Alan Viarengo is usually so obviously far out of touch with
reality as to be unworthy of comment. However, in a recent letter
to the editor Mr. Viarengo used statistical references that are not
common knowledge for the public.
Alan Viarengo is usually so obviously far out of touch with reality as to be unworthy of comment. However, in a recent letter to the editor Mr. Viarengo used statistical references that are not common knowledge for the public. Because the misuse of data and statistics may not be apparent to the general public, he needs to be called to task by someone who knows a thing or two about statistics.
Mr. Viarengo and I teach statistics. Mr. Viarengo teaches statistics at Gavilan College, I teach AP Statistics at Gilroy High School. Mr. Viarengo knows that his statements are not statistically valid. He is attempting to misuse statistics to fit his political agenda.
Mr. Viarengo proposes the use of the ANOVA test in evaluating teachers for merit pay. He states correctly, “if.students are divided up randomly, a comparison can still easily be made.”
Random sampling and random distribution are the foundations upon which most statistical analyses are based. The reality of K-12 education is that students are never randomly distributed. If students in California were randomly distributed, then students in Gilroy would be shipped all over the state to attend classes. This is obviously not reality, so comparisons between school districts that call for a random sample of the population are not valid.
Within a school district, students are often not randomly distributed. In Gilroy students attend neighborhood schools. Neighborhoods do not contain random members of the population. Comparisons between schools within a district that require a random sample of the population are not valid.
Within a school, students are not randomly distributed. The realities of K-12 class scheduling make it impossible to randomly assign students to teachers. Comparisons between teachers within a school that require a random sample of the population are not usually valid.
Since a random sample of students can not be assigned to each teacher, by Mr. Viarengo’s own admission, the ANOVA test is not a valid statistical measure.
Mr. Viarengo’s insists that California be compared to Utah in educational data. It is obvious that California and Utah have very different populations. Some of the differences known to show variation in educational success are:
• persons over 25 who are high school graduates – CA: 76.8 percent, UT: 87.7 percent;
• persons below the poverty level, CA: 14.2 percent, UT: 9.4 percent;
• households that speak a language other than English at home, CA: 39.5 percent, UT: 12.5 percent.
There are many other data points where Utah should have an educational advantage over California in both outcome and cost. The question becomes, “What makes Mr. Viarengo think that a comparison between California and Utah is statistically valid?”
Mr. Viarengo’s use of statistics and data analysis is out of touch with reality. Mr. Viarengo knows better than to use data and statistics in the manner in which he does. Mr. Viarengo’s statements about data and statistics are to be dismissed out of hand like almost everything else he says.
Wayne Scott, GHS Mathematics Teacher