Homeschoolers Score High on TerraNova 2nd Edition
from Howard Richman
[From Issue 85, Winter 2003-2004, of the PA HOMESCHOOLERS® newsletter]
In the fall of 2003, Susan and I administered the new TerraNova 2nd Edition Survey Test at group test sites throughout Pennsylvania. This test was published at the beginning of the 21st Century and represents a growing emphasis in American Education upon math concepts and language expression, and a reduced emphasis upon math computation and grammatical terminology. The idea is that math education should be emphasizing the kinds of problems that a calculator can’t do and language education should be teaching the skills that are necessary for articulate and organized writing. The new test is more similar to the SAT college entrance test than was the old test that we gave previously.
The home educated students scored very well on this test. In statistics there is a term called the “median” which roughly means “middle”. The median score falls exactly in the middle because half the scores are lower and half are higher. When the TerraNova test was normed, it was given to school students. Half of them scored above the 50th percentile and half scored below the 50th percentile (that’s how the 50th percentile was calculated).
The middle score for the 915 homeschooled students was much higher. The middle “Total Score,” a score which combined reading, language and math was the 77th percentile. The distribution of the homeschooled students total scores is shown below:
The middle score in reading was the 81st percentile, the middle score in language was the 75th percentile, and the middle score in math was the 65th percentile.
Only 535 of the students stayed for the optional science and social studies tests which we gave in the afternoon. The middle homeschooler scored at the 77th percentile in science and the 77th percentile in social studies.
Homeschoolers in the required grades (3rd, 5th, and 8th) did quite well with reading. The middle third grader scored at the 83rd percentile in reading. The middle fifth grader scored at the 73rd percentile, and the middle eighth grader scored at the 79th percentile. Their distribution of scores is shown in the graphs below:
Similarly, homeschoolers in all three grades did quite well in the language test. There appeared to be a trend in which the homeschooled students language scores improved at the higher grade levels. The middle third grader scored at the 66th percentile in language. The middle fifth grader scored at the 80th percentile in language, and the middle eighth grader scored at the 81st percentile in language.The distribution of their scores is shown in the graphs below:
Homeschooled students in Pennsylvania and Washington have not done much better than school-educated students in third grade math, no matter what achievement test is used. This is true because in both PA and WA, the compulsory school age is 8, and some homeschoolers tend to take a very relaxed approach toward formal math education until their children get to compulsory school age. In both states, the math scores of homeschoolers rise rapidly so that by fifth grade, most homeschoolers are scoring well above the average school student. This was again true with the TerraNova 2nd Edition test. The middle third grader scored at the 57th percentile in math in third grade, but at the much higher 72nd percentile in fifth grade, and the 70th percentile in eighth grade. The distribution of scores is shown in the graphs below:
Although homeschoolers, overall, did very well, there is always room for improvement. Next year we will invite parents to fill out a questionnaire while their kids are being tested Then in a future issue of this newsletter we will tell you which homeschooling practices and curricular materials correlated with higher achievement in reading, language, and math on this new test.