Guide to Applying for a Ph.D in Big Data
By Kat Campise, Data Scientist, Ph.D.
Ph.D. programs, in general, are a strenuous undertaking. You’ll spend between 4 to 7 years, on average, in deep and highly structured research on one topic with specific writing requirements. These won’t be blogs or superficial articles waxing poetic about the trials and tribulation of AI. You’ll be expected to publish and present your research to the highest levels of academia who will undoubtedly relish (at least some scholars will) in debating — if not outright challenging — every aspect of the research you conducted.
None of this is meant to scare you away from embarking on the Ph.D. journey. Rather, this is to prepare you for many years of sacrifice and, to be forthright, stress. Ph.D. completion rates hover around 50%. However, this statistic may be more promising depending on the graduate school you choose to attend and the program you intend to complete. For example, Duke University has Ph.D. completion rates as high as 95%.
By the conclusion of your Ph.D., however, you’ll be positioned as one of the leading experts in your chosen area of research. While this doesn’t make you omniscient or omnipotent (too many scholars conflate expertise with being downright arrogant), you will have more knowledge about a given subject than those at the bachelor’s or master’s degree levels. This knowledge is granular, meaning that through your applied research, you will have accrued greater understanding of the nuances involved in the problems you’ve studied at great length.
A Ph.D. is creational. The expectation is that you’ll create or discover something new in your research area. For example, if you’re in the midst of a Ph.D. in Data Science, deriving a brand new AI system, and then discussing how you arrived at this via your dissertation — which you will defend — is what a Ph.D. program will demand of you.
Should You Apply to a Ph.D. Program?
Most Ph.D. programs require full-time study. This will leave very little room for additional employment responsibilities, e.g., having a part-time job or attempting to work full-time. You won’t merely be reading others’ research and then repeating or summarizing it. You’ll critically analyze the strengths and weaknesses of their research, and then use it to inform your research design, development, and implementation. You’re building a brand new solution to a particular problem.
Many Ph.D. programs have a stipulation that you will be part of a teaching cadre, meaning you’ll be teaching either bachelor’s or master’s level students in your discipline. This is in addition to your research and writing. While these may be paid, the teaching assignments don’t tend to be as lucrative as jobs within private industry. For instance, a data scientist in the U.S. can earn approximately $120,000 per year. It’s extremely unlikely that you’ll earn that type of salary within your Ph.D. in Data Science program.
The flip side of this is that you can reach a six-figure salary once you complete your Ph.D. if you’re willing to take on the opportunity cost during your Ph.D. program.
So, should you apply?
If you are certain of the program, which includes having an idea as to what you want to research, you enjoy focusing on a problem (almost endlessly) and creating new solutions, and you’re willing to spend around 6 years of your life constantly reading, analyzing, writing, publishing, and presenting, then start by reviewing the next steps of the application process.
Step 1: Finalize School and Program Choice
Although there are a growing number of online programs, Ph.D. programs are still primarily an onsite experience for the sciences, technology, engineering, and math (STEM) disciplines. So, most Ph.D. applicants will need to take the university location into consideration along with the availability of the specific Ph.D. program.
Regarding program choice, ideally, you should have either a bachelor’s or master’s degree in a related discipline. In many cases, one of the application requirements is for you to have completed specific courses (or a directly relevant degree). Using data science as an example, all Ph.D. programs in data science currently require the completion of Calculus (at the very least, Calculus I), Linear Algebra, and advanced statistics.
Some programs go further and have programming requirements (Python, Java, R, etc.) along with coursework in data structures and algorithms. It’s rare to jump from a B.A. in English to a Ph.D. in Computer Science (or Data Science); not because someone isn’t capable of doing so, but due to the major “catch up” required in terms of extensive practicum in the subject. A Ph.D. is already rigorous without you needing to take a series of prerequisite courses.
Review the current professors’ research interests and publications. One of them is likely to be your advisor and you’ll need to invite others to be a part of your dissertation committee (if the Ph.D. is structured in that way). This will also help you to generate research ideas of your own while also helping your application to “connect” with the department’s goals and objectives.
Additionally, peruse the required courses. If you can find the syllabi for those courses, read through them thoroughly. Note the journals and journal articles they reference. If you can find them (many are locked away in pay for view gateways such as JSTOR, but Google Scholar may have them available for free via PDF), then start reading! Doing so will clue you in on both the professor’s research area — especially if they are an author for one or more of the articles — and the focus of both the particular course and the Ph.D. program.
Remember, the department and its constituents want a high Ph.D. completion rate (which also holds true for master’s and bachelor’s degrees). The prestige factor attracts more students and more students translate into more funding. It’s not all about the money, of course. But, they do strongly prefer candidates who will successfully complete the program and earn their Ph.D.
While you should read the program requirements carefully. Don’t hesitate to gather questions that you can’t find answers to (specifically about the program itself rather than “how do I apply”) and send an email to the Department Chair. Keep in mind that if this is in the middle of a semester, it may take them time to respond to you as they also have teaching, research, and other bureaucratic duties.
Step 2: Review the Application Process
Depending on the department’s website layout, usually, it’s pretty easy to find their “How to Apply” section. Wherever that is located, make sure you find it and review the materials you’ll need to send along with your application. Thus far, just about every U.S. university has an option for applying online (we’ve yet to find one who doesn’t accept online applications). An overwhelming majority of Ph.D. programs require the documents discussed in the steps below. As such, you’ll need to set aside additional time, and money, so that you’ll have all of the requisite materials.
Step 3: Gather Your Transcripts
All U.S. universities are going to ask for official transcripts. During the online application process, you may be asked to upload unofficial transcripts for review by their admission committee. Subsequently, the Graduate Department will request your official transcripts upon admission acceptance. If you have any gaps in education or there was a semester or two where you weren’t performing very well academically, this can be briefly (and professionally) addressed in your Statement of Interest or Letter of Intent; more will be included on this topic below.
Step 4: Test Scores
Some Ph.D. programs are moving away from the GRE testing requirement. Others will accept GMAT test results in lieu of GRE scores. But, STEM programs aren’t likely to abandon the GRE as part of the application process. You’ll need to pay close attention to any cutoff scores listed by the department and whether you should take the General GRE or its Subject Tests.
Depending on where you are located in the world, GRE fees range from $205 to $230. Subject Tests are $150 per subject. That aside, you’ll also need to spend time in test preparation mode which can be as little as 50 hours and as high as 120 hours. Your test preparation needs are unique and depend on many different factors. Most students perform better on one section over the other, e.g., if you have a Bachelor’s Degree in Math, the Quant section may be a breeze but your performance on the Verbal section may not be as stellar.
Also, keep the application due date in mind when scheduling your GRE test. Give yourself time to retake the test if need be while also ensuring that your test scores are received by the university before the application due date.
Step 5: Writing Samples, Resumes/CVs, and Letters of Intent
It cannot be overstated that scholarly work at the Ph.D. level requires a mind-numbing amount of writing (and research!). The department admission committee wants to determine if you can write at an academic level and if you have begun to form research interests. Essentially, they want to understand why you want to enter the Ph.D. program and how your studies will align with your career goals. All of this is part of determining not only your commitment but also your readiness.
Having industry experience is a bonus which is one of the reasons they ask for a resume or CV. As much as a Ph.D. seems to be “ivory tower” pontificating — admittedly, it can be — students who have some hands-on experience in the particular research area tend to have more successful outcomes — as do students who have a set of clear goals and objectives.
If you don’t have an academic writing sample, then this is the time to reach out to the Department Chair to determine what you should write about for application purposes. If you’ve completed a master’s degree, you should have your thesis to send. Some departments will explicitly state what the writing sample should contain. Summarily, if for some reason you don’t have a sample readily available, be prepared to create one.
What the department committee is likely not seeking is for you to have an already formed dissertation topic. If they’re seasoned academics, as they should be, they’re keenly aware that research interests evolve over time. But, as long as you have some direction, e.g., “I’m interested in researching how AI facial recognition can be accurately and equitably deployed in determining the likelihood of criminal activity”, then you’ll have a higher probability of making it to the acceptance pile.
Step 6: Letters of Recommendation
Sometimes referred to as “Letters of Reference” department requirements vary on the number and type of recommendation letters to include with your application. Usually, you’re required to send 3. Since you’re applying for admission into academia, recommendations from prior professors are the prevailing preference. However, an increasing number of universities also accept references from employers if they can include how your employment experience has prepared you for your intended academic studies.
The “how” of routing the reference or recommendation letters differs between universities. Some will still require that the letters are sent via postal mail directly to the department. But, there’s a shift towards simply uploading the letters as a PDF directly to your online graduate application.
Remember Self Care
Your application is viewed from a holistic perspective. Although GRE scores can be part of the admission consideration equation, most universities don’t view you as merely a test score number (which is one reason some are foregoing that requirement). As mentioned elsewhere, the department does want a high graduation rate along with generating scholars who are well-regarded in their expertise. The department admission committees are aware of the blood, sweat, and tears that committing to a Ph.D. program requires.
There is a high probability that you’ll experience disorienting moments including imposter syndrome. Life doesn’t always flow smoothly and definitely doesn’t stop just because you’re in the middle of your Ph.D. in Statistics (or whichever discipline you’ve chosen). It’s perfectly feasible to speak with your advisor about taking a short break from your studies so you can enact self-care. Only you can know and determine if that’s an action (or inaction) you need to take so you can return to your program revived and ready for the next set of challenges.