significance

significance

Clinical medicine A finding to be weighed in establishing a diagnosis, or influencing management of, a clinical state, which may be expressed as a finding of significance  Statistics A measure of deviation of data from a statistical mean, defined by a probability–p value, where a p of 0.05 indicates a 5% possibility or 1 chance in 20 that a dataset differs from a mean and 19 chances that it will not. See Clinical significance, Statistical significance.

McGraw-Hill Concise Dictionary of Modern Medicine. © 2002 by The McGraw-Hill Companies, Inc.

significance

(statistics) a description of an observed result that shows sufficient deviation from the result expected to be considered different from the expected result. Significance tests such as the CHI-SQUARED TEST can be carried out to produce a value that is converted into the probability that an observed result will match the result expected from a theory. In biology there is a convention that, if there is more than a 5% chance (P < 5%) that the observed result is the same as the expected, it is possible to conclude that any deviations are ‘not significant’, i.e. have occurred by chance alone. If, however, there is less than a 5% chance (P < 5%) that observed and expected are the same, then it is concluded that the deviations are ‘significant’, i.e. have not occurred by chance alone. For example, tossing a coin 100 times gives 58 heads and 42 tails. The probability that 58:42 is similar to the expected 50:50 is greater than 5%, thus we can conclude that there is no significant deviation between observed and expected results.

Collins Dictionary of Biology, 3rd ed. © W. G. Hale, V. A. Saunders, J. P. Margham 2005

significance 

In statistics, an indication that the results of an investigation on a population (e.g. patients) differ from those of another population (e.g. general) by an amount that could not happen by chance alone. This is evaluated by establishing a significance level, that is the probability, called p value, which leads us to reject or accept the null hypothesis H_o (there is no significant difference between two populations and the difference is attributed to chance) and accept or reject the alternative hypothesis H₁ that there is a statistically significant difference between two populations. A p value p < 0.05 is often considered significant, but the lower this figure, the stronger the evidence. See randomized controlled trial.

Millodot: Dictionary of Optometry and Visual Science, 7th edition. © 2009 Butterworth-Heinemann