4 Mistakes Aspiring Data Scientists Make and How to Avoid Them

1. Miscalculating How Much Time Is Involved

To be frank, it takes a lot of time and a lot of work to become a data scientist. How much really depends on each individual and how much background you already have in the field. For example, if you have a couple of mathematics degrees or several years experience in Information Technology or software development, then you will get up to speed a bit faster.

But the reality is that regardless of relevant experience, you will need to spend a lot of hours outside of your day job trying to learn all the ins-and-outs of data science, and it can be daunting. 

To adequately measure the time/work commitment, you will need a strategic view of data science and all of the constituent components that make up data science (e.g. inferential statistics, visualizations, machine learning, big data, R/Python/SAS/Excel,etc.). Do some research on the time it takes to master each area, and only then will you begin to understand how big of a time commitment data science really is. 

2. Underestimating the Commitment Needed

Everyone who has ever become an expert understands commitment is needed to achieve just about anything. Data science is no exception. It requires topical knowledge, commitment, and attention to detail. You will need to know descriptive statistics, inferential statistics, data visualization, and how to handle data sets. You will need patience, and the ability to think analytically. You’ll need an understanding of relational databases, SQL, Excel, machine learning, and so much more.

For a full list of the hard and soft skills, read the article, “What Skills Do I Need to Become a Data Scientist?”

To gain realistic expectations for the amount of work involved, you will need to extract resources from everywhere. This can include books and blogs, videos, podcasts and Data Science courses. You’ll need to spend some amount of time to determine the total outlay of time and effort (usually over months/years) to become a data scientist.

3. Lacking a Plan

One of the most common mistakes I see in aspiring data scientists is a total lack of planning. How on earth do you know if you are making progress, or if you are still going in the right direction if you don’t measure yourself against a strategy? 

Using whatever your preferred method (a piece of paper, spreadsheet, Google Doc, etc.) literally lay out your goals, and include dates, rain dates, and milestones.

There is too much to learn and see and practice to simply “shoot from the hip.”

Here is a small example of an annual plan:

• January: 
  • Start to brush up on Introductory Statistics
  • Install and start R or Python
  • Watch some videos on R or Python
  • Learn the data type, data structures, packages and libraries of R or Python
  • Start using R/Python and make a commitment to use it every day
• February:
  • Read a data science book
  • Start listening to data science podcasts
  • Continue learning statistics holistically 
    • Descriptive statistics
    • Inferential statistics
    • Predictive statistics
    • Nominal, ordinal, interval and ratio values
    • Various statistical tests: 
      • T-test
      • ANOVA
      • Chi-Square test for independence
    • Load a dataset into R or Python
    • Continue using R/Python and use it every day
• March:
  • Continue the Introduction to Statistics
    • What is Hypothesis Testing?
    • What is correlation?
  • Continue using R/Python and use it every day
• April:
  • Review topics in the first three months
  • Continue the Intro to Statistics
    • What is the t-test? F test?
    • What is ANOVA?
    • What are non-parametric tests?
  • Continue using R/Python and use it every day

Continue with composing your plan and doing the tasks and objectives outlined in the plan. Then, revisit it bi-weekly. See how well you are staying on target.

Sometimes you will fall behind. It happens. Sometimes you will veer from your plan. Realign and recompose if necessary but do not be discouraged. 

4. Limiting experience to auditing online courses 

This is a big mistake unless you can verify and validate that you actually took the course and absorbed the content. 

At some point, you will need to prove that you have made this commitment and that you have thought everything through. Just saying that you audited a course or two (or even 10 or 20) won’t be enough to sway any hiring manager from even granting you a phone interview.

There is nothing inherently wrong with auditing courses. You can learn a lot by doing so. But when you need to prove that you have the knowledge (especially when you start to float your CV to prospective employers), it will be very difficult to convince anyone that you learned the material and did the work. A certificate ameliorates this potential roadblock by telling those employers that you are more than familiar with the concepts and language of the field and that you can follow through on a commitment.

So take a Data Science Bootcamp—one that blends exploratory data analysis, R programming or Python, basic statistics, introductory machine learning and visualizations into one course. A course that covers these Data Science basics will help you learn how much time you’ll need to devote to each area. This will help you to plan out a strategy and timeline.

You will also have enough information at this point to decide if you really want to commit to this process. You will be able to plan and come up with a strategy for achieving this hard-fought goal of becoming a data scientist. It won’t be easy and it won’t be quick. But if it’s a role you enjoy and have a passion for, it will be worth it.