This section will address how to select which program you should attend.
MOTIVATIONS FOR DEGREE
Data Science is a multidisciplinary field that primarily combines Mathematics and Computer Science. In essence, a data scientist must have statistical knowledge and computer skills in order to solve complex problems. In this degree and career, you will be expected to use descriptive, predictive, inferential and causal models in order to explore and anticipate problems and come up with a solution based model. For these reasons, you should be looking into this degree if you have a strong grasp of and a strong interest in mathematical and scientific concepts.
Here is an article that explains the career of a data scientist. Note that this advice is catered to the USA, and may not completely apply to Pakistan.
TYPES OF DEGREE
In a nutshell, the following degrees are available for Data Science:
- Master’s in Data Science (this will take 1-2 years to complete)
- PhD in Data Science (this will take 4-7 years to complete)
Deciding whether to do a postgraduate degree in Data Science
Since Data Science is so heavily reliant on technological advancement and computer science, it is possible to become a data scientist without getting a postgraduate degree. Many successful data scientists have spent time self-teaching and learning the newest programming languages and working on data projects themselves in order to become expert data scientists. Many students also enroll in “bootcamps”, which are often programs that typically last 1-4 weeks (or longer) and train you in specific data science skillsets. They often offer you employment if you pass/do well. Ultimately, you will have to weigh the pros and cons of going through this sort of route or instead doing a postgraduate degree.
If you feel you are not too self-disciplined and know that you will be unable to teach yourself the right skills, then perhaps look at postgraduate degrees as an option. Also look at the types of jobs you want to be landing in the future and see if they require you to have a postgraduate degree under your belt. Lastly, think of any time constraints you have and how you can work around them if you go for a postgraduate degree.
You should read the following links to help you decide whether a postgraduate degree in Data Science is worth it for you:
Choosing between a Master’s and a PhD
- Due to the rapid advancements in technology, data science is a discipline that evolves extremely quickly. For this reason, many aspiring data scientists prefer the shorter Master’s program over the much longer PhD program. For example, something that you begin working on at the start of your PhD could become obsolete by the time you graduate — if you go for a PhD you will have to constantly keep track of new advancements.
- Look at the sorts of jobs you want to go for and see if they specifically require PhDs or if a Master’s will do. Often, a PhD can give you an edge in job applications, which is why it can be a good idea to go for them. Typically, research related jobs will require PhDs, while industry jobs will accept Master’s.
- Make sure that if you’re going for a PhD, that you know exactly what topic you are going into and you like your potential supervisors. You need to make sure you get the “right” PhD from the “right” supervisor in order to learn as much as possible.
- Note that in a PhD you will almost always have to take a qualifying examination.
- Here are some resources you can look at to make your decision:
How to select the best option for yourself (among specializations and sub-fields within this field)
There are numerous options to explore within the field of Data Science. You can combine the discipline with a host of other fields, such as business and psychology, depending on the electives and research opportunities that your desired universities offer. It is critical that you look at course descriptions for each university that you want to apply to when you are researching, and see if it is a good fit for you personally.
In order to decide what’s best for you, always think of the sort of career you want as a Data Scientist and where you want to end up. Discuss options with current mentors and future advisors when you are at that university, and look up jobs that you are interested in doing. Look up topics you are interested in and whether you can see yourself spending 5+ years on said topic (in the case of a PhD). If yes, actively try to find universities that offer electives or specializations in these subfields.
If you are interested in Data Science, you might also like::
- Computer Science
- No. of Universities to Apply: 5-8
Complement the above field-specific tips with general tips on program selection (under the tab of ‘selection’).
A lot of our tips talk about how you can strengthen your application, but you can build a stronger application when you’ve done the things this program values in the years prior to the application. The application itself is the communication part (in which you communicate what you've done to the admission committee); but this section gives guidance on the substance part (what you can actually do before you apply). In this section we talk about what you can do in the years leading up to applying that can make you an ideal candidate. Supplement the following tips with general tips (under the tab of ‘Pre-Application’) to become a competitive applicant.
Students interested in applying for Data Science in country do not need to have a specific degree in order to be eligible for admission, as committees look for people from a wide range of backgrounds. Universities do, however, expect you to have a background in Mathematics and Computer Science (see CourseWork section), and getting admitted without this is extremely difficult.
Do note, some universities require you to have a mathematically inclined degree, although this is extremely rare.
CourseWork and Transcripts:
For both a Master’s and PhD in Data Science, you need a strong foundation in mathematics, statistics, and computer science. Each program can have different coursework requirements that you must fulfill before applying, so make sure to read these carefully. Generally, you will need some of the following courses to be able to apply.
For both PHD and Master’s you are usually required to take the following:
- Probability and Statistics
- Data Structures (or Computer Programming)
- Advanced Calculus
- Linear Algebra
Some schools may admit you without you having fulfilled these requirements, but in that case they will require you to complete these courses (or equivalent) once you have enrolled into their program.
Most universities will not mention a cut-off GPA for their programs besides a 3.0, but having low grades is going to put you in an extremely difficult spot, especially if these grades are in Mathematics and Statistics courses. You should aim for 3.5-4.0, especially if you are going for a PhD.
For PhD Applicants: Research experience is sometimes a requirement, but even in places where it is not, it is highly recommended that you have some research experience under your belt. Publishing a paper in a refereed journal, as a co-author or author, is the best indication of great research experience. If this is not possible, do not get disheartened. You can also show your interest in research by perhaps being a Research Assistant for a professor working on a paper, or even being part of a smaller project and present your findings at a conference. Always keep an eye out for conferences and attend as many as you can, even if you are unable to present at them, as they are excellent for networking with the right people. Having a specific direction in your research can be helpful, especially if you wish to pursue this during your PhD and have specific supervisors that you want to work with who have also worked on the topic.
For Master’s Applicants: Research experience is usually not a requirement, but it can boost your application. Doing at least one research project, whether it be an independent study, senior project, or a project undertaken in a Statistics/Mathematics/Computer Science class, will elevate your application considerably. A research project is a great idea over the summer, or even during a gap year after you have completed your undergraduate studies.
Having prior professional experience is not always a strict requirement, but can be extremely useful in adding weight to your application. A lot of programs will ask for a Resume/CV and for these having prior Data Science experience becomes extremely important. It is particularly useful in the personal statement, where you will need concrete examples of how you have interacted with Data Science and how you know you are right for the degree. More than anything, work experience will help you understand if this is the right field for you. Try to find research opportunities or internships with places such as CERP (Columbus Energy Resources plc) or IDEAS (Institute of Development and Economic and Alternatives). Another good option would be to speak to your Statistics/Mathematics/Computer Science professors and ask what sort of opportunities they are aware of.
Complement the above field-specific tips with general tips on building a resume (under the tab of ‘Resume/CV’).
This is not a requirement for this program, but any volunteer work (such as helping organizing conferences relevant to the field) that can be linked to the field can make your application better.
The field of Data Science is becoming increasingly technological, and learning computing languages before you apply to a postgraduate program is always ideal. If you can, try to take courses that help you learn languages such as R, C++, and Python. If you are unable to take classes, you should look up online tutorials and classes to get yourself acquainted with at least one of these languages. Also get yourself acquainted with Linux.
This section provides an overview of general guidelines pertaining to the application process. It also delineates the key components of the application process.
Is this component required?
How important is this component (in the overall review of the application for admission)?
Standardized tests or entry exams
For PhD: GRE required, Mathematics GRE sometimes required
For Master’s: GRE often required
TOEFL/IELTS equivalent often required
Transcripts (past academic records)
Letters of recommendation
2-3 letters required
Resume or CV
Do not usually take place
- At this point, if you are seriously considering graduate school, begin your search by reading this guide and by searching the websites from the following links:
Complement the above field-specific tips with general tips on overview (under the tab of ‘overview’).
Pakistani applicants suffer most because of inadequate information -- or wrong information -- about essays and personal statements. This section will address those inadequacies specifically in relation to applying for this program. Supplement the following field-specific tips with general tips (under the tab of ‘essay’) to craft a stellar personal statement.
You will be required to enter at least one personal statement, or statement of purpose into your application. Some universities will have a prompt or question that you will have to answer, as well as instructions on what the word or page limit is, and what format the essay needs to be submitted in. Make sure you read these instructions extremely carefully and follow them. If you are given no instructions on the word or page limit, try to aim for 1-2 pages and do not go beyond 3 pages.
- It is recommended writing your own statement, and not using some pre-prepared format. Just give yourself enough time to do it.
TIPS ON GOOD AND BAD STATEMENTS
What is essential in the statement:
- Include a description of your education and experience as it relates to your future graduate career. Mention all research experience you have or special projects you may have done in a course, and try to combine this with your interests in Data Science. Mention any specific faculty members you want to work with at the university you are applying to, and why your background and interests suit this professor. Note that this advice is more important for PHD applicants.
- Mention, and explain clearly, any abnormalities in your application or in your path to graduate school. You may have a low GPA, or a low GRE score, or you may have taken a gap year or switched colleges or majors in the middle of your degree. You will need to give context to any of these issues or they could become a red flag or cause of suspicion for the committee. You do not have to go too far into your personal issues to explain, just give a big picture.
What are some elements of exceptional statements:
- More applicable to PHD applications: Be aware of the research going on in the universities that you are applying to, and show it in your statement. Highlighting potential areas where your prior experience overlaps with the research going on in the departments you are applying to will be golden and will make the committee consider you more closely.
- Indicating that you have experience and fluency in the following:SAS, SPSS, MATLAB R, Python, Java, C/C++, Hadoop Platform, SQL/NoSQL Databases.
- Communication skills are extremely important, and you can show this both through the way you write your personal statement and by talking about any experiences where you communication skills have shone.
- Showcasing that you have experience in:
- Math (g., linear algebra, calculus, and probability)
- Machine learning tools and techniques
- Data mining
- Data cleaning and munging
- Data visualization and reporting techniques
- Unstructured data techniques
- Focus on your aspirations and have a direction. This is the most important part as it helps the committee see whether you are the right fit for them or not.
What are bad statements/ what things to avoid:
- Spelling mistakes/grammatical errors.
- Too much information about your high school achievements and/or activities.
- Repeating things from the rest of your application. The essay is a space for you to provide information that is not apparent in the rest of your package.
This section will cover the basics about recommendation letters, which are one of the most important parts of the application process. Supplement the following field-specific tips with general tips (under the tab of ‘recommendations’) to ensure you have strong letters of recommendation.
You are typically required to have 2-3 recommendations for this degree. The process usually takes place online after you have submitted your recommenders’ contact information within your application. They are then contacted and asked to fill out/write your recommendation.
The best sort of letter of recommendation will come from a professor who knows you well and with whom you have a connection, especially in the field of Mathematics, Computer Science or Statistics. Having a minimum of one recommendation from a professor who has taught you one of these subjects, who can speak of your abilities in data science, is crucial. Other good referees in this degree would be research mentors, and supervisors. If you have worked in a Data Science related job, having someone recommend you from your work environment is also going to look extremely good on your application.
TIPS ON GOOD AND BAD LETTERS
What is essential in the LoRs:
- This is where the referee should talk about any low grades or shortcomings in your application and justify why that is so. They should also explain why you should still be considered despite these shortcomings.
- Descriptions and indications of your commitment and passion towards Data Science.
What are some elements of exceptional LoRs:
- Detailed representation of your statistics and mathematical skills from someone knowledgeable in the field. This should be coupled with reasons why you are fit for this degree and school.
- A candid assessment of your abilities — your strengths should be highlighted, and weaknesses should also be mentioned with clarification. Adding how graduate school would help in alleviating these weaknesses would be ideal.
- A description of your potential and what you could add to the program, such as research interests (especially for PHD admissions).
What are bad LoRs/ what things to avoid
- Vague and generic descriptions of your work and abilities.
- Repetition of aspects from your personal statement. There should of course, be confirmation of your passion for the subject and your abilities, but if your recommendation is just a duplicate of your personal statement, it will not give the committee a good impression.
This section will cover everything else related to the application process; including transcripts, interviews, resumes, and standardized tests.
The General GRE is almost always required for the PHD, and often required for the Master’s. Go through the list of Master’s programs here to see where it is required, optional, and recommended. Go through the list of PHD programs here to view the same information.
The Mathematics GRE is sometimes required by PHD programs. Rarely, it may be required by Master’s programs.
The TOEFL or equivalent are often required to prove English language proficiency.
Complement the above field-specific tips with general tips on preparing for standardized tests (under the tab of ‘tests’).
FINAL COMMENTS ON APPLICATIONS
Complement the above field-specific tips with general tips on final comments on applications (under the tab of ‘overview’).
This section will cover approximate costs of the program and provide information of resources that may help with funding. Complement the following field-specific tips with general tips on finances (under the tab of ‘finances’).
TIPS ON FUNDING OPTIONS
Research assistantships and teaching assistantships are great ways to fund your postgraduate studies in Data Science. Opportunities can be found with specific professors and in specific departments, so make sure to keep a look out constantly, and don’t be afraid to send out emails. This is a great resource that can help you in trying to successfully get an RAship.
The following sources were consulted in developing this tip-sheet and we encourage you to consult these sources for additional information and guidance on your application.
Furthermore, the following sources were also consulted in developing this tip-sheet: Discover Data Science (1), Medium (1) (2) (3) (4) (5) (6) (7) (8), New York University, Brown University, Yale University, Andreas Kretz, Data Science Central, Chapman University