Per-case descriptive features
Usage
compute_case_features(
data,
outcome_var = "outcome",
phase_var = "phase",
case_var = "case",
time_var = "time",
intervention_goal = c("increase", "decrease")
)Arguments
- data
Data frame containing SCD data.
- outcome_var
Name of the outcome variable.
- phase_var
Name of the phase variable.
- case_var
Name of the case variable.
- time_var
Name of the time variable.
- intervention_goal
Direction of a desirable change:
"increase"(default) or"decrease". For"decrease",NAPis reported as1 - NAPandPNDcounts phase-B points below the phase-A minimum.
Value
A data frame of summary features with one row per tier (each
combination of the grouping columns of tier_vars(), so a case
identifier repeated across studies or a case measured on several
behaviors gives separate rows), or NULL if requirements are not
met. case holds the case identifier and tier the full
tier label (e.g. "S1 / P1 / reading"). Within each tier, the
_A features use the first run of a baseline-role phase (see
phase_role()) and the _B features the next intervention-role
run, so labels such as "Baseline 1" and later reversal phases do
not change which phases are compared. The nonoverlap
descriptives NAP and PND are computed in the direction
given by intervention_goal.
Examples
d <- simulate_scd_data(n_cases = 2, n_phase_a = 6, n_phase_b = 6)
compute_case_features(d)
#> case tier mean_A mean_B sd_A sd_B slope_A slope_B
#> 1 C1 C1 46.64336 48.05030 9.124774 6.483815 0.3550227 -0.09015397
#> 2 C2 C2 57.23262 48.13144 10.194156 11.155911 -3.4646165 -1.01825202
#> cv_A cv_B autocor_A autocor_B mean_diff pooled_SD NAP
#> 1 0.1956286 0.1349381 -0.18254586 0.1115179 1.406941 7.915218 0.5833333
#> 2 0.1781179 0.2317801 -0.09699311 -0.1550718 -9.101179 10.685859 0.2222222
#> PND phase_length_ratio trend_accel_A trend_accel_B
#> 1 0.1666667 1 2.6576965 -13.22389
#> 2 0.0000000 1 0.4468388 -13.09183