Public school vs. private school ACT scores

This is a tale of 54,041 high school diplomas. That’s the number of public high school diplomas awarded in Tennessee last year (2006-2007). There are 324 public high schools in Tennessee. The public high schools are operated by 119 public school systems. There are 137 public school systems in Tennessee, but only 119 of them operate high schools.

I got curious this week about tracking down median ACT scores for Public vs. Private vs. Homeschool high school graduates. It turns out, even in the age of public data on the internet that this is not an easy question to answer. If the data to answer this question already exists somewhere on the internet, it’s extraordinarily well hidden. I spent several days searching for it… and I’m pretty handy with google. I did discover a blog in Kentucky which contained interesting articles commenting on the meaning of median ACT scores released for that state. Kentucky’s scores, released by ACT, Inc. of Iowa, give the median for ALL high school seniors, public, private, and homeschool. From the ACT data alone, you cannot tell how the public schools are performing, because ACT will not disagregate the data. Tennessee ACT scores are released in the same format as Kentucky.

But, it turns out, in Tennessee at least, there is a way to calculate median ACT score for the public schools. And if we know the number of public school students who took the ACT, and their median score, then we can calculate the median score for the remaining non-public school students.

In 2007, the median ACT for all students in Tennessee taking the test was 20.7. This is slightly below ( a half a point) the national ACT median score of 21.2. A half a point difference between two individual scores is probably not terribly significant. There are too many variables that can’t be controlled between two individual scores to ever be able to know why one student scored a half a point higher than another. BUT, comparing the median scores of two significantly sized groups IS meaningful… because all the individual variations offset and cancel each other out. 48,113 students took the ACT in Tennessee in 2007. 1,300,599 students took it nationwide. Comparing the averages for those two very large populations does tell us, with a pretty high degree of confidence, that Tennessee students did not perform quite as well as the national average.

But those 48,113 Tennessee students include public school, high school, and homeschool students. I have an inquiry in to ACT, Inc. asking them for the disaggregated data for those three groups, but they haven’t responded to me. The data would be very helpful in discussing some pretty pressing public policy questions about education. I don’t think it’s an accident that ACT doesn’t make the data readily available. I have the feeling that the data are not very flattering to public school administrators. And I suspect that’s why ACT hasn’t made them available.

But in Tennessee, there is another source of data about public school ACT scores – the Tennessee Department of Education itself. The Department has an online database that reports the number of students who took the ACT and the median composite score by school system. Actually, the online database has a great deal more information than that, but the median composite ACT scores are what I was interested in.

I don’t know whether it’s intentional or not, but the Tennessee DOE does not report the statewide median ACT score, nor does it make it easy to calculate, but all the pieces are there, on their website – they just have to be assembled.

So, I spent about four or five hours today, using the free wifi at University Pizza & Deli in Chattanooga, to pull up and copy off the median composite ACT scores for all 119 public high school systems in Tennessee. 35,725 public school students (out of 54,041 who graduate) too the ACT in 2007 – about 66.1% of the graduates. The median composite ACT score for all of them was 20.30. Since there were a total of 48,113 students who took the ACT in Tennessee, we can subtract out the public school students and the remaining 12,388 students were non-public school (private schools and home schools). And since we know the median composite ACT score for ALL students in Tennessee was 20.7 and the median for the PUBLIC school students was 20.30, we can calculate what the median composite score for the non-public schools was: that median composite ACT score in 2007 was 21.85.

So, we can now end the speculation and report with confidence that in 2007, in Tennessee, ALL students averaged a 20.7 composite ACT score, PUBLIC SCHOOL students averaged a 20.30 composite ACT score, and PRIVATE SCHOOL students averaged 21.85 composite ACT score. In other words, in 2007 private schools and home schools averaged 1.15 points higher on the ACT than the public schools. But of course, it’s the private school diplomas that the Department of Education thinks are suspect.

Since I had to compile the data for all 119 systems in a spreadsheet, I’ll post all of the data here – so that others can check my calculations, and so that the data will be available to everyone interested.

There are a number of other interesting observations about the public high schools that can be made from the data.

For example, here are the 10 public school systems in Tennessee with the HIGHEST median composite ACT scores:

TENNESSEE REGULAR % TAKING 2007 ACT Composite
SCHOOL SYSTEM DIPLOMAS ACT n median
1 Maryville City 321 76.9% 247 23.67
2 Oak Ridge City 321 68.8% 221 23.53
3 Kingsport City 400 82.8% 331 22.74
4 Greenville City 209 66.5% 139 22.68
5 Williamson Co. 1,966 80.6% 1,584 22.54
6 Tullahoma City 239 77.4% 185 22.35
7 Johnson City 398 73.1% 291 22.34
8 Pickett Co. 46 58.7% 27 22.11
9 Alcoa City 107 74.8% 80 22.01
10 Knox Co. 3,257 66.6% 2,168 21.97

And here are the 10 public school systems with the LOWEST median composite ACT scores:

TENNESSEE REGULAR % TAKING 2007 ACT Composite
SCHOOL SYSTEM DIPLOMAS ACT n median
1 Fayette Co. 187 65.2% 122 15.80
2 Memphis City 5,741 67.9% 3,898 17.56
3 Hancock Co. 62 38.7% 24 17.96
4 Haywood Co. 170 71.2% 121 17.98
5 Lake Co. 51 70.6% 36 18.11
6 Grainger Co. 241 53.1% 128 18.41
7 W. Carroll 79 54.4% 43 18.47
8 Campbell Co. 299 58.2% 174 18.63
9 Union Co. 196 53.1% 104 18.63
10 Hardeman Co. 234 56.0% 131 18.66

Here are the Here are the 10 public school systems in Tennessee with the HIGHEST percentage of graduating seniors who take the ACT:

TENNESSEE REGULAR % TAKING 2007 ACT Composite
SCHOOL SYSTEM DIPLOMAS ACT n median
1 McMinn Co. 292 92.5% 270 20.33
2 Union City 77 88.3% 68 19.93
3 Kingsport City 400 82.8% 331 22.74
4 Williamson Co. 1,966 80.6% 1,584 22.54
5 Bradford City 41 80.5% 33 19.18
6 Oneida City 83 79.5% 66 20.58
7 Shelby Co. 2,561 78.5% 2,010 21.72
8 Madison Co. 679 78.2% 531 19.27
9 Tullahoma City 239 77.4% 185 22.35
10 Huntingdon City 70 77.1% 54 20.20

And here are the Here are the 10 public school systems in Tennessee with the LOWEST percentage of graduating seniors who take the ACT:

TENNESSEE REGULAR % TAKING 2007 ACT Composite
SCHOOL SYSTEM DIPLOMAS ACT n median
1 Hancock Co. 62 38.7% 24 17.96
2 Fentress Co. 60 41.7% 25 19.92
3 Sequatchie Co. 116 44.8% 52 19.71
4 Greene Co. 488 45.7% 223 20.06
5 Trousdale Co. 91 47.3% 43 19.12
6 Johnson Co. 156 47.4% 74 19.81
7 Meigs Co. 94 48.9% 46 20.37
8 Washington Co. 656 50.8% 333 20.68
9 Bledsoe Co. 102 51.0% 52 20.73
10 Jefferson Co. 449 52.1% 234 20.52

Here are the 10 public school systems in Tennessee with the LARGEST number of graduating seniors who take the ACT:

TENNESSEE REGULAR % TAKING 2007 ACT Composite
SCHOOL SYSTEM DIPLOMAS ACT n median
1 Memphis City 5,741 67.9% 3,898 17.56
2 Davidson Co. 3,601 64.1% 2,307 19.11
3 Knox Co. 3,257 66.6% 2,168 21.97
4 Shelby Co. 2,561 78.5% 2,010 21.72
5 Rutherford Co. 2,328 66.1% 1,539 20.91
6 Hamilton Co. 2,322 68.0% 1,580 19.60
7 Williamson Co. 1,966 80.6% 1,584 22.54
8 Sumner Co. 1,691 62.9% 1,063 20.81
9 Montgomery Co. 1,644 59.9% 984 21.23
10 Wilson Co. 1,040 67.9% 706 20.70

And here are the 10 public school systems in Tennessee with the SMALLEST number of graduating seniors who take the ACT:

TENNESSEE REGULAR % TAKING 2007 ACT Composite
SCHOOL SYSTEM DIPLOMAS ACT n median
1 S. Carroll 31 58.1% 18 20.28
2 Van Buren Co. 37 62.2% 23 18.83
3 Richard City 37 70.3% 26 20.15
4 Bradford City 41 80.5% 33 19.18
5 Pickett Co. 46 58.7% 27 22.11
6 Hollow Rock-Bruceton City 47 57.4% 27 20.22
7 Lake Co. 51 70.6% 36 18.11
8 Fentress Co. 60 41.7% 25 19.92
9 Hancock Co. 62 38.7% 24 17.96
10 Huntingdon City 70 77.1% 54 20.20

The only significant sized sample of homeschoolers with ACT scores that I could find were 1997, 1998, and 2004 data released by ACT (cited on the HSLDA website). ACT reported that in 1997, 1,926 homeschoolers had a median composite ACT score of 22.5. ACT reported that in 1998, 2,610 homeschoolers had a median composite ACT score of 22.8. ACT reported that in 2004, 7,858 homeschoolers had a median composite ACT score of 22.6. These data are remarkably consistent over time AND they are significantly ABOVE the national averages. But remember, according the the Tennessee Department of Education, it is the homeschooler’s diplomas that are suspect.

Now, don’t you feel like you know the public school system in Tennessee much better?

Feel free to discuss amongst yourselves. Comments encouraged and solicited. Once again: here is the data. Or should that be, “here ARE the data…”

– Rob Shearer
Director, Schaeffer Study Center

7 thoughts on “Public school vs. private school ACT scores”

  1. The same old, lame excuses will be used to explain the higher scores in the private sector:

    1) Not all of the private schools are Category IV or independent homeschoolers. They are probably the low end of the spectrum. No data available to prove.

    2) The private schools can accept or reject whatever students they want, so they get all the “good” students (although the public system does not test all their students – the near exception being McMinn County who is getting it done on a shoestring budget compared to Memphis, et al).

    3) The private schools don’t have to deal with broken homes, parents that don’t care, urban youth, etc., etc. Soft bigotry code speak for “they don’t have to deal with the black kids”. The skin pigment level excuse. Proven wrong by the private sector.

    4) The private schools don’t have to deal with lower income, working poor, single parent, uneducated, etc. families. The income/education level excuse. Proven wrong by the private sector.

  2. So, if the State chooses to use Eric’s “excuses”, then they are in effect saying the following:

    1) They don’t want “good” students as government employees.
    2) They don’t want employees with supportive families.
    3) They don’t want employees from middle or upper income familes, or employees who grew up with both parents in the household, or employees with educated families, etc.

    In my book there can only be one of two reasons for this twisted thinking:

    1) They don’t want intelligent, self-thinkers in government positions, because they are the the intelligent elite and already have all the answers, or
    2)They know that they are NOT the intelligent elite and don’t want any intelligent, self-thinkers in government positions who might expose their inabilities.

Leave a Reply

Your email address will not be published. Required fields are marked *