neuropsychology / neuropsychology.r Goto Github PK

in

We're going to deprecate select_vars() eventually, so it's a good idea to remove it anyway.

I think there might be a bug in the calculation of dprime in the dprime function. This function adjusts for extreme values (e.g., n_miss = 0) using the Hautus (1995) correction, which is summarised as:

"A method that has found much use in log-linear analysis requires the addition of 0.5 to each cell in the two-by-two contingency table that defines the performance of the observer (e.g., Fienberg, 1980; Goodman, 1970; Knoke & Burke, 1980). Row and column totals are increased by one." Hautus (1995, pg. 46).

dprime performs this adjustment in the following way:

hit_rate_adjusted <- (n_hit + 0.5)/((n_hit + 0.5) + n_miss + 1)
fa_rate_adjusted <- (n_fa + 0.5)/((n_fa + 0.5) + n_cr + 1)
dprime <- qnorm(hit_rate_adjusted) - qnorm(fa_rate_adjusted)

But if I'm not mistaken, this is adding 1.5 to the row/col totals, when it should only be adding 1.

https://www.r-bloggers.com/automatic-tools-for-improving-r-packages/

Try to implement this.

Using the ppcor package.

This function really needs improving (support more statistical functions, enhance the output).

Add a remove_outliers(df, variable, treshold=1.96) function.

Goal: returns the dataframe without the outliers (based on one variable)

Problem: I don't know how to use dplyr::filter() (or any dplyr function) inside a function, for it does not take string variable as an argument. There must be a way to do it though...