Career Specific Information Requested

I have a job offer with government agency and as a part of the restrictions in the interview process I am not allowed to interview someone that actually works at the agency for which the job offer is from. I am looking for some more information on what the day-to-day responsibilities would be for this position. I am a relative novice with R programming and want more information as to what I would be getting myself into. Here are the listed duties in the job posting. Any information would be greatly appreciated, thanks.

  • Provide support for statistical software tools – particularly R packages – including but not limited to package documentation, function/package testing, and writing/updating R scripts to support package updates and new functionality;
  • Utilize version control software for tracking updates to code in software tool and data analyses;
  • Manage, manipulate, and assess scientific data (new and existing) for statistical analyses and incorporation into analysis tools (e.g. R packages) supporting researchers in the center;
  • Contribute to documentation/manuscripts to communicate results, software updates or features, statistical methods/models, and data analyses with stakeholders and researchers; and
  • Contribute to the design and implementation of methods that address goals including, but not limited to, developing best practices for data reporting, model assessment, package management, and similar efforts to increase transparency, reproducibility, and efficiency.

Job descriptions may, or may not, be reflective of what the new hire actually does. Much depends on the scale of the overall organization, how the unit fits in and what the organization does. So here’s my guess based on having worked in small and medium agencies, small and large A/E firms, Big Law and large and jumbo (JPM) banks.

  • Provide support for statistical software tools – particularly R packages – including but not limited to package documentation, function/package testing, and writing/updating R scripts to support package updates and new functionality;

Translation: help desk with the potential for an unpaid side hustle of doing other people’s work after hours. Because the Agenda package has to go out in the morning, and I’m going to be watching the big game tonight.

  • Utilize version control software for tracking updates to code in software tool and data analyses;

Translation: Go to person for all of things GiTHUB. “GOTO Considered Harmful.

  • Manage, manipulate, and assess scientific data (new and existing) for statistical analyses and incorporation into analysis tools (e.g. R packages) supporting researchers in the center;

Translation: Cleaning up sloppy spreadsheets because the IT department can always find an excuse why it’s impossible to have a simple SQL database on an existing SQL server so that it would be harder to create random errors. Use scripting tools, especially regular expressions to convert strings into date types and numeric types. Fix inconsistencies in spelling and abbreviations, in use of various signifiers that must be shown as NA and alerting principals when remaining data may prove insufficient for lack of power, say. Reconcile dueling spreadsheets using forensic tools, including tracing versions through email chains. Surreptitiously checking data with tools such as Benford’s law to detect faked data.

  • Contribute to documentation/manuscripts to communicate results, software updates or features, statistical methods/models, and data analyses with stakeholders and researchers

Translation: A two-fer, running rocygen on functions and trying to extract a coherent explanation from author as to what the function actually accomplishes. That’s documentation. Manuscripts includes graphics, of course, and proofreading and running edits. Becoming close friends with Quarto, pandoc, LaTeX, Word and the style sheets of each publication and a house style derived from committee work that pleases no one.

  • Contribute to the design and implementation of methods that address goals including, but not limited to, developing best practices for data reporting, model assessment, package management, and similar efforts to increase transparency, reproducibility, and efficiency.

Translation: Make sure you know what everyone else in the center’s professional circles is doing. Support your claims that you follow [non-existent] “industry standard” practices.

Unsaid: getting up to speed in the scope of domain knowledge, keeping an eye on developments from bioconductor.org if in life sciences. Monitor CRAN taskviews for center’s areas of expertise, review the “40 top packages” drops. Maintain list of must have packages and check release notes for breaking changes and dependency changes. Impose discipline on version upgrades to avoid stumbling into disasters and getting left behind on functionality.

I could go on in this vein, but I hope you get the idea.

This topic was automatically closed 42 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.