Difference between revisions of "Statistical significance"
Jump to navigation
Jump to search
m |
|||
| Line 1: | Line 1: | ||
In [[statistics]], it is often required to know whether two groups of data, both subject to random variation, should be reqarded as distinct or not. That is, are the observed differences between the groups due to chance or not? This is tested by initially assuming the [[null hypothesis]] that there is no difference, then computing a test [[statistic]] that has a known [[probability distribution]] under the null hypothesis. If the probability of the observed value of the test statistic is very low this suggests that the null hypothesis is untrue. | In [[statistics]], it is often required to know whether two groups of data, both subject to random variation, should be reqarded as distinct or not. That is, are the observed differences between the groups due to chance or not? This is tested by initially assuming the [[null hypothesis]] that there is no difference, then computing a test [[statistic]] that has a known [[probability distribution]] under the null hypothesis. If the probability of the observed value of the test statistic is very low this suggests that the null hypothesis is untrue. | ||
| − | + | ||
[[category:Probability and Statistics]] | [[category:Probability and Statistics]] | ||
Revision as of 22:07, December 14, 2009
In statistics, it is often required to know whether two groups of data, both subject to random variation, should be reqarded as distinct or not. That is, are the observed differences between the groups due to chance or not? This is tested by initially assuming the null hypothesis that there is no difference, then computing a test statistic that has a known probability distribution under the null hypothesis. If the probability of the observed value of the test statistic is very low this suggests that the null hypothesis is untrue.