Bias is best defined as:

• A. Random sampling error
• B. Systematic deviation from truth due to errors in design, conduct, or analysis
• C. Natural variation in data
• D. Randomization bias only in trials

Answer: (B)

Bias is a systematic deviation from the truth in study results caused by flaws in how the study is designed, conducted, or analyzed. It means the findings are consistently distorted in a particular direction, not just spread around by chance. This is different from random sampling error, which affects precision and arises from the inherent variability of taking a sample rather than from a flaw that pushes results in one direction. It’s also different from natural variation in data, which is random and expected. Bias can come from various sources, such as who is selected for the study (selection bias), how information or outcomes are measured (information or measurement bias), or how data are collected and analyzed. While randomization in trials helps reduce bias, bias can still occur through other aspects like design choices or analytical methods. Recognizing bias helps explain why a study’s results might systematically overestimate or underestimate true effects.