JAMAevidence Glossary

Terms are derived from Users' Guides to the Medical Literature: A Manual for Evidence-Based Practice, 2nd Edition and The Rational Clinical Examination: Evidence-Based Clinical Diagnosis. Updated December 2009.
Download a PDF of the glossary (179 KB). (Adobe Reader is required to open or print PDF files.)





Law of multiplicative probabilities
The law of multiplicative probabilities for independent events (in which one event in no way influences the other) tells us that the probability of 10 consecutive heads in 10 coin flips can be found by multiplying the probability of a single head (1/2) 10 times over; that is, 1/2 × 1/2 × 1/2, and so on.

Lead time bias
Occurs when outcomes such as survival, as measured from the time of diagnosis, may be increased not because patients live longer, but because screening lengthens the time that they know they have disease. See also Bias.

Leading hypothesis
The clinician’s single best explanation for the patient’s clinical problem(s).

Length time bias
Occurs when patients whose disease is discovered by screening also may appear to do better, or live longer, than people whose disease presents clinically with symptoms. Screening tends to detect disease that is destined to progress slowly and, therefore, has a good prognosis. See also Bias.

Levels of evidence
A hierarchy of research evidence to inform practice, usually ranging from strongest to weakest.

Likelihood functions
Functions constructed from a statistical model and a set of observed data that give the probability of that data for various values of the unknown model parameters. Those parameter values that maximize the probability are the maximum likelihood estimates of the parameters. See also Likelihood ratio.

Likelihood ratio
For a screening or diagnostic test (including clinical signs or symptoms), the likelihood ratio (LR) expresses the relative likelihood that a given test would be expected in a patient with, as opposed to one without, a disorder of interest. An LR of 1 means that the posttest probability is identical to the pretest probability. As LRs increase above 1, the posttest probability progressively increases in relation to the pretest probability. As LRs decrease below 1, the posttest probability progressively decreases in relation to the pretest probability. An LR is calculated as the proportion of target positive with a particular test result (which, with a single cut point, would be either a positive or negative result) divided by the proportion of target negative with same test result.

Likert scales
Scales, typically with three to nine possible values, that include extremes of attitudes or feelings (eg, from totally disagree to totally agree) that respondents mark to indicate their rating. See also Visual analogue scale.

Linear regression
The term used for a regression analysis when the dependent or target variable is a continuous variable, and the relationship between the dependent and independent variables is thought to be linear.

Linkage
The tendency of genes or other DNA sequences at specific loci to be inherited together as a consequence of their physical proximity on a single chromosome.

Linkage disequilibrium
A measure of association between alleles at different loci.

Local consensus process
A strategy for changing clinician behavior. Inclusion of participating clinicians in discussions to create agreement with a suggested approach to change provider practice.

Local opinion leaders
A strategy for changing clinician behavior. These persons are clinician peers who are recognized by their colleagues as model caregivers or who are viewed as having particular content expertise.

Locus/loci
The site(s) on a chromosome at which the gene for a particular trait is located or on a gene at which a particular SNP is located.

Logistic regression
1. A multivariate regression analysis that analyzes relationships between multiple independent variables and categorical dependent variables. 2. A regression analysis in which the dependent variable is binary.

Lost to follow-up
Patients whose status on the outcome or endpoint of interest is unknown.