SCI SCI education expert discusses AI, women in STEM

Five years after the foundation of the inclusive IT of the Northeastern University, the center can boast with its aim of making computer education more accessible. In its partner institutions, which have more than 100, the number of women and people of color who study IT have increased much more strongly than those of men and whites.

Led by Executive Director Carla Brodley, the center has spent its first half-decade paying particular attention to supporting students who did not take computer lessons in high school and therefore do not have the reference knowledge with which some of their peers enter the college. Brodley and his team have developed a multitude of best practices to help these students succeed in their introductory lessons and to attract students who may not consider IT as a major in the field. One of the centers Most recent projectsFor example, will help a group of universities to test the way in which integrated IT majors – which merge composition composition programs with various other majors, such as statistics, graphic design and even English – attracting new students to discipline.

But the computer landscape has also changed since the launch of the CIC. In particular, the generator has become a point of pain for teachers because students use technologies for cheat coding assignments. But the generating AI has also become an increasingly important and interesting part of the technological landscape, which means that institutions must find the best ways to integrate the study into their existing programs.

While the center provides for the rest of its first decade, ED inside spoke with Broadley by phone of the place where IT teaching, one of the fastest growth majors in recent decades, is – and where it goes. The interview was published for duration and clarity.

1 and 1 IT women's rates of ICC partner institutions have soamed in recent years, thanks to the work of the CIC. What key elements have found contribute to these figures?

The strongest key element is to do it as someone who is completely new in computer science does not feel late from the first day and has a fair start. There are some sort of three components to this.

(First,) Ensure that they do not feel bad in a classroom with people who already know everything and that they are sitting there talking about what they have passed on the AP exam and the way it is easy. It is a terrible feeling. And they generally lie, but you don't necessarily know that only 18 years old. The second component ensures that your heaps understand that not everyone will come with the same level of experience and do not think that these people are stupid because they may not understand something that the majority of students who have had coding and some experience. So TA training is really important.

And the third thing is in a way in weeds; It is this idea of ​​a common evaluation, which is, if you have several sections of a course, by ensuring that they have the same assignments and the same exams, because then when they go in the next course, they all learned the same things. This is important, because if you do not have that and you have someone who is completely new in computer science and he gets the easy teacher for the first class, he suffers in the second class, while the experienced student does not do so.

We haven't done anything in particular for women of color. It turns out that women of color are more likely not to have a prior coding experience, because it is an optional course in high school, and it is not taught in each high school.

2 The warmest subject of AI is fears concerning the cheating of students. How do you see the institutions attacking this now?

We have heard of our partner schools that CAS, in particular, are fed up with the rating programs generated by AI, and teachers see great disparities in what people get their missions in relation to the way they do their exams. They could get an A on their assignments, then they fail the exam. And so it is a kind of clear indication that they may have had a little too much help.

I think we have to develop rating policies (place) that make sure you cannot take a course by simply using generating AI, and I will give you an example of a rating policy that would. So you take the note that a person had on the exam – say that he obtained a B on the exam. As long as written duties are in a note from a letter of this, so if they have obtained a B and they have obtained an A on written duties, then yes, it will increase their overall note. But if they obtained a D on exam and an A on written duties, then they just get a D for the course. It would be the policy that I would personally establish in my lessons.

This is just an example. And I don't think anyone was content in the best way to manage this.

Trying to find a way to do so, I think, is actually a really interesting research question: how do you assess the progress of a student while encouraging the use of generator for their learning, but not for cheating? There has always been cheat in the coding classes, and there is software that can take two programs and see whether or not they are structurally and logically identical – even if a student modifies indentation, formatting and all variable names, you can always say that logic is the same. We have been using it for decades to catch human cheating to human. It is more difficult to ask for the cheating from the computer to humans. So I think it's an open problem and it's fascinating. And how can we encourage students to do the work themselves?

3 and 3 Obviously, in recent years, a story that we have seen a lot in the news has been an increase in the number of technological and coding layoffs replaced by the AI. Is it a just concern? Is it something that students feel worried when they go into computer science at this stage?

I would say that students are very concerned about this, and we start, not necessarily to the northeast, but in some of the other schools with which we associate each other, we are starting to see a little less demand. IT was fair Grow a giant in recent yearsTo the point where it is difficult for universities to take even the endowment of the necessary courses. This could give us a little respite if the inscriptions decrease a little.

But I think that the demand for people who understand the technical AI will be important enough, and I simply do not see the request disappear for computer science. That there are so many entry -level jobs, I think that is what students concern. But I have not seen the data nationwide for that. Unless (The Burning Glass Institute) or indeed.com will publish something on this subject, I do not trust anecdotal (evidence).

4 Some of the ICC projects and programs have been funded by the National Science Foundation. What do you think of the cancellations of subsidies that occur there and how does it affect the teaching of STEM and IT?

So, I do not know the details (cancellations), but I know that there are publicly published lists that have been canceled, and I think many of them are worth it and did important things. And I am in distress of the cuts of scientific research in general which occur in the country. I think it will be a challenge for us, as a country, to make very significant progress in the resolution of the challenges of the world, and some of the challenges of the world concern STEM education.

I think that the work that CIC does, where we are working to make computer accessible to people without prior IT experience, does not take place on any of the recent decrees and certainly helps to expand participation in IT, simply because of which has previous IT experience before their arrival at the university. Less than 60% of our secondary schools teach computer science. So, whatever the identity of a student, if they are in a school that has no computer science, he did not take it before going to university. Make sure that they can be able to cross a computer program and, certainly, pass through the intro sequence without having the impression of being late from the day they start is a kind of mission of what we did in the first cycle. I think that these types of initiatives will continue to be very important as we move forward.

5 What are the main problems of computer education that the CIC will work to tackle in the future?

Two new initiatives: one is to facilitate access to education in technical AI. This contrasts to know how to use AI and the use of AI in learning environments in other subjects. It is actually a question of really understanding the AI ​​algorithms from a computer point of view.

We are currently working on a landscape study throughout the country of which offers a master's degree, which offers minors, which offer concentrations in AI, which needs courses. We also examine the prerequisites, because one of our work hypotheses that we just have anecdotal evidence for the moment is that it can take students to their seventh or eighth semester to have completed the necessary courses to be able to take the technical courses of AI. And this is a problem, because we really need them to be able to access them earlier, but it causes a second related problem, that is to say that if a student can only go to AI lessons during their last year, you cannot really set up a prior structure within AI.

The second major initiative we are examining is really examined and working on what is called the problem of credit loss between community college and four -year universities in STEM. In computer science, when we look at this, yes, there are articulation agreements between community colleges and universities of four years, but often these are not updated on a very regular schedule. What is happening is that, to a certain extent, with strict and progressive requirements such as IT, where you have to take IT 1 before taking IT 2, students will end up not having something accepted in this sequence, which means that they really go up and take more introductory equipment during the transfer to four years.

We are working on a small pilot project around this, because in my opinion, loss of credit is a really urgent problem and an urgent problem. This means that many people cannot specialize in STEM when they arrive at the four -year -old university, because they simply cannot afford another semester of school.