Types of Bias and Ways to Reduce.

An ex-prisoner was contacted months after he contracted norovirus in prison. He seems unable to remember what his symptoms were.

• Selection Bias
• Recall Bias
• Differential misclassification bias
• Pathological Bias

An Asian man conducts a case-control study. The demographic of the cases
reflect the general population, but all the controls selected are Asian men.

• Loss to followup
• Selection Bias
• Diagnostic Bias
• Information Bias

A study put up fliers for those who wanted to
participate in a study as controls. However, the researchers found that
those who responded were much different demographically than the cases.

• Self-selection bias
• Case control bias
• Demographic bias
• Hawthorne Bias

In a study of the effects of hypertension on heart disease, a nurse performs
blood pressure measurements on the subjects using a faulty
sphygmomanometer (real thing) that always records blood pressures on the high side.

• Bias Bias
• Hawthorne Bias
• Information Bias
• Reporting Bias

What is a double blind study?

• A study in which the subjects are unknowingly watched and observed.
• A method of calculating data in which one studies a sample while simultaneously preventing the intrusion of Bose-Einstein like quanta by connecting the ends of the two pairs of pretarsal orbiculari.
• A study in which the subject and tester do not know who is receiving the experimental treatment.
• An error in epidemiology in which the tester unknowingly treats the experimental group as a control group.

*¡Random Trivia!*
(For Arnav's sake)
In 2011, US special ops forces raided Osama Bin Laden's flat and killed him, marking a significant event in the US fight against terrorism. Where did this attack occur?

• Qatar
• Kazakhstan
• Pakistan
• Iraq

A study falsely concluded that alcoholism led to coronary heart disease.
However, the real association lies in the fact that smokers are more likely
to be alcoholics as well.

• Diagnostic Bias
• Late look Bias
• Loss to follow-up
• Confounding Bias

A study of a new form of cancer requires that scientists interview several patients with varying stages (degrees of severity) of that cancer. Because they started interviewing the stage 1 (most healthy) patients first, the scientists were unable to collect the perspectives of several stage 4 patients as they had died.

• Berkson's Bias
• Interrogative Bias
• Nonresponse bias
• Late look bias

An epidemiologist seeks to find out whether working in a nuclear power plant increases the likelihood of dying from anything. His cohorts are nuclear power plant workers, but he compares them to the general population who may or may not have jobs. As a result, his data is below the true value.

• Healthy Worker Effect
• Measurement Bias
• Nonresponse Bias
• Confounding Bias

Is using a placebo pill good, or bad?

• Bad, because it leads to confusion over the causality of an outcome.
• Bad, because it is unethical to experiment on people with placebos, even if they consent beforehand.
• Good, because it allows scientists to assess a person's body by giving a dud pill, and determine the people who are ready to be administered real pills.
• Good, because it allows the case and control group to be comparable at baseline, and rids of any bias of knowing that one is part of the case or control group.