NAEP Technical DocumentationMathematics Assessment Estimation Variables for 2015

Variables used in the population-structure model for the mathematics 2015 assessment are listed and described below.   

Variables used for main effects and interactions in the population-structure model:

• DSEX – School-reported gender. This variable has three levels: male, female, and missing.

• DRACE10 – Student-reported race. This variable has eight levels: White, Black, Hispanic, Asian, American Indian/Alaska Native, Native Hawaiian or other Pacific Islander, multiracial, and missing.

• SRACE10 – School-reported race. This variable has eight levels: White, Black, Hispanic, Asian, American Indian/Alaska Native, Native Hawaiian or other Pacific Islander, Two or More Races, and missing.

• SCHTYPE – Type of school. This variable has six levels: public, private, Catholic, Bureau of Indian Education, Department of Defense, and missing.

• CENSREG – Regions as defined by the Census Bureau. This variable has five levels: Northeast, Midwest, South, West, and missing.
  

• UTOL4 – Type of location. This variable has five levels: city, suburb, town, rural, and missing.

• PARED – Parent education level. Information for this variable is collected from students in grades 8 and 12 only. It has six levels: did not graduate from high school, high school graduate, some education after high school, college graduate, unknown, and missing.

• SLUNCH1 – School-reported free or reduced-price school lunch. This variable has four levels: eligible, not eligible, not available, and missing.

• SD3 – School-reported student with disability. This variable has four levels: Yes - IEP, Yes - 504, no - not SD, and missing.

• EL3 – English learner. This variable has four levels: yes, no, formerly, and missing.

• ACCOM2 – Accommodations. This variable has three levels: accommodated, not accommodated, and missing.
  

• CHRTRPT - Charter school flag. The variable has three levels: yes, no, and missing.
  

Main effects not used in interactions:

• UTOL12 – Type of location. This variable has 13 levels: large city, midsize city, small city, large suburb, midsize suburb, small suburb, fringe town, distant town, remote town, fringe rural, distant rural, remote rural, and missing.
  

• SLUNCH – School lunch. This variable has seven levels: not eligible, reduced-price lunch, free lunch, information not available, school refused to provide information, not participating, and missing. 
  

• Additional Accommodations Variables –  Each of the following variables is scored as yes/missing: bilingual dictionary, Braille presentation, Braille response, breaks, calculator, cue to stay on task, directions read aloud in English, directions read aloud in Spanish, extended time, large-print booklet, magnification device, one-on-one, read aloud (all), read aloud (occasional), read aloud in Spanish, school staff administers, scribe, Sign language presentation, Sign language response, small group, Spanish-English booklet, and special equipment.
  

• CENSDIV – Census division. This variable has 10 levels: New England, Middle Atlantic, East North Central, West North Central, South Atlantic, East South Central, West South Central, Mountain, Pacific, and missing.
  

• PCTBLKC – Percentage of Black students. This variable has six levels: 0, 1–5, 6–25, 26–50, 51+, and missing.

• PCTHSPC – Percentage of Hispanic students. This variable has six levels: 0, 1–5, 6–25, 26–50, 51+, and missing.

• PCTASNC – Percentage of Asian students. This variable has six levels: 0, 1–5, 6–25, 26–50, 51+, and missing.

• PCTINDC – Percentage of American Indian students. This variable has six levels: 0, 1–5, 6–25, 26–50, 51+, and missing.

Two-way interactions included in the population-structure model: All main effects listed above were used to examine two-way interactions with the exception of Hispanic x Hispanic and CENSREG x REGION. (Note: Hispanic x Hispanic interaction means that Hispanic [yes, no, or missing] is not crossed with Hispanic from DRACE10 or SRACE10.)

Three-way interactions included in the population-structure model: Hispanic x 5 race variables x (each of the) main effects listed above.

Variables obtained from survey questionnaires:

• All categorical student, school, and teacher survey questions are also included in the model.
  
• Student questionnaire (PDFs) – Grade 4  (353K), Grade 8  (371K), and Grade 12  (280K).
  
• School questionnaire (PDFs) – Grade 4 combined Mathematics, Reading, and Science  (346K), Grade 8  (347K), and Grade 12  (372K).
• Teacher questionnaire (PDFs) – Grade 4 combined Mathematics, Reading, and Science  (323K), and Grade 8  (243K).

Printer-friendly Version