Chapter 2: Observational studies
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Guidelines and tips
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Observational Studies vs Controlled Experiments: Understanding Confounding Factors, Study notes of Statistics
Observational studies and their importance in scientific research, despite the availability of controlled experiments. It provides examples of confounding factors, such as smoking and liver cancer, and explains how to make groups comparable by dividing them into subgroups. The document also covers the concept of controlling for a factor and provides examples of confounding in the context of pellagra and cervical cancer.
Typology: Study notes
Pre 2010
1 / 114
This page cannot be seen from the preview
Don't miss anything!
Related documents
Partial preview of the text
Download Observational Studies vs Controlled Experiments: Understanding Confounding Factors and more Study notes Statistics in PDF only on Docsity!
Chapter 2: Observational studies
Last chapter • Controlled experiments (ch 1)
Last chapter • Controlled experiments (ch 1)
Best design: double-blind randomized controlled study • Take a representative sample of study units
Last chapter • Controlled experiments (ch 1)
Best design: double-blind randomized controlled study • Take a representative sample of study units - Randomize the study units into a treatment/control group( randomized controlled
Last chapter • Controlled experiments (ch 1)
Best design: double-blind randomized controlled study • Take a representative sample of study units - Randomize the study units into a treatment/control group( randomized controlled
Use placebo and don’t let study units know that whether they gettreatment or placebo. Don’t let evaluators know who gets treatmentor placebo ( double-blind
Evaluate the response in both groups
Last chapter • Controlled experiments (ch 1)
Best design: double-blind randomized controlled study • Take a representative sample of study units - Randomize the study units into a treatment/control group( randomized controlled
Use placebo and don’t let study units know that whether they gettreatment or placebo. Don’t let evaluators know who gets treatmentor placebo ( double-blind
Evaluate the response in both groups
Why is this such a good design?
Last chapter • Controlled experiments (ch 1)
Best design: double-blind randomized controlled study • Take a representative sample of study units - Randomize the study units into a treatment/control group( randomized controlled
Use placebo and don’t let study units know that whether they gettreatment or placebo. Don’t let evaluators know who gets treatmentor placebo ( double-blind
Evaluate the response in both groups
Why is this such a good design? • Because of randomization, two groups will be similar except for thetreatment. - Hence, if there is a difference in the responses, then this is likely caused by the treatment.
New chapter • Observational studies (ch 2)
New chapter • Observational studies (ch 2)
Why look at these when controlled experiments are so good?Because controlled experiments are sometimes hard to do or evenimpossible. Examples:
New chapter • Observational studies (ch 2)
Why look at these when controlled experiments are so good?Because controlled experiments are sometimes hard to do or evenimpossible. Examples: • Does smoking cause cancer?
New chapter • Observational studies (ch 2)
Why look at these when controlled experiments are so good?Because controlled experiments are sometimes hard to do or evenimpossible. Examples: • Does smoking cause cancer? - Does pollution cause global warming? - Problem with observational studies: often the groups are notcomparable.
New chapter • Observational studies (ch 2)
Why look at these when controlled experiments are so good?Because controlled experiments are sometimes hard to do or evenimpossible. Examples: • Does smoking cause cancer? - Does pollution cause global warming? - Problem with observational studies: often the groups are notcomparable. - We’ll look at various examples.
The basics
Treatment and control • We again use the method of comparison: we compare a treatment (orexposed) group to a control group. Examples: