Granite School District has begun a process to improve achievement test scores. The district was disappointed with the results of recent statewide tests that showed Granite students below national and state averages in many areas.

Principals in the district have been asked to set up school committees to analyze test results and develop strategies to improve them. The Stanford Achievement Test Eighth Edition is being administered annually to all fifth-, eighth- and 11th-graders throughout the state. The 1990 results will form a base line to determine how schools increase or decrease performance."None of us wants to have this happen again," Deputy Superintendent Riley O'Neil said. "We played a poor first game."

Earlier this week, a Magna parent contacted the Deseret News to express outrage at the low scores in his neighborhood. Of the 10 lowest-scoring schools in the district, five were in the Magna area, Steve Harris said. He said the scores are indicative of the district's tendency to favor east-side schools.

Trish Hull, a PTA official in Magna, said the low scores were surprising. "It does distress us. We are determined to change it for next year." She said she had been asked to serve on a committee at Pleasant Green School to study the issue.

School officials deny the allegations that any favoritism is shown to east-side schools. Funds are distributed on a per-child basis. There are, however, variations in socioeconomic factors, parental support and other indicators that put schools at risk.

However, the test results showed that some schools perform better than others, even with the same risk factors.

Granite School Board member Lynn Davidson said during board meeting Tuesday that there needs to be a study of the effect federal funds infused into low-income schools to see whether they favorably affect performance. The additional money doesn't seem to be having much effect, he said.

The district needs to find ways to "reach out to kids who are expected to fail," he said.

O'Neil said each school is being challenged to bring scores to the upper end of the level of expectations set by the state Office of Education. The state office created a range of scores for each school by considering the number of children in the school who receive free or reduced-price lunch or whose families receive state aid - indicators of the school's socioeconomic base.

Schools will be urged to share the results of their self-studies and strategies they develop to raise scores, he said.

Assistant Superintendent Briant J. Farnsworth told the board that teachers will get help in administering the test, including in-service training. They will be encouraged to help children prepare for the tests, although no actual test material will be taught.

Some children don't understand the vocabulary used to explain the tests or the seriousness of the tests, Farnsworth said. The standardized tests are biased against children who are educationally at risk to begin with, he said.