Is it possible/appropriate to apply the bayesfactor_models() function to to sequential designs, where the number of observations depends on an interim assessment of the evidence collected so far (Schönbrodt & Wagenmakers, 2018; Schönbrodt et al., 2017)? This would be useful for starting a human experiment with a given sample size N, and choosing a stopping rule based on the output of bayesfactor_models(), e.g.:
bayesfactor_models(m1, m2, denominator = m1)
The paper by Schönbrodt et al papers and other ones I perused indicate a method for sample size estimation and stopping rule criteria that apply to t-test designs, while I am testing linear mixed-effects models (m1 and m2 above).
Is it possible/appropriate to apply the bayesfactor_models() function to to sequential designs, where the number of observations depends on an interim assessment of the evidence collected so far (Schönbrodt & Wagenmakers, 2018; Schönbrodt et al., 2017)? This would be useful for starting a human experiment with a given sample size N, and choosing a stopping rule based on the output of bayesfactor_models(), e.g.:
bayesfactor_models(m1, m2, denominator = m1)
The paper by Schönbrodt et al papers and other ones I perused indicate a method for sample size estimation and stopping rule criteria that apply to t-test designs, while I am testing linear mixed-effects models (m1 and m2 above).