and the survey answered by the administrator is made.Also, no real conclusion summarizing the experience and main gaps in the training of doctoral researchers is provided.4.4.Over the years, several recommendations have been put forward that address many if not all of the topics focused by the present study.It is unclear why the authors have not considered them when discussing their results and proposing recommendations for improving doctoral training.The authors systematically refer to doctoral researchers as students.Granted, in the UK, doctoral researchers have the status of students.However, internationally, the acknowledgement of doctoral candidates as researchers at the initial stages of their training is growing.Any reflection on doctoral training, its purpose, scope, strengths, and limitations, should consider such good practice and acknowledge those researchers working at the doctoral level as researchers and not as students.This is a minor but significant issue as it goes directly into our understanding of what doctoral training is and what it means to acquire a Ph.D. 5.2.When presenting the background of the study, the authors state that the Taskforce on Training consisted of first-and second-year doctoral researchers across Wellcome Trust-funded PhD programmes.Why only first-and second-year doctoral researchers were included in the task force and why were the early findings only shared with these researchers?Considering the aim of the study was to examine the experiences of doctoral researchers attending Wellcome Trust Ph.D. University of Michigan, Ann Arbor, Michigan, USA 2 University of Michigan, Ann Arbor, Michigan, USA 3 University of Michigan, Ann Arbor, Michigan, USA Thank you so much for the opportunity to review the manuscript "A survey of the training experiences and needs on Wellcome Trust PhD programmes" and to be able to provide the following comments.Understanding the training and professional development needs of PhD students continues to be an important area of research.I think this manuscript could benefit by providing a high-level summary of research from Australia and the United States that has addressed this same issue.

it cite the current literature? Partly Is the study design appropriate and is the work technically sound? Yes Are sufficient details of methods and analysis provided to allow replication by others? Yes If applicable, is the statistical analysis and its interpretation appropriate? Not applicable Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.Reviewer Expertise: Graduate education: professional development, diversity and inclusion, peer mentoring, and graduate student engagement.Program evaluation and assessment.I confirm that I

have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above. Version 1
https://doi.org/10.21956/wellcomeopenres.21668.r68645