Skip to contents

Calculates the count of students at each performance level based on percentages and total valid scores. This is a derivation by design: NJDOE publishes performance-level shares (`pct_l1`..`pct_l5`) but not the underlying counts, so this function exists to compute them from each row's own published percentage and its own published `number_of_valid_scale_scores` (the same-cell denominator those percentages are shares of).

Usage

parcc_perf_level_counts(df)

Arguments

df

dataframe, output of fetch_parcc

Value

df with counts of students by performance level (`num_l1`..`num_l5` = `round(pct_lN / 100 * number_of_valid_scale_scores)`) and a `value_source` column: `"derived_from_pct"` where `number_of_valid_scale_scores` is non-NA, `"published_pct_only"` (with every `num_lN` NA) where it is NA. A performance level a test does not have (eg L3-L5 for NJGPA) carries `pct_lN`/`num_lN` NA regardless of `value_source`, which describes the derivation method, not applicability.