DEAR EDITOR:
Columnist Doug Meier has once again misused data. Mr. Meier
abuses data with regularity. I wonder if he mishandles data
intentionally in an attempt to mislead readers or out of mere
ignorance?
DEAR EDITOR:
Columnist Doug Meier has once again misused data. Mr. Meier abuses data with regularity. I wonder if he mishandles data intentionally in an attempt to mislead readers or out of mere ignorance?
On June 11, Mr. Meier gave us the following, “[Parents and homeowners are] seeing the correlation between steadily declining SAT test scores over the last seven years and the reign of multicultural hogwash during that time.”
First, let’s examine the decline in SAT scores. Mr. Meier is fixated on this piece of data and has yet to demonstrate that the data is statistically valid.
In order for a drop in SAT scores to be valid, Mr. Meier must answer the following:
1. Is the “brain drain” reported by the Alliance for Academic Excellence real or imagined? If the drain is real, what has been the effect on SAT scores?
2. Have the demographics of SAT test takers in Gilroy changed?
3. Has the percentage of students taking the test changed?
Until Mr. Meier answers these questions, and a few others, his use of a decline in the SAT scores must be ignored in the making of policy. Making policy with data that has not been demonstrated to be statistically valid would be foolhardy.
Even if one assumes that the decline in SAT scores is valid, the use of a correlation between SAT scores and multicultural literature shows political bias not statistical savvy. A correlation does not show a cause and effect relationship.
To demonstrate that a correlation does not show a cause and effect relationship consider an example where the number of leaves on a tree is correlated to the probability that it will snow.
It can be shown as the number of leaves on a tree decreases the probability there will be snow increases.
One can draw fancy charts and graphs to convince others of such a correlation. That does not mean there will be an increase in the probability of snow in July when all of the leaves on trees are burned off in a forest fire.
To show a cause and effect relationship usually requires an experiment.
Mr. Meier has not done an experiment and thus fails to make the cause and effect relationship between SAT scores and multicultural literature; therefore, the correlation between SAT scores and multicultural literature is invalid until demonstrated otherwise.
Mr. Meier has once again misused data.
Does he abuse data from a lack of knowledge or for political purposes?
Wayne Scott,
GHS Math Teacher
Submitted Saturday, June 12 to ed****@************ch.com
