DATE:     4/6/01
CONTACT:  Dr. Gregory Hawkins, (864) 656-0217
WRITER:   Diane Palmer, (864) 656-4741

Study accounts for factors 
influencing SAT scores

     CLEMSON -- Do the effects of poverty, class size, racial
composition and teacher characteristics affect Scholastic Aptitude 
Test (SAT) scores? 
     They do, if you ask Gregory Hawkins, assistant director and
research coordinator with the Jim Self Center on the Future, housed 
at the Strom Thurmond Institute at Clemson University.
     Standardized test scores are now relied upon as benchmarks of
school performance. Those scores may gain even more importance if 
many of the Bush Administration's education initiatives move 
forward and if South Carolina enforces its "meet the standards or
repeat the grade" policy.
     "Raw scores alone are meaningless," Hawkins said. "There are many
other factors that go far in explaining raw scores on the SAT."
     The Jim Self Center did a study at the request of the Office 
of Access and Equity at Clemson University, based on the 1998-99 
SAT scores, comparing actual and predicted scores. The purpose of 
their inquiry was to better enable them to identify South Carolina 
high schools that appear most successful in stimulating high-level
performance from students in challenged environments.
     "We looked at SAT performance at every high school in South
Carolina, while accounting for the effects of poverty, racial
composition, class size and various teacher characteristics -- such 
as education levels, salaries and other factors," Hawkins said. 
"In doing so, we level the playing field because we know, for 
example, that poverty has a tremendous negative relationship 
with academic performance".
     The Center developed statistical models that determine what 
a school should average on the SAT, given its circumstances. They 
then compared these "expected" scores to the school's actual 
average scores.  
     "By doing so, we were able to identify those schools that do 
far, far better than expected and those schools that didn't quite 
meet expectations, despite their more favorable circumstances," 
Hawkins said.  
     The study identified the top 30 South Carolina high schools 
that performed higher than expected on the SAT for the 1998-99 
academic year and the 30 schools that failed to meet expectations.
     Academic Magnet High School in the Charleston School District 
was number one with an average SAT score of 1,208, 179.94 points 
better than the 1,028.1 that had been predicted.
     Harleyville-Ridgeville High School of Dorchester 4 District, 
on the other hand, made an actual average SAT score of 719, 136.8
points less than the predicted 855.8.
     This research suggests that school-level influences -- the
practices within individual schools -- may greatly impact student
academic performance, according to Hawkins. "We recommend field 
studies to identify those practices that either advance or inhibit
student performance relative to expected performance levels," he 
said.  "With this knowledge," Hawkins added, "we may be better 
able to help other schools facing similar circumstances."
     Hawkins and his office have a grant proposal under consideration
with the U.S. Department of Education, to further this work and 
perform these field studies.
     According to Hawkins, additional work is needed to identify
possible latent variables associated with racial composition and
poverty and more fully comprehend their affects on academic
performance.  
     "We don't think for one moment that the color of one's skin
determines their academic abilities, but rather the circumstances 
and experiences of being impoverished or a minority member may 
present obstacles to performance," Hawkins said.
     The full report can be viewed at the Jim Self Center on the 
Future website at: 
http://www.scfuture.clemson.edu/education/sat/SATperf.pdf. 
Hawkins also has these scores available for each of the state's 
193 high schools.
                              END