Sponsored by:


Daphné Vanessa

Shamil Rodriguez



Stay Up to Date With The Latest and Not-So-Greatest News About Student Loans and More.

About This Episode

Alejandra Acrosta (@AleAcostaEd) is a policy analyst with the higher education initiative at New America. She conducts research and analysis on the equitable and ethical use of predictive analytics in higher education. Acosta previously interned at Lumina Foundation and the Kresge Foundation, where she informed investments and strategies that aimed to make higher education more equitable. Prior to that, she was a graduate researcher at the National Forum on Higher Education for the Public Good, where she analyzed institutional policies that affected undocumented students in Michigan’s public higher education institutions. She is also a proud to have served as a college adviser for low-income first-generation students in Silicon Valley. The product of public education from K–12 to college, Acosta holds a bachelor’s from UCLA and a master’s in higher education from the University of Michigan. Her immigrant family’s education success story and work with marginalized communities are the foundation of her work.

Today’s episode goes into the intersection between predictive analytics and higher education. Alejandra does a wonderful job of breaking what is predictive analytics exactly and how it applies to higher education and the student experience.


  • The basics of Predictive Analytics and how we interact with it everyday
  • Algorithmic bias and how data should not be seen as neutral or objective
  • How colleges and universities can use data to improve student completion rates
  • Thoughts on college and university accountability for student success
  • And much more…

Enjoying the show? Leave us a rating and review. Every comment helps! Drop in your IG handle so we can thank you personally!


Alejandra Acosta (00:00:00): When you're on Amazon, you buy one vacuum and then they're like, look at all these other vacuums and cleaning supplies you might like. So how is it that these platforms figure out, you know, how are they reading my mind? Um, and the short answer is predictive analytics.

The Student Loan Podcast Intro (00:00:18): Welcome to the student loan podcast. Here. You'll find practical advice on tackling student loan debt, paying down your higher education expenses and inspiring stories about paying off student loans. We're your hosts, Daphne, Vanessa and Shamil Rodriguez.

Shamil Rodriguez (00:00:36): Welcome to another episode of the student loan podcast. We are excited today to bring you a special guest. That's going to speak about something that we all interact with every single day, but may not realize that we're doing it. And that's predictive analytics. Our guests, Alejandra Acosta with the new America think, and action tank is actually going to break that down for us. And she's also going to teach us how predictive analytics is being used by higher education in ways that impacts your everyday life as a student. So if you're ready to learn, put on your thinking caps, because today is going to be a great episode. So without further ado, here we go. Okay. Alejandra, tell us a little bit about yourself.

Alejandra Acosta (00:01:20): Sure. Um, so I am a policy analyst at new America. New America is a think tank based in DC. And what that means is that that is we are a nonprofit organization that studies a whole wide range of topics. And so in other words, my job is to think, which is really cool and new America covers topics ranging from education to technology, to domestic policy, to all, all kinds of other things. Um, and so I'm in our education program, specifically our higher education program. And as a policy analyst, I study different things. Um, I don't always read actual policy, but I do a lot of research on different topics, including largely predictive analytics use in higher ed, as well as other topics in federal policies, such as financial aid, offer letters, um, accreditation and all kinds of other things. Okay.

Daphné Vanessa (00:02:19): It sounds super interesting. Can you tell us how you got there a little bit more about your background and how you got to where you are today?

Alejandra Acosta (00:02:28): Sure. Um, so I feel like a relatively new policy person, even though I've been in this role for a couple of years now, but, um, I would say my journey started when I was, um, a college advisor in the San Francisco Bay area in California. Um, there, I was essentially a college counselor, so I helped students with their college application process with their financial aid process. Um, did that for a couple of years and absolutely loved it, but realize that if I wanted to stay in that profession, that I would be helping students navigate a system that changed every year and that didn't necessarily work for them. So once I realized that I thought, well, let's go change the system. And I think you do that through policy. Um, so from there I pursued graduate school, um, in higher ed to understand how the system worked better.

Alejandra Acosta (00:03:24): Um, and again, without actually really understanding what policy meant, um, looked for opportunities in DC. Um, I had interned at the Kresge foundation while I was in graduate school. A foundation is, um, an organization that gives, um, that gives money out to different entities to support certain strategies and goals. Um, and so there, I learned a lot more about higher ed learned what our foundation was and therefore was able to apply to an internship with the Lumina foundation. And I know that the podcast has had Dr. Katherine wheedle from the Lumina foundation on the podcast. So she was my colleague for a little bit of time from, and so in that position, I was able to learn about the higher ed policy world. Um, and from there, all the people that I met and interviewed just to understand how things worked, um, from there, I met the director of our program at Numerica and, and the rest is history.

Shamil Rodriguez (00:04:21): Great, great. And that's wonderful. And I, and I like that we share that path because a lot of times our listeners aren't aware that this is a job that you can actually get paid to think. Right. And I like that new America says that they are a think an action tank. Can you tell us a little bit about what that distinction really means? Just the, just so that people understood.

Alejandra Acosta (00:04:42): Yeah. So the concept of a think tank often leads people to believe that the organization only does research and does really like rigorous reach research. I imagined it as essentially academia outside of academia. And so often folks in think tanks, the notion is that you just do research, but you don't actually know anything or like have any real connection with the communities that you're talking about, or you just like publish research and put it out into, onto the internet or outer space or whatever, and just leave it there without doing anything. Um, new America wants to differentiate itself by, um, taking action on the research that we conduct. So as an example, with our higher ed team, how our team conducted research on financial aid offer letters found that it was really difficult for students to understand them because each one of them was different and sometimes not very transparent, so we could have left it at that. But my team has done a lot of work taking that research to the Hill to tell, um, members of Congress like, Hey, there's a problem here. We need to do something about it. Um, and since then legislation has introduced to start to consider how we can fix that problem. Um, and so that is what would I think would differentiate new America as a think and action tank or research and action tank, as opposed to just a place where people do research and don't actually do anything about it.

Daphné Vanessa (00:06:14): That sounds so cool. And it's pretty cool. How involved are you with the private sector in terms of action? Because often solutions aren't, um, solutions could be multi-faceted right. So is there, does new America think about collaborating with the private sector, NGO community international organizations, or is it strictly with the United States government?

Alejandra Acosta (00:06:38): Yeah, um, I, I assume it it's different for each program. Um, but for us, at least in the higher ed program, I will only speak to that because that's what I know best. Um, I, I don't think that we have too much collaboration with the private sector. Um, I, I think my team's expertise is more along the lines of working with government or within government. So that's where we have largely focused on what we do collaborate with, um, other nonprofits. So for example, that same financial aid offer letter work, um, was that in collaboration with another nonprofit called you, excuse me, you aspire. Um, and so we often collaborate with other groups, um, either to get data, to do the research with them, um, or to strategize about what we're going to do next. Very cool. I agree. I think that that's a very big distinction.

Alejandra Acosta (00:07:33): Um, at least for us, we, I like the idea that you guys are taking action and trying to collaborate with folks and, and really not just leaving the idea, sit there as an idea. Uh, so, so kudos to you guys for that. Thanks. Um, so just joined, but you know, you're, you're, you're adding your part too, right? Yeah, I sure try. Thank you. So we can get, uh, before we move forward, I actually want to continue to go a little bit backwards, some more, uh, to give the listeners a little bit more background, uh, about where you went to school. Uh, I know on your profile, um, you know, you speak about how your immigrants, uh, background your parents' immigrant story to, to the States has had an impact in your life. Um, you know, and we've had podcasts guests that have talked intimately about their experiences coming here to the States.

Alejandra Acosta (00:08:22): Um, so why don't you just share that if you don't mind, uh, with some of the listeners, you know, your undergrad, your, your, your, your background, and like what motivates you to do this? Because I can imagine that not everybody just decides, you know, wakes up one day. I was like, I'm going to go change policy. You've worked at a think tank today, you know? Yeah. Um, I'm happy to share. I think sometimes in the policy world, um, intros go by really quickly. So I didn't dive much into that, but I'm happy to share. Um, I grew up in the San Francisco Bay area and for college, I went to UCLA there. I studied sociology, Spanish and Chicano studies. I was very much an overachiever back then and trying to recover. Um, and really my, my motivation for all of the work that I've done in my life is like you said, it's is my immigrant background.

Alejandra Acosta (00:09:15): So I'm of Mexican heritage. Um, some of my family was born here and some of my family was, um, born in Mexico and immigrated over here. Um, for many of the reasons that other families immigrate to the United States. Um, but specifically for me, I have my, from my grandparents, my grandparents were, um, musicians who played music at, um, you know, different parties and events to make a living, um, homemakers who, you know, made sure that there was food on the table and that everyone was taken care of. Um, and what really, what I hold very closely to my heart is, um, my grandpa on my dad's side was about, I said also he was a migrant farm worker, um, in the States he would come over from Mexico, um, to work in the fields, picking crops and picking the food that, um, not that we eat now, because of course it was a very long time ago, but the food that people put on their tables, um, and on both sides of the family that very humble work, set up a foundation for, for my parents to be able to pursue their education, as challenging as it was, they were certainly, um, non-traditional students who were working at the same time that they were going to school learning how to navigate everything, very challenging.

Alejandra Acosta (00:10:35): Um, and so for me that, you know, those two generations were just like figuring out how everything works. Um, set up a very, very stable foundation for me to pursue more of a traditional education, where I was able to leave home to live in a residence hall, um, to have like the typical college experience. And I, I saw throughout my life that that wasn't the case for everyone that even my peers didn't have that opportunity. And I was like, well, that's not okay, so let's do something about it. Um, and so that has always been, um, I would say like the foundation of my mission and the work that I do, um, you know, just seeing how across two or three generations things have changed so much and how that's not always the case. Um, and so the work that I do is in pursuit of making what I had more available to, hopefully everyone someday, but at least more people or as many people as possible.

Shamil Rodriguez (00:11:35): Well, thank you. Thank you for sharing. That's um, it's a very common theme that we've heard from a lot of our guests that take action. Um, and a lot of us can relate to that, that story, uh, or that motivation. So thank you for sharing that with us. I hope that if anybody has any comments that they want to share on that, you know, when we post this podcast, we'd love to hear it, uh, because it is, it is a, it is not easy for folks to come to the States and like your grandfather did work here. Um, but you know, you do that with the promise of helping the next generation. And it's, it's really amazing to see how you, you know, you literally are the fruit of that labor.

Alejandra Acosta (00:12:15): Yeah. That's a great way to put it. Thank you for that

Shamil Rodriguez (00:12:18): Course. So let's, um, let's, let's, let's jump into, um, predictive analytics. I, I think, um, smooth transitions, right? Uh, no, but I think it would be great, uh, just because like, I, I, I, I was reading your bio again and, and really trying to, to, to see, I was like, you know, how, how are they using predictive analytics to really, uh, impact higher education? I think it would be really interesting for the user or listeners too, to hear how you guys are doing that. I knew America and how you got into that, that space.

Daphné Vanessa (00:12:50): Totally agree. And can we backtrack first to just, not everybody in the audience knows what predictive analytics are? So if we could give, uh, an overview

Alejandra Acosta (00:13:03): For the people that don't say, show me the data

Daphné Vanessa (00:13:05): Every day,

Alejandra Acosta (00:13:08): Happy to the, and really, I think my quick transition from being a, basically a college counselor to working in policy, um, has made it so that I always want to make this information as accessible as possible. Like, I didn't know what this was before I started working at new America. I didn't know what a foundation was. I didn't know what a think tank was. And so I'm happy to define things throughout, throughout the conversation, um, because the world that I work in can be very, um, very closed. And I think that that's part of the problem and it's something that I want to somehow fix it in my career as well. So happy to explain that, um, I think pretty much everyone who listens to this podcast is probably used Amazon or Netflix sometime in their life. I'm assuming you two have, um, or Spotify or YouTube or anything like that.

Alejandra Acosta (00:14:02): And so if you notice, I mean, you, you have certainly noticed that there's so many memes about it, or like Twitter threads about it. Um, but when you use those, those platforms and websites, um, you'll often watch one TV show or one video, and then the platform will say like, next step is this video, or like, maybe you would also like this other TV series or this other movie when you're on Amazon, you buy one vacuum and then they're like, look at all these other vacuums and cleaning supplies you might like, so how is it that these platforms figure out, you know, how are they reading my mind? Um, and the short answer is predictive analytics. Um, and so I say that so that you can like imagine and see that like, Oh, okay, I have already used this in my life, or I have already, um, you know, like been involved with predictive analytics in my life.

Alejandra Acosta (00:15:00): Um, but predictive analytics has essentially, um, the use of prior data to predict future outcomes. And so what that means in regular people, terms, regular people, like all of us here, and those of us listening is that, um, platforms or schools or companies will use information on what has happened. So for example, um, I only watch, let's say like, um, romcoms on Netflix. So they'll use that information about me and the thousands of other people that watch romcoms on Netflix. See what other things those people have watched create a mathematical equation with a bunch of numbers and factors and variables and things that I truly don't understand that then spits out, um, a prediction. So out of the thousands of people that watch romcoms on Netflix, what other types of romcoms did they watch after watching? I don't know, crazy rich Asians or something like that. Um, and so then that is how we, the users see these other recommendations for other movies or for other series. So that's essentially what predictive analytics is and looks like from a user perspective. Do you two feel like that kind of explains it a little bit more than I can do there?

Daphné Vanessa (00:16:30): The analogy, I think that's so much better than how I would've explained it, the reference to something familiar. I think I hope, uh, the audience is going to get a glimpse into how it factors in everyday life. So I think that that was a great, great definition. Yeah.

Alejandra Acosta (00:16:45): Cool. I'm glad.

Shamil Rodriguez (00:16:48): Yeah, I agree. I agree completely. That was good. Um, it compares, everybody has touched some sort of platform that does it. Uh, and so that was really well said in social media too, right? Like everybody else's social media in some way or another even LinkedIn counts. Um, and they do the same thing, right? You, you like a photo or a certain type of photo. Um, and then all of a sudden you see, you know, dozens of recommendations for that similar thing that other people are, are also doing at the same time. So really well said, that was really good. So, so here's the question. How do you, how do you tie that to higher education? I mean, we know our, so we saw it for Netflix to see it for Amazon. We see it for, you know, dozens and dozens and dozens of other companies is higher education using that, or are they going that way? What do you, what have you seen so far?

Alejandra Acosta (00:17:40): Yeah, so I think it, what's funny is that this is not something that I think a lot of people are aware of the fact that colleges and that we know and love, um, are using data to help predict student outcomes. Um, I, I think about like half, at least half of all colleges use predictive analytics in some way. Um, and there are different ways that they use it. So, um, we have a report at new America that one of my colleagues wrote, um, that my colleague Iris Palmer wrote, I think with somebody else as well, that's a landscape analysis or like a really broad overview of what the use of predictive analytics in higher ed looks like. Um, and they found that, you know, it's used for some like less student facing things like predicting, like when should we turn on the sprinklers for how long, how do we use water efficiently, which is not a huge concern to students. Um, and they also call it does also use it for I'll talk about, I guess, like two primary places where colleges use this, um, primarily in admissions and in retention and retention means keeping, um, students enrolled in school all the way, all the way through graduation. And so those are the two places where predictive analytics is, is used the most in higher ed. I would say it's most common or has been used the longest in admissions, but it's increasingly being used, um, in retention to see how we can help students stay in school. And

Daphné Vanessa (00:19:15): Where are they getting the data from? So my question is predictive analytics relies on some foundational data sets. College is such a fluid experience way back when I was there. I know there's a lot more technology these days. Um, but how are they gathering data to make these analyses?

Alejandra Acosta (00:19:39): Yeah, so I think it depends on, um, whether you're at a college is using this for admissions or for retention. Um, and so I guess the short answer is from a lot of different places and it depends for what, um, I'll speak specifically to retention. Um, because I think that's, what's most related to student loans. Um, the predictive analytics use in admissions is really interesting and pretty crazy. Um, so I'm happy to talk about that later if you're interested. I also encourage listeners to look into it.

Daphné Vanessa (00:20:17): Yeah, no, you can't, you can't have say a crazier than interesting and then not tell us.

Alejandra Acosta (00:20:24): Okay. Short answer is colleges use, um, predictive analytics to, to like decide where they will recruit students, um, how much like contact they have with students to convince them to apply and then to predict who will actually, um, who will actually apply. And finally, who will actually enroll, um, sometimes there's some financial aid stuff going on in there. So, you know, sometimes maybe when we were applying to college or when I was a counselor, you're like, how is it that this student got in, but not the other one. It's very complicated. There's very much a human and financial aspect to it. But now that I've learned a little bit more, it's very possible that they were like secret algorithms going on there, um, that like helped colleges make those decisions.

Shamil Rodriguez (00:21:13): What data sets are they using?

Daphné Vanessa (00:21:15): Like where are the reports coming from? Is my question? Is it the portal that you apply into or are they taking demographic information? Are they looking at drop-off like, what, what are they looking at to make these decisions?

Alejandra Acosta (00:21:28): Yeah, for admissions, a lot of the data comes from at least for freshmen admission, admissions comes from, um, from testing. So like the sat and the act, um, whether or not you took it, there's often ways that, um, that these testing companies can get more information. Sometimes if you took a practice test or filled out a survey, they just have huge, um, databases that, that colleges can buy and then use for admissions. Um, in addition to their own data that they collect over time, um, for retention purposes, um, that one for, um, colleges also collaborate with, um, private companies or private firms either to help them with data collection, to help them make the model, to help them have some sort of platform to house all of this on. And so it it's been a little while since I've looked at the retention space now, so I'm not as clear on it. Um, I think for retention, a lot of schools collect their own data. Um, and from there start to figure out like, what are the patterns, um, you know, like what students are more likely to graduate if you fail this prereq class, like what happens after? Um, and so I believe that that one is a bit more internal. Um, but since it's been a little while it it's possible that I'm incorrect about that.

Daphné Vanessa (00:22:55): No, that that makes a lot of sense. So it sounds like data is sourced both internally and from third parties from the moment at which somebody is even considering going to college by sourcing information from testing companies and other sort of, uh, pre college entities. And then once they're in the university analytics on student behaviors that can be tracked such as student tenants at events or whether or not they enrolled in a certain course or what was a voluntary choice versus a required choice and how those were made. Anything that can be true.

Alejandra Acosta (00:23:31): Yep. That's exactly right. Very interesting. Yeah.

Shamil Rodriguez (00:23:33): So what was the, I guess the, the arounds, do you want to go a little bit more into how this plays into student loans? I know that was something that you've mentioned before. How are these retention data sets or I guess, predictive analytics, how has that applying to student loans and outcomes?

Alejandra Acosta (00:23:52): Yeah. And I want to start answering that question just by clarifying. Like, I, I know that like hearing so-and-so entity, like has my data, or is tracking X thing can sound a little bit scary. Um, especially when it, when now we're hearing like, Oh, this college that I know and trust and love is doing the same thing as like Amazon, like that's a little bit weird and it makes me uncomfortable. Um, and it's both a good and a bad thing. Um, it's a good thing because colleges historically have operated off of like a feeling. So like your professor or your counselor, like, Oh, I've kind of noticed that, like you're not showing up as much. I've kind of noticed that students that come from, um, from this class and had a hard time there, aren't doing as well in like class. And so like, let me act upon that feeling.

Alejandra Acosta (00:24:50): Um, and that doesn't always work. And so it's a good thing that colleges are actually collecting data on what's happening with students and hopefully doing something and doing the right thing about it. Um, so in that sense, using predictive analytics is a really good thing, at least for retention, in my opinion, um, because you don't just like admit the student and then just like not know what's happening with them for the next several years. Um, the other side of that is we have to be really careful with how we use, um, data and predictive analytics because they are both imperfect. Um, and they both reflect our real world. So I think there's often a misconception that data is neutral and data is fair that numbers are neutral. Um, but that's not necessarily true. Um, our numbers, our algorithms were created by humans, um, and were structured and coded and defined by humans who have their own biases. Our, our systems, our society has biases entrenched in that. And so that can also pose a problem when you're using data and analytics in higher ed. Um, so I just want to say before answering the question, clarify that this is both a good and could be a bad thing. It's certainly a challenging thing.

Shamil Rodriguez (00:26:18): The bias, can you just, uh, just elaborate on what that might mean? Or like, just so that people are understanding like an example that people can kind of take home of what that might look like.

Alejandra Acosta (00:26:28): Yeah. Um, so in the retention space, an example could be like, let's see, I have now been admitted to my college. Um, and I am a first-generation student. I am low income and I'm also a student of color. Um, I have X GPA from the other college that I went to or the high school that I went to. Um, and so like with this whole fancy algorithm, the college could try to figure out, okay, what is this? Student's likelihood of graduating from our school? Um, since me as a student, I have a lot of cards stacked against me. Um, it could be that the model then says like, Oh, you don't have a very high likelihood of graduating from this college if that's in the wrong hands. Well, I guess there's, there's two places where the bias can happen. One in the data. And one when you're acting upon the data, the data in and of itself is biased in the sense that, like I said, um, our society, our schools, our systems have systemic racism entrenched in them.

Alejandra Acosta (00:27:45): So just like in terms of the racial, ethnic, um, part of this, um, that sort of demographic data could say like, Oh, Hey, a Latino student is less likely to graduate. A black student is like less likely to graduate compared to their white peers. Um, even if you take that data out of the models, a lot of the time you can still see the same kind of outcomes because that systemic racism is so entrenched in our system, that you don't even need to put that data into a model to basically see the same results. So for example, take that information out, but you still have zip code or you have, um, first-generation status, you have GPA, all of those other things are still so closely tied to race that the model could still have, um, those similar kinds of outcomes where they say like, so-and-so kind of student, student of color is less likely to graduate. Um, another example, sorry, is that

Daphné Vanessa (00:28:47): No, no. I, I think I am, I'm in agreement with you. I think that algorithmic bias has a lot of challenges, especially as it relates to predictive analytics. And then you have to think about even the questions that are being asked, right. And in whether there is bias in the types of questions that are being asked, are those being framed for a particular type of model student, instead of asking questions that may or may not be beneficial towards, you know, students from different socioeconomic classes or students from different race or ethnicities. So I think there is a lot of there's a lot. It relates, they, they, it all relates and needs to be considered by really appreciate that you saying data is not perfect. That is something that people often forget.

Alejandra Acosta (00:29:37): Yeah. And it's really dangerous to think that it's, that it's neutral objective because it's not. So what I was going to say next is like, think about like how you code, um, I don't know, something like GPA, let's say like, Oh, let's see what the students like. Let's put GPA into buckets, like high GPA and low GPA. And that's like a little block inside the algorithm or the model. If I, as the person creating or like looking at the data, coding the data, making the algorithm. If I decide that high GPA is just a 3.8 or higher, and that low GPA is 3.8 and lower, what does that mean? Right. Like, that's basically my opinion in this, in this fake scenario. Right. And let's say that, like with that one building block within the model, if anything below a 3.8 is a low GPA, then that particular block could make it so that more students are shown as being at risk of not graduating, but just because you have a 3.7 doesn't mean you have a low GPA doesn't mean you're not going to graduate and you could have a 2.7 GPA and still graduate.

Alejandra Acosta (00:30:54): Right. So the way that you define things within this, like very long algorithm, that is also one of the ways that bias comes in, in predictive analytics where bias can come in. Is that like, once you have like the outcome spit out by the model, let's say, Allie has a 60% chance of graduating from this college. If you have a counselor or a professor or a college administrator looking at that outcome on some sort of like online platform or dashboard, then that's another place where there could be bias where I could see as like the accounts that would be like, Oh, this student only has a 60% chance of graduating because they failed this class or whatever. Um, let's move them to an easier major so that they're more likely to graduate. Um, and so there's just so many places where this could go wrong. Um, and I don't mean to like freak people out. Um, there's also a lot of places where, where things can go. Right. Um, but I will finally get back to the question that shimmy lasts me a while back. Um, I think it was, how does this relate to student loans? Is that right?

Shamil Rodriguez (00:32:08): Uh, yeah. No. So I think a good, a good transition there when you're talking about outcomes, I just was curious to see, do you find in your research that a lot of colleges and universities are, uh, actually using this data or are they really, I mean, cause from my experience, when we were making decisions about data sets that we would like our platforms that would use to collect data, and this is on a community college level, it's not on the same level as some of these institutions, but I remember we would have, you know, one admin that actually knew how to use and create the reports, but like there were still conversations about like what should be on the dashboard, like what information needs to be pulled and how should we pull that data? So, I mean, we're, you know, at that, my experience at that time was still over, I guess, surprising to say over a decade ago at this point. But, uh, but at that time it was still bird, it was a new field that wasn't as relevant as it is today. So I'm curious to see, have you seen that a lot of colleges and universities are utilizing this data or are they really kind of like not using it to its maximum potential to actually help students graduate and like shift, you know, catch them before they fall off track?

Alejandra Acosta (00:33:18): Yeah, it's definitely increased in the past, um, decade or so. Um, I, I saw a report with actual numbers about those, but again, a while back until I couldn't spit the numbers out at you right now without being completely inaccurate. But I would say, um, my, my guesstimate, from what I remember from the report, I think it was a survey, um, through some sort of membership organization, I think like close to 50% of colleges are, are like thinking about using it or preparing to use it or actually using predictive analytics to help, um, students stay in college.

Shamil Rodriguez (00:34:00): Wow. It's, it's just sad to hear that because, and I say sad in the sense of, I would want it to be more right. Because I know always spoke about the scary stuff. Like, you know, big brother is watching you, the collecting all your data. Right. And you know, it can, it can go down such a scary path. But when you think about the idea of like, as a parent or like the person who's sending your child to college and university, right. Or as an admin, who's working there, who's trying to help like as a counselor trying to help students get to that finish line. Um, you know, you do want to know that there's something keeping up with that student to kind of help make sure that they don't go too far off the beaten path. Like what if somebody missed three or four consecutive classes, has anybody checked in with that person

Alejandra Acosta (00:34:41): Right. For a family emergency

Shamil Rodriguez (00:34:44): Or is, has something occurred in their personal life? That's gotten them off path. Are they feeling disheartened? You know, like they feel like they can't make it anymore. So like those things, uh, you, you would hope that that's where, you know, use this super powers, I guess we'll say for good, right. Where you can help keep students on the right path. So, so if let's say that the 50%, we'll just guesstimate at this point, and we can always add that on the show notes afterwards, where, where do you think this applies in terms of, uh, you know, uh, you know, student loans, as soon as if pay or, you know, completion and student loans and predictive analytics, where's that crossroads that you see happening right now?

Alejandra Acosta (00:35:24): Yeah. It's all in the outcomes. And so outcomes, um, are, you know, could it be a lot of different things, you have an outcome from a predictive algorithm or model. Um, and so that could be like so-and-so, or so-and-so, student's likelihood of graduating from a college that is one outcome. Another outcome that is often referenced in higher ed is whether or not students graduate. Um, others could be like how much money you earn after you graduate, whether you're able to pay off your loans, et cetera. Um, and so where predictive analytics in retention in higher end and student loans come together is that the use of predictive analytics in higher ed is one of the primary uses uses is to help students actually graduate. Um, and so I think you have talked about this in your podcast before, but, um, newsflash, unfortunately, a lot of students don't actually finish college once they start.

Alejandra Acosta (00:36:27): And so that poses a really big problem. If students are borrowing money to finance their education, um, you know, you go into college, you say yes to taking out X amount of loan so that you can go to school, but then once you get to your college, it's really lonely. Nobody's checking up on you. Um, you have some sort of hardship, um, and you say, well, this is too much. I don't want to do this anymore. Or I have to go take care of other things. And so you're left with your student loan debt and no degree. And so that's where these two things come together. If more colleges use predictive analytics to keep track of how students are doing and reach out proactively when they start to see like, Hey, so so-and-so student, hasn't shown up to class a few times, um, they've failed the past couple of quizzes or whatever other piece of information at that point, that that's a critical moment where a college can reach out and say like, Hey, what can we do to help you and hopefully help the student graduate so that then they have the means with which to pay back their loan.

Alejandra Acosta (00:37:34): That's like the short way to explain it. Um, and this is where predictive analytics has a lot of power because you can, I mean, how many of us have, um, you know, gone through college or any sort of new experience and felt very alone. Like nobody knew what was going on. Nobody knew what you were going through. And you were just trying to figure out all of this new stuff by yourself. The predictive analytics is one way to help alleviate that for so many students who are in our call in our higher ed system now who are adults who are returning to school or first-generation or low income and are navigating this really complicated process with little support. Um, and so that's where, um, predictive analytics and student loans come together

Daphné Vanessa (00:38:20): Makes a lot of sense. Has there been policy discussion surrounding the transparency of predictive analytics? For example, if it's one thing to reach out when you notice that the data and the analytics are you, that it's time to reach out to somebody so that they can finish their degree, make sure that whatever external action, um, can be addressed, however, it's something else for the student to have that information themselves and be empowered to, to get themselves back on track. Has there been any conversation about transparency behind predictive analytics?

Alejandra Acosta (00:38:59): Yeah. The short answer is no, not really. Um, and that's part of something that I would, I would like to change. It's probably going to be quite a marathon though. Um, right now the use of predictive analytics in higher ed is, um, largely studied like at the institutional level, like a practice thing. Like here are best practices for how to use data or how to use predictive analytics, um, but not so much at the policy level, um, at the federal and state level for higher ed policy. Um, believe it or not, there's not very good data that exists, um, on what's happening in higher ed. Um, and so one example is,

Daphné Vanessa (00:39:46): Yeah, I think y'all have covered this before. No, no, no, no, please, please, please continue.

Alejandra Acosta (00:39:53): Yeah. I mean, it's really sad. It's, it's weird. It's just super outdated. There's data about so many other things. Um, like what type of romcom I like on Netflix, but like not whether or not if I'm a part-time student, I graduated from college. And so the example that I want to give is that a lot of the data that exists, um, at the federal or state level about students, like in different databases that are publicly accessible, um, they look at graduation rates, but only for what is called a first time full-time student. And so that, what that means in regular people terms is, um, a freshmen or this is the first time they've ever gone to college, that enrolls full-time. Um, and so whatever graduation rates that you see for colleges, a lot of the time on these databases are really inaccurate because they're only looking at the students that are freshmen and that enrolled full-time. And as we know, that's not what the story is for most students these days, most students, um, you know, have other responsibilities, they have children or dependents, they have a job they've moved to different schools, they're transfer students. And so that, that like very foundational piece of data excludes a lot of students. So I share that to exemplify how the data that we have about students at the federal and state level is improving. Um, but still doesn't tell the whole story that is very

Daphné Vanessa (00:41:34): Clear. And for students who don't fall into that category, the university isn't necessarily creating programs for that type of student. Right? So that data point is almost beneficial to universities.

Alejandra Acosta (00:41:49): Yeah. All of these data points and just the fact that schools pay attention to what's happening to students, I think is a huge shift from how college was like maybe like 10, 20 years ago. So just like knowing what's happening is so basic and foundational, but it's really pretty new. Um, and it's hopefully going to be for the better

Shamil Rodriguez (00:42:12): You can see that. I think the, I hope it could be the, for the better as well. I think before it used to be, once you get in it's up to you to finish and wrap it up and graduate and move on. But if your mission is to educate and to create productive members of society that are going to change the world in their own way, then wouldn't you, if you had the ability to improve your odds of that success, what do you want to take advantage of it? And I think that's where, uh, being more aware of what your students are doing and trying to help them get to that finish line. And that objective, uh, makes the most sense. Um, so, you know, that actually seems like a really good place to talk about. Like, who's, you know, we're talking about like the federal government state level, who's CA who's capturing data who sharing it.

Shamil Rodriguez (00:42:58): Who's transparent. Who's not, but you know what, I think a good it's maybe a good segue into, uh, they ended up about the idea of who's like keeping up, like who's doing that now. Right. Because we're, we're in the pandemic, you know, a lot of us believe we see the light at the end of the tunnel, right? Like we're still in it. Right. We're still not there yet, but there, but I know that we've had a recession before. Right. We've had dark times in America before, and we've seen how, um, how unfortunately people can be taken advantage of during those times. Um, do you, do you want to just talk about what that actually means? Am I being too vague? Like, would you mind just kind of elaborating on, uh, you know, our people, our schools helping to commemorate the decisions who are the watchdogs, like who's helping keep track of all this stuff now.

Alejandra Acosta (00:43:50): Yeah. Wash dogs was a really good phrase for that. And I'll, I'll aid the transition here to just because, um, I want to make the connection to student loans clear. So again, like if you take out loans from college and then you don't graduate, you don't have the degree to help you have higher earnings. And so you are essentially, I guess, like kind of stuck with whatever next tightest degree you have and the earning potential of that next, this degree without, uh, with, um, with loans that you have to pay back. Right. And so that's what it looks like at the, at the individual level. And that's, that's happening across all schools in the country. Um, and so, you know, let's think about this, like a kind of like a, I take it one step higher is any, like, does anybody care that these students don't graduate?

Alejandra Acosta (00:44:47): Does the college care about me? If I do not graduate? Does the college care if there's a huge group of us that doesn't graduate? Are there any consequences? If this is like a really big problem at a school? So like, if most of our students don't graduate or if most of our students graduate, but they don't make enough money to pay back their loans. Is there anybody who is paying attention to that? Is that a problem? Does anybody do anything about it? That's not a question that I ever asked myself in my college journey. Um, but if you think about it, there should be some sort of like mechanism to keep colleges accountable, um, to actually fulfilling their job of having a student graduate. Right. I think like a lot of us probably think about that kind of like subconsciously, but like, don't quite have the words to talk about it.

Alejandra Acosta (00:45:45): Um, and so this is one of my weirdly favorite topics in higher ed. Um, it's called accreditation. And what that means is that, um, there are entities, um, aside from accreditors as well, but I'll keep it specific to accreditation just for the sake of making this make sense. Um, there are these nonprofit membership organizations that colleges are a part of, um, that give a college, a stamp of approval to receive often to receive, um, federal financial aid. Um, and so if you, if you think about it, you're like, okay, well, how does a college even come to exist? One and two, maybe if, if any, if either of you two, what if any listeners applied to scholarships you could often see, um, in the application it would say, um, you have to have graduated, maybe have this GPA and be attending an accredited college or university.

Alejandra Acosta (00:46:48): And you're like, I dunno what that means, but I'm pretty sure my college is accredited. So whatever I'm going to play anyway, that's certainly happened to me. Um, I was like, I'm UCLA as a good school. I'm pretty sure they're accredited, but I have no idea what that means. Um, and so there are these, these entities, they give colleges a stamp of approval to operate and for many to access financial aid dollars. So what does that mean for the individual? Um, if I choose to enroll in college for most students in the States, um, their college is accredited. Um, so if I choose to enroll in a school, I probably can't pay for this completely out of pocket. So I need some sort of help to pay. Right. Am I alive? Like the question then comes is, am I allowed to use this money at my college? And so that's what the accreditor does. They give a college, a stamp of approval. They say you are good enough to be worthy of, um, basically an investment, um, from the federal government, from taxpayers, from the individual student, um, that chooses to send their financial aid money to pay for school. Does that make sense? Cause this can get pretty confusing

Shamil Rodriguez (00:48:06): And I think, I think it does. I think it was, uh, especially when you said like how does the school become a school? Right. Um, it's, it's in my opinion, when this, when I first thought of those types of questions, the idea was, Oh, okay. Like it's probably going to be like state government, you know, like, or whatever, or maybe some federal program that they have to apply for, and then they get approval. And my great, but then when you learn about accreditation and you're like, Oh, wait a minute. There is a group out there and they're not there. And, and their job is to make sure that, that, uh, these schools, you know, qualify to meet those needs, uh, so that they can actually, you know, kind of function in that way. Uh, I, I found that to be really intriguing, but it also made me wonder, I was like, well, who sets the standard right back to that bias perspective. It's like, well, who sets the standard? You know, how often are they being regulated? What are they checking? Right. Like then it just opens up another set of questions.

Alejandra Acosta (00:49:02): No, just all kinds of questions that I make. You can definitely go like into a black hole of questions about all of this. Um, but yeah, so like we were talking about how like, like how outcomes can affect the individual. This is just like another level of like looking at this more of a macro level. Like what happens if a college, like if most of their students don't graduate in campaign off their loans. Um, so then they are these entities that are several entities that look at colleges to make sure that they're good quality and doing what they should be doing. Accreditors are just one of them. So creditors are looking at whether colleges are good enough to keep operating and to keep having really the privilege of accessing students, financial aid. Um, you know, you're choosing to take out a loan to go to school, but if your school has historically like not let student not help students graduate or not giving them the tools, they need to actually make money to live and pay off their loan afterwards.

Alejandra Acosta (00:50:03): And that's a problem. Um, and you had mentioned the recession, um, you know, back in 2008 and what we're going through now in both of these times, um, there has been, and there was an increase in, um, students enrolling in for-profit colleges. Um, I will not judge any student's decision to go wherever they want to go. However, there is a, a lot of evidence that for-profit colleges don't often, um, give students the outcomes that they need to be successful to, you know, again, make enough money to pay back their loans to actually graduate. Um, and so I wanted to bring this up because in this really difficult time, there are a lot of people that are desperate to go back to school, to get some sort of degree and be able to get back into the workforce. Um, and B schools can often target those really desperate people who tend to be low income, single mothers, people of color immigrants, um, to make money off of students, financial aid off of their loans. Um, and if they don't have good outcomes, then that's not fair to the student who is making a huge commitment to take out thousands of dollars to pay for school, to eventually pay them back. Um, when the school itself is not doing their job of actually helping a student graduate,

Shamil Rodriguez (00:51:30): I think you bring up a good point there because, um, I, it's hard because I don't want to poopoo on, on the for-profit, uh, colleges, because the reason I want to say it that way is because I still think that there are a lot of nonprofits that are not acting in the, in the good right. In the same way. So I think to me, it's a, um, it, it seems to be easy to identify those for-profits and to just say like, Hey, they're doing bad and that's good that they, they have been caught. And I know there have been schools that have been shut down and people haven't gotten their money back for, for a such poor results. But I do think that there are some non-profit colleges and universities that need to who have that magnifying glass placed on them, uh, so that they, they all, they too can be held accountable. And, you know, I think it's something that needs to exist.

Alejandra Acosta (00:52:27): Yeah, for sure. I completely agree with you. I use the for-profit example because I think that's, what's easiest for us to really start to understand this topic. Um, and it's complicated. Outcomes are complicated. If you look at community colleges, most students technically don't graduate, um, for a lot of different reasons, either they choose to just take a few classes for fun. Um, the students that enroll in community colleges just have a lot of other things to worry about. And so it can look like, Oh, Hey, my, my community college has terrible graduation rate. Like at, at that point, even though you have a terrible graduation rate, are you really going to take away the, um, the opportunity to access financial aid there so that students aren't able to use their grants and loans to go to community college? No, you're not going to do that.

Alejandra Acosta (00:53:15): So it's certainly a very complicated topic. Um, but I just wanted to take, I want to take that idea of like, okay, there's predictive to help like an individual student graduate from a college. And then if you look at it at the macro level, like this is kind of a pervasive problem. Um, and it really affects student loans because so many students don't graduate actually graduate from college and is anybody taking note of, of that happening and doing something about it? So it affects the individual, the individual outcomes. And then at a macro level, it has implications as well.

Shamil Rodriguez (00:53:53): Well, well, well said, well said. So I think for, to kind of recap, uh, everything that we're hearing here. So, you know, there's the data that's being collected, which also isn't necessarily, uh, I guess the, we use the word clean. Um, I want to make sure I'm using the right terms here. Right. But isn't necessarily, uh, objective data, right. Because of the systemic issues that we mentioned, and that you've mentioned, uh, and put out here, and then in addition to that, the algorithmic bias that definitely brought up, which is so true, right? The programmers that are putting all of these things together that are putting these equations out there, have their own, uh, decisions and perspectives that they have to make, right. In order to make these decisions they have, or at least the make this process work, they do have to eventually make a decision as to what is a cutoff like for your example on the GPA.

Shamil Rodriguez (00:54:43): Uh, you know, and so that, that needs to be filtered out or at least monitored, right. To see how that's actually impacting people. And then another part is transparency, right? So, you know, your data is being collected as a student, but like how far does it go? Uh, can you request that your data not be collected or utilized in that way? You know, what rights do you have over your own privacy and data? There sets too. I think these are some policy topics that definitely should come up. And then after that, it's, you know, after all that's done, when you graduate, or if you even get to that point, are you going to be saddled with student loan debt and then not have a degree? Uh, you know, and if you aren't graduating, but you've got to tell a student loan debt, then who's holding the school accountable for doing that right.

Shamil Rodriguez (00:55:28): For getting into that position, because there is, there's a two way street, right? You are still a student and you still need to actually do your work to get there. But, uh, to me, it's a dance, right? It's, it's, it's two, two partners coming together to make this work. And I think that, uh, um, I've been really happy to have you on to talk about all these different areas. And I think we, I hope I hope we'll find out, right. But I hope that we've done a good job of connecting the dots between the moment you were applying to school. Right. And how you can graduate, still have student loan debt, but let's say have the degree so that you can pay that off in a reasonable amount of time. Right. So we're not even saying like that pseudomonas shouldn't exist and predictive analytics is somehow going to eliminate student loans. No, we're at least saying that they should be utilized by schools in an efficient way to allow for individuals to graduate and have their, have their degree and have their sights set on tackling that student loan debt, because that university was a partner with them in their journey, not just getting them in the door and then forgetting who they are.

Alejandra Acosta (00:56:36): That was perfectly said, thank you so much for summarizing it like that.

Shamil Rodriguez (00:56:40): No, I feel like we covered so much ground. And I was like, Oh man, this is really great stuff, but I'm over here, nerding out on. And I'm like, I need to make sure that everybody else's is, is where we're going here.

Alejandra Acosta (00:56:51): I mean, it's a pipeline from the day that you enroll in school to when you're done. Um, you know, there's, there should be accountability along the way. Of course there's personal responsibility. Um, but it should, like you said, this is a dance. It shouldn't be just one person putting in all of the effort. Um, and I think that's the message that I'm trying to get across is, you know, not graduating and therefore not being able to pay off loans is a huge problem. And so there needs to be accountability at different levels of the higher ed system to make sure that more students are able to pay off the loans that they do decide to take out

Shamil Rodriguez (00:57:29): Very well said, Daphne,

Speaker 4 (00:57:31): I would just say, you know, what, it was more student loan journey. I think if you can share that in your email, not everybody has a student loan during your journey with the cost of higher education, if you could share that, I think that'd be helpful to the audience.

Alejandra Acosta (00:57:47): Sure. Um, so I think, well, it's funny. I think a lot of people maybe heard like, Oh, she's not Latina from an immigrant family. She, you know, it's, it's likely that she had a hard time. I am an I'm an exception. Um, my student loan journey has been, um, very privileged, um, because of the foundation that two generations before me set up. So this is certainly not the norm, um, regardless of race or income or anything. Um, so I, I want to recognize how privileged I am in terms of my student loan story. Um, so for me, both my parents and I took out loans to finance my, um, my undergraduate education. Um, but at the end of, of my undergraduate career or a couple of years afterwards, um, my parents with, um, you know, extra jobs that they did with the, the care that they took to take care of their finances, a lot of privilege, a lot of hard work on a lot of floods, um, were able to pay off my loans so that I did not have, um, any debt that I had to pay back once I was starting my career with the idea that, um, I would have the financial freedom to make my own choices, to not depend on anybody or anything, um, and to choose what I truly wanted instead of what I needed.

Alejandra Acosta (00:59:24): Um, and so I recognize the huge privilege and that, um, I did take out loans for graduate school and that's a slightly different story. Um, but for undergrad, that is how that went for me. So I, I am very lucky. Um, I know that that is not most people's story, but if there is any sort of glimmer of hope here is both my parents took out loans for their education. Um, my mom went to a private law school, um, while she had three children, she is crazy. Um, but, uh, after many years after selling a house, um, paid off her loans and that is what enabled me to have the story that I have. So I think if there's any glimmer of hope in that story, it's that, you know, maybe for, for a current student, student loan, borrowers, things won't be easy or pleasant.

Alejandra Acosta (01:00:23): Um, but your hard work, um, could end up benefiting future generations. I think I'm a story of, um, of, you know, the goal that immigrants have when they come here or even just folks who are struggling with the goal that they have for future generations. I think I'm the product of that. And I hope that that, um, can serve as some sort of like little bit of inspiration or hope not because I did anything right, but because just of the work that was done before me, um, and I truly believe that that's something that other folks can, can do as well. Um, so that is my story to certainly very privileged. Um, but also part of the reason why I do this work is again, throughout my entire educational journey, I have been very, very lucky, mostly because people before me worked really, really hard. And so I feel a responsibility and a calling to make the system work better for more people. So that my story is not an exception.

Daphné Vanessa (01:01:24): I love that. And just because you had a beautiful ending, doesn't take away the hard work that your parents put in. So, um, I appreciate you recognizing your privilege, but I also want to just call out that, that, um, that's why your parents came here. Your grandparents came here, that's that? That's the beauty of this entire thing. So thank you for sharing that.

Shamil Rodriguez (01:01:45): All right. On that note, I, I have nothing else. Uh, that was wonderful. Okay. So if you guys want to hear more information about any of the, of the topics here, uh, feel free to visit the show notesPage@thestudentloanpodcast.com forward slash episode 25. That's the student loan podcast.com forward slash episode 25.

powered by

Related Episodes


Submit a Comment

Pin It on Pinterest

Share This