Washington Times Op Ed - Home School Results on Standardized Tests

[ELaurie]

Washington Times Op-ed—Outstanding Results on National Tests

 

by J. Michael Smith

HSLDA President

The U.S. Department of Education (DOE) estimates there are more than 1.5 million children being taught at home. Furthermore, the DOE estimates that homeschooling has been growing at 7 percent a year for the last 10 years.

Two related questions many people ask are, “Why has homeschooling become so popular?†and “What is motivating parents to take on this daunting responsibility?†In the most recent report by the DOE, parents gave three basic reasons for choosing home education: to provide religious or moral instruction, concerns about the school environment, and dissatisfaction with academic instruction at other schools.

Regarding the third reason, there is new research showing that the average homeschooler who takes standardized achievement tests is doing very well. The study, commissioned by the Home School Legal Defense Association and conducted by Brian Ray, an internationally recognized scholar and president of the nonprofit National Home Education Research Institute (NHERI), is called “Progress Report 2009: Homeschool Academic Achievement and Demographics.â€

The study included almost 12,000 homeschool students from all 50 states who took three well-known standardized achievements tests—the California Achievement Test, the Iowa Test of Basic Skills and the Stanford Achievement Test—for the 2007–08 academic year. The students were drawn from 15 independent testing services, making it the most comprehensive homeschool academic study to date.

The results reinforced previous homeschool studies conducted over a period of 25 years.

Five areas of academic pursuit were measured. In reading, the average homeschooler scored at the 89th percentile; language, 84th percentile; math, 84th percentile; science, 86th percentile; and social studies, 84th percentile. In the core studies (reading, language and math), the average homeschooler scored at the 88th percentile.

The average public school student taking these standardized tests scored at the 50th percentile in each subject area.

Beyond academics, there were significant results regarding achievement gaps. It is common knowledge that gender, as well as parents’ income and education levels will greatly affect a public school student’s academic results. Public schools have invested greatly to try to close these achievement gaps. The study, however, shows the achievement gaps found in public school were greatly diminished for the home educated.

For example, homeschooled boys scored at the 87th percentile and girls at the 88th. Household income had little impact on the results of homeschool students: Children of parents with an income between $35,000 and $49,000 scored at the 86th percentile, whereas children of parents with an income over $70,000 scored at the 89th percentile.

As one would expect, the education level of parents did affect the results. For example, homeschool students of parents without college degrees scored, on average, at the 83rd percentile for the core subjects. When one parent had a college degree, those students scored at the 86th percentile, and when both parents had a college degree, those students scored at the 90th percentile. There was virtually no difference, however, between the scores of students whose parents were certified teachers and those who were not.

In summary, the results were slightly better than the most recent large academic study regarding homeschoolers (the 1998 Rudner study), and the average homeschool test results continue to be 30-plus percentile points higher than their public school counterparts.

In my opinion, there are two main factors for these outstanding results: the educational environment where learning takes place, and the individualized, one-on-one instruction. Most homeschool students are directly taught by their parents, who love their children enough to make the sacrifice to stay at home to make sure their child is taught in a safe and loving learning environment. Second, one-on-one instruction emphasizes the best interests of the child rather than the best interests of the group.

In a sentence, homeschooling is a recipe for academic success.

[Sandy in Indy]

Very interesting...and validating.

[Corraleno]

Washington Times Op-ed—Outstanding Results on National Tests

 

by J. Michael Smith

HSLDA President

 

I've been trying to find this on the Washington Times webpage, to send the link to some relatives, but I can't find it. Could you post the link?

Thanks,

Jackie

[tex-mex]

My son participated in this study -- via BJU Press. I'm thrilled that the data shows HSing is successful. Yay!

 

Here is the link from HSLDA:

 

http://www.hslda.org/docs/news/washingtontimes/200908100.asp

 

I love this quote from the link:

"There was virtually no difference, however, between the scores of students whose parents were certified teachers and those who were not."

Edited September 2, 2009 by tex-mex

[Caraway]

Can anyone link to the original?

[Laura Corin]

Can anyone link to the original?

 

But I can't find it in the online paper.

 

Laura

[tex-mex]

But I can't find it in the online paper.

 

Laura

 

Hidy,

The author -- Michael Smith is the president of the Home School Legal Defense Association. He may be contacted at (540)338-5600; or send email to media@hslda.org. The office can help you.

 

The group that ran the study -- NHERI -- has a website:

 

http://www.nheri.org/Latest/Homeschooling-Across-America-Academic-Achievement-and-Demographic-Characteristics.html

 

The official study paper will be released in November 2009. HSLDA got the press release out early -- as they recruited many members to be a part of the study. That's how we got involved.

Edited September 2, 2009 by tex-mex

[Reya]

I've been trying to find this on the Washington Times webpage, to send the link to some relatives, but I can't find it. Could you post the link?

Thanks,

Jackie

 

The WT is really supportive of homeschooling, BTW--they've sponsored homeschooling lecturers!

[JudyJudyJudy]

Thanks for posting this. It's very interesting.

[Pamela H in Texas]

as they recruited many members to be a part of the study.

 

Not to be a killjoy, but....

 

The above is the reason this sort of thing is not taken seriously by many educated people. It looks good to say, "our kids score 30% higher than their kids," but it's actually just not true. Our self-selected kids, on average, score 30% higher than all of their kids.

 

I really don't understand why a better study with scientific results can't be done. I mean, we couldn't do one in Texas because there'd be no way to do so. But in a state that requires testing, why is it not possible. The state gets information from each legal homeschooler so why can't that information be used to say that in X state, homeschoolers score A% above public schoolers. Do that for enough states and you have a good idea about how well homeschoolers REALLY do.

 

Now, I think homeschoolers on average will still score better than public schoolers on average. I just don't think the gap will be 30 percent.

 

Now, there are still issues with doing it my way. First, it's still a self-selected group. It would only be those legally homeschooling and in states that require testing. I don't think there is an issue with multiple tests being considered as tests are normed at 50% for the average anyway. And it will be somewhat limited in what it can offer, especially if there is a difference in scores between states which require only testing 3 times over the years and those who have lots of regulation. Then you wonder how homeschoolers in states like Texas (which has about 1/6th of the country's homeschoolers) and Illinois would play in.

 

I wish there were a way to do a COMPLETELY random test, especially including a big sampling from ALL states, so that regulation could be considered. Do states with a bit more oversight have higher test scores than those without? Do states ranked higher/lower in terms of education in general follow those same trends when homeschooling is the method or does homeschooling even the playing field? What is the percentage of kids scoring well or poor (how does our bell curve look?)?

 

Anyway, I just worry how we look when we spew such data which is obviously scientifically flawed. I wish there were a way we could have more accurate information. Sure, it might not look quite as good, but I think accuracy is more important than inflated numbers.

[creekland]

Not to be a killjoy, but....

 

The above is the reason this sort of thing is not taken seriously by many educated people. It looks good to say, "our kids score 30% higher than their kids," but it's actually just not true. Our self-selected kids, on average, score 30% higher than all of their kids.

 

 

 

:iagree:

 

Parents that are like those of us on this board take schooling seriously and have children that tend to do well. Parents who 'homeschool,' but don't (and I'm not meaning method, but those that really DON'T) won't have kids who took these tests. In public schools, all kids take the tests at this level - including those whose parents don't give a hoot.

[Carrie12345]

While I find the information on the (lack of) achievement gaps extremely interesting, am I the only person who finds it unfair to use these numbers as an overall comparison?

 

Unless my understanding of math is worse than I thought, "the 50th percentile" will always be the average for public school students, whether the majority of students are illiterate or rocket scientists. 50% will always be smack dab in the middle of 0% and 100%.

 

If homeschooled students were calculated independently, THEY would get slapped with the 50th percentile average, regardless of how smart or dumb any of the individual students were.

 

I do believe the statistics are telling, just not nearly as much as we're supposed to believe.

 

I also don't think it's fair to compare *my homeschooled children to kids without families or kids without food or kids without books in their house or kids whose parents who don't value education. Obviously, you're going to find far fewer of those kids being homeschooled, which is definitely not the public schools' "fault".

 

I also think standardized tests are a horrible measuring stick. When I was in school, I almost always scored in the 99th percentile, and I'm fully aware that my children are receiving a much better education than I did (in middle class suburbia).

 

Like I said, the numbers are interesting. But they really don't mean all that much.

[Carrie12345]

Not to be a killjoy, but....

 

 

 

Man! I should have just waited a few more minutes, lol!

[Ferdie]

My son participated in this study -- via BJU Press. I'm thrilled that the data shows HSing is successful. Yay!

 

Here is the link from HSLDA:

 

http://www.hslda.org/docs/news/washingtontimes/200908100.asp

 

I love this quote from the link:

"There was virtually no difference, however, between the scores of students whose parents were certified teachers and those who were not."

 

Ditto. I remember filing out survery.

 

Very cool to see the results. Thanks for the link.

[Ferdie]

:iagree:

 

Parents that are like those of us on this board take schooling seriously and have children that tend to do well. Parents who 'homeschool,' but don't (and I'm not meaning method, but those that really DON'T) won't have kids who took these tests. In public schools, all kids take the tests at this level - including those whose parents don't give a hoot.

 

True to a point, but I thought in certain states testing was mandatory. My friend in CO told me she has to test every other year.

[Renee in NC]

True to a point, but I thought in certain states testing was mandatory. My friend in CO told me she has to test every other year.

 

And in NC, too, but this study doesn't take all hs students who test - it only studies those who test *and* volunteer to send their scores to NHERI. This study isn't scientifically valid, unfortunately, so it won't be a great help to convince people who are anti-hs.

[Pamela H in Texas]

True to a point, but I thought in certain states testing was mandatory. My friend in CO told me she has to test every other year.

 

BUT those aren't the results we're seeing! We're seeing a self-selected group of results. The parents decide if they want Johnny's scores (and the family's questionnaire answers) to count. I would LOVE to see how every legally homeschooled 5th grader in CO scored versus how every 5th grader did.

 

But even THAT wouldn't be fair because as was brought up, the average homeschooled 5th grader doesn't LOOK LIKE the average public schooled 5th grader. You're still not comparing apples to apples. But at least we'd be comparing fruit to fruit instead of dogs to apples.

[JudyJudyJudy]

Wow, I skimmed the article quickly, so I guess I missed something: I didn't realize that the only scores of homeschoolers that counted were scores of people who wanted to send them in! How can anyone think that is a valid study? I'd expect better of "an internationally recognized scholar." Even my 10-year-old knows better than that.

[JudyJudyJudy]

But even THAT wouldn't be fair because as was brought up, the average homeschooled 5th grader doesn't LOOK LIKE the average public schooled 5th grader

The same argument could be made when comparing scores of students in upper middle class suburbia to scores of students in poor rural areas or areas filled with projects. It simply isn't a fair comparison.

[TravelingChris]

There have been other studies that showed that there was no link at all between how well homeschooled students did and whether their state was highly, moderately, or not regulated. Other studies have looked at things which don't depend on parent self selection like ACT or SAT scores. You can say that not all students take these tests and that is true but in my state, not all students take the school tests too. Last year, they started to allow portfolio reviews for students who don't know English well or had certain handicaps. The controversy here is are some of those children being allowed portfolio reviews improperly. I remember when I was in school, they took some of us out of our honors classroom and gave us an idiotic test all because we spoke more than one language. Now the funny thing was not that we passed the test, but they labeled one of my neighbors as a limited English speaker. She didn't have issues because of another language since she had been in the US since infancy. She had issues because she was limited in intelligence, not in English. They mislabeled her and didn't give her the right classes.

 

 

In my fifteen years of homeschooling in six states and two countries, I have not run across poorly educated homeschoolers except for ones that had obvious problems and I couldn't assess whether they were educated to their abilities or not. I found the children who were using the lamest curriculum and had the most unfortunate backgrounds to still be doing at least as well as publicly educated children and usually better. How do I know- I taught in lots of Sunday Schools and VBS, helped in lots of co-ops and homeschool programs, worked with Scouts and 4H,etc, etc. I don't need convincing that homeschool works. Could a school environment work for these same kids- some yes, some no. I think about adopted children I have seen who came from horrendous orphanages and needed lots of attention and one on one tutoring to get up to speed in our country and our language. I think about dyslexic students who graduate and when you hear what they did with their homeschooling education, you just feel amazement. I know that my own daughter who is shy and suffers from medical issues would not be doing the advanced work she is and has done at a school where they would focus on her medical and other issues and not focus on her abilities.

 

We participated in this study. My middle one helped raise the averages, my youngest lowered them. This year, my youngest topped out on math scores and even her lower scores improved a lot. I focus on what my kids need, nobody did that in either my dh or my schools.

[JudyJudyJudy]

It shows that homeschoolers can be very successful. I don't think it shows much else.

 

But to be fair, states manipulate the data as well. I just read an article about NY (I am trying to find it). Basically it said not to get too excited about the recent awesome improvements in test scores because the numbers needed to be considered passing (or proficient) have been lowered. So nothing has really changed. I suspect the same is true in other states. What makes me crazy is that people seem to be ignoring this fact. No miracles have been performed, they are just manipulating the numbers.

No, in Georgia, they just cheat. ;) Seriously, though, at least two administrators were arrested for cheating, and others are being investigated.

[Stacy in NJ]

of homeschoolers in general, and self selection is taking place, obviously that's true for public and private schooled children as well.

 

The drop out rates in our urban school districts are sometimes near the 40% mark. Those kids will never get near a PSAT, ACT or SAT test.

 

College entrace exam are not going to be representative of all kids in ps's or homeschoolers. While more kids than ever are taking the exams, there are still large numbers of ps and private schooled children NOT taking the exams at all.

[JudyJudyJudy]

In my fifteen years of homeschooling in six states and two countries, I have not run across poorly educated homeschoolers except for ones that had obvious problems and I couldn't assess whether they were educated to their abilities or not.

I wish I could say the same.

[angela in ohio]

Because so many homeschoolers will not allow testing (either because they are private about their information or because they don't believe in testing their child,) a true study is impossible. This is the closest we can get. This test is slightly more accurate than the numbers homeschoolers have been using for years, when just those using both the full BJU package and standardized testing scored well.

 

Yes, it is flawed, but the masses do not understand statistics in the least. The media uses that to their advantage every day, so why not let it be good for us once. ;) :lol:

 

No, I don't think the average homeschooler is anyone near the 80-something percentile. I think studies like this damage homeschoolers themselves as well as our credibility with those smart enough to see the flaws. Many homeschoolers I know rest on the statistics, and it makes them put in less than their full effort. Between this and the "all public school kids are getting a horrible education" mentality, I think some homeschoolers get a warped view of the situation.

[Pretty in Pink]

I remember when this survery came around through HSLDA. Self-reporters are not a reflection of the general public. The numbers look good. Too bad the method is flawed.

[EKS]

Not to be a killjoy, but....

 

The above is the reason this sort of thing is not taken seriously by many educated people. It looks good to say, "our kids score 30% higher than their kids," but it's actually just not true. Our self-selected kids, on average, score 30% higher than all of their kids.

 

I really don't understand why a better study with scientific results can't be done. I mean, we couldn't do one in Texas because there'd be no way to do so. But in a state that requires testing, why is it not possible. The state gets information from each legal homeschooler so why can't that information be used to say that in X state, homeschoolers score A% above public schoolers. Do that for enough states and you have a good idea about how well homeschoolers REALLY do.

 

Now, I think homeschoolers on average will still score better than public schoolers on average. I just don't think the gap will be 30 percent.

 

Now, there are still issues with doing it my way. First, it's still a self-selected group. It would only be those legally homeschooling and in states that require testing. I don't think there is an issue with multiple tests being considered as tests are normed at 50% for the average anyway. And it will be somewhat limited in what it can offer, especially if there is a difference in scores between states which require only testing 3 times over the years and those who have lots of regulation. Then you wonder how homeschoolers in states like Texas (which has about 1/6th of the country's homeschoolers) and Illinois would play in.

 

I wish there were a way to do a COMPLETELY random test, especially including a big sampling from ALL states, so that regulation could be considered. Do states with a bit more oversight have higher test scores than those without? Do states ranked higher/lower in terms of education in general follow those same trends when homeschooling is the method or does homeschooling even the playing field? What is the percentage of kids scoring well or poor (how does our bell curve look?)?

 

Anyway, I just worry how we look when we spew such data which is obviously scientifically flawed. I wish there were a way we could have more accurate information. Sure, it might not look quite as good, but I think accuracy is more important than inflated numbers.

 

I completely agree with this.

 

Homescoolers who take education seriously are more likely to test. And those kids are more likely to score well. Also, many kids being tested using independent testing services are being tested *at home by their parents*. I know for my son this had a huge impact on his scores. And, frankly, under those conditions, you can't be sure that standard procedures were followed. For a real scientific study, standard procedures are vital.

 

As for the curve, if you were to graph the scores from the kids at our homeschooling group (not that I know what they are, but I do know the *kids*), I would suspect that you would find a somewhat bimodal distribution, that is, a high peak and a low peak.

 

I agree that this would be a very interesting study.

[Orthodox6]

Does anyone know of "validating research" on this topic, which is published by anyone OTHER THAN the HSLDA ? Although their data may be accurate, I would be much more inclined to give credence, if there were supportive reports issued by organizations without such entrenched -- and political -- ties to the homeschooling community. Also, I don't know their statistical methodology -- (Is it included with the published report ?) -- to assess whether or not these reported numbers are "monkeyed numbers".

 

For me, this is very similar to placing my trust in medical studies issued by one, and only one body -- and nobody else ever stands up to voice agreement.

Edited September 2, 2009 by Orthodox6

[Ferdie]

And in NC, too, but this study doesn't take all hs students who test - it only studies those who test *and* volunteer to send their scores to NHERI. This study isn't scientifically valid, unfortunately, so it won't be a great help to convince people who are anti-hs.

 

Just for clarification we didn't send in our test scores. The testing facility sent them in with my consent. When I ordered the test they asked if we would be willing to be part of the sample and I completed the survey before my son took the test and long before I received the test scores. So my son's test scores were part of this survey regardless of the results.

 

But I see what you mean. If I am testing in a mandatory state and I predict that my child will have low scores then I would decline to be part of the survey. I guess that could happen.

Edited September 2, 2009 by Ferdie

[kalanamak]

I completely agree with this.

 

Homescoolers who take education seriously are more likely to test. And those kids are more likely to score well. Also, many kids being tested using independent testing services are being tested *at home by their parents*. I know for my son this had a huge impact on his scores. And, frankly, under those conditions, you can't be sure that standard procedures were followed. For a real scientific study, standard procedures are vital.

 

I agree that this would be a very interesting study.

 

:iagree:

 

And, on a more ominous note, I'm sure the HSLDA people are bright enough to know this was a selected group, and therefore are knowingly "lying with statistics". :glare:

Sophistry, and not even very clever sophistry. We need better stats, especially comparing apples and apples, e.g. hs LD kids vs. public school LD kids.

I once read that the upshot of the Kansas City MO court-mandated change in public schools showed that the only school that improved was the one that required the parents sign a statement saying they would make their kids do their homework....not the schools with extra computers etc. I read it in a ?Newsweek? article, and I'd like to see something more substantial. Anyone know anything about this?

[creekland]

As for the curve, if you were to graph the scores from the kids at our homeschooling group (not that I know what they are, but I do know the *kids*), I would suspect that you would find a somewhat bimodal distribution, that is, a high peak and a low peak.

 

I agree that this would be a very interesting study.

 

This would be my best guess prior to an actual study as well. It would be very interesting to see results if a study could be done.

[Denise in Florida]

Unless my understanding of math is worse than I thought, "the 50th percentile" will always be the average for public school students, whether the majority of students are illiterate or rocket scientists. 50% will always be smack dab in the middle of 0% and 100%.

 

If homeschooled students were calculated independently, THEY would get slapped with the 50th percentile average, regardless of how smart or dumb any of the individual students were.

 

I had the same reaction to the 50% :lol: When I read the average public school student scored 50% I actually said 'well duh'. It is a useful comparison though because the point was that homeschool kids were scoring higher than ps kids. This addresses one of the concerns the public has with homeschooling.

 

I don't know how they could select the data but it would be interesting to compare homeschool kids with ps kids who had parents who were involved with their education. Is the difference due to teaching or parenting? Is it even possible to draw that distinction? We are fairly unschoolerish so for us it is the same thing.

 

I loved the article, thanks OP.

[Perry]

I all depends on how they got their sample, and what the participation rate was. If they picked a true random sample and had a high participation rate, the results could be valid.

 

If they cherry picked the sample, or had low participation, the results are useless.

 

I'm guessing useless, but we'll see. It's awfully difficult to do these studies well.

[Susu]

The study does give us some information.

 

If the average public schooler scores at 50%...which of course, by default, they do...at least you do have a base line to compare your child's scores, whether you are in the public school system or at home. I can individually look at my child's scores and see how they are doing compared to their public schooled counterpart. And this is the reason I test...to see where I have both strengths and weaknesses in my curriculum choice. If my child is scoring 95% in math compared to his peers, then I can assume I am on the right track with math. If he scores 1O%, then obviously I must make a major change!

 

I would be interested in a more broad based result including more home schoolers, but I have a real problem with "standardized" testing anyway, and am not sure this could be done in a way that would satisfy us independent home schoolers! (Would we want to be required to test based on the public school agenda? How about the fact that as classical home schoolers, our curriculum differs significantly from many public schools?) What is a "standard" test anyway?

 

1. There is no "standard" curriculum. And things like history/social studies, and science vary greatly from state to state....not to mention how varied the curriculum is within home schools.

 

2. The public system actively and systematically teaches to the test. I have seen it in action both as a former teacher and as a parent who sent her children to school for a semester. In our homeschool, I do not prep my child in any way for the test, and I see numbers of questions on the test that my children did not even touch that year. (For example...we might have focused on biology and the science portion of the Iowa for that year might be heavily focused on physical science.)

 

3. Who are these test writers??? lol I have seen too many illogical and poorly written test questions. Many of them assume experiences in a classroom that would bewilder a home schooled child...or even a rural public schooler. And many questions list answers that don't include another equally valid and logical choice. (My children have pointed out several questions that they had a different answer to (and their answer was perfectly legitimate.) This drives me crazy! And on top of it all, some of the questions are unwieldy. Have you looked at the grammar section on the Iowa? Four sentence paragraph, with a, b, c and d on the sides...and the child is supposed to find the mistake somewhere in that mix? (Some of them were hard for me...a college graduate and certified teacher!) It is not surprising that the school system spends a lot of time training the children to answer the style of question found on many standardized tests.

 

I am not sure what kind of empirical study could be done to properly measure how well home schooled student do compared to public schooled students. For that matter, I am not convinced that standardized tests do that even within the public schools. There are just too many variables. Perhaps a better study would involve college graduation rates and/or success in adult life among both public schooled and home schooled populations...but most of that is more anecdotal..and subject to those same troubling variables of each child's home environment.

 

Susu

[Ferdie]

I all depends on how they got their sample, and what the participation rate was. If they picked a true random sample and had a high participation rate, the results could be valid.

 

If they cherry picked the sample, or had low participation, the results are useless.

 

I'm guessing useless, but we'll see. It's awfully difficult to do these studies well.

 

I participated in the above survey and I don't know how they would have cherry picked my family. I tested once before in the state of California, offsite through my hs group. I do not remember the name of the standarized test, but the results came directly to me and were not released to the survey group mentioned in the OP.

 

The second time I tested I was selected to participate in the survey after I ordered the test from BJU, but before I received the test. I filled out the survey before I gave my ds the test at home. At the time I remember thinking that my ds was at a disadvantage taking the test at home because my other two children were so loud. LOL!

[Perry]

I participated in the above survey and I don't know how they would have cherry picked my family. I tested once before in the state of California, offsite through my hs group. I do not remember the name of the standarized test, but the results came directly to me and were not released to the survey group mentioned in the OP.

 

The second time I tested I was selected to participate in the survey after I ordered the test from BJU, but before I received the test. I filled out the survey before I gave my ds the test at home. At the time I remember thinking that my ds was at a disadvantage taking the test at home because my other two children were so loud. LOL!

 

 

One way to do it is to look at neighborhood data- I forget what it's called but it's a database used for marketing surveys. If you only invite those in the higher SES areas, or people living in certain school districts, you're going to have extremely skewed data.

 

It would be very sneaky. But it could easily be done.

 

If they don't describe how the sample was selected in the report, it's suspect.

[angela in ohio]

Unless my understanding of math is worse than I thought, "the 50th percentile" will always be the average for public school students, whether the majority of students are illiterate or rocket scientists. 50% will always be smack dab in the middle of 0% and 100%.

 

If homeschooled students were calculated independently, THEY would get slapped with the 50th percentile average, regardless of how smart or dumb any of the individual students were.

 

If you norm the public-schooled students and the homeschooled students independently, you can no longer compare their results to each other. The 50th percentile homeschool child, in this data group, would still be more accomplished (as assessed by the standardized test) than the 50th percentile public school child, though.

[tex-mex]

Holy guacamole... looks like this is a hot button topic? ;)

 

I understand where people are coming from... but please take into account that basically all studies are flawed. Take for example, my hubby working for a prestigious university as a Clinical Research Scientist. He was in charge of an FDA (Phase II & III) drug studies for asthma back in the 80's and 90's. There was one drug that caused one of his subjects to die directly from. I recall the doctors being alarmed and sent in the data to the pharmaceutical company for review. Surprisingly, with no problem... the FDA approved the drug. Of course, that same drug has been recalled recently due to the same reason 18+ years ago. You can use the data to manipulate the study. Common knowledge.

 

Some of my thoughts:

1) As a former public school schoolteacher, I don't think I ever had a class average of 50th percentile (working at grade level ability= average). Working in many inner city schools in Southern CA and the Mojave Desert... usually I had kids who were not even at grade level. I remember one year as a 3rd grade teacher having 3/4 of the class at a first grade reading level -- I couldn't teach from the 3rd grade reader. Frustrating. With a lot of hard work -- at the end of the year, I had (out of a group of 34 students) many kids test at the high 30th up to the low end of the 40th percentile (2 standard deviations below grade level AND one standard deviation low average ability). However, 5-6 kids were usually at or below the 23th percentile (which is cause for an IEP/Special Ed evaluation) and 1-2 kids were exceptional at 60% percentile (high average ability). Our school's grade level average were even more depressing with norms in the low 40% group. Things changed when NCLB came into play and in my state (CA) soon came demands to teach to the test. Districts had us test quarterly and try to pass that annual state standardized test. (In reality, that was a HUGE no-no from a purist POV about standardized testing... the test is meant to be taken "cold" and with no prompting or pre-testing. NCLB... ironic.) But all of that was close to 10 years ago -- I have colleagues currently teaching telling me that it is even worse. How valid are those results we see (i.e. TAKS)? A sticky wicket. (NOTE: At the end of my teaching career, I finally did have class test averages at 55th to 70th percentile... but that was at a public charter school that "cherry picked" high ability students. This is what other teachers in my school district said against our results in comparasion.)

 

2) Recruiting for a study is commonly done (hubby did this for years for the FDA drug studies) by advertising. Someone said that it was an unfair % of students tested. If you look at the study, you will note there were K-12 test subjects from all 50 states, Guam and Puerto Rico. My hubby said that is a dream group from his POV as a former clinical researcher... it is difficult to gather a diverse group of subjects. Another point about study recruitment: Son and I have a rare genetic liver disease. I am participating in an NIH sponsored drug study. The recruiters of this study are all over the world. Unfortunately, they have yet to get 8 healthy female (X-linked) carrier subjects -- so far they have only been able to recruit 4 healthy subjects in a period of 3 years. I am test subject #2. Reason? The majority of females who have this disease are cognitively challenged (i.e. mentally retarded) or in such bad debilitating health they cannot travel to participate in the rigors of the study for Phase I and II. Does this mean we should toss out the results gathered so far for the newest line of wonder (rare orphan) drug to help that 1 in 80,000 child born with this disease? Of course not. Like I said, studies are flawed. But the data gathered can still help researchers. We attend the annual medical conferences for our rare disease to understand the results of the study.

 

Yes, HSLDA may have had alternative motives (ummm... they are a lobby group to protect HS rights like the current possible ratification of the UN Rights of the Child.) but if folks will note there has NOT been a study about HSing in a long time. Again, should we throw out the baby with the bath water? I say no. The results basically show that HSing is working. I know plenty of my former NEA teacher (anti-HSing) friends who will be seething hopping mad when this journal article gets released in Nov 2009. I ask you who are quick to criticize... why tear fellow HSers down? We need to support one another despite our differences.

 

3) To whomever brought up that LD students were not part of the study: My son has ADHD, Asperger's and a whole host of labels that in a public school would qualify him for Special Ed. Like someone said, I also did not turn in test results. With BJU Press, on the testing form (before the test was administered) there was a blurb if I wanted my results sent directly to NHERI from BJUP. I checked yes. I realized my son's scores (based on past ITBS tests) were low in some areas. I refuse to teach to the test and give it to my son "cold" using my old teacher (test proctor -- classroom environment, timed test, stick to the script... etc) methods of adhering to the rules of being an authentic test administrator. I know personally of 3 people (who were not HSLDA members) in my support group who also allowed their child's test data be sent to NHERI... which is 5 students who all had LD from dyslexia to ADHD to other issues. Again, I cannot speak for the 11,000+ test subjects NHERI looked at, but am willing to bet despite the LD label, our special needs kids still did better on average compared to the public schools. HSing is making a difference.

 

I don't mean to be a downer for those speaking their mind... but I do have issue with those who are not willing to see the other POV and are quick to criticize. We HSers need to be supporting one another.

Edited September 3, 2009 by tex-mex

[Catherine]

Just a quick FYI-recruiting method is a huge source of bias in any study. Eliminating bias may be impossible; minimizing it is desirable though. I don't think reasonable people on either side of this question could honestly say that all efforts were made to eliminate selection bias in this study.

[johnandtinagilbert]

Homescoolers who take education seriously are more likely to test.

hmmm...really? Take a look at our course load. Heavy. Serious. Likely to test...not so much. Am I unique in this?

 

In the past I have picked what I saw as "vital years" and used standardized tests, via professional providers, and each time I confirmed what I already knew about my students, so we've stopped that brand of testing. Last year, a portfolio evaluation, at $20/dc worked just fine.

 

For college prep, we will work through some SAT type Workbooks and probably have them start taking PSAT in 9th grade, but I found the above comment curious.

 

Is that really the case?

[tex-mex]

Just a quick FYI-recruiting method is a huge source of bias in any study. Eliminating bias may be impossible; minimizing it is desirable though. I don't think reasonable people on either side of this question could honestly say that all efforts were made to eliminate selection bias in this study.

 

:iagree:

[Paisley Hedgehog]

nm

[unsinkable]

:iagree: We were 'invited' to participate in this study as well, but declined. So, I know for certain there were *at least* two 'average' scores that were left out of the sample! LOL

 

I know plenty of homeschoolers who don't test (in our state it's not required). Some of those homeschoolers are rigorous in their studies and some don't seem to give a hoot. Most of the ones I know just rest on their belief that "well, at least my kids are getting a better education than in the public school!" with little emperical data to back that up.

 

But, hey, we all know that there's more to school than test scores and there's no way to gague that on one of these scales.

 

And for me, test scores are less important than many other things.

[Twinmom]

It's an imperfect study, for sure...but probably the best we've got right now. There's no perfect, unbiased study in any field, as much as we might like to believe one could be done. There are always complicating issues and those doing the report just have to do their best to make it accurate. There are so many factors involved in sampling a group of homeschoolers that, frankly, I'm surprised they were able to accomplish a study at all. We aren't really a group that likes to be studied, as far as I can tell! Better to fly under the radar and all that...;)

 

That said, I contributed my kiddos' test scores voluntarily, even though I knew I'd have a special needs student scoring low. My little attempt to make the results more universal! :D I think the study is a nice bit of information to have (and probably better than what we've had before!), but I'd not put a tremendous amount of weight behind it.

[Pretty in Pink]

Yes, HSLDA may have had alternative motives (ummm... they are a lobby group to protect HS rights like the current possible ratification of the UN Rights of the Child.) but if folks will note there has NOT been a study about HSing in a long time. Again, should we throw out the baby with the bath water? I say no. The results basically show that HSing is working. I know plenty of my former NEA teacher (anti-HSing) friends who will be seething hopping mad when this journal article gets released in Nov 2009. I ask you who are quick to criticize... why tear fellow HSers down? We need to support one another despite our differences.

 

Nope. Not trying to tear anybody down. There's nothing wrong with pointing out obvious flaws in a supposedly scientific study. I find the study mildly interesting. Nothing more.

[Mom0012]

I actually found this study kind of discouraging. I work very hard with my kids and yet, they don't score in the 85% across the board. But now I'm supposed to think that most homeschoolers do.

 

I pulled out my daughter's IOWA test and even though she is an excellent student and tests well, it really wasn't even possible for her to score that high in many areas. In her sentence reading section, she got 20 out of 20 correct, but her national percentile is 79.7 and the National curve equivalent is 67%. So 85% wasn't even possible. This is the case with many of the subtests where she got 1 or 0 wrong.

 

When I used to test my son at home with the CAT, he scored 90% and above in most areas. It was an easy test compared to the IOWA and it was done in the comfort of his own home without any distractions. Now that I send him to the school for testing (IOWA), he's nowhere near that. I suspect if I adminstered the IOWA at home instead of at the school he'd do much better, but like I said, in many areas, you can't even score above the 70% even if you got every single problem right.

 

I always wonder what test people are using when they say their kids are scoring so high.

 

Lisa

[Pamela H in Texas]

.really? Take a look at our course load. Heavy. Serious. Likely to test...not so much. Am I unique in this?

 

Ooops, you're backwards though. Testees are more likely to be conscientiously homeschooled students. The point wasn't that people who homeschool conscientiously test, but that the ones that use tests are ones that generally homeschool well (regardless of method). The mom who can't get her kids to the table 2 out of every 5 days per week, who knows her kids are 3 years behind because of her lack of homeschooling, etc are unlikely to test. They are also more unlikely, if they have to test per the law, to okay someone else look at the scores. Even without it being my fault, I'd be unlikely to send my son's test scores to someone else. I definitely wouldn't do it if it were my fault!

[nmoira]

Methodology issues aside -- I'm not impressed by what I've read so far, but I'll wait to read the report -- the following jumped out at me:

 

In my opinion, there are two main factors for these outstanding results: the educational environment where learning takes place, and the individualized, one-on-one instruction.

Most of these children would also perform well in a school environment. Children whose parents take an active interest in their education tend to perform better than their peers.

 

Most homeschool students are directly taught by their parents, who love their children enough to make the sacrifice to stay at home to make sure their child is taught in a safe and loving learning environment.

And those who don't homeschool don't love their kids enough? This is ludicrous and is patently offensive to non-homeschoolers.

[Pretty in Pink]

And those who don't homeschool don't love their kids enough? This is ludicrous and is patently offensive to non-homeschoolers.

 

I agree. I found the quote to which you are referring in the article to be...snarky. Not necessary.

[ElizabethB]

There is a pre-publication report that addresses some of those statistical concerns:

 

He began by asking whether children with higher test scores participated in the study in a higher proportion than those with lower scores. Ray found that 69.4% of the parents did not know their children’s scores before the study was conducted. The scores for this group and for the group of parents who did know their children’s scores were the same: the 88th percentile.

 

Ray then compared the scores of study participants with all the scores (study participants and non-participants) from three of the four major

testing services involved with the study. Nearly all of the 22,584 students tested by these three services were homeschooled. The all- scores average was between two and four percentile points lower than the participants-only scores.

 

Ray also found no difference in the scores of students from groups with lower-participation rates compared to those of students from

higher-participation groups.

 

That does still leave out families who don't test...but several states make all homeschoolers test, those scores should be available somewhere.