### Glossary terms for 'F'

Face validity | A term that describes how well a measurement appears to measure a particular phenomenon, based on whether it seems reasonable; generally not a very reliable method for assessing validity. For example, a measurement of popularity in adolescents was regarded as having face validity, because the investigators thought that it differentiated the popular students in their high school from those who were not. See also content validity, construct validity, and criterion-related validity. |

Factorial trial | A clinical trial of two or more treatments (e.g., A and B), sometimes with two unrelated outcomes, in which subjects are assigned randomly to receive either active treatment A and placebo B, active treatment B and placebo A, both active treatments A and B, or both placebos A and B. For example, the investigator performed a factorial trial to determine whether long-term use of beta carotene and aspirin B6 affected the risk of gastrointestinal cancer. |

False-negative result | Can be used in two different ways. In the context of a medical test, refers to a test result that is falsely negative in a patient with the condition being tested for. For example, though the patient had biopsy-proven breast cancer, her mammogram had given a false-negative result. In the context of a research study, refers to a study result that fails to detect an effect in the sample (i.e., the study result was not statistically significant) that is present in the population. For example, though subsequent studies showed that cigarette smoking increases the risk of stroke, an early case-control study had a false-negative result (P = 0.23). |

False-positive result | Can be used in two different ways. In the context of a medical test, refers to a test result that is falsely positive in a patient without the condition being tested for. For example, though the patient did not have breast cancer or develop it during 6 years of follow-up, her mammogram had a false-positive result. In the context of a research study, refers to a study result that detects an effect in the sample (i.e., the study result was statistically significant) that is not present in the population. For example, though subsequent studies showed that cigarette smoking does not increase the risk of Parkinson?s disease, an early case-control study had a false-positive result (P = 0.03). |

Fixed-effects model | A general term used in multi-level statistical analysis, discussed in this book only with respect to meta-analysis, where it describes a statistical model in which the study weights and the variance of the summary effect estimate is based only on the within-study variances of the included studies. For example, in a meta-analysis of clinical trials of the effect of practicing yoga on depression, the results of the trials were variable; the summary effect based on the fixed-effects model was dominated by one large study and the confidence interval was narrower than would have been estimated with a random effects model. See also random-effects model. |

Glossary material from Hulley SB et al. *Designing Clinical Research*, 4th ed. Philadelphia, Lippincott Williams & Wilkins, 2013.