With the formula for column P(Ln-1) in place, we now need to set up the first row of column P(Ln). Just temporarily, set column P(Ln-1|Result) to be the same as P(Ln-1). (This eliminates Bayesian updating, which is not correct within BKT...; we'll fix it later.) Based on the P(T) parameter, what should the formula be for P(Ln) so that it accurately represents learning?

As always, don't be shy to try out the others, and see whether they return you what you expected. If not, consider why they fail.

A. df3.loc[:, "P(Ln)"] = df3.loc[:, "P(Ln-1|Result)"]*t
B. df3.loc[:, "P(Ln)"] = (1-df3.loc[:, "P(Ln-1|Result)"])*t
C. df3.loc[:, "P(Ln)"] = df3.loc[:, "P(Ln-1|Result)"] + df3.loc[:, "P(Ln-1|Result)"]*t
D. df3.loc[:, "P(Ln)"] = df3.loc[:, "P(Ln-1|Result)"] - (1-df3.loc[:, "P(Ln-1|Result)"])*t
E. df3.loc[:, "P(Ln)"] = df3.loc[:, "P(Ln-1|Result)"] + (1-df3.loc[:, "P(Ln-1|Result)"])*t