Fetching contributors…
Cannot retrieve contributors at this time
172 lines (97 sloc) 4.92 KB
title subtitle author job framework highlighter hitheme widgets mode
Types of Data Analysis Questions
Jeffrey Leek, Assistant Professor of Biostatistics
Johns Hopkins Bloomberg School of Public Health
io2012
highlight.js
zenburn
 mathjax
selfcontained

## Types of Data Analysis Questions

In approximate order of difficulty

• Descriptive
• Exploratory
• Inferential
• Predictive
• Causal
• Mechanistic

Goal: Describe a set of data

• The first kind of data analysis performed
• Commonly applied to census data
• The description and interpretation are different steps
• Descriptions can usually not be generalized without additional statistical modeling

## Descriptive analysis

http://www.census.gov/2010census/

## Descriptive analysis

Goal: Find relationships you didn't know about

• Exploratory models are good for discovering new connections
• They are also useful for defining future studies
• Exploratory analyses are usually not the final say
• Exploratory analyses alone should not be used for generalizing/predicting
• Correlation does not imply causation

## Exploratory analysis

Liu et al. (2012) Scientific Reports

## Exploratory analysis

http://www.sdss.org/

Goal: Use a relatively small sample of data to say something about a bigger population

• Inference is commonly the goal of statistical models
• Inference depends heavily on both the population and the sampling scheme

## Inferential analysis

Correia et al. (2013) Epidemiology

Goal: To use the data on some objects to predict values for another object

• If $X$ predicts $Y$ it does not mean that $X$ causes $Y$
• Accurate prediction depends heavily on measuring the right variables
• Although there are better and worse prediction models, more data and a simple model works really well
• Prediction is very hard, especially about the future references

## Predictive analysis

http://fivethirtyeight.blogs.nytimes.com/

## Predictive analysis

http://www.forbes.com/sites/kashmirhill/2012/02/16/how-target-figured-out-a-teen-girl-was-pregnant-before-her-father-did/

Goal: To find out what happens to one variable when you make another variable change.

• Usually randomized studies are required to identify causation
• There are approaches to inferring causation in non-randomized studies, but they are complicated and sensitive to assumptions
• Causal relationships are usually identified as average effects, but may not apply to every individual
• Causal models are usually the "gold standard" for data analysis

## Causal analysis

van Nood et al. (2013) NEJM