Update QQ plot docs

aherbert · Aug 10, 2023 · 7e0dae7 · 7e0dae7
1 parent 0ed4bac
commit 7e0dae7
Showing 1 changed file with 2 additions and 2 deletions.
diff --git a/docs/analysis_plugins.rst b/docs/analysis_plugins.rst
@@ -1712,7 +1712,7 @@ When the plugin runs a dialog is presented that allows multiple datasets to be s
      - Random seed for the bootstrap repeats (hexidecimal input).
 
    * - Show QQ plot
-     - Display a quantile-quantile plot of the model against the data distribution. The quantile for the model is taken using the cumulative probability of the data as k / (n + 1).
+     - Display a quantile-quantile plot of the model against the data distribution. The quantile for the model is taken using the cumulative probability of the data as (k - 1/3) / (n + 1/3).
 
 The analysis extracts the tracks and filters them to stationary molecules. A histogram of residence times is created and fit using the exponential model (see :numref:`Figure %s <fig_residence_time_analysis_histogram>`). Fit progress is recorded in the ``ImageJ`` log window. Bootstrapping is optionally used on the data to produce a 95% confidence interval for the fit parameters. 
 
@@ -1821,7 +1821,7 @@ The following parameters can be set:
      - Random seed for the bootstrap repeats (hexidecimal input).
 
    * - Show QQ plot
-     - Display a quantile-quantile plot of the model against the data distribution. The quantile for the model is taken using the cumulative probability of the data as k / (n + 1).
+     - Display a quantile-quantile plot of the model against the data distribution. The quantile for the model is taken using the cumulative probability of the data as (k - 1/3) / (n + 1/3).
 
 The simulation allows experimenting with the exposure time and number of samples required to obtain satisfactory results for the populations of bound molecules.