ACE spatial cueing: For newer Explorer data that has a "neutral" condition, outputs costs for neutral - incongruent and neutral - incongruent
ACE filter: Outputs Snodgrass-corrected d'. Use with caution when trial counts are low!
ACE modules in general:
where response window summaries are output, now outputs minimum response window as well
For ACE Explorer data, now outputs practice_count, or the number of practice rounds that subject did for that module. Note that at the max value, 5 rounds, it does not currently differentiate between whether the subject maxed out practice and failed the last round, or passed the last round.
Bug fixes:
proc_by_module() failed when app_type = "explorer" and output = "wide" because it attempted to drop a column that didn't exist in demographics
load_ace_bulk() now returns all tap_.*_rt columns in forward/backward spatial span data as double instead of character
post_clean_low_trials() checks for the minimum trial count only in at-all-responded trials within each condition. Previously, was accidentally checking within early-responded trials, which is likely too strict of a response requirement.
Now requires dplyr >= 0.8.0 as well. Tidyverse updates are generally solid about backwards compatibility, so we boldly go for more sensible features!
Also now prevents installation with R >= 4.0.0 to be safe about backwards compatibility.
Notes copied from README.md:
New outputs of
proc_by_module()added:practice_count, or the number of practice rounds that subject did for that module. Note that at the max value, 5 rounds, it does not currently differentiate between whether the subject maxed out practice and failed the last round, or passed the last round.Bug fixes:
proc_by_module()failed whenapp_type = "explorer"andoutput = "wide"because it attempted to drop a column that didn't exist in demographicsload_ace_bulk()now returns alltap_.*_rtcolumns in forward/backward spatial span data as double instead of characterpost_clean_low_trials()checks for the minimum trial count only in at-all-responded trials within each condition. Previously, was accidentally checking within early-responded trials, which is likely too strict of a response requirement.Now requires
dplyr >= 0.8.0as well. Tidyverse updates are generally solid about backwards compatibility, so we boldly go for more sensible features!Also now prevents installation with
R >= 4.0.0to be safe about backwards compatibility.