📊 Biostatistics & Epidemiology🟡 High-yieldStatistical Inference

Tests of Significance (Biostatistics)

Statistical tests that judge whether a difference found in a study is likely to be real or could easily have arisen by chance.

Key points

A test of significance judges whether a difference seen in a study is likely to be real or could have arisen just by chance.
It starts with a null hypothesis, which says there is no real difference, and the test works out how likely the observed result would be if that were true.
The p-value is the probability of getting a result at least as extreme as the one observed if the null hypothesis were true; a small value (by convention below 0.05) leads to rejecting the null hypothesis.
The test chosen depends on the data: the t-test compares the means of two groups of numerical data, and analysis of variance compares more than two groups.
The chi-square test compares proportions or counts in categories, such as whether a disease is commoner in one group than another.
A significant result is not the same as an important one; a tiny, unimportant difference can be statistically significant in a very large study.

Memory trick

How likely is a result this extreme if the null were true? That is the p-value; below 0.05 (by convention), reject the null. t-test for two means, ANOVA for more, chi-square for categories; significant is not the same as important.

All of Community Medicine More Biostatistics & Epidemiology

Next in Community Medicine · Biostatistics & Epidemiology

Loading…

📊 Biostatistics & Epidemiology🟡 High-yieldStatistical Inference

Tests of Significance (Biostatistics)

Statistical tests that judge whether a difference found in a study is likely to be real or could easily have arisen by chance.

Key points

A test of significance judges whether a difference seen in a study is likely to be real or could have arisen just by chance.
It starts with a null hypothesis, which says there is no real difference, and the test works out how likely the observed result would be if that were true.
The p-value is the probability of getting a result at least as extreme as the one observed if the null hypothesis were true; a small value (by convention below 0.05) leads to rejecting the null hypothesis.
The test chosen depends on the data: the t-test compares the means of two groups of numerical data, and analysis of variance compares more than two groups.
The chi-square test compares proportions or counts in categories, such as whether a disease is commoner in one group than another.
A significant result is not the same as an important one; a tiny, unimportant difference can be statistically significant in a very large study.

Memory trick

All of Community Medicine More Biostatistics & Epidemiology