Construct Validity: Validity b/c the measure is accurately "labeled"

Content Validity: extent to swhich the content of the measure contains the full scope of the phenomenon under study (completeness)

Criterion Validity:  Is your measure statistically associated with some external standard or other measure already shown to be valide

selection bias: Due to subjects, not study: An error in the individuals & groups participating in the study / Systematic differneces exist b/t those selected and those not particpating in study 

confounding bias: Occurs when an observed association b/t an exposure and an event is mixed up with some third factor

information bias: A bias resulting from measurement error of the study that may result in misclassification of subjects into study. e.g. defective lab equipmetn, bias in responding to questionersy (i.e. not remembering), interverier effects

What is confounding bias?

How do you reduce confounding bias?

Confounding bias is error due to the interpretation of evidence, reather than the measurement of it. (i.e. a confounding bias would be sayiny carrying around matches causes diseases. In this situation, "smoking" is the confounder.

How to decrease confoundign bias:

• Increase experimental evidence
• Attempt to control for all possible "third" factors
• Focus on randomization, ensures confounders are equally distributed

test-retest reliability: Concordance of results when the same instrument is administered to the same individuals/groups on separate occsions.

inter-rater reliability: degree to which different raters/observers give consistent estimates of the same phenomenon/subject gorup 
 

intra-rater reliability: degree to which the same raters/observers give consistent estimates of the same phenomenon/subject gorup