Type a .spv result row by its statistic-name CELL VALUE, not a header
Source:R/stato-map.R
dot-spv_stato_type_label.RdJASP/jamovi/R result tables are WIDE: one row per test, one column per
statistic (t, df, p), so stato_type_column() types a column by its
header. .spv tables decode LONG/tidy instead (see R/spv.R): one row
per cell, with the statistic's own NAME sitting as a value inside a
dimension column (e.g. a row where the "Statistics" column reads
"Sig. (2-tailed)" and value is 0.037) – there is no header to type at
all. This types that row from its OWN category value instead, reusing the
SAME .STATO_MAP/.MC_STAT_MAP dictionaries stato_type_column() draws
on (SPSS's row labels and JASP/jamovi's column headers name the same
statistics, just positioned differently in the table), plus SPSS-specific
normalisation stato_type_column() has no reason to apply (stripping a
leading "Type III "/"Type I "/"Type II " sum-of-squares qualifier, and a
[%1:*...]1-style templated placeholder SPSS sometimes leaves unresolved
in a repeated-measures error-term label).
Value
the same shape stato_type_column() returns:
list(annotationValue, termSource, termAccession).