✪ Casey O'Neal | Harvard University. [Introduction] Recently, law schools in the three prestigious universities in Harvard, Yale and Berkeley, California have successively announced their withdrawal from the US NEWS university rankings, which has attracted widespread attention. P

✪ [US] Kathy O'Neal | Harvard

Mac (translation)

[ Introduction ] Recently, the law schools of the three prestigious universities in the United States, Harvard, Yile , and Berkeley, California, successively announced their withdrawal from the US NEWS university rankings, attracting widespread attention. Previously, Chinese universities such as Renmin University, Nanyang University , Lanzhou University , etc. withdrew from the world university rankings, which caused considerable controversy. Some comments pointed out that in recent years, various university rankings have emerged one after another. The construction of universities and the choice of students seem to be bound by university rankings and their standards, and the essence and real standards of universities have been partially obscured. This may be an important reason why some universities choose to withdraw from the world rankings.

The author of this article, Cathy O Neil, dissected the history of the generation of US NEWS rankings. In 1983, the magazine "US NEWS" which was on the verge of bankruptcy created the world-renowned US NEWS ranking system in order to reverse the decline. However, in the early stages of its inception, the magazine lacked the necessary reference standards, so the magazine editors established university evaluation standards in accordance with the well-known and well-known universities recognized in the United States. For ease of evaluation and statistics, standards prefer certain specific and measurable data. The author of believes that from the beginning of this ranking, it cannot well measure the entire achievements of university construction. However, this ranking system has been enthusiastically sought after by American society. In order to obtain a higher ranking, some universities do not hesitate to compare the ranking standards to build universities and even provide false data. False data has led to false prosperity, and false prosperity has further intensified the vicious cycle of ranking industries. is under the US NEWS ranking system, many universities and students are unable to bear the burden. To make up for the necessary costs required by ranking standards, many schools have high tuition fees, further exacerbating the issue of educational equity. A large number of intermediaries and commercial companies make profits in it, but pass on the costs to the middle class and the poor. , this standard has long been out of the United States. In China, the trend of university rankings and the general utilitarianization in higher education hit it off, and there have been dangerous clues between rankings and university construction. What is the future development trend of ? It remains to be observed continuously.

This article was originally published in "Algorithm Hegemony" by Casey O'Neill, reproduced from "Wenhui Scholar" , which only represents the author's views and is for your reference.

How was the university ranking born?

▍html tml7University ranking: Self-rescue moves of magazines on the verge of bankruptcy

This story began in 1983, a magazine on the verge of cessation of publication in the United States, "U.S. News (US NEWS) decided to launch a large-scale project: evaluates 1,800 colleges and universities across the United States and ranks these schools by excellence. If progress goes well, the results of this project will become a useful tool to guide millions of young people to make the first major choice in their lives. For many, this choice will determine their career path, a lifelong circle of friends, and usually includes a lifelong partner. In addition, magazine editors also hope that this university ranking project can drive magazine sales - maybe in the week when the university rankings were launched, the sales of "U.S. News (US NEWS) " will catch up with "Time" and " News Weekly ".

But what data should be used for university rankings? At first, the staff of "US NEWS" relied entirely on the feedback they received from the questionnaire they sent to the presidents of various universities for rating. results, Stanford University ranked first among comprehensive universities in the United States, and Amherst College is the first-ranked college of liberal arts. Although the ranking results are very popular among readers, they also make many university leaders very angry. magazine received a scattered complaint, all of which accused the rankings of unfair results. Many university presidents, students and graduates insist that their schools should be ranked higher, and magazines should take a closer look at the relevant data.

( Law schools of three universities, Harvard , Yale, and Berkeley, California, announced their withdrawal from the US NEWS University Ranking)

▍ ranking rules: How is the evaluation model established?

In the next few years, the editors of "US NEWS" have been thinking about what specific data they can measure. Many models were born, but a large number of the evaluation factors came from intuition only. The process of establishing the model is not rigorous, and statistical analysis lacks basis. The basis for establishing the model is only the most important factor for education that people imagine out of thin air. Then, look for the relevant variable that can be measured, and finally give each variable a certain weight in the formula at will, and the model is completed.

In most fields, the process of model establishment is usually relatively rigorous. For example, researchers in agricultural disciplines compare inputs (soil, sunlight, and fertilizer) and outputs (the yield of crops with specific characteristics after harvest). Then, they can conduct next experiments and optimizations according to goals, such as certain costs, taste or nutritional value.

However, what the editor of "US NEWS" did is the " Education Excellence" ranking, which is a more abstract and vague value than the cost of grain or the protein content of each wheat grain. These edits do not have a direct way to quantify how the 4-year college learning process affects a certain student, let alone tens of millions of students. They cannot measure all aspects of a student's study, happiness, confidence, friendship and other aspects in four years of college life. Former US President Lyndon Johnson positioning higher education is: "Higher education is a way to deepen self-realization, expand personal productivity and increase personal returns", but no one of them is suitable for being placed in the university ranking model. The editor of

"US NEWS" just selected some variables that seem to be relevant to the evaluation target. They studied SAT ( Academic Ability Proficiency Test ) grades, school student-teacher ratio, and admission rates. They counted the percentage of freshmen who successfully entered sophomore year and the percentage of students who successfully graduated. They count the percentage of the total number of people who are still alive graduates donating to their alma mater, based on their donations to their alma mater likely suggesting that they like their alma mater’s education. The scores that account for 3/4 of the weight in the ranking result are all from an algorithm that contains the above variables; the scores that account for 1/4 of the weight are from the subjective evaluations of university presidents across the United States.

▍ ranking game: How does the evaluation model become a vicious circle?

"US NEWS" was released in 1988. When this ranking developed into the national standard, a vicious cycle occurred. Ranking will consolidate itself. If a university ranks lower in the rankings released by the U.S. News, its reputation will decline and its student source will worsen. Excellent students will avoid this university, and so will excellent professors. Graduated alumni will reduce their contributions. As a result, the ranking of this school will continue to decline. Simply put, ranking determines the fate of a university.

Before, universities had various ways to promote the success of school education, many of which were based on rumors and anecdotes. For example, some professors have received unanimous praise from many students; some students embarked on an outstanding career path after graduation and became diplomats or successful entrepreneurs; and some students published first-class novels. These positive deeds are widely known through word of mouth, and the school's reputation has also been improved. But is McAllister College better than Reed College ? Or is Iowa better than the University of Illinois? This is hard to say. Different universities are like different types of music or different eating habits. There are many different opinions on a certain university, and there are sufficient reasons for both good and bad. And now, the overall reputation of the university is clouded by a set of numbers.

If you think about this from the perspective of a university president, you will find that the university rankings are actually very bad. There is no doubt that most presidents cherish their university experience because to some extent it is the university experience that motivates them to climb the academic ladder and become presidents of a university. But now, these principals who are at the peak of their careers need to invest huge energy to improve the scores of the 15 assessment items related to the excellence of school education, and these 15 assessment items are defined by a group of editors from a second-rate magazine. They seem to be back in their student days, begging the teacher to give them high scores every day. If the university rankings published by "US NEWS" are only popular on a small scale, it would not cause any trouble. However, the influence of this ranking has developed rapidly and quickly became a national standard. The education industry suddenly became nervous and quickly set a strict task list for both university presidents and students. The university ranking model of "US NEWS" is huge in scale, causing a large-scale damage and leading to an almost endless vicious cycle.

Some university leaders try their best to improve their rankings. Baylor University set up a bonus to encourage freshmen to take the SAT exam again, hoping that taking another exam will improve their grades and Baylor University rankings. Some prestigious schools, including Bucknell University in Pennsylvania and McKenna College in in California, gave fake data to "US NEWS", exaggerating the admission scores of freshmen in their schools. In 2011, New York-based Aina College admitted that its school teachers fabricated almost all the data: test scores, acceptance and graduation rates, freshman retention, teacher-student ratios and alumni donations. But the lies took effect, at least for a while. According to the estimates of "US NEWS", the fake data has promoted Aina College from 50th to 30th in Northeastern universities.

More school leaders are trying to find a more conventional way to improve their school rankings—working to improve the school’s performance on every variable that affects the final score.They might think this is the most efficient way to use resources. After all, as long as they work hard to cater to the model of "US NEWS" and get a higher ranking, they can raise more funds, attract better students and professors, and then further improve their rankings. Are there any other options besides this?

▍html l7 ranking "Arms Race": How can we emerge in the ranking?

Robert Morse has been working in the magazine "US NEWS" since 1976. He is the organizer of this university ranking project. In an interview, he said that ranking colleges will help promote universities to set more meaningful goals. If they can therefore work to improve graduation rates or divide students into smaller classes to improve teaching effectiveness, then ranking is a good thing. He admitted that the magazine could not obtain the most relevant data for the excellence of university education, that is, the learning content of students in each school. However, the "US NEWS" university ranking model based on alternative variables is enough to reflect the problem.

However, when you build a model based on alternative variables, it becomes much easier to drill the model's vulnerability. This is because alternative variables are easier to manipulate than the complex facts it represents. For example, suppose a website that wants to hire a social media expert, and the human resources manager decides to choose a variable—focusing on candidates with the highest number of Twitter followers. Twitter followers are one of the hallmarks of social media engagement, right? But imagine that candidates will soon be able to use everything they can to increase Twitter followers. Someone will spend $19.95 directly “buy” a large number of fans who are controlled by the machine. Because people exploit loopholes in the recruitment system, alternative variables lose their effectiveness.

In the "US NEWS" university ranking event, people quickly accepted that the ranking was a reflection of the quality of university education. Therefore, each university can only choose to cooperate, and they have to work hard to improve the scores of each assessment item involved in the ranking. In fact, what many schools are most anxious about is the uncontrollable factor that accounts for 1/4 of the ranking results, namely reputation scores, and the feedback from the questionnaire leaders from various universities and colleges.

In 2008, the ranking of Texas Christian University (TCU) in Fort Worth City dropped sharply. Three years ago, the school's ranking was 97, and the rankings were reduced to 105, 108 and 113 in the following three years. The school’s alumni and supporters were angry about this, and President Victor Bosini was also in an embarrassing situation. He insists that Texas Christian University is improving on every metric, “our freshman retention is improving, our fundraising and everything else is improving.”

Bosini's defense has two problems. First of all, the ranking model of "US NEWS" is not an isolated judgment of each university. Even schools that improve scores in each indicator will lag behind other schools whose scores improve faster in rankings. In academic terms, the evaluation model of "US NEWS" is a distribution model. This led to an "arms race" between schools. Another problem is that Texas Christian University cannot control the reputation score that accounts for 1/4 of the weight. Admissions Director Raymond Brown pointed out that reputation is the most weighted variable in the model, "it's ridiculous because it's totally subjective." Freshman Admissions Director Wies Vargona pointed out that in order to improve reputation scores, universities are advertising themselves.

Nevertheless, Texas Christian University decided to move forward with a controllable score of 3/4 weights. The university launched a $250 million fundraising campaign.By 2009, the school had raised US$434 million, far exceeding the target quota. Since fundraising is one of the ranking indicators, this alone improves the ranking. Texas Christian University spent much of that money on campus facilities improvements, $100 million spent on building a central mall and student activity center in an effort to make the Texas Christian University campus look more attractive. There is nothing wrong with 's practices, but its original intention is to cater to the ranking model of "US NEWS".

Perhaps more importantly, Texas Christian University built a sports training venue at the time, and devoted a lot of resources to football projects. In the next few years, Texas Christian University’s Horned Frog football team became a national strong team. In 2010, they defeated the veteran team Wisconsin in the Rose Cup soccer game and won the All-American Championship.

This victory brought the so-called " Flotie effect" (the Flutie effect) to Texas Christian University. In 1984, in an extremely exciting college football game, Boston University quarterback Doug Flotel defeated the University of Miami in a last-second after a very long pass that turned the game.

Flotel is made from and this becomes as a legend. Within two years after the end of this game, the college application rate of Boston University rose by 30%. Georgetown University has also had legends that brought the same publicity effect. The school's basketball team led by Patrick Ewing has reached the National Championship three times. , winning sports competitions is a key factor in attracting students to apply to a certain university. When a large number of senior high school students from sports schools watch university sports games on TV, the strong school of the team has formed a great attraction to them. These students will be proud that they are students of the school and wear team uniforms with the school name written on it. The admission applications received by these universities soared. As more students apply for admission, the admissions office can raise the admission threshold to increase the average score of college freshmen in SAT, which helps improve the university rankings. In addition, the more students the school rejects, the lower the admission rate and the more beneficial it is for ranking.

Texas Christian University strategy worked. By 2013, the university had become the second-ranked university in Texas student selection, with the first being the famous Houston Rice University. This year, the average scores of freshmen in the college entrance examination and entrance examinations at Texas Christian University both reached the highest level in history, and their ranking in the United States also rose sharply. In 2015, the school ranked 76th in the United States, which means that in just 7 years, the school rose by 37 places.

▍ ranking leads to a significant increase in university tuition fees?

Now, we finally discovered what is the biggest flaw in the university ranking model of "US NEWS". We cannot say that the alternative variables selected by the editor of US NEWS to judge the "educational excellence" are invalid, but the greater mistakes they made come from variables that they did not include: tuition and fees, student bursaries. These variables are missed by this ranking model.

This brings up a key question that we will discuss frequently: What is the goal of the modeler? In the case of university rankings, you need to consider it from the perspective of the editors of 1988 "US NEWS".

✪ [US] Kathy O'Neal | Harvard

Mac (translation)

[ Introduction ] Recently, the law schools of the three prestigious universities in the United States, Harvard, Yile , and Berkeley, California, successively announced their withdrawal from the US NEWS university rankings, attracting widespread attention. Previously, Chinese universities such as Renmin University, Nanyang University , Lanzhou University , etc. withdrew from the world university rankings, which caused considerable controversy. Some comments pointed out that in recent years, various university rankings have emerged one after another. The construction of universities and the choice of students seem to be bound by university rankings and their standards, and the essence and real standards of universities have been partially obscured. This may be an important reason why some universities choose to withdraw from the world rankings.

The author of this article, Cathy O Neil, dissected the history of the generation of US NEWS rankings. In 1983, the magazine "US NEWS" which was on the verge of bankruptcy created the world-renowned US NEWS ranking system in order to reverse the decline. However, in the early stages of its inception, the magazine lacked the necessary reference standards, so the magazine editors established university evaluation standards in accordance with the well-known and well-known universities recognized in the United States. For ease of evaluation and statistics, standards prefer certain specific and measurable data. The author of believes that from the beginning of this ranking, it cannot well measure the entire achievements of university construction. However, this ranking system has been enthusiastically sought after by American society. In order to obtain a higher ranking, some universities do not hesitate to compare the ranking standards to build universities and even provide false data. False data has led to false prosperity, and false prosperity has further intensified the vicious cycle of ranking industries. is under the US NEWS ranking system, many universities and students are unable to bear the burden. To make up for the necessary costs required by ranking standards, many schools have high tuition fees, further exacerbating the issue of educational equity. A large number of intermediaries and commercial companies make profits in it, but pass on the costs to the middle class and the poor. , this standard has long been out of the United States. In China, the trend of university rankings and the general utilitarianization in higher education hit it off, and there have been dangerous clues between rankings and university construction. What is the future development trend of ? It remains to be observed continuously.

This article was originally published in "Algorithm Hegemony" by Casey O'Neill, reproduced from "Wenhui Scholar" , which only represents the author's views and is for your reference.

How was the university ranking born?

▍html tml7University ranking: Self-rescue moves of magazines on the verge of bankruptcy

This story began in 1983, a magazine on the verge of cessation of publication in the United States, "U.S. News (US NEWS) decided to launch a large-scale project: evaluates 1,800 colleges and universities across the United States and ranks these schools by excellence. If progress goes well, the results of this project will become a useful tool to guide millions of young people to make the first major choice in their lives. For many, this choice will determine their career path, a lifelong circle of friends, and usually includes a lifelong partner. In addition, magazine editors also hope that this university ranking project can drive magazine sales - maybe in the week when the university rankings were launched, the sales of "U.S. News (US NEWS) " will catch up with "Time" and " News Weekly ".

But what data should be used for university rankings? At first, the staff of "US NEWS" relied entirely on the feedback they received from the questionnaire they sent to the presidents of various universities for rating. results, Stanford University ranked first among comprehensive universities in the United States, and Amherst College is the first-ranked college of liberal arts. Although the ranking results are very popular among readers, they also make many university leaders very angry. magazine received a scattered complaint, all of which accused the rankings of unfair results. Many university presidents, students and graduates insist that their schools should be ranked higher, and magazines should take a closer look at the relevant data.

( Law schools of three universities, Harvard , Yale, and Berkeley, California, announced their withdrawal from the US NEWS University Ranking)

▍ ranking rules: How is the evaluation model established?

In the next few years, the editors of "US NEWS" have been thinking about what specific data they can measure. Many models were born, but a large number of the evaluation factors came from intuition only. The process of establishing the model is not rigorous, and statistical analysis lacks basis. The basis for establishing the model is only the most important factor for education that people imagine out of thin air. Then, look for the relevant variable that can be measured, and finally give each variable a certain weight in the formula at will, and the model is completed.

In most fields, the process of model establishment is usually relatively rigorous. For example, researchers in agricultural disciplines compare inputs (soil, sunlight, and fertilizer) and outputs (the yield of crops with specific characteristics after harvest). Then, they can conduct next experiments and optimizations according to goals, such as certain costs, taste or nutritional value.

However, what the editor of "US NEWS" did is the " Education Excellence" ranking, which is a more abstract and vague value than the cost of grain or the protein content of each wheat grain. These edits do not have a direct way to quantify how the 4-year college learning process affects a certain student, let alone tens of millions of students. They cannot measure all aspects of a student's study, happiness, confidence, friendship and other aspects in four years of college life. Former US President Lyndon Johnson positioning higher education is: "Higher education is a way to deepen self-realization, expand personal productivity and increase personal returns", but no one of them is suitable for being placed in the university ranking model. The editor of

"US NEWS" just selected some variables that seem to be relevant to the evaluation target. They studied SAT ( Academic Ability Proficiency Test ) grades, school student-teacher ratio, and admission rates. They counted the percentage of freshmen who successfully entered sophomore year and the percentage of students who successfully graduated. They count the percentage of the total number of people who are still alive graduates donating to their alma mater, based on their donations to their alma mater likely suggesting that they like their alma mater’s education. The scores that account for 3/4 of the weight in the ranking result are all from an algorithm that contains the above variables; the scores that account for 1/4 of the weight are from the subjective evaluations of university presidents across the United States.

▍ ranking game: How does the evaluation model become a vicious circle?

"US NEWS" was released in 1988. When this ranking developed into the national standard, a vicious cycle occurred. Ranking will consolidate itself. If a university ranks lower in the rankings released by the U.S. News, its reputation will decline and its student source will worsen. Excellent students will avoid this university, and so will excellent professors. Graduated alumni will reduce their contributions. As a result, the ranking of this school will continue to decline. Simply put, ranking determines the fate of a university.

Before, universities had various ways to promote the success of school education, many of which were based on rumors and anecdotes. For example, some professors have received unanimous praise from many students; some students embarked on an outstanding career path after graduation and became diplomats or successful entrepreneurs; and some students published first-class novels. These positive deeds are widely known through word of mouth, and the school's reputation has also been improved. But is McAllister College better than Reed College ? Or is Iowa better than the University of Illinois? This is hard to say. Different universities are like different types of music or different eating habits. There are many different opinions on a certain university, and there are sufficient reasons for both good and bad. And now, the overall reputation of the university is clouded by a set of numbers.

If you think about this from the perspective of a university president, you will find that the university rankings are actually very bad. There is no doubt that most presidents cherish their university experience because to some extent it is the university experience that motivates them to climb the academic ladder and become presidents of a university. But now, these principals who are at the peak of their careers need to invest huge energy to improve the scores of the 15 assessment items related to the excellence of school education, and these 15 assessment items are defined by a group of editors from a second-rate magazine. They seem to be back in their student days, begging the teacher to give them high scores every day. If the university rankings published by "US NEWS" are only popular on a small scale, it would not cause any trouble. However, the influence of this ranking has developed rapidly and quickly became a national standard. The education industry suddenly became nervous and quickly set a strict task list for both university presidents and students. The university ranking model of "US NEWS" is huge in scale, causing a large-scale damage and leading to an almost endless vicious cycle.

Some university leaders try their best to improve their rankings. Baylor University set up a bonus to encourage freshmen to take the SAT exam again, hoping that taking another exam will improve their grades and Baylor University rankings. Some prestigious schools, including Bucknell University in Pennsylvania and McKenna College in in California, gave fake data to "US NEWS", exaggerating the admission scores of freshmen in their schools. In 2011, New York-based Aina College admitted that its school teachers fabricated almost all the data: test scores, acceptance and graduation rates, freshman retention, teacher-student ratios and alumni donations. But the lies took effect, at least for a while. According to the estimates of "US NEWS", the fake data has promoted Aina College from 50th to 30th in Northeastern universities.

More school leaders are trying to find a more conventional way to improve their school rankings—working to improve the school’s performance on every variable that affects the final score.They might think this is the most efficient way to use resources. After all, as long as they work hard to cater to the model of "US NEWS" and get a higher ranking, they can raise more funds, attract better students and professors, and then further improve their rankings. Are there any other options besides this?

▍html l7 ranking "Arms Race": How can we emerge in the ranking?

Robert Morse has been working in the magazine "US NEWS" since 1976. He is the organizer of this university ranking project. In an interview, he said that ranking colleges will help promote universities to set more meaningful goals. If they can therefore work to improve graduation rates or divide students into smaller classes to improve teaching effectiveness, then ranking is a good thing. He admitted that the magazine could not obtain the most relevant data for the excellence of university education, that is, the learning content of students in each school. However, the "US NEWS" university ranking model based on alternative variables is enough to reflect the problem.

However, when you build a model based on alternative variables, it becomes much easier to drill the model's vulnerability. This is because alternative variables are easier to manipulate than the complex facts it represents. For example, suppose a website that wants to hire a social media expert, and the human resources manager decides to choose a variable—focusing on candidates with the highest number of Twitter followers. Twitter followers are one of the hallmarks of social media engagement, right? But imagine that candidates will soon be able to use everything they can to increase Twitter followers. Someone will spend $19.95 directly “buy” a large number of fans who are controlled by the machine. Because people exploit loopholes in the recruitment system, alternative variables lose their effectiveness.

In the "US NEWS" university ranking event, people quickly accepted that the ranking was a reflection of the quality of university education. Therefore, each university can only choose to cooperate, and they have to work hard to improve the scores of each assessment item involved in the ranking. In fact, what many schools are most anxious about is the uncontrollable factor that accounts for 1/4 of the ranking results, namely reputation scores, and the feedback from the questionnaire leaders from various universities and colleges.

In 2008, the ranking of Texas Christian University (TCU) in Fort Worth City dropped sharply. Three years ago, the school's ranking was 97, and the rankings were reduced to 105, 108 and 113 in the following three years. The school’s alumni and supporters were angry about this, and President Victor Bosini was also in an embarrassing situation. He insists that Texas Christian University is improving on every metric, “our freshman retention is improving, our fundraising and everything else is improving.”

Bosini's defense has two problems. First of all, the ranking model of "US NEWS" is not an isolated judgment of each university. Even schools that improve scores in each indicator will lag behind other schools whose scores improve faster in rankings. In academic terms, the evaluation model of "US NEWS" is a distribution model. This led to an "arms race" between schools. Another problem is that Texas Christian University cannot control the reputation score that accounts for 1/4 of the weight. Admissions Director Raymond Brown pointed out that reputation is the most weighted variable in the model, "it's ridiculous because it's totally subjective." Freshman Admissions Director Wies Vargona pointed out that in order to improve reputation scores, universities are advertising themselves.

Nevertheless, Texas Christian University decided to move forward with a controllable score of 3/4 weights. The university launched a $250 million fundraising campaign.By 2009, the school had raised US$434 million, far exceeding the target quota. Since fundraising is one of the ranking indicators, this alone improves the ranking. Texas Christian University spent much of that money on campus facilities improvements, $100 million spent on building a central mall and student activity center in an effort to make the Texas Christian University campus look more attractive. There is nothing wrong with 's practices, but its original intention is to cater to the ranking model of "US NEWS".

Perhaps more importantly, Texas Christian University built a sports training venue at the time, and devoted a lot of resources to football projects. In the next few years, Texas Christian University’s Horned Frog football team became a national strong team. In 2010, they defeated the veteran team Wisconsin in the Rose Cup soccer game and won the All-American Championship.

This victory brought the so-called " Flotie effect" (the Flutie effect) to Texas Christian University. In 1984, in an extremely exciting college football game, Boston University quarterback Doug Flotel defeated the University of Miami in a last-second after a very long pass that turned the game.

Flotel is made from and this becomes as a legend. Within two years after the end of this game, the college application rate of Boston University rose by 30%. Georgetown University has also had legends that brought the same publicity effect. The school's basketball team led by Patrick Ewing has reached the National Championship three times. , winning sports competitions is a key factor in attracting students to apply to a certain university. When a large number of senior high school students from sports schools watch university sports games on TV, the strong school of the team has formed a great attraction to them. These students will be proud that they are students of the school and wear team uniforms with the school name written on it. The admission applications received by these universities soared. As more students apply for admission, the admissions office can raise the admission threshold to increase the average score of college freshmen in SAT, which helps improve the university rankings. In addition, the more students the school rejects, the lower the admission rate and the more beneficial it is for ranking.

Texas Christian University strategy worked. By 2013, the university had become the second-ranked university in Texas student selection, with the first being the famous Houston Rice University. This year, the average scores of freshmen in the college entrance examination and entrance examinations at Texas Christian University both reached the highest level in history, and their ranking in the United States also rose sharply. In 2015, the school ranked 76th in the United States, which means that in just 7 years, the school rose by 37 places.

▍ ranking leads to a significant increase in university tuition fees?

Now, we finally discovered what is the biggest flaw in the university ranking model of "US NEWS". We cannot say that the alternative variables selected by the editor of US NEWS to judge the "educational excellence" are invalid, but the greater mistakes they made come from variables that they did not include: tuition and fees, student bursaries. These variables are missed by this ranking model.

This brings up a key question that we will discuss frequently: What is the goal of the modeler? In the case of university rankings, you need to consider it from the perspective of the editors of 1988 "US NEWS".When they build the first statistical model, how do they know if the model is valid? First, if the model can reflect some established university rankings, it shows that it has a certain degree of credibility. For example, if Harvard University, Stanford University, Princeton University, and Yale University rank at the forefront of the university ranking model, this confirms to a certain extent that the university ranking model designed by the editor of "US NEWS" is effective. To establish such a model, they only need to study those first-class universities and think about what is special about these universities. What are the common points of excellent universities? What is the difference between these schools and the guaranteed schools in the town next to them? They found that the SAT scores of excellent universities are very high, and most of them can graduate smoothly; the alumni who have graduated are very rich and will continue to donate to the school; etc. In this way, the university ranking project team of "US NEWS" has established an evaluation index system to measure the excellence of education by analyzing the advantages of famous universities.

Now, if the project team incorporates education costs into the algorithm, its model output may change strangely - universities with cheap tuition fees are likely to break into the ranks of excellent universities, and this result will be widely questioned. Since the public may regard the university rankings finally released by "US NEWS" as particularly important, it is safer to adopt conservative and conventional algorithms to ensure that first-class universities are at the forefront of ranking output results. Of course, high costs may be the price of excellence, and this is not unreasonable. The ranking model of

"US NEWS" excludes costs from the algorithm, which is like handing over a gilded check book to the university presidents. The only instruction that the latter needs to follow is to maximize the score of 15 assessment indicators, while reducing costs is not included. In fact, increasing tuition fees will give them more resources to improve their performance in assessment projects.

From then on, tuition fees have soared. From 1985 to 2013, the tuition fees for higher education increased by more than 5 times, almost 4 times the inflation rate. In order to attract top students, universities, like Texas Christian University, have begun to invest heavily in campus infrastructure, building student centers with glass walls, luxurious dormitories, and gyms with climbing walls and whirlpool bathtubs. On the surface, this is a good thing for students, and these facilities can enrich their college experience – provided they don’t have to bear these costs in the form of student loans, and the pressure to repay the student loan may follow the students for decades. But we can't blame everything on the university rankings of "US NEWS". Our entire society not only agrees with the idea that university education is essential, but also readily accepts the fact that diplomas from top-ranked schools can help students quickly enter the privileged class. The ranking model of "US NEWS" has grown into a behemoth based on the fear and anxiety that arise from this. The ranking model has strongly stimulated continuous investment in education among all parties, while soaring tuition fees have been ignored.

▍ ranking leads to the intensification of university management methods?

For the need to improve ranking, each university manages its students like a portfolio. This is very common in the big data field, from small advertising to large political fields. In the eyes of school leaders, every prospective college student represents a set of assets and one or two debts. For example, a high school student's excellent performance in sports events is regarded as an asset, but at the same time his or her performance may be at the mid-to-lower level, and the latter is debt. He may also need to apply for a bursary, which is another debt.To balance the portfolio, they should find other candidates who can go to school at their own expense and have excellent grades. But those ideal candidates may choose to go to other better schools even if they are admitted. This is also a risk that must be quantified. Given that the entire evaluation system is very complex, the education consulting industry has emerged in order to "optimize enrollment".

Education consulting firm Noel-Levitz has developed a predictive analysis software package called Forecast Plus. The package allows admissions teachers to evaluate the situation of college students based on geographical location, gender, race, research field, academic status, and "any other characteristics." Another consulting agency called RightS students is dedicated to collecting, buying and selling relevant data to help university clients find the best candidates for admission, including students who can pay full tuition and those who may be eligible for off-campus scholarships. In this sense, learning disabilities may be an advantage for university admission.

All of this happens in this huge ecosystem centered on the university rankings of "US NEWS", and the ranking model actually acts as the Supreme Court within the system. If the editor of "US NEWS" re-arranges the weight of some alternative variables in the model, such as reducing the weight of test scores, or increasing the weight of graduation rates, the entire education ecosystem must re-adapt to the new rules. This change will affect consulting companies, high school entrance guidance departments, and all students.

▍ rankings: Who is the real winner?

In 2010, the magazine "US NEWS" was discontinued. However, the entire ranking industry was not affected at all and continued to grow and grow, and the ranking extended to medical schools, dental schools, graduate schools of literature and engineering, and even high schools. With the development of the ranking industry, the means of exploiting model loopholes are becoming more and more abundant. In the global university rankings of "US NEWS" in 2014, Saudi Arabia King Abdulaziz University (KAU) ranks seventh in mathematics department, second only to Harvard. However, the school's mathematics department has only been established for two years, and no one knows how it suddenly jumped to the top 10 in the world, even surpassing the mathematics department of the University of Cambridge and the University of MIT. Leo Pacht, a computer biology expert at

, University of Berkeley, studied this problem. He found that KAU had contacted many mathematicians with extremely high citations and hired them as visiting professors of at the school for an annual salary of $72,000. According to a recruitment letter found by Pacht, the cooperation agreement stipulates that the mathematicians must work in Saudi Arabia for three weeks a year. The university will undertake their business class tickets and arrange for them to stay in five-star hotels. It is conceivable that their work in Saudi Arabia adds value to the school. But more importantly, the university also asked the mathematicians to change the mailing address they recorded on the Thomson Reuters academic citation website to KAU, which is a key reference factor in the US NEWS ranking model. This means that KAU can declare that the academic papers and monographs of their many new visiting professors are their results. Since the number of paper citations is an important reference data in the ranking model algorithm, KAU's ranking has soared.The vicious cycle of

ranking model production and its widespread anxiety have also led to the booming development of the entire college entrance tutoring education industry. An education company called Top Tier Admissions has launched a four-day "university application bootcamp" training program, which costs up to $16,000 (excluding accommodation and diet). During this period, these sophomore students will learn how to write applications, learn how to "win" interviews, and create an "activity list" to summarize all the awards they have won and all sports, club activities and community volunteer services they have participated in, because these are the elements that the university admissions office is concerned about.

Educational companies professionals understand each school’s enrollment model, so they know how to get a prospective college student to be included in the “portfolio” of their target school. Ma Zhenyi, founder of Star Tengke International Education Group in California, USA, has brought market analysis to the extreme in the education industry. He used the model he developed to evaluate prospective college students and calculated the likelihood of their admission to the target college. He told a reporter from Bloomberg Business Weekly that assuming that an American high school student has an average credit score of (GPA) of (GPA) of 3.8, SAT score is 2,000 points, and extracurricular activities time is 800 hours, then the probability of he being admitted to New York University is 20.4%, and the probability of being admitted to USC is 28.1%. Then, StarTenko will provide a secured portfolio of advice. If the student receives counseling from a consulting firm and is eventually successfully admitted to New York University, the student will be required to pay $25,931 to the consulting firm, and if he is finally successfully admitted to the University of Southern California, he will be required to pay $18,826. If his application is rejected by both schools, the consulting firm will not charge any fees.

All or at least part of the enrollment model of each university comes from the university ranking model of "US NEWS". This creates huge business opportunities for professional consultants like Ma Zhenyi. By cultivating their connections in various universities to obtain first-hand information or reversely deducing the enrollment model algorithms of various schools, they have cracked the enrollment models of most schools.

Of course, the main victims are still the majority of the United States, i.e. the poor and middle class, who don’t have that much money to spend on courses and consulting firms. They missed precious internal information. The result is that the education system deviates from the poor and the middle class, pushing them to the path of poverty, further aggravating the solidification of social classes.

However, even those students who try their best to enter prestigious universities are not winners. The university admissions system is profitable only to a few people and has no educational value at all. It just reorders and categorizes a group of 18-year-olds in some novel way. During the exam preparation stage, mastering more basketball skills or writing applications that meet the target university standards with the help of professional tutors does not allow them to master truly meaningful skills. Not to mention that many people pass the test by trash. All these students, whether from the rich or the working class, were trained to adapt to a huge machine.

So, is there any solution? During his second term as President Obama, he proposed a new university ranking model that is more in line with the interests of the majority of people in the United States than the university ranking model of "US NEWS". His secondary goal is to undermine the influence of for-profit colleges (which is a money-sucking scourge). Obama’s idea is to link the university ranking system to a different set of metrics that include affordability, the proportion of poor and minority students, and the employment of students after graduation. Like the ranking model of "US NEWS", the model will also consider graduation rates. If a university performs below the minimum standard on these metrics, it will be kicked out of the $180 million annual federal student loan market (for-profit colleges have always been there).

Obama's modeling goal sounds valuable, but every ranking model has vulnerabilities to drill. Once loopholes are exploited, the model will have a new round of vicious cycles and a large number of unexpected harmful consequences. For example, it is easy to increase the graduation rate, just lower the graduation requirements. Many students cannot pass the exams in mathematics, science majors, and foreign languages. So relax the requirements in this regard so that more students can graduate. But wouldn’t this be ironic if the goal of our education system is to develop more scientists and technicians? It is also easy to improve the income level of graduates. All universities need to do is reduce liberal arts majors and remove education and social services departments because teachers and social workers earn less than engineers, chemists and computer scientists.

Perhaps, it is also a good thing that the Obama administration failed to come up with a re-adjusted ranking system. University presidents strongly resist the new ranking system. After all, they have been working hard for years in line with the ranking model of "US NEWS". The new ranking model proposed by Obama involves other variables such as graduation rate, class number, graduate employment placement and income level. If universities are evaluated strictly according to indicator standards, the scores they obtain will seriously damage the rankings and reputations of many universities.

So, the government finally made concessions. Instead of re-ranking universities, the Ministry of Education has published a large amount of survey data on its website. In this way, students can check the indicators they care about, including the number of class members, graduation rates, and the average debt of fresh graduates. they no longer need to understand the weight of any statistical method or variables. is like a travel website, and now everyone can develop their own personal models.

This article is excerpted from Kathy O'Neill, Ma Qingling Translated by "Algorithm Hegemony", reproduced from "Wenhui Scholars". pictures are from the Internet. If there is any infringement of , please contact us to delete it. Personal sharing is welcome. Please contact this official account for reprinting by the media.