Bioinformatics Career Survey 2011/2012

Question

[Pierre Lindenbaum]

In 2008, Michael Barton conducted the Bioinformatics Career Survey whose results are still available on openwetware: http://openwetware.org/wiki/Biogang:Projects/Bioinformatics_Career_Survey_2008

The results of this survey have been used in biostar:

• Post not found and
• Salary For A Bioinformatics Programmer In Europe?

and elsewhere (http://genomebiology.com/2008/9/12/114)...

We would like to repeat this survey and, with the rise of Twitter & co, we could receive more answers than previously. Here, I'm asking the BioStar community what are the questions that you would like to see in this survey, what do you want to know about the community.

Here are a few suggestions based on the previous survey. Feel free to suggest some new questions or to complete the following ones:

• profile: name, latitude & longitude, (this should be optional, I remember some people didn't want to be identified)
• academics/industry
• Happiness
• subject (shall we use a controlled vocabulary=SW)
• Research Areas (SW?)
• programming languages (SW?)
• favorite tools
• social web (biostar, twitter...)
• journals
• do you contribute to wikipedia?
• what the best way to build this survey? etc...

PS: hashtag: #bioinfsurvey

PS: community wiki.

Edit: Michael started a repository on github: https://github.com/michaelbarton/bioinformatics-career-survey

Edit2: the survey has started, take it here: http://bioinfsurvey.org/

[Michael Barton]

Thanks for writing this up Pierre! I didn't even know the data had been used in those biostar questions. Now I see further why a repeat would be useful.

[Casey Bergman]

Also discussed on BioStar here: Economics Of A Career In Bioinformatics

[Michael Barton]

Thank you for the useful suggestions so far. Up votes all around.

[Giovanni M Dall'Olio]

Do you have a deadline for preparing the questions? When would you like to start the survey?

[Pierre Lindenbaum]

@Giovanni: No, I don't even know how we shall build this survey (surveymonkey?). I would like to write a dynamic HTML form but in that case, I would need a server to receive the POST queries and store the data.

[Pierre Poulain]

@Pierre. Maybe a google doc form? All answers are then gathered in an (google doc) spreadsheet.

Answer 1

[Michael Barton]

• Include all the original questions from the first survey so that they can be compared over a four year period.
• A better choice of controlled vocabulary for describing bioinformatics research areas. Previously I used topic areas from the journal "Bioinformatics". This is somewhat limited however too large a vocabulary might be confusing.

Answer 2

[Ido Tamir]

• educational background: CS / biology / molecular biology / other?
• income: this I guess will be lies and denials, but many find this interesting
• cooperation/integration with wet-labs: scale 1-5 (difficult to answer, vague)

Time spent on:

• service: analysing other peoples data
• research: developing algorithms, analysing own data, more independent than service
• IT: (building warehouses, maintenance of programs)

But maybe shorter surveys will give more answers.

[Michael Barton]

Educational background was quite interesting last time around. It allowed me to produce this figure:

[Ido Tamir]

Very nice (altough a bias of 2% might not be significant - do you have p-values for this ;- just joking)) So the last survey anyway had these questions. I was to lazy to look up the last survey. The questions should have been included here completely.

Answer 3

[Travis]

Also useful I think:

a) How many times a year are you permitted to attend conferences?

b) How many times per year do you receive external training in topics relevant to your role?

[Michael Barton]

Both good questions.

[Travis]

Perhaps better asked as "Do you feel you receive adequate external training?" etc.

Answer 4

[Simon Cockell]

It would be nice to be able to track who has answered in both surveys, and see how their responses have evolved. I realise this probably isn't possible, but maybe there could be a 'did you participate in 2008?' question, which would at least allow the proportion of repeat participants to be determined.

[Larry_Parnell]

I agree with the 'did you participate in 2008?' question as it can allow us to uncover trends.

Answer 5

[Giovanni M Dall'Olio]

Does your group make use of any of the following technologies to support software development:

• Version control software
• Issue tracking systems
• Wikis & GoogleDocs/Zoho documents
• Software Project Management System (which include all of the above. E.g. trac, bugzilla, mantis, github, etc..)
• other...

Software Development process:

• Is there a project manager in your group, apart from the group leader?
• How frequently do you meet with your group leader each week, to plan what to program and what to implement?
• And how frequently do you meet with your group leader each week, to discuss how to implement what is planned?
• From which background does your group leader comes from? Is he a biologist or a computer scientist?
• Is your group leader able to help you on programming problems in case you need?
• Do you have systematic Code Reviews in your group? Is there anyone in charge of reviewing the code before submitting a paper?
• Are you the only person in charge of your code, or is the code ownership shared among the members of the lab? How frequently do you use code written by your labmates?
• Do you adopt any agile programming technique in your group?

Online Communication skills and miscellaneous:

• Do you make use of open source libraries, such as BioPython, BioPerl, BioConductor, etc?
• Where do you look for help about technical/programming issues? (e.g. Biostar...)
• Do you have a blog, where you discuss about your research?
• Do you use twitter/facebook/google+/etc.. for your research?

Collaboration

• On average, with how many laboratories have you collaborated on each paper you have published?
• ...

[Larry_Parnell]

All good stuff, but may need to be condensed in order to avoid "respondent fatigue."

[Giovanni M Dall'Olio]

@Larry: I agree, I am just making a list of what would be interesting to ask.

[Niallhaslam]

I think the average number of collaborating institutions is very different to the tools question. So, whilst I'd agree with the condensation of some of the praxis questions I think its important to include the collaboration question.

Answer 6

[Larry_Parnell]

Michael's suggestion to use the questions from the first survey is important. However, those questions must be reviewed for accuracy (perhaps an older question now needs to be split into more and with finer detail). Including those older questions is great but should not be done blindly. We deal with lots of survey data for our genetic studies and so I can confidently state that this review is relevant. Lastly, as Simon suggests, the survey needs to ask if the respondent completed the first survey.

Under "academic/industry" there needs to be a "government" option. This could also include "freelance."

Educational background should be expanded to include (bio)statistics, genetics, chemistry, biophysics/physics, medicine, chem. engineering, etc, etc

Include: "Do you write a blog" "Do you read/leave comments on other blog sites - never, seldom, often..."

Journals: Which do you read? Have you in the last X years served as reviewer, editor, etc of any journal?

Do you belong to any professional society?

I like Ido's "cooperation/integration" question. With how many wet labs do you collaborate? Rate the interaction from smooth/well integrated to rough/a struggle. This can be repeated for "in-house" and "external" collaborations.

HRanjeev's "employability" questions are tough to formulate in my mind. These would not be relevant for me as I have been in my current job for 8+ years and received my degree at a time when a complete genome was something like phage lambda. Remember, it is fun to ask questions, but trying to analyze and interpret the responses can be difficult.

I like what Travis suggests on conferences and training. The survey can offer a few conference suggestions as to which you attended, perhaps as a table of conference name X years (2009, 2010, 2011). This would be very informative in selecting where to attend/present next time.

Answer 7

[Travis]

Probably relevant:

Level of freedom with IT resource (1-5).

5 being complete freedom, 1 being serious lack of permissions and reliability on centralised IS groups for installations/updates etc.

[Travis]

Fair point. In that case the question could be split into things like: Were you permitted to choose your own OS? Do you have admin privileges on your system?

[Michael Barton]

The most useful questions (my opinion of course) previously were those that were more concrete and quantitative. E.g. it was easy to count the number of instances of people describing themselves as using Ruby or what their salary was. Level of freedom here is subjective. Perhaps instead it could be "Were you allow to choose your own computer set up?"

[Michael Barton]

Good ideas, thank you.

[Larry_Parnell]

This is a very relevant situation/freedom/obstacle for many research groups. However, while I have some freedom over my machine, I have no control over the compute cluster we use. The question then becomes, What work environment do I have in mind when I answer such questions? Hmmm.

[Travis]

Perhaps a generic - Do you feel you have sufficient computational resource and access rights to do your job properly? - would be less verbose whilst capturing the required information.

Answer 8

[Russh]

as mention has been made of educational background: if your background (degrees/work experience prior to your bioinformatics work) was not in bioinformatics - how did you retrain?

Not sure if anyone else is interested in this

[Michael Barton]

Asking questions about field of undergraduate degree, masters degree, and PhD could determine whether people retrained.

[Larry_Parnell]

Hmmm. My PhD was in Cellular & Molecular Bio, but I added research that is now called computational biology or bioinformatics. Those terms did not exist at the time. Half my thesis was computational, but one would not know that simply from the dept. that granted me the degree.

[Russh]

but you'd know and hence wouldn't need to answer the question. I'm more interested to see if people have had to fund courses for themselves to get into bioinformatics from another field.

Answer 9

[Khader Shameer]

I would like to add the following 2 Q's:

Where do you work

• Industry
  
• Bioinformatics research lab
• Bioinformatics core in a University/Research institute
  
• Biology/Clinical lab

This will help us to figure out where/how we are distributed

How do you perform your day-to-day computing:

• Cluster / HPC maintained by lab
• Cluster / HPC maintained by University/Institute/Company
• Cloud (Amazon / Private / SDSC etc.)
  

I believe cloud will take a second place here :)

Answer 10

[Hranjeev]

Employability - from school to your first job

Edit#

• Employment status
• Employability index perhaps a combination of
• Ease of Employment - how fast you secured a job after completion of your studies
• Relevance of your bioinformatics job role to your studies

[Michael Barton]

"How many months have you been unemployed since becoming qualified?"

[Michael Barton]

"How many months after finishing higher education did it take you to find a job in bioinformatics?"

[Larry_Parnell]

I find that this question may be difficult to interpret or even irrelevant for many, especially for those who have been either in the field for a long time or at their current position for a long time. Perhaps, the question can be formulated as "Have you been at your current position for more than 5 years?" If yes, skip to next question (automatically with branched logic). If no, then ask the above.

[Michael Barton]

Could you be more specific? Provide an example question?

Answer 11

[Jeroen Van Goey]

After having taken the survey, I feel that it is too focused on researchers.

Some feedback to make future surveys more balanced:

• In "What is your current position", add some industry positions like "commercial software developer", etc.
• The software I write in my free time is open source. The software I write for my work is commercial, closed source. Somehow allow me to make this distinction.

Also, I have a personal blog, twitter account, and use stackoverflow, BioStar and github. But none of these individually I use every week. So I have checked none of the options, but that seems like I never use any of these services. So I would prefer a more options to answer there, like: uses every week / uses every month / ...