Fkkkkkkk

Topics: Bias, Selection bias, Hypertension Pages: 11 (2421 words) Published: May 10, 2013
Key Points
Measurement issues
Chance
P value
Confidence intervals
Measurement Bias and Error in Study Design Bias
Selection bias
Measurement bias
Confounding

Estimation
Process of using calculated sample values to determine the probable value of a population parameter • Point estimate
• Confidence interval estimation
Range of values that has a known probability of capturing the parameter

Chance
• From estimates in the sample population make inferences about the risk in the total population • Random error is when the value of the sample measurement diverges – due to chance from that of the true population value. P values

• Is the result because of chance?
• Measured by P value (statistical tests)
• P value less than 0.05 P<0.05 probability of obtaining the observed value by chance is less than 1 in 20 Selection bias
Statistical error that causes a bias in the sampling portion of an experiment. The error causes one sampling group to be selected more often than other groups included in the experiment. This may produce an inaccurate conclusion if the selection bias is not identified. For example, if an experiment only selects people of a certain race with similar characteristics and excludes any group that deviates from these elements. "The students at the University had to start their experiment over because they discovered a selection bias in the species gathered."