Graduate Campus - We Data Survey 2024
Présentation
Le présent document résume les résultats du sondage concernant les besoins en soutien statistique des doctorantes et doctorants de l’Université de Genève. En filtrant les réponses incomplète nous avons 89 répondants. Les résultats initiaux montrent qu’il y a un réel besoins de soutien pour ces étudiants et qu’il doit se faire dans différentes dimensions:
- Mentorat
- Communauté
- Ressources
- Rencontres ponctuellesProfils des doctorant(e)s
In which language do you prefer to express yourself?
La plupart des doctorant(e)s parlent au moins l’anglais (80.9%), mais il y a tout de même un petit pourcentage qui ne parle que le français (19.1%)
Which specific subjects within statistical or scientific research pique your interest?
En ce qui concerne les sujets qui intéressent notre population, nous avons dans le top 3 et dans l’ordre décroissant Data Visualization, Statistical Tests (parametric and non-parametric) et Linear Modeling / Regression analysis (glm, multilevel, etc.) avec respectivement les proportions suivantes 86.52%, 74.16% et 73.03%. Ces résultats ne sont pas surprenants puisque ce sont généralement des sujets populaires, surtout auprès des débutants.
Which programming language(s) interest you?
Sans surprises, R et Python sont les langages les plus populaires avec un intérêt marqué pour R.
On a scale from 1 (no experience) to 5 (expert) what is your current proficiency level in programming …?
Il est important de noter que la proportion de personnes sans expérience pour le langage Python est élevée, mais pas tous les répondants souhaitent utiliser Python dans leur recherche. Du côté de R, cette proportion est élevée, mais il y aussi pas mal de débutants.
Besoins
Communauté et mentorat
Nous pouvons souligner les résultats principaux suivants:
- La quasi-totalité des répondants affirme être intéressé à rejoindre une communauté d’entre-aide (96.63%)
- La majorité est prêt à partager des ressources pour l’entre-aide (61.80%)
- Il y a bien plus de personne cherchant un mentor (70.79%) que de personnes prêtes à faire mentor (21.35%)
Data meetup
Concernant les potentielles rencontres entre doctorant(e)s, il y a clairement une demande pour ce genre de projet. Plus précisément, nous pouvons souligner les informations suivantes:
- La vaste majorité des répondants sont intéressés par ce concept (97.80%)
- La majorité souhaiterait avoir ces rencontres une fois par mois (62.64%)Analyse thématique
L’analyse des commentaires a permis de dégager plusieurs informations :
- Il y a une demande pour des leçons de base en R et Python.
- Même si beaucoup n’ont pas d’expérience, quelques personnes savent exactement ce qu’elles cherchent (ex. méthodes). Ces personnes sont généralement issues de disciplines très quantitatives, comme la psychologie ou les sciences.
- Les mentors sont assez demandés, mais généralement, les personnes moins débutantes cherchent des mentors avec des compétences spécifiques, et de préférence dans le même domaine qu’elles.
L’analyse des commentaires révèle également les raisons d’un plus grand intérêt pour la communauté plutôt que pour les mentors. Nous pouvons citer les raisons suivantes : - Flexibilité dans le temps - Autonomie - Faible investissement
La plupart des personnes recherchent avant tout une communauté. Il apparaît que certaines souhaitent simplement avoir la possibilité de poser leurs questions et de participer librement à des activités. Certains répondants ont suivi ou suivent encore des cours, que ce soit en ligne ou en présentiel, mais cela ne les empêche pas de s’intéresser à la potentielle communauté ou à la recherche d’un mentor.
Lorsque nous examinons les commentaires des répondants, nous pouvons identifier trois thèmes récurrents : la recherche de références, le manque de confiance et la recherche d’un environnement adéquat.
Recherche de référence
En accord avec la proportion de personnes intéressées par des rencontres mensuelles, il semble que celles cherchant des mentors veulent principalement quelqu’un qui leur montre la bonne direction et les rassure dans leur processus d’apprentissage. Par exemple, les commentaires suivants utilisent des termes comme approprié, sur la bonne voie ou bon(ne) [analyse|code] :
I am a phd in the political science department and my research is in the field of political economy. I started taking quantitative methods classes at the beginning of my PhD but would love some mentorin, i.e., to disucss things like what models would be the most appropriate etc. - basically someone i could ask some questions to, to get some feedback about whether or not I am on the right track
It could be interesting to discuss about data I collected to be sure that i’m doing the good analysis; also to check if what i’m doing seems to be good (in a script).
Nous voyons aussi que pour certain(e)s, avoir un mentor c’est aussi trouvé un moyen de se motiver et de se discipliner dans l’apprentissage.
A mentorship could be a motivation to discipline oneself into learning coding/data science.
Ici la réponse au problème peut être aussi simple que d’être dans la même pièce que quelqu’un à qui nous pouvons poser des questions, d’où l’importance des rencontres:
It would be great to work on my analysis in a space where there is someone who can answer my questions et guide me
Manque de confiance
Le manque de confiance transparaît dans les affirmations des individus quant à leur capacité à endosser le rôle de mentor. Ce manque de confiance n’est pas surprenant, étant donné que la majorité des personnes admettent avoir une expertise limitée en R et Python. Surprenamment, un grand nombre souhaitent devenir mentors, mais se sentent insuffisamment confiants en leurs propres compétences pour assumer cette responsabilité :
knowledege on basic-intermediate statistics so I can later on try to tackle other more advanced and specific topics/techniques I’d love to become a mentor eventually, but first I must feel confident that I can actually mentor someone into some particular skills.
Not proficirnt enought to ‘mentor’ somebody, but happy to share
I don’t feel I have enough knowledge and competencies to mentor another person. But would love to do so if I improve in these areas
I’m open to help, but right now I don’t have the skills yet. However, if I can help I would like to, I think we learn better with other people
I am open but I don’t have any experience on this :(
I am not sure if I have enough knowledg to mentor someone else at te moment, but in the future once I feel confortable enough in the languages I would be happy to.
I feel I need to develop my skills before I can mentor someone.
I would love to, however I think I simply don’t have the capacity to do it (not enough knowledge). However, I tend to write down and source everything I find/use (great to create manuals and checklists that I share with great enthusiasm).
Not yet, but why not when I would have gained expertise, just to give back the help I might have received
Cela révèle également un désir d’aider et de contribuer. Une communauté devient donc essentielle pour permettre aux personnes d’apporter leur aide sans devoir assumer entièrement le poids du rôle de mentor. Il est important de noter qu’il existe des répondants confiants, prêts à endosser la responsabilité de mentor. Toutefois, ces derniers sont peu nombreux par rapport à la demande, et ils sont souvent limités par des contraintes spécifiques, telles que leur domaine d’expertise ou leur disponibilité.
Recherche d’environnement
Plusieurs participants ont suivi des cours pour compenser leur manque de compétences. Malgré cela, l’intérêt pour un mentor et une communauté demeure vif pour tous. Ce qui ressort des commentaires est, d’une part, le besoin d’interagir avec d’autres personnes :
I have only taken some introductory courses on R and Python (Google Colab). Unfortunately, I find these are skills difficult to develop and much more difficult to keep without proper constant practice. I lack a context for practice, so eveytime I try to train in these skills, I feel like climbing down to ground level.
For R I’m taking an online course in Codecademy platform and I started using RStudio to write very basic codes in my own data but I would like to practice more and interact with people who are also using it. For Python, I started an introductory course but I’ve never used it.
I would like to interact with other users and discuss different ways to write codes as well as getting help when I can’t figure out why my code is not correct
Comme mentionné précédemment, il existe également un désir marqué de contribuer au sein d’une communauté. Cet intérêt se manifeste aussi chez les personnes désireuses de partager des ressources ou de mentorer :
I love trying by myself. I’m interested in sharing our experience with a community of practice
Yes, I am [interested in mentorig]. It’s a way to pay it forward. Also, when you mentor/teach others you also improve your own expertise and knowledge
Ainsi, à travers nos trois thèmes principaux - la recherche de références, le manque de confiance et la recherche d’un environnement adapté - il apparaît clairement que le besoin d’une communauté est une préoccupation sous-jacente. La communauté pourrait résoudre de nombreux problèmes en permettant aux doctorants d’être soutenus et de s’entraider. Toutefois, cela ne couvre pas tous les besoins ; une part significative des personnes requiert un suivi personnalisé (mentorat) et nous ne disposons pas d’assez d’offres de mentorat pour répondre à cette demande.
| mentee_info_cmt | mentor_info_cmt | share_resources_cmt |
|---|---|---|
| Dear Team, I would like to be mentored, since I wanna improve my data analyzing skills. I want to become an expert in the field of biological data analysis, so I can use my skills for my job after finishing my PhD | NA | NA |
| As I am starting my PhD, I don't have a need currently but I would be interested in being mentored in the future for the analysis of sequencing data (metagenomics and RNA-seq) | NA | NA |
| I would like to know more about code structure, specifically how to clearly and quickly organize code when starting a project | NA | Some youtube vides about reinforcement learning |
| NA | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| Data analysis using R | Research design | NA |
| I am a PhD student in psycholinguistics. I have a master's degree in linguistics from the Faculty of Humanities in Geneva. I have no experience in methodology and statistics, and I have very little time to take the courses normally provided for master's students. I would need to be guided in a targeted manner to learn quickly what I really need | NA | I have no teaching materials to share for the moment |
| NA | NA | NA |
| I will like to learn about Stata, R, and Data Analysis as Trade Researcher. As I have a law background, these are research skill required when I tru to apply for trade resaerch opportunities. | Yes, I am. It's a way to pay it forward. Also, when you mentor/teach others you also improve your own expertise and knowledge | NA |
| NA | NA | NA |
| NA | NA | NA |
| I am a phd in the political science department and my research is in the field of political economy. I started taking quantitative methods classes at the beginning of my PhD but would love some mentorin, i.e., to disucss things like what models would be the most appropriate etc. - basically someone i could ask some questions to, to get some feedback about whether or not I am "on the right track". | I do not feel I have the level to mentor someone at this stage. | NA |
| I work in the field of neuroscience and as a part of my work data analysis is essential. I use to use Matlab a lot but now I would like to explore other lenguajes and discover new tools. | Yes, I feel confortable mentoring people. Specifically, from the biology background or Neurosciences as me. | NA |
| NA | Not for now, may be later | Many links to share (from basic data viz to bayesian modeling) |
| Genomics and proteomics data analysis and visualization | NA | I am keen to share resources that I find interesting for myself |
| I am doing a PhD in the Life Sciences department and for now I would like to learn how to use Matlab/Python for linear regression analysis, plotting and data representation. With the years I will maybe want to explore more. | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| I am working on EEG, fmri and behavioural datasets analysis. | NA | NA |
| NA | NA | NA |
| I am really interested in reproducibility and overall good research practices and quality assurance of my own work, but I am not sure how to approach this. | Overall, my background is in Neuroscience, I have worked in the pharma industry before my PhD, so I might be helpful for students who are unsure about their future career. I have been told that I give great feedback for oral presentations, slides and posters. Having studied in an English-speaking country, I am also comfortable with academic writing in English, but of course I am no expert :) | NA |
| NA | NA | NA |
| Je m'intéresse à la recherche en éducation médicale. | Oui c'est très intéressant dans le domaine médical sur ces thèmes : diabète, syndrome métabolique, obésité. Et en éducation médicale ,sur le thème de : la pratique du Feedback avec les étudiants en médecine . | Oui je souhaiterais mais je vais d'abord me documenter pour préciser quel type de ressource . |
| It would be great to work on my analysis in a space where there is someone who can answer my questions et guide me | Maybe later... | NA |
| statistics and data visualization for clinical research (I am an orthopedic surgeon) | NA | NA |
| basic biology research | NA | NA |
| NA | NA | NA |
| I would like to consolidate my knowledege on basic-intermediate statistics so I can later on try to tackle other more advanced and specific topics/techniques | I'd love to become a mentor eventually, but first I must feel confident that I can actually mentor someone into some particular skills. | Nothing to share yet, but if I find something interesting I'll consult if it is shareable |
| NA | NA | NA |
| i work with regular and spatial transcriptomics in the context of alzheimer's disease on rstudio | NA | the Seurat vignette for anything RNA related |
| NA | NA | NA |
| R for data analysis (biomarker selection, data visualisation, ...) need to learn about PCA, UMAP, volcano plot, heat map, data filtering, transformation, ... | I do not have the skills to mentor someone | Again, I do not have the knowledge to do this, unfortunately, ... |
| NA | NA | NA |
| I am currently a PhD student in Translation Studies. I will be conducting a questionnaire (closed-ended questions) and interviews (open-ended questions). | NA | NA |
| NA | NA | NA |
| Longitudinal analysis | Social Sciences | NA |
| NA | NA | NA |
| NA | Not proficirnt enought to 'mentor' somebody, but happy to share | NA |
| I would like to develop my skills in more complex analyses for instance fMRI analyses (PCA, ICA...), EEG analyses, as well as factor analysis, structural equation modeling. I would also like to develop better habits and methods and learn how to use version control properly. I learned code in R on my own during a research project and I am trying to follow some online courses (MOOC) to get back to the basics but I would very much like to be mentored to have more advice. I work in psychology & neuroscience | I don't feel I have enough knowledge and competencies to mentor another person. But would love to do so if I improve in these areas | I found nice MOOC courses online for tidyverse on R and a book linked to that |
| I would like to interact with other users and discuss different ways to write codes as well as getting help when I can't figure out why my code is not correct | I'm open to help, but right now I don't have the skills yet. However, if I can help I would like to, I think we learn better with other people | When I come across freely and legally available resources I will share them with the community |
| NA | NA | NA |
| NA | NA | NA |
| NA | I'm pretty good with R in general and would be happy to share some knowledge provided it doesn't take up too much of my time. | NA |
| Start from scratch to really understand. | I am open but I don't have any experience on this :( | NA |
| NA | NA | NA |
| Possibly (there was only yes or no), but only so much so as to specifically help with my project goals, as I do not have much bandwidth outside of that at the moment. | I am not sure if I have enough knowledg to mentor someone else at te moment, but in the future once I feel confortable enough in the languages I would be happy to. | NA |
| It could be interesting to discuss about data I collected to be sure that i'm doing the good analysis; also to check if what i'm doing seems to be good (in a script). | I'm not good enough | NA |
| I would like to hear more about this. I'm currently taking a intermediate R course to continue learning, but I would like to apply what I'm learning on research. Right now, I work mostly on epidemiological studies, focusing on Global Health. | I feel I need to develop my skills before I can mentor someone. | I found a really nice video on gtsummary (https://www.youtube.com/watch?v=ko9vCHYJD7Q). For me, it was mind-blowing |
| Economics, previous knowledge of Stata | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| I am currently trying to realise a doctorate in psychology. I try to use both frequentist and bayesian stats. I mainly carry out descriptive and inferential stats with emphasis on Cluster Analysis, Latent Class Analysis, Factor Analysis/Principal Component Analysis, Structural Equation Modelling. Ideally, I would like to get rid of (or use as little as possible MS Word) in favour of R/RStudio/Quarto Markdown with APA 7 and Zotero integration (latter two are essential). | I would love to, however I think I simply don't have the capacity to do it (not enough knowledge). However, I tend to write down and source everything I find/use (great to create manuals and checklists that I share with great enthusiasm). | As mentioned earlier, I tend to write down and source everything. This material can be used to create instruction manuals and/or checklists (I am currently trying to write with the help of another student a step-by-step checklist on how to prepare a systematic review of the literature). |
| I need to analyse a lot of quantitative data from different biological experiments. I need more knowledge to know which is the best test to use everytime. | NA | NA |
| NA | NA | NA |
| Ecology and parasitic interaction depending on temperature. I would like to improve basic statistical analysis and gain skills in mechanistic models building. | NA | NA |
| Working in Cognitive Psychology, I need to develop an efficient way of writing code. I can write lines of code that will do the job, but that are clearly way too long and not clear enough for an efficient code-sharing. I tried a few different courses, but they never quite answered this question and it is often that when I dig into my data, I have very specific questions that emerge after the courses. I can search/find responses on the web (chatGPT or other), but it just give a specific answer, no help on how to more globally think about the structure of a code. Would be very helpful to have a reference expert to whom I could ask those kind of questions, or who could give me refs or advices on which course would be best for my needs. | Not yet, but why not when I would have gained expertise, just to give back the help I might have received | First that come to my mind is https://statisticsglobe.com/, I would need to go back into my notes to find others I have used in the past |
| NA | NA | NA |
| Biology, molecular biology | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| I am a PhD student in developmental neuroscience and I would be needing to delve into spatial transcriptomics and sc/snRNASeq data analysis and related statistics in general. | NA | NA |
| I would appreciate been mentored as a beginner in the subject | I am not skilled enough (in the subject) to mentor another student | I would love to share the resources |
| Advanced quantitative data analysis (linear mixed models and latent variable modeling for longitudinal data, latent class analyses) | NA | NA |
| NA | NA | NA |
| I'm interested in R for analysis of data related to political science, international relations and international political economy fields. | NA | https://r4ds.hadley.nz/ R for Data Science is a very useful free tool too learn about R and statistics. |
| NA | NA | NA |
| I am a PhD in environmental sciences and I'll need help for data analysis because this is a new field for me. I'll need help to get knowledge on basic steps. | I have not enough knowledge to be a mentor. | NA |
| analyses sequencing data | NA | NA |
| A mentorship could be a motivation to discipline oneself into learning coding/data science. | I don't have experience in coding, so I can't mentor someone else. | I haven't found any teaching material yet, but if I do, I'd be happy to share. |
| not at this time, but maybe in the future | NA | NA |
| I would like to learn how to improve my coding practices | NA | NA |
| I need to improve my R programing skills | NA | Didn't find something yet but I can sure share when needed |
| NA | NA | NA |
| I would like to be mentored in using R | NA | I don't think I have free resources to share (yet). |
| NA | NA | NA |
| In python | NA | NA |
| NA | NA | NA |
| NA | NA | NA |
| It would be intereting to be mentored by someone in the same discipline | I don't have enough experience | NA |
| NA | NA | NA |
| I love trying by myself. I'm interested in sharing our experience with a community of practice | NA | I'm an absolute beginner so far ! |
| might be, especially for multilevel modeling | I am more into standard social science analysis, regression models, but also longitudinal fixed & random effect models | I could share some mixed-method practices |
| NA | NA | NA |
Conclusion et action
À travers cette analyse, nous avons identifié quatre problèmes que l’association pourrait partiellement résoudre. Toutefois, pour répondre à tous les besoins, l’association ne dispose pas actuellement des moyens suffisants.
- Ressources :
- Besoin : Ressources (gratuites ou payantes) pour apprendre les bases méthodologiques ou en programmation, ainsi que des ressources spécifiques pour les personnes plus avancées.
- Majoritairement solvable : L’association dispose de ressources personnelles et d’une liste de ressources en ligne gratuites et payantes. De plus, de nombreux participants indiquent vouloir partager leurs propres ressources. Nous pouvons organiser une plateforme pour collecter et organiser ces ressources.
- À améliorer : Explorer la possibilité de demander à des enseignants de l’université de partager leurs ressources.
- Communauté :
- Besoin : Être en contact avec d’autres personnes intéressées par les méthodes statistiques et la programmation pour pouvoir s’entraider.
- Majoritairement solvable : L’association peut fournir une plateforme en ligne pour permettre à chacun(e) d’échanger sur la thématique. Nous envisageons d’utiliser Discord pour faciliter l’organisation.
- À améliorer : Cette plateforme (Discord) est externe à l’université et non officielle. De plus, elle est liée à notre chaîne YouTube, ce qui permet la présence de personnes externes.
- Mentorat :
- Besoin : Sept participants sur dix ont exprimé leur intérêt pour avoir un mentor.
- Actuellement insolvable : Il y a trois fois plus de personnes intéressées par le mentorat que de personnes désireuses de mentorer. De plus, les deux groupes ont des restrictions concernant la matière à traiter et le temps disponible. L’association ne dispose actuellement d’aucun membre pouvant exercer ce rôle bénévolement. Il est donc impossible de répondre à ce besoin sans moyens supplémentaires.
- À améliorer : Cette situation suggère qu’il pourrait être judicieux pour les départements d’externaliser ce service, compte tenu du besoin de suivi personnalisé.
- Rencontre :
- Besoin : Les doctorants sont intéressés à se rencontrer régulièrement (une fois par mois) dans le cadre de data meetups.
- Partiellement solvable : L’association organise déjà des R-Lunches (rencontres autour de R chaque premier mardi du mois). Cependant, ces rencontres portent sur des sujets qui changent souvent. L’association pourrait organiser une rencontre annuelle ou semestrielle. Mais cela ne suffira pas à répondre à la demande.
- À améliorer : En l’état actuel, l’association ne peut pas faire plus que ce qu’elle fait déjà. Des solutions simples pourraient être envisagées, car pour certaines personnes, le simple fait de pouvoir travailler dans une salle où d’autres travaillent avec la possibilité de poser des questions est déjà bénéfique.
Glossaire
| Original Name | Shortened Name |
|---|---|
| lang_pref | in_which_language_do_you_prefer_to_express_yourself |
| interest_ | which_specific_subjects_within_statistical_or_scientific_research_pique_your_interest_ |
| prog_lang_ | which_programming_language_s_interest_you_ |
| int_ | interest_ |
| stat_tests | statistical_tests_parametric_and_non_parametric |
| linear_modeling | linear_modeling_regression_analysis_glm_multilevel_etc |
| long_analysis | longitudinal_analysis_survival_panel_time_series_etc |
| datavis | data_visualization |
| bayesian | bayesian_statistics |
| multivar_stats | multivariate_statistics_pca_cluster_sem_etc |
| survey_design | survey_design_and_analysis |
| network | network_analysis |
| spatial | spatial_statistics |
| sci_sim | scientific_simulation |
| report_prod | report_production |
| sci_pub | scientific_publication_reproducibility_version_control |
| lang_ | prog_lang_ |
| prof_ | on_a_scale_from_1_no_experience_to_5_expert_what_is_your_current_proficiency_level_in_programming_using_the_ |
| prog_exp_context | would_you_like_to_provide_any_additional_information_or_context_about_your_programming_experience |
| meetup_freq | how_often_would_you_be_interested_in_participating_in_data_meetups |
| mentee_info | would_you_like_to_be_mentored_if_so_we_would_love_to_hear_more_about_your_interests_and_needs_in_the_comment_section_discipline_needs_and_any_useful_information |
| _cmt | _comment$ |
| mentor_info | are_you_open_to_the_possibility_of_mentoring_another_ph_d_student_your_participation_is_entirely_voluntary_and_you_may_decline_later_if_you_do_not_feel_it_aligns_with_your_current_commitments_please_let_us_know_more_about_you_in_the_comment_section_subject_or_field_you_are_comfortable_mentoring_in |
| platform_intro | we_are_excited_to_introduce_a_new_initiative_aimed_at_fostering_collaboration_and_knowledge_sharing_within_our_university_community_we_are_creating_a_local_online_communication_platform_where_individuals_can_freely_engage_in_discussions_seek_answers_to_questions_and_exchange_ideas_related_to_statistics_and_research_your_participation_will_enrich_this_platform_and_contribute_to_a_vibrant_community_of_learners_and_researchers |
| join_platform | would_you_like_to_join_this_local_online_communication_platform_at_the_university_where_everyone_is_free_to_ask_and_answer_questions_and_exchange_ideas_about_statistics_and_research |
| share_resources | the_communitys_aim_will_also_be_to_bring_together_as_many_free_educational_resources_as_possible_so_that_everyone_can_learn_for_themselves_if_they_wish_you_dont_have_to_be_the_author_of_these_resources_to_share_them_as_long_as_they_are_freely_and_legally_available_would_you_like_to_share_free_teaching_materials_if_so_let_us_know_in_the_comments_section_what_type_of_resource_you_found |
| contact_email | if_you_have_expressed_int_in_mentorship_seeking_a_mentor_mentee_sharing_resources_joining_an_online_community_or_wish_to_be_contacted_later_we_would_love_to_stay_connected_please_feel_free_to_leave_your_email_address_below_your_email_will_only_be_used_for_the_purposes_youve_indicated_and_we_respect_your_privacy_thank_you_for_your_willingness_to_engage_further |