Oct 24th 2012, 14:27 by R.L.G. | NEW YORK
LAST YEAR we looked at the first-ever global survey of English-language skills by EF Englishtown, an education company. This year, EF has produced its second study of the same subject. It's worth revisiting for the changes between last year and this one.
First, I'll repeat my caveat from last year: "This was not a statistically controlled study: the subjects took a free test online and of their own accord. They were by definition connected to the internet and interested in testing their English; they will also be younger and more urban than the population at large. But Philip Hult, the boss of EF, says that his sample shows results similar to a more scientifically controlled but smaller study by the British Council."
The test will obviously not reach poor and rural folk who lack internet access. So if a country has a urban elites who are good with English, and lot of rural poor people who cannot take the test, its score might be relatively inflated. In another country where nearly everyone is online but English skills are mediocre, the scores might be relatively depressed.
Despite that, the index has value. It is based on the test results of a huge sample: 1.7m people over three years in more than 50 countries. For the first time this year, gender, age, industry and job-level are broken out for those who want to get further into the data. (Spoiler alert: women do better than men, and the 30-35 set does best in terms of age. Those working in tourism do better than those working in mining and energy.) Fascinating individual country reports, including regional maps, are here. We learn, for example, that Moscow compares with Austria in its English skill, while Russia's Urals region compares with Qatar or Mexico.
Last year, the biggest surprise to me was that China and India were ranked alongside each other, despite India's much better reputation for English skill. That has changed this year, owing to a methodological tweak. India is now well ahead. Michael Lu of EF explains in an e-mail:
The 1st EF EPI report was based on four tests and in the 2nd report, we removed one of the four tests that didn’t fully test listening skills (it was optional). This change was made to ensure the EPI gets more accurate over time. Countries that are better at vocabulary/grammar/reading and weaker at listening had their rankings slightly inflated in the 1st report, as was the case with China. Conversely, countries such as India which are better at listening moved up in rankings.
In addition, he notes that 12 countries were added to the survey, and that the scores are very close to each other. For that reason, countries may seem to have moved quickly up or down the rankings despite no great real-world change in their English skills in the year. Many of the newly added countries ranked ahead of Brazil, for example, and that plus a slight score change caused Brazil to fall 15 places in the rankings. The index, Mr Lu says, should get more accurate over time as data continue to come in and the methodology is refined.
各国英语熟练指数排名
去年大家都看到了英孚(一家教育公司)发布的首份全球英语技能调查报告。今年它发布了同一主题的第二份报告，值得去看看一年来都有哪些变化。
首先，得重提一下这项调查的局限性。去年说过：“这不是一项统计学规范意义上的调查。它采取的是让参与者在方便时参与免费在线测试的形式。参与者是可以访问互联网且有兴趣测试其英语水平的人，通常属于比较年轻的、以城市居民为主的人群。但是根据英孚教育的菲利普?赫尔特经理的介绍，其结果与另一项由英国文化协会(British Council)完成的较小规模的统计学规范研究结果相近”。
这项测试明显不可能覆盖贫困的、农村地区的人群，因为他们访问互联网不方便。因此如果某个国家的城市精英群体擅长英语，而大量的农村贫困人群无法参与该测试，那么该国的得分可能会有点虚高。有的国家如果几乎所有人都能上网但是英语水平一般，则该国的得分可能会相对偏低。
尽管如此，这一指数仍是有价值的。它是基于数量巨大的测试结果样本得出的，有来自50多个国家的1700多万人参与了测试。今年的调查报告中首次按照性别、年龄、行业和职级划分进行了数据分析，使我们的认识更为深入。(插报：女性得分比男性高、在所有的年龄组中30-35岁这一组表现得最好、旅游行业人员比矿业和能源业得分高）这份引人注目的国别英语熟练指数报告（包括地区地图）可在以下网址中找到：链接。有了它我们可以回答许多问题，比如：莫斯科和奥地利比哪个英语水平高？俄罗斯乌拉尔地区和卡塔尔或墨西哥比又如何？
去年的报告最让我觉得惊奇的是中国和印度竟然排名相近，因为传统上大家都认为印度的英语水平比较高。今年经过调查方法上的调整，结果变了样：现在印度大幅领先于中国。英孚教育的卢迈克（Michael Lu）在一封邮件中解释道：“第一份英孚英语熟练指数报告基于四项测试，而在第二份报告中我们剔除了其中一项测试。这项测试原来是可选的、没有对听力进行全面的测试。”这一调整使得英语熟练指数变得更加准确了。原来在词汇、语法、阅读等方面比较强，而在听力方面比较弱的国家如中国在第一份报告中排名稍为有点偏高。调整之后，象印度这样听力比较好的国家排名上升了。
另外，卢先生提醒说今年的报告中新增加了12个国家，他们的得分非常接近。由于这个原因，某些国家的排名可能在榜上急剧提升或降低了好多位，但是实际上他们的英语水平并没有大的变化。比如说好几个新增的国家排名在巴西之前，加上巴西本身得分有小小的变化，因此导致它的排名下降了15个位置。卢先生说，随着时间的推移，越来越多的数据将会加入、方法会更加完善，排名也将会变得更加准确。
原文链接 精益求精
2012.10.24 各国英语熟练指数排名 http://www.ecocn.org/thread-163083-1-1.html
2012.10.20 真实的交易 http://www.ecocn.org/thread-159341-1-1.html
