API test scores are poor indicators

Staff writer Dan Johnson deserves credit for an informative article in the April 4 issue of your newspaper (Most test scores in local districts lower than similar state schools?). He not only provides information about API scores of some Petaluma schools, but goes beyond to explore some of the real influences on these scores that some, especially at the state level, hold so dear.

I wish the Press Democrat would do more to inform citizens about what exactly the API scores it annually publishes actually tell us, or rather do not tell us, about our schools. I believe, as a high school teacher for over 30 years at Rancho Cotate High School, that API and other test scores actually tell us very little about the quality of the education provided to students and the efforts of principals and teachers at any particular school.

First of all, there is a big problem with using test scores to measure school improvement or lack thereof from one year to the next since each year?s score represents a different cohort of data (students). A different group of students will simply produce a different result on any test. Such comparisons are meaningless yet they are used to measure a school?s improvement.

Furthermore, one of the greatest predictors of the API is the PEL (parent educational level) of a school. Old Adobe superintendent Diane Zimmer-man is exactly correct when she cites the influence of the socio-economic level of a community and the PEL of the school on the API. My personal research of the data available at the state department of education Web site has shown that there is an incredible positive correlation between the API scores of Sonoma County and North Coast high schools (and probably elementary schools as well) and the PEL of a school.

The above two factors may be such powerful influences on API that a school?s test score may decline if the student body changes due to families with limited English or less parental education increasing in a district or parents of a higher socio-economic level leaving a school for some reason, despite the effort of a school?s staff working harder to raise scores at their school. Try as they might, it may not be enough to raise scores as much as might be desired.

As mentioned in the article, the grouping of schools with others supposedly ?similar? is also questionable, especially when PEL is not a factor listed as a criterion. I did a study of the group of schools our high school is placed in and found the correlation between PEL and API as well, despite supposedly being ?similar? schools. Our school is rated in the bottom 10 percent of schools in our similar group, but then again, that bottom sub-group has the lowest PEL of the 100 schools the state thinks we are similar to. The highest two sub-groups had the highest PEL.

And how does one justify the inclusion of Rancho Cotate with charter schools across the state? How similar are we to Galileo High School in San Francisco, an urban school with a student body that is over 80 percent Asian-American? Does this make sense? I would predict that the same might be true for the Petaluma area schools mentioned in the article as well. Are they really all that similar to the other schools in their group?

If a school like ours, or Elsie Allen and Piner in Santa Rosa sees a significant number of students of a higher PEL transfer to another school, the API score will decrease. Our school has seen a decline in PEL since the opening of Technology High School in our district. That school has a 2006 API score of 839 while ours was 696, but their PEL is 3.67 while ours is 2.96, having declined each year since Technology High opened.

What irks me is that the public sees the API scores and evaluates schools accordingly. I know there is a disconnect between our API scores and the quality of education we provide. I would imagine it is the same elsewhere in other schools. Realtors use test scores to sell homes, but does a school in a particular neighborhood have a high API because it really is an especially good school with a good staff or does it score high because of the socio-economic makeup of the community it resides in?

There are simply much better indicators of the quality of a school than test scores or the API. It is time this notion becomes a greater part of the discussion about school quality at all levels so we can end the mania with testing our students to death. Thanks to the Argus-Courier for doing its part.

(Mark Alton is a teacher at Rancho Cotate High School in Rohnert Park.)

UPDATED: Please read and follow our commenting policy:
  • This is a family newspaper, please use a kind and respectful tone.
  • No profanity, hate speech or personal attacks. No off-topic remarks.
  • No disinformation about current events.
  • We will remove any comments — or commenters — that do not follow this commenting policy.