eelbrain.normalize_in_cells(y, for_dim, in_cells=None, data=None, method='z-score')

Normalize data in cells to make it appropriate for ANOVA 1

  • y (Union[NDVar, str]) – Dependent variable which should be normalized.

  • for_dim (str) – Dimension which will be included as factor in the ANOVA (e.g., 'sensor').

  • in_cells (Union[Factor, Interaction, NestedEffect, str]) – Model defining the cells within which to normalize (normally the factors that will be used as fixed effects in the ANOVA).

  • data (Dataset) – Dataset containing the data.

  • method ('z-score' | 'range') – Method used for normalizing the data: z-score: for the data in each cell, subtract the mean and divide by the standard deviation (mean and standard deviation are computed after averaging across cases in each cell) range: for the data in each cell, subtract minimum and then divide by the maximum (i.e., change the range of the data to (0, 1); range is computed after averaging across cases in each cell).

Return type



This method normalizes data by z-scoring. A common example is a by sensor interaction effect in EEG data. ANOVA interaction effects assume additivity, but EEG topographies depend on source strength in a multiplicative fashion, which can lead to spurious interaction effects. Normalizing in each cell of the ANOVA model controls for this (see [1] for details).


See Compare topographies.



McCarthy, G., & Wood, C. C. (1985). Scalp Distributions of Event-Related Potentials—An Ambiguity Associated with Analysis of Variance Models. Electroencephalography and Clinical Neurophysiology, 61, S226–S227. 10.1016/0013-4694(85)90858-2