Data science is the practice of using data to answer questions, make predictions, and support decisions. It sits at the overlap of statistics, programming, and domain knowledge. Statistics helps us reason about uncertainty, programming lets us work with large or complex data, and domain knowledge keeps the analysis connected to the real world.
This matters because organizations use data science in medicine, climate science, business, sports, engineering, and many other fields.
Key Facts
- Data science combines statistics, programming, and domain knowledge to extract useful information from data.
- A common workflow is question, collect data, clean data, explore data, model data, interpret results, communicate findings.
- Mean = (sum of all values) / n.
- Sample proportion = x / n, where x is the number of successes and n is the sample size.
- Prediction error = actual value - predicted value.
- Correlation measures association, but correlation does not prove causation.
Vocabulary
- Data Science
- Data science is the field of using data, statistics, computing, and subject knowledge to answer questions and solve problems.
- Dataset
- A dataset is a collection of observations, measurements, or records organized for analysis.
- Model
- A model is a simplified mathematical or computational representation used to describe patterns or make predictions.
- Feature
- A feature is an input variable or measurable property used by a model to analyze or predict an outcome.
- Bias
- Bias is a systematic error in data collection, analysis, or interpretation that can lead to unfair or inaccurate conclusions.
Common Mistakes to Avoid
- Treating messy data as ready to analyze is wrong because missing values, duplicate records, and measurement errors can distort every result.
- Confusing correlation with causation is wrong because two variables can move together due to coincidence, a hidden third variable, or reverse cause and effect.
- Using a model without checking its accuracy is wrong because a model can fit old data well but perform poorly on new data.
- Ignoring domain knowledge is wrong because a statistically strong pattern may be meaningless, impossible, or misleading in the real situation being studied.
Practice Questions
- 1 A data scientist records daily website visits for 5 days: 120, 150, 130, 170, and 180. Find the mean number of visits.
- 2 In a survey of 200 students, 86 say they use a fitness app. Find the sample proportion who use a fitness app and write it as a decimal.
- 3 A model finds that ice cream sales and drowning incidents both increase during summer. Explain why this correlation should not be interpreted as ice cream causing drowning.