Leveraging your PhD Data Skills.
Understanding how to manage, interpret, analyse and visualise data is a hot skill for a range of jobs. Learn how to leverage your PhD to emphasise that these are skills you already possess.
Throughout your PhD you’ll be obtaining a range of skills that apply to many different contexts. One of the more common skills, especially in STEM subjects, includes exceptional data skills. However, this doesn’t just apply to your ‘traditional’ STEM disciplines, as it’s also relevant for non-STEM PhDs too. PhD’s in Geography, Sociology, Psychology, Economics and many more also provide you with good data skills. If you’re unsure as to whether you have data skills all you need to ask yourself really is if you know what SPSS or STATA is and whether you have had any ‘statistics’ training during your PhD, Master’s degree, or your undergraduate degree. For those who have had more advanced training in Python, R, MATLAB, SQL, or anything along those lines – you’re in a great place.
Data itself is becoming an increasingly popular skill within the current job market. As we move to a more digital age with things like social media, online shopping, the improvements of AI and Machine Learning data crops up in almost every business. What’s also staggering is how much data is being collected that companies just simply don’t know how to use. Due to this lack of knowledge, there is less emphasis on significance testing and hypothesis development - otherwise known as A/B testing. Of course, this varies from industry to industry, but in general – most people haven’t had as thorough data training as PhD students, placing you ten steps ahead of the competition. In the working world, the use of data typically falls under two roles, a data analyst and a data scientist.
Data analyst roles are simply to be given data sets and then provide some ‘insights’ that can be used for a range of purposes relating to a business strategy. The reason PhD students are more than ready to take these roles on is because ‘insights’ have nothing to do with significance testing in any way shape or form. The vast majority of the time it just consists of ‘describing’ and ‘interpreting’ data. For PhD students, you would usually gravitate to using the mean and standard deviation or normal distribution to get a general idea of your data. However, in industry this isn’t even the main focal point. If you can take your data, stick it into a nice-looking bar chart or line graph – you’re good to go and can consider yourself a well-equipped data analyst. The parts which come more challenging are learning how to interact with wider industry practices. Things like data warehousing and databases are less likely to be on your skill set as an academic, however having general data skills and a desire to learn about these and programming languages like SQL will set you up for success.
In fact, the further and further you get away from STEM disciplines these data analyst roles become less complex. Industries that specialise in marketing, commerce, retail, or anything that collects some sort of data will need ‘data analysts’, even companies that don’t provide anything data related, such as airlines or charities will employ data analysts internally to help with their own business goals. The ironic thing about this process is that senior staff typically don’t have much statistics or data training (hence the need to dedicated data teams). In turn, mangers or project leads actually end up asking for really simple, or rudimentary ‘insight’s’ which you could do in Microsoft Excel – although it just doesn’t sound as cool.
Because of this, the most important skills you actually need as a data analyst when coming out of a PhD isn’t necessarily data skills (as you already have this), it’s actually skills relating to data management, interpretation and presentation. If you have good presentation skills, know how to create a narrative with data, and support this narrative with a couple of cute graphs/visualisations you’re in business.
Another role that relies on data skills is a data scientist. These roles require a slightly more advanced set of skills – but again, there’s a high chance you have these skills already. More often than not data scientist positions include some sort of A/B (or hypothesis) testing. This requires you to (on occasion) look at a p-value. Extending this further, data scientists are also employed to do ‘machine learning’ – it’s a real buzz word. For PhD students, machine learning is a really intimidating word. However, you actually have no reason to fear it. Often, machine learning just refers to a range of modelling techniques such as regression models. Here, you’re ‘insights’ are to just provide information about ‘features’ or ‘things’ which might influence (consumer) behaviour.
In other words, what variables of your linear regression model have the most weight or explain the most variance. If you can do that, hey you can put machine learning on your CV. Typically, these aren’t difficult things to do for the vast majority of PhD students, especially those who have had statistics training. The part that get’s more challenging is the wider set up on how you process, interact with and manipulate big data. Again, knowledge of data warehouses, data lakes, cloud systems, endpoints and so forth can make this more challenging. Having said that, general statistics and programming/data knowledge is more than enough to get you into the world of data as a data scientist or a slightly different data role that still requires these skills.
Building on from this, you’ll also be expected to do these types of analyses within a pre-determined programming language. If you can do it in Python or R, you’re good to go as these are the most commonly used tools. If you don’t have these skills – you can learn them relatively quickly, they’re both free to use and because of their open source nature you can troubleshoot any problem online. If you are interested in these tools, check out this post as there are some links for great resources to get you started.
Up to this point, we’ve highlighted how possessing a good understanding of data, statistics, and some understanding of programming languages can be leveraged to find a role in data, either as an analyst or scientist. Additional skills you’ve acquired from your PhD such as presentation skills, data visualisation (for them cute graphs remember), communication, data interpretation, data management, working with people, problem solving, critical thinking (when interpreting data), and just a general understanding of how research is conducted really do aid these career options.
Once you get to the end of your PhD and you want to transition out of academia, spend some time thinking about data as a potential career option. Due to the variability of these types of roles it’s possible to find a company/career that aligns with your values. Furthermore, data skills can be applicable to a range of job titles. A few examples include:
Data analyst and data scientist as mentioned
Product analyst
Intelligence roles
Insight analyst
Business analyst
Roles within marketing and advertising
Psychometrician
Researcher (whether that be insights, operational, market, or UX)
Don’t be shy to throw your hat in the ring! If you’re still midway through your PhD and enjoy working with data, try to learn as much as possible on the side. Often universities provide additional training in statistics and data, or even with programming languages like Python and R so make the most of these opportunities! You never know when these skills will come in handy.
