Do they need to drill on them? Anyway, this is not a general pedagogy thread and it seems pretty clear to me that we are unlikely to see eye to eye.
1 Like
So you canāt explain it with your own words, but want it to be taught in high school?
I donāt know. How else you learn to recognize patterns? We do it in machine learning exactly because it works with humans.
I cite you to a comprehensive library of resources and articles, that includes digestible summaries from a multitude of experts, and your criticism is I should paraphrase!
Got it.
Since I have a PhD in Computer Science, I believe I have some vague knowledge what ādata scienceā is.
And since I see a lot of people that like talking about ādata scienceā donāt know much about it, I was trying to check if that is true in this case.
The core CS topics remain but at many colleges and for many students the core is just the beginning - ātable stakesā for where the true value of a college education starts.
Our S knew from HS that he was interested in ML, so he did the smart thing and doubled down on math finishing HS after taking multivariate calculus at Cal over the summer. This helped him greatly as he dove into ML at Stanford.
So, my answer to the question ādo you need calculusā for CS remains the same: "Yes, especially if you are interested certain CS areas of expertise. How math is taught is a whole new topic.
Maybe we are asking different questions then. Suppose the question is āIf you plan to go on to a career that is similar to 90% or more of all the other people with CS degrees, do you need it?ā My answer is that you will probably never need calculus again. This is true not only of application development (including backend) but even research-level work. As noted, I have a PhD and can state this from experience. In fact, part of my dissertation probably touched on a little more calculus than most. I am pretty sure I at least used a linear approximation of an infinite series to prove a bound on an approximation (itās been a while, so I would have to check).
Good for you. Why do you feel the need to aggressively challenge someone like that? You are trying to set some kind of idiot poser trap rather than have an earnest conversation? How does that further the conversation?
I am referring to the movement to reassess the applicability of the current math curriculum to what the modern world demands of its citizens. Data science is a work in progress, but at its core it IS the very question posed in this discussion: what should high schools teach? It has been discussed in various other threads before, didnāt think it needed to be defined as a test for my qualifications to participate in the thread. I didnāt realize we had to divulge our degrees before we could post.
For those who genuinely donāt know what it is, I guess that doesnāt include you, hereās a quote from the site explaining it:
āThe ability to work with, understand, and use data has become an essential life skill and requirement for an ever-expanding range of jobs and careers. Data is everywhere around us. Ninety percent of the worldās data has been created in the last two years (Marr, 2018). This new data intensive world can be difficult to navigate; decisions that used to be straightforward are now more complex, requiring individuals to be constantly separating fact from fiction. In short, the need to analyze and interpret data is no longer confined to engineering or computer programming; it has become an essential life skill. Everywhere we turn, data is telling and weaving stories about our world.
Yet, the K-12 education landscape has been slow to recognize the changes the data explosion has made to society (Wilkerson & Laina, 2017). We are woefully under-preparing our students to successfully navigate the 21st century. The mathematics we teach is rooted in the 1950s space race and offers little practical utility in the 21st century. There is a distinct and widening gap when it comes to the skills and competencies students need in life compared to what is taught in schools.
To prepare learners adequately and succeed as an educational reform, a fundamental shift is needed when it comes to the standards and competencies students are taught. However, we need to build a common definition of data science in K-12. What should be taught to students in K-12 education? What competencies will produce individuals who are data literate when they graduate from grade 12? ā
2 Likes
Because saying ādata scienceā is about the same as saying ālife scienceā. It is a collection of topics that are very vaguely connected with the extremely vague term ādataā.
You CANāT teach ādata scienceā in high school, as you canāt teach ālife scienceā. You can teach some parts of statistics, but nothing as vaguely defined as ādata scienceā.
Regarding calculus and computer science, I expect the overwhelming majority of kids entering the colleges that are the subject of this thread will have taken calculus during HS and been quite successful at it. For example, Harvardās freshmen survey reports 95% of students took calculus during HS, and I expect the remaining 5% are unlikely to be prospective CS majors entering as new students. This prevents calculus from being used as an effective weeder type course.
There is a common issue that different HSs have different quality and rigor of their HS class, which relates to why the referenced colleges generally have all new students take a math placement test to help determine what math starting point and level is most appropriate. Students with weaker math backgrounds may need to retake calculus, and the referenced colleges generally offer slower versions of their standard math classes for kids that need to take calculus again . This may be less common at highly selective LACs since the smaller class size can limit the number of different math levels available.
Calculus is generally not required to take CS classes. Students from a wide variety of backgrounds often take basic programming classes For example, in the 2019-20, 1800-1900 students at Stanford took 106A, which is Stanfordās intro programming class ā the vast majority of undergrads and a good portion of grads. This class has no prerequisites and requires no past programming experience.
Math is almost required for a CS bachelorās degree, often including some courses beyond just calculus. For example, Stanford requires 26 credits of math (usually 6 or 7 courses) including Calculus, Probability for Computer Science, and Foundations of Mathematical Computing. The remaining courses are self selected electives. Students are able to count HS calculus towards this requirement, with sufficient AP exam score.
Among the kids who need to repeat HS calculus in college, extremely few kids at āprestigiousā colleges do not pass the class. Grades below B are usually rare at such colleges. For example, Harvardās recent senior survey lists a median self-reported GPA of ~3.87/4.0. Getting a B or in very rare cases a C may discourage some kids from continuing to higher math levels, but Iād expect the calculus requirement rarely prevents kids from pursuing CS at the discussed colleges.
Such colleges generally do not have a problem with too few kids being able to do a CS major. Instead CS major enrollments are often growing to huge sizes. In recent years the growth has been especially large in previously underrepresented groups, including women, URMs, and lower income. For example, the most enrolled majors at Stanford are listed below, as of 2020-21. CS has several times more enrolled students than anything else. 217/622 = 35% women may not sound high, but thatās a lot better than in the past.
Most Enrolled Majors at Stanford
1 . CS ā 622 students (407 men, 217 women)
2. Econ ā 185 students (107 men, 81 women)
3. Hum Bio ā 181 students (45 men, 136 women)
4. Engineering* ā 163 students (70 men, 93 women)
5. MS & Engineering ā 147 students (81 men, 66 women)
6. Mathematics ā 127 students (93 men, 34 women)
7. Symbolic Systems ā 118 students (58 men, 60 women)
8. Biology ā 113 students (54 men, 59 women)
*General engineering ā often indicates has not yet decided on specialization
2 Likes
There are some very smart math education experts who vehemently disagree with you. I suggest you read up on it. It may not relate to what you do for a living as a phD in comp sci, but that is really the point.
Thatās their job: chasing the next buzzword.
As I said, I know enough about ādata scienceā, and that term always brings smile to my face. 
Might I remind members of the forum rules: āOur forum is expected to be a friendly and welcoming place, and one in which members can post without their motives, intelligence, or other personal characteristics being questioned by others."
and
āCollege Confidential forums exist to discuss college admission and other topics of interest. It is not a place for contentious debate. If you find yourself repeating talking points, it might be time to step away and do something elseā¦ If a thread starts to get heated, it might be closed or heavily moderated.ā
http://talk.qa.collegeconfidential.com/guidelines
The users debating have been here long enough to know the rules, so I am once again putting the thread on slow mode until tomorrow.
1 Like
I love slow mode! Itās like watching the sloth in the DMV office in Zootopia!
Now if only we could have a slow mode for real world conversations!
3 Likes
Agree for the 90%, but how then should prospective students proceed? If you know for sure that you are interested in a specialization that need Calc (ML, CV, etc) taking calc makes sense because you will be using it. If a student is unsure, they should still consider taking calc because it opens more options (and options are good). Only if a students is sure that they donāt want to play in the deep end of the pool should they skip calc (if they change their mind they could take it in college but would lag their peers).
Iām primarily reflecting what my recently graduated son has recounted me. I only have a MSEE and I did take quite a bit of math (MV Calc etc), but the depth at which my son uses it today far exceeds what I used.
True enough. This raises other questions such as how much value is added by the āprestigious collegeā beyond the selection criteria they applied when accepting students in the first place.
Also, how much do these criteria correlate to intrinsic value in later careers and how much is simply expected and applied in a circular definition of āmeritā? E.g., at one time it would have been expected to read Latin or ancient Greek. While this was within the capabilities of most top students, that doesnāt mean that it was a good use of their time or that students whose talents lay elsewhere were fairly excluded.
Not really on topic, I admit. To tie in (repeating myself) ROI measured financially can be achieved in many ways without an elite degree. People are still going to want these degrees if for nothing else, the personal sense of accomplishment or the recognition from others. A lot of it is about dreams. You can be a multimillionaire and still feel like you bought a āone way ticket to Palookavilleā if you turned down Harvard for an affordable choice. There are also probably very direct social consequences including who you eventually marry, etc. Some Americans like to pretend we live in a classless society, but this is far from the case.
I apologize if I offended anybody.
I absolutely hate it when people complain that this or another subject in school is useless in their future job. Or when they propose the current fad to replace something that works.
I apologize, too.
It is a valid question, though, whether calculus in high school āworksā for most people who take it in high school. I donāt think anyone is suggesting replacing it. It makes sense for a small subset of students. Before about, what, a decade ago, it was never assumed to be a requirement to get into an āeliteā school, and now it basically is. Which is why the discussion is relevant to this thread. Should it be? Can we serve high school kids better?
I am no fan of Jo Boaler, the Stanford prof who is one of the main drivers behind adding ādata scienceā into the K-12 curriculum. But on this, I think she has a point. I am fairly math savvy compared to my attorney colleagues. I come from a very accomplished STEM-oriented family, and am the freak for choosing something outside of math/engineering/computer science. We can get into a long conversation about the gender implications behind that, but that is my reality. I took graduate level math and statistics, before I changed fields. Learning to read and understand research has been incredibly valuable to me. Calculus, not so much. I am around smart people in my career every day that are virtually innumerate, and I have an advantage over them. I am amazed with how people are unable to question or critically analyze the data we are all bombarded with daily. Fundamental data analysis is an important life skill, yet it does not exist in the math curriculum. We can debate what should be included inādata scienceā or whether that is the right term. but my point is that for people not going into math-heavy fields, even like law, a high school course in data (most people donāt get to stats until after calculus) is a really good idea.
3 Likes
I resemble that remark! 
I would argue that it should for STEM students. Because thatās what they study in college.
I am absolutely not against studying some of the subjects that are included in the ādata scienceā buzzword. Statistics, for example, should have much more emphasis in high school, because it is essential to so much of anybodyās life. The problem with ādata scienceā is that itās vagueness will make it very easy to put some useless stuff that nobody will care about.
My claim is that it doesnāt really matter what particular area of math is studied. They all make it easier for the student to pick and learn another math area. In the machine learning field it is called transfer learning: you teach the computer using examples in one area (e.g. recognize cats), and it can perform fairly well in another (e.g. recognize cancer).
1 Like