Introduction to data science

Data is everywhere

The PPDAC cycle¹

• A general approach to data-driven problem-solving
• Data literacy and data science as opposed to statistics

1. (Spiegelhalter 2019)

Replication and Reproducibility

A fully reproducible¹ study has

• Available data.
• Computer code (software) that produces the results of the study.
• Documentation that describes the software and data used in the study, and
• ways to share the data and code.

1. (Peng, Dominici, and Zeger 2006)

Reproducibility and transparency outside science

• Data are valuable and often hard to get
• Data may guide good decision-making, if understood
• Efficient management and use of data requires data literacy and data science skills

Tools in data science

We want software where analyses can be:

• Human- and computer-readable, meaning that we want to be able to write scripts or computer programs that execute the analyses.
• Documented, meaning that along the code, we want to be able to describe what the code does.
• Available and able to share with others, meaning that our analyses can be run on open and free software to maximize the ability to share them.

• Human- and computer-readable, meaning that we want to be able to write scripts, or computer programs that execute the analyses.
• Documented, meaning that along the code we want to be able to describe what the code does.
• Available and able to share with other, meaning that we analyses can be run on open and free software to maximize ability to share them.

Tools in data science

• R and RStudio
• Quarto
• Git and GitHub

Tools in data science ¹

1. (Wickham and Grolemund 2017)

How to learn how to code

• There are (almost always) multiple solutions to a problem.
• Someone else has already had the same problem
• Find your motivation
• “Microdosing”

• There are (almost always) multiple solutions to a problem. When faced with difficulties, do not give up trying to search for a perfect single solution. Instead know that there are multiple ways of defining the problem and therefore multiple ways of making stuff work.
• Someone else has already had the same problem. The internet is full of questions and answers, also related to what ever problem you might have. Learning how to write “googleable” questions is a great skill. By adding “in R” to your problem in a google search term often helps finding R related solutions.
• Find your motivation. The skills that you will learn in this course are transferable to countless potential work related roles for the future you! To be able to showcase these skills may lead you to your dream job! Find your motivation for learning how to analyze data and communicating insights!
• “Microdosing” statistical learning. Replace your social media influencers with R users and data scientists! I find R people on Twitter and mastodon. Tweets and posts in this format keeps your R brain going!

Data in practice - Storing data for everyday use

• Spreadsheets can be used for efficient storage of data for everyday use
• Spreadsheet software contains functions that adds information on top of the data…

All happy families are alike; each unhappy family is unhappy in its own way.

Data Organization in Spreadsheets: Empty cells¹

1. (Broman and Woo 2018)

Data Organization in Spreadsheets: A tidy version (happy family)

Data Organization in Spreadsheets: Nonrectangular layouts

Data Organization in Spreadsheets: Data dictionary

Data Organization in Spreadsheets: Plain text

Minimize the risk of failure by adopting good habits!

From wikipedia: Maria Spelterini (1876) is walking across a tightrope across the Niagara Gorge, from the United States side to Canada, with her feet in peach baskets. In the background is the Niagara Suspension Bridge, which is full of spectators. In the distant background is the Niagara Falls.

source: https://commons.wikimedia.org/wiki/File:Maria_Spelterini_at_Suspension_Bridge.jpg#/media/File:Maria_Spelterini_at_Suspension_Bridge.jpg

References

Broman, Karl W., and Kara H. Woo. 2018. “Data Organization in Spreadsheets.” Journal Article. The American Statistician 72 (1): 2–10. https://doi.org/10.1080/00031305.2017.1375989.

Peng, R. D., F. Dominici, and S. L. Zeger. 2006. “Reproducible Epidemiologic Research.” Journal Article. Am J Epidemiol 163 (9): 783–89. https://doi.org/10.1093/aje/kwj093.

Spiegelhalter, D. J. 2019. The Art of Statistics : How to Learn from Data. Book. First US edition. New York: Basic Books.

Wickham, Hadley, and Garrett Grolemund. 2017. R for Data Science: Import, Tidy, Transform, Visualize, and Model Data. 1st ed. Paperback; O’Reilly Media. http://r4ds.had.co.nz/.