The QAICc Weight (called an Akaike Weight by Burnham and Anderson 1998) of a model is exp( -1/2 * Delta QAICc ) divided by the sum of this quantity for all models:
QAICc Weight of Model Mi = exp(-1/2 * Delta QAICc of Model Mi) / [ sum for all models of exp(-1/2 Delta QAICc ) ] .
Thus, the sum of the QAICc weights is 1. These weights are used in Model Averaging. More details can be found in http://neota.cnr.colostate.edu/fw663/lecture7.pdf.
QAICc Weights are also used to create the Model Likelihood value, the likelihood of this model given the set of models.
Equivalent weights are computed using BIC model selection