Tuesday, March 6, 2012

Latin and the SAT (pars secunda)

Magis utile nil est / artibus his quae nil utilitatis habent.  (Ovid, Epistulae ex Ponto, Liber I, V (Cottae Maximo) ll. 53-4)

Last year I put together a few observations and thoughts about Latin and the SAT.  After I noticed the 2011 Total Group Report was out some months ago, I got to looking at it and noticed that things seemed to be about the same as the year before.  This time I thought I'd see what was going on with the Subject Test data, and that led me to looking at the correlation between years of study and various subject areas.  Eventually that got me into wondering where the three (four if you count a distance learning course offering) languages at my school fit with the national data.  You can download copies and check the data at the CollegeBoard Data & Reports page.  For earlier years look under Archived SAT Data & Reports.

Caveats: The SAT report says that the data in it is for high school graduates in the year 2011.  It claims a total number of test-takers of 1,647,123.  What percentage of all high school graduates this represents I don't know.  The 2011 ACT report claims "About 49% of all 2011 high school graduates in the United States took the ACT during high school, or about 1.62 million graduates."  Were there then about 3.3 million high school graduate?  There must be some percentage of graduates who took neither test, those heading for the military, into technical/vocational training programs, or directly into the workforce.  Presumably there are some stats on this, but I don't know where they are.  The ACT report does not mention foreign languages.

Subject tests have been going on for what seems like forever.  I think I took them in English and French back in 1966 or 1967.  At least I picked up 6 credits at Penn State for them.  In 2011 there were some 312,000 subject test takers.  This is up from about 279,000 in 2006.  Of these about 3,000 have taken the subject test in Latin.  That one percent of the subject test taking population does better on the three aspects of the general SAT than any other group, and of course rather considerably better than the norm for all students (which has been totaling about 1500 the last five years - see below).  Note that I have just added up the three scores for Critical Reading, Mathematics and Writing.


year
total of the mean scores of Latin students (about 1% of the total number of test takers)
mean of the total mean scores for other students
next highest set of mean scores to Latin
subject of the test with the next highest mean scores
lowest set of mean scores
subject of the test with the lowest mean scores
2006
2011
1833.8
1922
German
1698
Spanish listening
2007
2020
1831
1907
German
1676
Spanish listening
2008
2026
1827
1912
German
1665
Spanish listening
2009
2019
1833
1929
German
1675
Spanish listening
2010
2028
1838.25
1927
Physics
1685
Spanish listening
2011
2026
1838.05
1917
Chemistry
1674
Spanish listening

As I did last year, I wonder why the fact that these Latin students have done about ten percent better than the average subject test taker, and about 35 percent better than the average SAT taker doesn't lead to larger numbers of Latin students.  One minor curiosity is that these Latin subject test takers score better (by an average of 15 points) than those who take the Math 2 subject test on the math part of the general SAT, where they score about the same as Chemistry students and only some 35 points, on average, below Physics students who do the best on that part of the SAT.  I suppose these Latin subject test takers were mostly among the students who took two or three subject tests.  A large majority of subject test takers took two or more tests.  Whatever that may mean, there seems to be something about Latin itself, or about the sort of student who takes Latin seriously enough to do a Subject Test in it, or both, that sets this small group of subject test takers distinctly apart from all the rest.

Each annual report breaks down subject areas into six major groups: English, mathematics, natural sciences, social sciences & history, foreign languages, and arts & music (which also includes computer related courses).  I thought I'd see how foreign language students in general compared to others over the last six years in the total of their scores for Critical Reading, Mathematics and Writing per number of years of study in any particular one of these subject areas.  That led to the production of these six charts, including one for students taking AP or Honors courses.  In these charts the subject areas are sorted in descending order by the mean of the scores for the six years.

One year of study:

Subject
Total mean scores


2011
2010
2009
2008
2007
2006
mean
Arts and Music
1495
1503
1502
1494
1505
1504
1500.5
Social Science & History
1386
1371
1379
1356
1378
1412
1380.33
Natural Sciences
1395
1380
1373
1349
1354
1379
1371.67
English
1365
1343
1310
1307
1353
1454
1355.33
Mathematics
1358
1334
1286
1266
1326
1421
1331.83
Foreign Languages
1322
1321
1314
1318
1321
1328
1320.67

Two years of study:

Subject
Total mean scores


2011
2010
2009
2008
2007
2006
mean
Arts and Music
1503
1509
1507
1495
1519
1522
1509.17
Social Science & History
1429
1429
1435
1423
1432
1439
1431.17
English
1426
1411
1438
1402
1423
1454
1425.67
Foreign Languages
1407
1412
1414
1413
1420
1424
1415
Natural Sciences
1409
1404
1402
1390
1400
1410
1402.5
Mathematics
1382
1365
1368
1346
1372
1397
1371.67

Three years of study:

Subject
Total mean scores


2011
2010
2009
2008
2007
2006
mean
Foreign Languages
1540
1545
1544
1536
1548
1548
1543.5
Arts and Music
1509
1519
1516
1509
1533
1531
1519.5
Social Science & History
1465
1474
1473
1463
1475
1477
1471.17
Natural Sciences
1433
1440
1442
1438
1451
1454
1443
English
1396
1400
1403
1401
1408
1397
1400.83
Mathematics
1385
1386
1390
1391
1400
1396
1391.33

Four years of study:

Subject
Total mean scores


2011
2010
2009
2008
2007
2006
mean
Foreign Languages
1657
1673
1669
1654
1681
1681
1669.17
Arts and Music
1588
1603
1596
1586
1612
1616
1600.17
Natural Sciences
1555
1577
1564
1561
1580
1586
1570.5
Social Science & History
1541
1553
1548
1538
1554
1557
1548.5
Mathematics
1521
1543
1540
1533
1550
1554
1540.17
English
1528
1539
1535
1523
1541
1545
1535.17

Five or more years of study:

Subject
Total mean scores


2011
2010
2009
2008
2007
2006
mean
Foreign Languages
1681
1681
1704
1681
1706
1699
1692
Natural Sciences
1644
1652
1683
1653
1671
1665
1661.33
Mathematics
1638
1643
1669
1640
1657
1652
1649.83
Social Science & History
1607
1605
1640
1609
1621
1609
1615.17
Arts and Music
1592
1585
1597
1584
1600
1595
1592.17
English
1585
1570
1607
1580
1587
1566
1582.5

Students who reported taking an AP or Honors course in these subject areas:

Subject
Total mean scores


2011
2010
2009
2008
2007
2006
mean
Foreign Languages
1721
1731
1733
1714
1738
1736
1728.83
Mathematics
1704
1713
1715
1704
1729
1731
1716
Natural Sciences
1705
1714
1715
1702
1723
1724
1713.83
Arts and Music
1695
1704
1711
1689
1710
1707
1702.67
Social Science & History
1676
1687
1690
1677
1698
1701
1688.17
English
1663
1674
1678
1665
1686
1687
1675.5

Foreign language students start off at the bottom of the barrel in year one.  By the third year, they have taken the lead and stay in it solidly the rest of the way.  The numbers clearly indicate that the long-term, serious pursuit of a foreign language is associated with significantly higher critical reading, math and writing scores on the SAT to a greater extent than the long-term, serious pursuit of any of the other five subject areas.  The fact that this information has been out there for years, needing only to be looked at and put together, makes a person wonder about either the bona fides or the intelligence of those who are behind such things as NCLB, CCSS, STEM, and other schemes for what passes for school reform and improvement in this day and age, all of which mention foreign language instruction either not at all or only as a kind of amusing hobby draining energy and time from important subjects like English.  It is indeed curious that if one were to be limited to five choices of subject areas to teach or to study, this particular set of data logically would dictate the elimination of English.

Probably the distinction between required and elective courses and the prevailing policies of most schools have much to do with these results.  Most schools probably require things like mathematics and English for all four high school years, natural sciences and the social sciences for maybe three or four, while foreign languages and the arts are either pure electives or something in between (e.g., perhaps two years are required for kids intending to go to college; or maybe everybody has to take a semester or two of music or art or technology or human development or something of that sort).  English and math have been the focus of the high-stakes testing program in Pennsylvania known as the PSSA so they tend to get an awful lot of attention from the education department, school administrators and teachers.  Foreign languages don't get a lot of attention.  And that is a good thing, given the tendency of the officious know-it-alls in PDE, PSEA and colleges of education to pump out enormous quantities of obfuscatory balderdash on whatever half-baked fad their beady little brains happen to have become focused on.  In other words foreign language teachers have the advantage of not being impeded by the burden of figuring out how to deal with a confusing array of such things as "assessment anchors" (with their reporting categories, references, descriptors, cross-walking, eligible content, sample items, etc., etc.) unlike English teachers who have to deal with that whole mess and somehow either use it or pretend to use it in order to teach in such a way as to reduce the odds of the sort of student failure on the PSSA tests that could result in unpleasant consequences for the teacher.  Perhaps if good English teachers had more time to actually teach English, students who have to take English for four years might do a bit better on the SAT.  Another big factor behind these results is probably the self-selection thing.  Naturally subjects which all students are constrained to take will have a higher concentration of the unwilling and unmotivated type of student who doesn't read, think or learn very well. Non-mandatory subjects will tend to have more students characterized by willingness and personal motivation who have some confidence in their ability to do well because they have acquired some reading, thinking and learning skills, or in the case of the arts and music because they have more than ordinary artistic and musical talent.  A third part of the probable explanation for these kind of results is simply that there is something about the study of foreign languages that tends to improve a person's mind to a greater extent and in more ways than the study of any other subject.

The next logical question for me then was to work out which language is associated with the probability of scoring above average on the three SAT tests.  Here I've limited my inquiry to the languages offered at my school, one of which (German) is only offered by distance learning, and may in fact have fallen by the wayside by now.  [Just as a side note: Chinese and Hebrew students tend to do about as well and sometimes a bit better than Latin students: the Chinese students tend to be really good on the math part of the test, while the Hebrew students, like Latin students, tend to do about the same on all three segments of the test].  A few years ago, one of my better Latin students took this course and did well; he majored in German in college and is currently doing post-graduate work somewhere in Germany.  Here are two charts showing first the sets of critical reading, mathematics and writing scores for the past six years for Latin, German, French and Spanish, with the basic national average added so as to make it obvious where each stacks up relative to it, and second the enrollment trends for each of these four languages as measured by the number of students who reported taking at least some quantity of coursework in each language.


Subject
Total mean scores


2011
2010
2009
2008
2007
2006
mean
Latin
1655
1667
1670
1657
1678
1668
1665.83
German
1583
1595
1599
1594
1600
1596
1594.5
French
1557
1569
1571
1562
1582
1575
1569.33
National average
1500
1509
1509
1511
1511
1518
1509.67
Spanish
1496
1503
1505
1494
1510
1507
1502.5


Subject
total number of students in thousands


2011
2010
2009
2008
2007
2006
mean
Latin
76
77
77
72
86
87
79.17
German
53
53
54
52
61
64
56.17
French
220
214
214
205
236
247
222.67
total test takers
1647
1547
1530
1518
1494
1465
1533.5
Spanish
905
890
882
820
870
852
869.83

There is perhaps some irony in the circumstance that the number of Latin students (who routinely score about ten percent better than the national average) has decreased by about 11,000, and the number of Spanish students (who routinely score a little under the national average) has increased by about 53,000.  Pretty obviously something besides a familiarity with this data and an appreciation for the value of high SAT scores has been driving curriculum, staffing and enrollment decisions. 

In my school (with a total student population in four grades of about 600), the enrollment in Latin IV is -0-, in Latin III -3-, in Latin II -6-, in Latin I -11-, in French IV -0-, in French III -0-, in French II -10- and in French I -25-, in Spanish IV -6-, in Spanish III -16-, in Spanish II, -60-, in Spanish I -90-.   We have no AP or Honors courses in foreign languages.  Half, or slightly more (at least last year), of our senior class goes to college.  With the rather low enrollment in the third and fourth years it appears that some of our students are not making the most of their opportunities.