A while ago, I attended an event full of data scientists. A panel had been convened and I asked them if there was a need for a professional organisation or body to set out a list of standards and competencies of a data scientists in order to remove some ambiguity about the role. Some said yes. Others said no.
One said that there should be an organisation with such a responsibility because there are some core skills necessary to do that job and to call yourself a data scientist without having covered the fundamentals of it might be disingenuous. However, another said that the discipline might be too immature at the moment for a professional and/or governing body.
A third panellist gave a perspective from the education side of training people to be data scientists. This person said that with the increase in the number of data science master’s courses, some guidelines probably are needed. This is because the people who turn up to do such a course will want to know what they are getting and the assurance that they will graduate with the skills they need to
do the job they want to do.
The person added that there is an organisation that accredits courses in software development which gives a stamp of approval and some consistency to the curricula of software engineering master’s courses. Such an organisation that delivers accreditation would be useful for universities that deliver data science courses.
One possible cause of the problem is ‘data scientist’ being used as a blanket term that covers many areas of specialism. This leads to people who are outside of the discipline to believe that all data scientists are experts in all of those specialisms. One panellist said that the data science team has got to be good at everything whereas not every individual has to be an expert.
One issue commonly faced by data scientists is having to deal with recruiters and people from other business units who get caught up in the hype and buzzwords. One data scientist said they had received a job spec from a recruiter that called for the candidate to be a restrictted Boltzmann machine expert. A few questions to the company revealed that they didn’t know what a restrictive Boltzmann machine was.
Going forward, as more people gain a better understanding of what data science and data scientists can do, and as the discipline matures, job titles can become more specific and better describe the person’s actual job and tasks. Also, if companies that want to become data-driven but have a FOMO (fear or missing out) mentality, they could be bedazzled by the fancy methods and processes instead of thinking about the problem they want to solve.
It seems that time and greater understanding will be the things that relieve the pressure from data scientists from the expectation of being data polymaths who can solve and do everything. Also, a greater comprehension of the importance of the team wouldn’t go amiss. As one said: “We don’t want to lose the sense that we are in this together and this is a collective problem-solving activity.”