Veronica noticed a statistical issue in the d-prime analysis where we're grouping the pairwise similarity data frame by game1 to compute average across-game similarity for each game, but because each game pair only appears once (i.e. the game pairs are derived from the upper-triangular matrix), games that appear later in the list will have fewer comparisons and therefore less reliable estimates.
Veronica noticed a statistical issue in the d-prime analysis where we're grouping the pairwise similarity data frame by
game1to compute average across-game similarity for each game, but because each game pair only appears once (i.e. the game pairs are derived from the upper-triangular matrix), games that appear later in the list will have fewer comparisons and therefore less reliable estimates.