A*StarSM Experience Completed A*StarSM Reviews

Review of Elementary Schools

The following reviews were conducted as demonstration projects in three East Coast and Midwestern states.

2008 Statewide Grade 5 Math

Audit conducted during the spring of 2010
Audit review of response patterns by school; 1,702 schools reporting 104,530 student records.

2008 Statewide Grade 4 Math and Science

Audit conducted during the summer of 2009
Audit review of response patterns by school; 1,311 schools reporting 87,161 student records.

2000 Statewide Grade 3 Math

Audit conducted during the fall of 2001
Audit review of response patterns by school; 2,317 schools reporting 140,203 student records.

Review of Classroom Test Administrations

The following reviews were conducted as part of a multi-faceted research and applications development project in a large, northeastern urban school district.

1998 – 2001 Elementary & Middle School Reading and Math

Exploratory reviews of assessments in an urban school district. Full reviews of 37,063 classes with a total of 910,783 students and partial review of 1,828 classes with a total of 47,654 students.

Reviews of Assessment Center Testing at Occupational Training Schools

The following reviews were prepared to meet the triennial requirements of the U.S. Department of Education under 34 CFR 668.150(b)(8). Termed "Ability-To-Benefit" testing, the test administrations reviewed were conducted to qualify, in part, applicants for federal financial assistance to attend post-secondary training programs.

2006 A Review of Ability-to-Benefit Test Administrations

A review of testing conducted to qualify applicants for federal financial assistance to attend post-secondary training. Reviewed 6 tests involving 145,509 applicants, 831 test administrators and 885 test locations.

2002 A Description of Ability-to-Benefit Test Administrations

A review of testing conducted to qualify applicants for federal financial assistance to attend post-secondary training. Reviewed 6 Tests involving 161,920 Applicants, 711 Test Administrators and 661 Test Locations.

Review of Employer Administered Job Applicant Testing

The following reviews were conducted as research support for the evaluation of Ability-To-Benefit testing. Employer administered testing has provided an exceptionally stable set of norms against which to compare norms developed in other assessment settings.

1995 – 1998 Job Applicant Testing

Exploratory reviews of response patterns by employer job applicant groups. Reviewed two forms of a Verbal Skills test based on 38,667 job applicants in 692 groups and two forms of a Quantitative Skills test based on 35,319 job applicants in 632 groups. All groups arose from job applicant testing by 87 employers at various locations across the United States.

Support for A*StarSM Interpretations

Monitor test administrations: A*StarSM personnel have monitored a wide range of test administrations, including those in elementary and middle school, at occupational training schools, and at employer hiring locations. This experience supports the interpretation of many different forms of misadministration, including testing site problems, inadvertent proctor errors, and deliberate manipulation.

On site audits: On site audits of assessment programs where irregular response patterns have been found support conclusions of misadministration as the cause. Specific misadministration practices are found to result in characteristic irregularities.

Review of answer documents: Many forms of improper influence involve changing or adding to answers recorded on answer documents. Reviews of answer documents find that patterns of 'bubbling in' or erasing answers may confirm certain forms of response pattern irregularity and yet are unaffected by other forms of pattern irregularity. A review of answer documents may be helpful, but is more often inconclusive.

Study of sequential test administrations: Analysis of multiple test administrations by a large number of teacher/proctors allows an analysis of the frequency of irregularities by test administrator. The analysis finds that proctors who are found with irregular response patterns are more than 20 times more likely to have irregular response patterns on subsequent test administrations.

Study of response pattern irregularities across institutions and test taker groups:
A*StarSM has conducted response pattern reviews in multiple school districts and across several hundred occupational training schools. The nature of response pattern irregularities are found to be highly similar, reflecting method of misadministration rather than test taker group or institution where the testing is conducted.