Browse the glossary using this index

Special | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | ALL

Page:  1  2  3  4  5  (Next)
  ALL

A

Alpha-Level (α-level)

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

  • Significance level
  • Probability of rejecting H0 when H0 is true
  • Usually set at α=0.05 or α=0.01

Alternative Hypothesis (H1 or Ha)

(Last edited: Tuesday, 14 March 2017, 8:55 AM)
  •  Experimental hypothesis
  • The population parameter differs from H0
  • When H0 can be rejected, it does not mean that H1 can be accepted, because of the remaining probability of α- and β-error

 


B

Bias

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

  • Systematic error or deviation in results
  • Example: selection bias

Systematic differences in the groups that are compared: group A has 50% more smokers than group B 

→ groups are not comparable



Blinding

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

 

 

  • Process of hiding which comparison group a particular participant belongs to
  • Used to minimize risk of bias
  • Common methodology in clinical trials
  • Types:
  1. Single blind: participants unaware
  2. Double blind: participants+ outcome assessors/ testers unaware
  3. Triple blind: participants+ outcome assessors/ testers+ data analysts unaware

 


C

Confidence Interval (CI)

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

 

 

  • Measure of uncertainty around main finding of statistical analysis
  • Usually set at CI=95%, also CI=99% or CI=90%

 


Continuous Data

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

  • Information that can take any value within a certain range
  • Opposite of discrete data
  • Example: Weight of a person: The weight can be any value within the range of people´s weight (e.g. 60.3 kilogram).


D

Dependent or Outcome Variable

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

 

 

Clinical trial

Dependent variable= Outcome (ill or healthy)

Independent variable= Treatment arm (new drug)

 


Discrete Data

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

  • Information that can only take certain values
  • Opposite of continuous data
  • Example: Number of people in a room: It is impossible to have 40.3 people in a room.


F

False Negative

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

= Negative test when individual actually has the disease



False Positive

(Last edited: Tuesday, 14 March 2017, 8:55 AM)

= Positive test when individual does not actually have the disease




Page:  1  2  3  4  5  (Next)
  ALL