Skip to main content

Introduction to Data Science

Before we begin

1.1 Expectations

Setting reasonable and clear expectations are a cornerstone of adult education. This is partly the responsibility of the course administration and already began with the announcement of the course. Nonetheless, setting expectations remains largely the responsibility of the course instructors.1 Ideally, the instructors’ expectations of the students are both reasonable & clearly communicated. Students should anticipate that an instructor’s expectation are higher than they imagined. More on this below.

It’s also the instructor’s responsibility to make clear what the students can expect from both the instructors and the class itself. This is the process of managing expectations. Motivated but inexperienced students ofter have unrealistic expectations of what can be achieved, even in 3 months.

1.2 Challenges to Setting Expectations

In this course, the expectations of both the instructors and the students are affected by 3 considerations.

1.2.1 1 - Many moving parts

First, data science, as a STEM2 discipline, is very technical. That is, there are many moving parts to keep track of.

You’ll learn new computing tools, programming languages, mathematics & algorithms, on top of some design, philosophy, probability theory and ethics – and there’s still more!

This has the consequence of increasing:

  • The difficulty of the subject matter
  • The likelihood of missing a hastily-made false conclusion, or
  • The chance of an outright system failure, leading to delays and increased costs.

To be sure, non-technical disciplines are also difficult, but the pace of change, unexpected incompatibilities, debilitating bugs, difficulty & diversity of the material pose a particularly difficult challenge for technical disciplines.

1.2.2 2 - A wealth of material, a paucity of time

Second, there is wealth of material that we must cover in a limited amount of time. The amount of material, its diversity, depth & difficulty already poses a challenge, but we must cover enough material in 12 weeks that you can continue your journey on your own.

There is no official certification to become a data scientist. This is more worrisome when we consider the high demand and poor understanding of data science in companies. Thus, there is base level of proficiency that we require before you successfully pass the course. We don’t want you to misrepresent yourself are knowning more than you do, but there

1.2.3 3 - Large spread of backgrounds

Third, the participants are likely to have a wide spread in both their academic backgrounds and capabilities or familiarity with the new tools used in this course. Although we will do everything we can to help each student success, there are limitations. Given the breadth of the material, we are forced to move on to the next topic even if not all students are full comfortable with the material.

1.3 What can students expect of …

1.3.1 … their instructor>

Student’s can expect that, barring illness or other unforeseen event, the instructor will:

  • Be available for, on average, 4 hours a day from Sunday - Thursday for the duration of the course.
  • Address student questions in the group discussions, when possible, or,
    • If relevant to overall discussion, research the solution, or,
    • If not relevant, attempt to direct the student to appropriate resources
  • Make themselves available at periods throughout the course to assisstat

Students should not expect:

  • Prompt responses to non-essential questions posted on the Slack workspace.

1.3.2 … their TA?

1.3.3 … this course?

1.4 What do the instructors expect of the students?

This is a 12-week full-time immersive course.

1.5 Resources

We’ll use …

  • x
  • y
  • z

1.6 Our goals

This course will introduce you to an abundance of technical skills, both practical and theoretical. You’ll learn the two programming languages of data science, R & Python,

The thing to remember about technical skills is that they can all be mastered, eventually. Just like a difficult puzzle, what is first impossible becomes possible, what is first unknown becomes known. To be sure, combinations of innate talent, prior knowledge, strong work ethic, curiosity, genuine interest & motivation, complementary instruction and a supportive environment all help! Nonetheless, there are some things that are more difficult to learn.

1.7 The Hard Parts

There are three broad areas that are addressed throughout the course that are not

  1. Input -> How to learn - i.e. How do we receive & retain new knowledge?
  2. Processing -> How to think - i.e. How does a data scientist approach their work?
  3. Output -> How to talk - i.e. How do we transmit knowledge & make an impact?

1.7.1 How to learn

  • Think about the big picture
  • Think about the baggage you’re carrying
  • Think about repetitive patterns

1.7.2 How to think

  • What is the process of doing data science?

  • How can I contribute to the task at hand? What does my background, training, position, network, etc. allow me to do, that no one else can?

  • There may be no such thing as a dumb quesion, but what is a “good” question?

  • How can think of informatvie and useful question,

  • How do I know what I know?

  • What does our data look like?

  • What form can and should our data take?

  • Which questions can our data really answer? Which can it not? And which are easily mistaken between the two groups?

  • Can we ever really know the Truth? i.e. Can we know what reality actually looks like?

  • What are the limits of scientific knowledge? What is the scientific method?

  • What is the relationship among variables? and why? Is it useful to know about it?

  • How do they influence each other?

  • Is an extraordinary event noteworthy, or just due to chance?

  • How certain are we about our results?

  • Can we falsify our hypothesis?

  • Where did our hypothesis come from?

  • What forms of bias may be present?

    • Which are avoidable? How?
    • How can we deal with influencef our work

  • What is a model?

    • What are its limitations?
      • How can we uncover and know for sure?
      • How can we overcome them?
    • What does our model get right?
      • What does it get wrong? Why?

  • What are assumptions?

    • Why are they important to recognize?

1.7.3 How to talk

  • What is communication, really?
  • What are the basic & universal principles of good communication?
  • How can I offere critique my self and others? What makes critique constructive?
  • Who is the audience and what are their needs?

1.8 Some Dangerous Fallacies

You may have encountered subject matter that was intuitive and easy for you, but confusing and difficult for your peers. Sometimes this leads to the erroneous conclusion that we all get it or don’t. In reality, there are a few important points to remember:

Fallacy Truth
You’ll know a subject is for you because it will feel easy. Even if you love a topic, you may still encounter difficultly learning it well. This doesn’t mean you aren’t cut out for that field! But also, you are the best judge of your capabilities.
STEM folks just get it and the rest of us should just stick to the humanities, social sciences and creative fields. and Technical skills the only underst people are just that Some of It so, you may th others the situation where ahad the experience that experienced that learning
If I don’t understand what was taught, there’s something I did wrong (e.g. not paying attention, not taking notes or reviewing material). Sure, we all have to take responsibility in our own education, but not every teacher, explanation, or exercise will click with every student. The larger the class, the less personalized instruction is possible. You are encouraged to seek other sources of learning material. Find a resource that speaks to you and share it with the class, other students may find it useful. If many students didn’t understand the lesson, the instructor needs to be informed.
We each have our own learning style. Some of us are visual learning, others are hands-on learners. We all learn by engaging with the material. We are learn better when there are the following.

A list

  • Live demonstrations
  • Engaging class discussions
  • Appealing & informative supporting visuals
  • Clear and simple text appropriate to the skill level of the class
  • Practical exercises that reinforce the material
  • Positive feedback for success & suggestions for improvement otherwise

instructor nees ’s time for We each not sEach of us has some subjuect area that We all have topics that we find personally easy that of us find some topics easier than others,

In an academic sense, the Science is our first and formost our interest. Both instructors and students work in service of the work, and must not purposely do anything to compromise it. This means you grade is not the priority. If you