Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
22 Cards in this Set
- Front
- Back
What is bias?
|
Systematic deviation of results or inferences from the truth
|
|
What are the problems with bias?
|
create spurious association
mask a true association |
|
When does bias occur?
|
in all study types and designs
usually do to investigator or study participants |
|
When can bias be fixed?
|
cannot be fixed, should be minimized in the design phase but analysis cannot get rid of bias
|
|
What is random error?
|
equal misclassification of cases and controls, will even out in a large enough sample
|
|
Which type of error leads to imprecise results? Which type leads to inaccurate results?
|
imprecise = chance (random error)
inaccurate= bias (systematic error) |
|
What are the major types of bias?
|
selection
information confounding |
|
What is selection bias?
|
animals are not representative of the population
(selection is not independent of exposure or outcome) |
|
What is self-selection bias?
|
people choose to enroll their animals, and may have sicker or healthier animals than the rest of the population
|
|
What is response bias?
|
usually seen with questionnaires, where the responders are more willing to fill out the survey because it is close to them or something like that
|
|
What is loss to followup bias?
|
animals lost to follow-up are different than the study population as a whole
|
|
How can you minimize selection bias?
|
clearly define the study population
explicitly ID case and control definitions Take cases and controls from the same population |
|
What is information bias?
|
The method of gathering information is inappropriate and leads to systematic error in measurement of the exposure or outcome
|
|
What are examples of information bias?
|
recall bias
reporting bias interview bias |
|
What is recall bias?
|
Exposed subjects have a greater sensitivity of recall
ie owner looking for ways to explain the animal's condition etc |
|
What is reporting bias?
|
reports of information are more detailed and complete for sicker animals, may be able to find greater association
|
|
What is interview bias?
|
The interviewer has a systematic bias in soliciting, recording and interpreting information (especially problematic if exposure or outcome status is already known)
|
|
How can you minimize information bias?
|
blinding
questionnaires with multiple questions asking the same information increased accuracy by double checking records and gathering data from multiple sources use multiple controls |
|
What is a confounder?
|
a factor related to both exposure and outcome which accounts for some or all of the relationship between the two
|
|
Why is confounding an issue?
|
describes a true association, but may be misleading as to the causation
|
|
What are common confounders?
|
age, species, sex
|
|
How can you deal with confounding?
|
randomization and matching of participants
restriction of potential confounders use stratified or multivariable analysis *better to try and reduce problems during study design rather than analysis |