THE公佈2023年世界大學排名,香港5校入前100名

ANChan59

annie40 發表於 22-11-17 20:55
放諸四海，工程學系乃國之根本，永遠都是需求中的頂流。香港受地域人口所限，產業只能歸邊求存，無奈之舉吧 ...

有時間另外開貼分享一下，或者在日誌分享！

ANChan59

https://www.reuters.com/legal/legalindustry/top-ranked-yale-law-school-shun-flawed-us-news-rankings-2022-11-16/

這個新聞報道耶魯大學及哈佛大學法學院退出US News Ranking 的法律學院排名。

耶魯大學法學院在這個排名連續排名第一32年，哈佛大學今年排第四。

問題在學生測考成績及畢業之後的人工，大學目標跟評比重點有衝突。

現在先提出來？幸好是前列大學，不然的話，冇人會報導，或者俾人酸。

beastiebistro

ANChan59 發表於 22-11-18 08:33
https://www.reuters.com/legal/legalindustry/top-ranked-yale-law-school-shun-flawed-us-news-rankings- ...

我認為最後一段最有意思："defaulting to third-party data would be a positive change because schools can't manipulate those numbers"。雖然係跌落地揦番拃沙... 但係都直接講咗呢啲榜嘅最大問題，就係數據係可以manipulate嘅。有得manipulate，自然就會有人manipulate。所以排行榜都係得啖笑。
如果做榜嘅公司都知啲數據係manipulate 過嘅，點解一路都用呢？啲做榜公司自己又點manipulate 啲數據呢？

ANChan59

beastiebistro 發表於 22-11-18 08:44
我認為最後一段最有意思："defaulting to third-party data would be a positive change because schools  ...

因為他們有附屬顧问公司教大學manipulate 數據，這是公開的秘密！

部分評分準則係主觀成份重嘅，他們的assessors 係可以做好多嘢。

shadeslayer

ANChan59 發表於 22-11-18 08:33
https://www.reuters.com/legal/legalindustry/top-ranked-yale-law-school-shun-flawed-us-news-rankings- ...

內容有講評分和大學的 diversity values 有分歧。哈佛比人告完仲咁高調要 diversity.

但 Columbia 一時話排第四，一時18，一時又話 Harvard 第四

beastiebistro

ANChan59 發表於 22-11-18 09:24
因為他們有附屬顧问公司教大學manipulate 數據，這是公開的秘密！

部分評分準則係主觀成份重嘅，他們的as ...

真係有少少we know they're lying, they know they're lying, they know we know they're lying, we know they know we know they're lying, and yet they continue to lie嘅感覺...
哥倫比亞唔會因為由第4變第18而唔再係名校，反而啲由百幾嘅打入百大以為自己升晒喱。除咗啲校友自我感覺良好，就喺呃吓啲唔識嘅家長學生咁...

都話得啖笑

shadeslayer

beastiebistro 發表於 22-11-18 08:44
我認為最後一段最有意思："defaulting to third-party data would be a positive change because schools  ...

Yes. That is funny.  If it is a positive change to use third party data and they know it, why didn’t they use third party data earlier for all universities in the ranking.  Why did they wait until two major law schools quitted the ranking?

shadeslayer

beastiebistro 發表於 22-11-18 11:34
真係有少少we know they're lying, they know they're lying, they know we know they're lying, we know t ...

Calling them liars is a bit unfair.  They fine tune data presentation to achieve the best possible scores in the ranking.  

Data is correct, it is the interpretation and data selection that give them the edge.

beastiebistro

shadeslayer 發表於 22-11-18 11:58
Calling them liars is a bit unfair.  They fine tune data presentation to achieve the best possible  ...

Semantics... Can't change a quote...

One may "interpret" the misrepresentation of data as lying. Doesn't really matter. What I'm trying to say is that these so called "authoritative" rankings are really a bit of a joke. And top universities are top in spite of the rankings.  

shadeslayer

我看排名，不會追著看，且只看看大約位置，每十個二十個，一個區間咁看可以了。

shadeslayer

beastiebistro 發表於 22-11-18 12:05
Semantics... Can't change a quote...

One may "interpret" the misrepresentation of data as lying. Do ...

Agree it is the universities that make the ranking, not ranking makes the universities.  People taking ranking too literally are doing it at their own detriment.

ANChan59

shadeslayer 發表於 22-11-18 10:42
本帖最後由 shadeslayer 於 22-11-18 10:48 編輯

內容有講評分和大學的 diversity values 有分歧。哈佛 ...

哥倫比亚大学、哈佛大学商学院係双第四。New York Times 在这个讲得清楚d。

ANChan59

beastiebistro 發表於 22-11-18 12:05
Semantics... Can't change a quote...

One may "interpret" the misrepresentation of data as lying. Do ...

Misrepresentation is cheating.

shadeslayer

ANChan59 發表於 22-11-18 14:53
Misrepresentation is cheating.

I am not saying these ranking is not cheating.  I am saying in general statistics are so open to adjustments, you can almost create any results you want without technically lying.  Through sampling size, sampling time, slicing and dicing data, inclusions and exclusions etc.

For example number of students, if the ranking organisation don’t define when to measure the student population the. The timing you so the sampling makes a huge difference.   Are they lying or cheating by choosing the best possible timing for sampling students on university records? No.  This is just one example.  They might have done some real cheating. But things like these are not one of them.

ANChan59

shadeslayer 發表於 22-11-19 06:43
I am not saying these ranking is not cheating.  I am saying in general statistics are so open to ad ...

Data is genuine, interpretations can be infinity. But what is common sense and integrity?

ANChan59

shadeslayer 發表於 22-11-19 06:43
I am not saying these ranking is not cheating.  I am saying in general statistics are so open to ad ...

我的经验在硕士及博士论文，用data mining tool 做分析，係会data massage。

硕士的課题有关基建，跨越十年八年，我用政府统计处二三十年数据，因为基数不同，我全部要用log ，run 完第一轮，得个桔。

及后发现我用的数据有leader，有lagger，我要重新考虑。我将所有数据分别做出time curve，发觉个别形态很接近，不過会推前或者移后1-3年。

我將有關聯性的數據按情況調整time period，再測試，開始見到有关连性，及后用不同SPSS的modules，有些冇乜发现，到后来终于揾到一些眉目。在模块内仍然要做不同reference frame rotations，最后找到一个我觉的statically significant又make sense嘅结果。

我当时扮演统计学家的角色，我massage data，我係因为academic curiosity 尋找一些知识，不是刻意去誤导教授及考试官，我每个对数据处理都在methodology部分詳细解释，可以在阳光下验证。

用7月份大学学生教师统计人数做师生比例，果位statistician 嘅common sense & integrity在那裏？

shadeslayer

ANChan59 發表於 22-11-19 09:30
我的经验在硕士及博士论文，用data mining tool 做分析，係会data massage。

硕士的課题有关基建，跨越十 ...

當然啦，陳sir 你從學術嘅角度尋找答案，目的就係要找出真相。有啲大學嘅目標好單一，就係要排名高。喺唔犯規底下有啲咩蠱惑招可以出，有啲大學做得出都唔出奇。

Common sense and integrity don’t help you climb the ranking ladder.  

poonseelai

shadeslayer 發表於 22-11-24 11:36
本帖最後由 shadeslayer 於 22-11-24 11:37 編輯

當然啦，陳sir 你從學術嘅角度尋找答案，目的就係要找 ...

"Common sense and integrity don’t help you climb the ranking ladder"

so true

ANChan59

shadeslayer 發表於 22-11-24 11:36
本帖最後由 shadeslayer 於 22-11-24 11:37 編輯

當然啦，陳sir 你從學術嘅角度尋找答案，目的就係要找 ...

If that’s the general issue of academia, so UGC is doing a great job to eliminate integrity and professionalism.

If chairmen and the committee not from academics and lack of integrity and professionalism. How to uphold or enhance the educational standards?

shadeslayer

Well, chairman and board of directors aren’t always subject matter experts.

The question is whoever be at the board they should respect the execution branch under the CEO for their subject matter expertise.  

If the board decides ranking is the most important thing for the university, then there is very little anyone can do.

But the board members and chairman also change once in a while.  So the strategic priorities also change over time.