Glossary terms for 'L'
Level of statistical significance | See alpha. |
Likelihood ratio | A term used to describe the quantitative effects of a medical test result on the likelihood that a patient has the disease being tested for. It is defined as the likelihood of that test result in a patient with the disease divided by the likelihood of that result in a patient without the disease (the mnemonic is WOWO: with over without). For example, the likelihood ratio for the characteristic symptoms of typical angina (exertional substernal pressure) is about 50 for the diagnosis of coronary artery disease. |
Likert scale | A set of answers (usually 5) to a question that provide a similarly spaced range of choices. For example, the potential answers to the question “How likely are you to return to this emergency room for care?” were as follows: Very likely, Somewhat Likely, Neither likely nor unlikely, Somewhat unlikely, Very unlikely. |
Logistic regression model | A statistical technique used to estimate the effects of one or more predictor variables on a dichotomous outcome variable, adjusting for the effects of other predictor and confounding variables. For example, in a logistic regression model, men were about twice as likely as women, and blacks about three times as likely as whites, to develop strokes, adjusting for age, blood pressure, and diabetes. |
Glossary material from Hulley SB et al. Designing Clinical Research, 4th ed. Philadelphia, Lippincott Williams & Wilkins, 2013.