Bias

(Difference between revisions)
Jump to: navigation, search
Line 2: Line 2:
  
 
1.  [[wikipedia:Selection bias]]
 
1.  [[wikipedia:Selection bias]]
This happens when the samples in an experiment (whatever they may be, usually people) are not typical of the overall  population being sampled   
+
    This happens when the samples in an experiment (whatever they may be, usually people) are not typical of the overall  population being sampled   
  
 
Example: sampling peoples' attitudes of gun control in Texas but only interviewing people who live in Dallas and nowhere else
 
Example: sampling peoples' attitudes of gun control in Texas but only interviewing people who live in Dallas and nowhere else
  
 
2.  Researcher bias
 
2.  Researcher bias
When the researcher(s) want a particular result to arise out of the experiment and affect   
+
    When the researcher(s) want a particular result to arise out of the experiment and affect   
 
      
 
      
 
Example: religious researchers who want to attribute a particular phenomena to God instead of assuming another, most-likely more plausible explanation
 
Example: religious researchers who want to attribute a particular phenomena to God instead of assuming another, most-likely more plausible explanation
  
 
3.  Personal bias
 
3.  Personal bias
In studies involving people, this usually happens when the people in the study know what the experiment is trying to explain and then try to affect the result in a particular direction.  This is why most experiments are ran [[blind study]]
+
    In studies involving people, this usually happens when the people in the study know what the experiment is trying to explain and then try to affect the result in a particular direction.  This is why most experiments are ran [[blind study]]
  
 
4.  Sample size bias  
 
4.  Sample size bias  
This is a form of selection bias in which the [[sample size]] is small (could be for many reasons and sometimes not on purpose) and results in skewed results
+
    This is a form of selection bias in which the [[sample size]] is small (could be for many reasons and sometimes not on purpose) and results in skewed results

Revision as of 22:04, 29 November 2010

When discussing the results of a particular experiment, typically you'll find scientists and the media talking about bias. Bias is an error in the results an experiment (usually because of the design) that favor one result or explanation. Bias in the scientific sense is different from personal preference. Bias can be intentional or unintentional when performing an experiment, sometimes it can't be completely avoided. There are different kinds of bias in scientific sense:

1. wikipedia:Selection bias

    This happens when the samples in an experiment (whatever they may be, usually people) are not typical of the overall  population being sampled  

Example: sampling peoples' attitudes of gun control in Texas but only interviewing people who live in Dallas and nowhere else

2. Researcher bias

    When the researcher(s) want a particular result to arise out of the experiment and affect  
    

Example: religious researchers who want to attribute a particular phenomena to God instead of assuming another, most-likely more plausible explanation

3. Personal bias

    In studies involving people, this usually happens when the people in the study know what the experiment is trying to explain and then try to affect the result in a particular direction.  This is why most experiments are ran blind study

4. Sample size bias

    This is a form of selection bias in which the sample size is small (could be for many reasons and sometimes not on purpose) and results in skewed results
Personal tools
Namespaces
Variants
Actions
wiki navigation
IronChariots.Org
Toolbox