© Michael Smith Laboratories, 2016-2018 About

Plot of the selected term's p-value and 95% confidence interval over its history.

This is calculated using *six month rolling regressions* on the historical changes in
'Pop. Hits' and 'Hits'. Given the results of these regressions we use their 95% confidence
intervals to calculate a 95% confidence interval around the p-value. This range is then used
to calculate the negative log of the coefficient of variation with respect to that edition's
significance cutoff to get the stability score.

The stability score typically ranges between -10 and 10 with 10 being more stable than -10.

A score of **No Change** means we saw no significant variation in both
Pop. Hits and Hits over its recent past (keep in mind this does **not** mean the p-value
is constant).

A score of **No Data** means there was not enough available.