This preview shows pages 1–12. Sign up to view the full content.

Click to edit Master subtitle style Unit 6 Questions and Answers

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Study guide for Unit 6: Precision/validity of an association Internal vs. external validity Generalizability Chance: population vs. samples Hypothesis testing ( p -values to reject or accept Ho) Estimation: Confidence intervals (95%, 99%) Computation and interpretation of a 95% confidence interval Statistical power: review of Type I and Type II errors Four features to power Example of a sample size and power calculation Overpowered samples: clinical vs. statistical significance
What are the 4 explanations for an association to be obser ved in a 1 – 2 – 3 – 4 -

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
What are the 4 explanations for an association to be observed in a study? n 1 – Chance = PRECISION n 2 – Bias - confounding n 3 – Bias - other n 4 – Truth (causal) We must rule out 1-3 and then infer 4 through causal criteria that we will learn more about in Segment 3 of the VALID ITY
Define validity and name and Validity: Two types of validity: 1 – 2 -

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Define the 2 types of validity n Validity: The extent to which what we measure in the study is TRUE in the real world. n Two types of validity: 1 – Internal validity: the extent to which the comparison between the two groups (cases/controls; exposed/unexposed) is FAIR and the results are a true representation of the real measure of effect (OR/RR/AR) 2 – External validity: the extent to which study data
Tr ue or False? External validity must be sought first before internal validity T F

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
True or False? n External validity must be sought first before internal validity T F n Internal validity is most important. When we are sure that a finding is internally valid, we can ask whether it is externally valid. If it is not, it can still stand as a valid study. n Ideally we will design our study to have both types
I nternal vs. exter nal validity {7418B5D8-F7CD-4DB4-8F48-EBEDB BE0B09} Sample {6DEF3 9C-A13C-4696-9479-2 1032DF97DE} Target population {8BD9C68A-D06C-42CF-846B-52E 321CD8D6} State/country/world Extent to which the  cases/controls or  exposed/unexposed  come from the  same underlying  Extent to which the  cases/controls or  exposed/unexposed  represent all cases/ controls and all

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Chance = precision Is our finding due to chance? Each sample will have some natural variability Each ‘study’ draws a new sample A case-control study attempts to draw a sample from the universe of cases and ‘non-cases’ in the population
Calculate the OR for each scenario and label each as ‘Type 150 100 175 75 D Ð E Ē 200 50 200 50 D Ð E Ē 125 125 125 125 D Ð E Ē 125 125 175 75 D Ð E Ē “TRUE TABLE”

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 07/15/2011 for the course PHC 6000 taught by Professor Staff during the Summer '08 term at University of South Florida.

### Page1 / 44

This preview shows document pages 1 - 12. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online