Sunday, October 9, 2011

What Works in Education?

HOMESCHOOLING (or some variation of it - cottage schools or co-ops). According to a survey conducted in 2009 by the Home School Legal Defense Association (HSLDA) that "surveyed 11,739 homeschooling students and their families from all 50 states through 15 independent testing services," homeschooled students perform well above their peers on standardized testing, a measure which is widely used to measure "success" in schooling. The students performed between 34-39 percentile points above their peers in all subjects, and the rate of success crosses socio-economic status, race, income and whether or not the parent or person providing the education is certified or not (see full report here).

The most interesting thing in this study is the absolute lack of correlation between certain negative factor and success - unlike public school, lower-income homeschoolers did just as well as their upper-income peers, and the education level of the home "teacher" did not change results significantly either (nor did the amount of spending per pupil).

So what made the difference? Why do homeschooled students do so well?

The keys are time, attention and a filter. Homeschooled stduents tend to think of the world as a classroom, a place where learning can, and does, happen at any time. Students in public schools are mandated to be in school 180 days a year from X a.m. to Y p.m., and that is where they learn. Rare is the publicly schooled stduent who, on their own, indulges in deep thought and exploration of a topic of study - school ends when they board the bus to go home. The backslide of reading level over the summer is widely documented in research, especially among lower-SES families; students can lose as many as two grade levels in reading over the summer because reading is something that only happens in school. Statistics for math computation are less widely known, but anecdotally math teachers can relate stories of having to re-teach entire units just to get started with the grade-level curriculum. Homeschoolers learn year-round, and although they may be forced to document a school-year based attendance, many school through the summer, holidays, etc.

The second factor is attention. A homeschooled student may be competing with 2.5 other siblings to get their questions answered (homeschoolers tend to have 3.5 kids per family, as opposed to the national average of 2), while students in a public school are competing with 15, 20, 25 students (or more - some states like North Carolina do not put a limit on class size in 4-12th grade) for a teacher's attention. A student may never get their question answered and may just stop trying.

The final factor for success in homeschool versus public school is the filter. Public schools, their teachers and students are bound by the whims of people not in the classroom - legislators, school district administrators, building level administrators and any number of other consultants, parents and "stakeholders" in the system. Programs come and go on a yearly basis, administrators require teachers to deliver fancy projects that may or may not be tied to instruction for "open house," superintendents have pet projects that must be accomodated. There is no filter for what each student needs, what works in the classroom or what just make sense for the age level and ability of the students. Homeschoolers don't need to follow directives or change what works for their students just because someone has decided it worked in another house, so it must be implemented in all houses. Parents and people who run co-ops and cottage schools (which can be structured as one class a week or part-time instruction in multiple subjects, or many other variations depending on what the community needs) can make decisions that directly improve the experience; they can filter out what doesn't work, and they are not bound by someone else's dictates (even governmental intervention - or lack thereof - made no appreciable difference in scores, according to the HSLDA study).

Homeschooling works, and here is the proof. Stay tuned for more.

No comments:

Post a Comment