Derive Advanced Analysis Variables for Student Data

Computes derived variables for research analysis, including GOLD assessment gain scores, K-readiness transition indicators, chronic absence flags, service density indices, and eDECA pre-post gains.

Usage

student_transform(
  clean_obj,
  add_gold_gains = TRUE,
  add_kready_transitions = TRUE,
  add_chronic_absence = TRUE,
  add_service_density = TRUE,
  add_edeca_gains = TRUE,
  chronic_absence_threshold = 18,
  school_days = 180
)

Arguments

clean_obj

An alprek_student_clean or alprek_student_panel object.

add_gold_gains

Logical. Compute GOLD fall-to-spring gain scores (6 domains x raw + scale = 12 columns)? Default TRUE.

add_kready_transitions

Logical. Compute K-readiness transition indicators (6 domains)? Default TRUE.

add_chronic_absence

Logical. Compute chronic absence flag and percentage? Default TRUE.

add_service_density

Logical. Compute service count and risk index? Default TRUE.

add_edeca_gains

Logical. Compute eDECA pre-post T-score gains (5 constructs)? Default TRUE.

chronic_absence_threshold

Numeric. Days absent threshold for chronic absence flag. Default 18 (~10% of 180 school days).

school_days

Numeric. Total school days for absence percentage calculation. Default 180.

Value

The same S3 class as the input object, with additional derived columns in $data. A $transform_log element is added to track which variables were created.

Details

This function is designed to be called after student_clean() and optionally before student_bind_years(). It does not modify the S3 class, so downstream functions (export, validate, linkage) continue to work without modification.

The function is idempotent: running it twice produces the same result (existing derived columns are overwritten).

Source columns that do not exist in the data are silently skipped with an informational message.

Derived Variables

GOLD Gains (12 cols)

gold_[domain]_gain_raw and gold_[domain]_gain_scale for 6 domains: literacy, math, se, physical, cognitive, language. Computed as spring - fall.

K-Readiness Transitions (6 cols)

gold_[domain]_kready_improved: 1 if fall = Emerging and spring = Accomplished, else 0.

Chronic Absence (2 cols)

chronic_absence (0/1 flag) and chronic_absence_pct (days_absent_total / school_days * 100).

Service Density (2 cols)

n_services (count of binary service indicators) and risk_index (count of risk indicators).

eDECA Gains (5 cols)

edeca_[construct]_gain: post T-score minus pre T-score for initiative, self_reg, attachment, tpf, behavior.

Examples

if (FALSE) { # \dontrun{
clean <- student_clean(student_read("data/FCPK Student Details 23-24.xlsx"))
enriched <- student_transform(clean)
# Or with panel:
panel <- student_process_years(configs)$panel
enriched_panel <- student_transform(panel)
} # }