Welcome to Educator Insights, where we discuss teaching in the digital age and embracing technology to enhance learning. I’m your host – Olivia Lara-Gresty and today I am thrilled to have a very special guest joining us. Please give a warm welcome to Angelica Spratley, a distinguished Senior Data Science Instructor at the Flatiron School. In this episode, we will embark on a thought-provoking journey, exploring the captivating realm of “Technology, Education and Diversity: Bridging the gap in Data Science.” Angelica’s expertise in unique perspective will undoubtedly enlighten us all as we are sure this discussion will out for some invaluable insights. So, with that want extend my gratitude to Angelica for gracing as their presence today and to our cherished listeners. So, Angelica thank you so much for being on our podcast.
I’m glad to be here with you Olivia.
Thank you and so Angelica Jelly is okay if I call you jelly, I love that and you know I don’t know what your students call you. But when I was a teacher I you know definitely it was Laura Gresty was kind of a mouthful for students. So, I turned it into miss LG, so I love a nickname that carries through with people.…
Angelica
Yeah, definitely call me jelly.
OHHH, Miss LG. Yeah I think K through 12 call me by my last name but my adult learners just call me jelly.
Olivia
I Love that. It makes students feel familiar like they really know us. That’s awesome. Um, yeah well I wanted to just kind of you know, give our listeners a chance to get to know you a little bit so I was hoping you could provide us to some insights and know that you’re working as a data science instructor now. But is hoping you give us some insights into your background and the path that led you to this involvement in Data science and in curriculum design.
Angelica
Yeah, I think I just mentioned that I did k through 12 as well. So, after I got my bachelor’s degree, I transitioned in becoming an engineering teacher at a high school. So I taught juniors and seniors, computer science as well as digital electronics and both of those were advanced placement courses. So, we did a lot of video game design. We did python code and we even talked about data science a little bit so that propel means to actually get a masters in data analytics where we did. Pretty much data analytics and data science from a through z as well as do a project while I was getting my master’s that was very purposeful in predicting substance abuse relapse in patients so that we can intervene and save lives so that was like my first exposure to a data science model. Loved it so much went into the consulting field both for higher education universities as well as pharmaceutical industries and then said hey I’m missing that education I think as a lot of teachers. Do they’re like how can I add education back into my day to day so decided to actually marry the 2 data science as well as education. So now I am making content making curriculum and actually helping adult learners transition into the industry to become data sciences machine learning engineers or somewhere in the AI space.
Olivia
That’s awesome and I love that you know you have a bit of a windy path it all kind of comes together beautifully as you say it. But you know for a lot of our listeners out there and just people in general you know paths don’t always have that linear path and so even for your students, I’m sure it’s great for them to see someone who’s tried a couple different you know fields but has found kind of this um, underlying thread that connects them all. So that’s awesome and you know just thinking at that a little bit and I think that something that’s pretty cool for both of us is that we share a background in the field of biology and so. I’m curious if you think that this might have this foundation has influenced your role as a data scientist in any way.
Angelica
Yeah, so go biology people you know biologists I think while I was getting my undergrad and biology, I didn’t equate anything that I was doing as data science right? It wasn’t until I became a data scientist and a consultant or data analyst. Where I was like hey that degree actually prepared me for this role because I know maybe with you too Olivia that you spent some time in labs we were collecting data using our great pipets put in stuff on agar plates things of that nature and we were collecting data and we were doing trials over and over and over again in the lab and so just the importance of data and how collecting data and analysing data is very important to data science I was actually on the trajectory of data science and didn’t even know it.
Olivia
No, that’s awesome and I think that’s really great for our students and our teachers to really think about because you know if you’re teaching whatever class you’re teaching. Maybe it’s math or whatever it is. You’re really exposing your students to more than just that 1 subject even you know. An art teacher that might be working with duke. Ah you know geometric shapes and whatnot might be getting into proportions getting into architecture. You know every subject is so interdisciplinary in certain ways and it’s great for our educators to hear from you thinking at your student perspective that you were learning about you know this whole field of data science while you were actually. You know, technically getting a biology degree.
Angelica
Yeah, I definitely agree and like you said even the arts are so important now, especially with which we’re going to talk about artificial intelligence with just generating art landscapes and using music wavelengths to predict out music and sounds so that we can do algorithms to help people that are hearing impaired. There are so many use cases in Data science that covers almost every subject. So, it’s great that you’ve mentioned, especially the arts as well because Arts can also prepare you versus just the heart sciences.
Olivia
Yeah, no, that’s great and yeah, I mean you mentioned it. Let’s dive into it so you know as a data scientist and as a teacher now, I’m curious about how you’re leveraging AI and technology to enhance the learning experience for your data science students. And how this technology might help you facilitate student engagement interaction.
Angelica
Yes, so I love AI, data science and how it all comes together and I know this has been a controversial topic for instructors like do I allow my students to use things like ChatGPT or generative AI in the classroom and what is the risk associated with that. I will say my use of AI is not representative of the thoughts of Flatiron school as an organization. These are just my thoughts and opinions and what I do in my own personal classroom. So yes, I’m all about using AI in the classroom. So, a lot of my students since they are in Data Science, they are more aware of the pros and cons of AI which we’re going to talk about later. So, they use it in a more responsible fashion and so definitely we’ll give pointers there on how your students can use it responsibly. But since they code all day there’s a lot of AI tools out here. So, they have copilot pair programmer as long as they have that .edu account they can use it for free. So really, really love the student accounts that were able to provide students or that you should be providing students to let them leverage some of these tools and technologies and that is just like their coding assistant. It’s going to help them troubleshoot their code, especially if they can’t get to me quickly right.
So, a lot of students if they’re working on an assignment at home or they’re in the class and I’m in a breakout room with another student and they really need help troubleshooting. They can just go into this AI assistant and it’s going to give them a hint about where they went wrong and explain why that code is wrong. It also allows them to learn multiple languages. So this co-pilot can actually translate their code from let’s say python into another language such as Java. So, if they really want to get more and more technical skills underneath their belts this is a great AI tool for them. Also, I encourage my students to always do self-learning. I know that people probably have heard of Khan Academy course things like data camp I’m going to highlight data camp where they have an AI assistant right inside of the e learning platform. So anytime the students get an answer wrong or they need a hint or they need the entire solution, they can just click on that AI assistant and type what the issue is and it’s going to help them solve that problem and also of course ChatGPT that’s probably the most popular AI tool that educators are familiar with. I do not allow them to regurgitate ChatGPT onto assessments. But I do encourage them when they need to get started to like brainstorm creative solutions to problems or to summarize a long research paper to enhance learning comprehension or reading comprehension or to get some quick facts about a package that they want to use in their project. Also Canva they have AI to help them generate slide decks to give them their own template of a slide deck and they also have AI tools to help generate images right. And so, they can actually enhance their creativity by generating a template that suits their personality suits the business problem as well as generating images that also represents them so these are some of many ways that I leverage that AI in the classroom Olivia.
Olivia
Yeah, and Jelly that you mentioned so many great platforms and just and models and I think you know, especially thinking about the learners that you have it makes sense. You know that these tools are out there. They should be practicing with them and of course we’ve all heard of ChatGPT but thinking about getting into more specifics that you know. Different models that are best for each industry and kind of allowing students to explore how to leverage these models in different ways is awesome and I do want to mention I love that you are calling out some of those free tools for any educators listening out there definitely thinking about which tools your students can have access to and kind of leverage that unique role of a student. I know personally when students find out that they’re kind of elite and have that exclusive access. It gets them really excited. So definitely finding as an educator, I think finding those resources for students is a really important role that you play because it just you know allows them. The internet has a lot out there, right? So, being able to point students in the right direction really can help and as you said help them really find the tools to enhance their creativity, I love that that framing of how you’re really encouraging them to use those tools.
Angelica
Yeah, definitely and yeah, Copilot is definitely one that is great for students accounts that are working in Computer Science. But there are free tools out here or at least free trials. Not all the tools that I mentioned are free. But they will give you some exposure that you can try it out as an educator and if you have to go and get the budget for it or do fundraising for it. Just so that you can utilize it in your classroom I will say that that is a great thing to do.
Olivia
Yeah, that’s awesome and you kind of touched on this before but I’m curious you know how do you approach your students’ use of AI for assignments. I know you mentioned not for tests but as data scientists do they think they have a different perception of AI is compared to students in other fields.
Angelica
Yeah, I think most definitely! So it’s important for me that before I introduce them to AI in all of these tools and technologies that I’ve mentioned for them to understand the bias in it right? AI is not going to always give them the right answer. But if they have the ability to critically think and see right from wrong and not just take AI at face value, do their own investigation become like their own investigators they can leverage AI in a more responsible way. So definitely what my data science students they really understand that just because it’s generating that code doesn’t mean that’s the correct code for your business problem. So how can you go through the output line by line to see if that’s what the output actually should be, so they are able to do a lot of Qa and Qc on generated output which I always encourage them. I tell them to think of these tools like Wikipedia. I know a lot of teachers at least when I was going through school was like hey Wikipedia or ask Jeeves I think ask Jes was like a long time ago when yes, there right?
Olivia
And we might be dating ourselves by saying we know ask me.
Angelica
Ah is fine to start there but it may not be the sound of truth so get other resources cite other sources and that’s the same thing with my students. I don’t want them to just cite AI right. I want them to be able to use it as a starting point like Wikipedia and then try to do research on other sources to make sure that is the expected code or the expected facts that they hope to give and to hope to get out of these tools. So yes, they definitely understand the responsible use of AI but like I said AI is great like some of my students hate reading is’s going to help them paraphrase white papers some of my students do not like given presentations so AI is going to give them presentation pointers. Some AI tools out here. You can upload a video Olivia and it’s going to tell you how to enhance that video right? Some of them get writers block and they don’t even know where to start so AI is going to help them brainstorm activities as far as logos or slogans that they can use in their business and yeah, they are very much aware of the inaccuracies of these tools yet. They are still can be considered enhancements to kind of get them at least starting the race and then they can finish the race themselves.
Olivia
Yeah, that’s great and I think you know just keeping in mind that you know students want to think about what’s actually best for them just because it’s there doesn’t mean that it should be. You know what they’re taking and you know AI is great. But then if you can’t perform the test or do the actual presentation or whatever it is you might realize, Hey I didn’t really learn this. I didn’t really check my sources. This isn’t actually accurate. So, I think this is you know in the classroom is the place for that failure for students to try it out realize that hey I got called out I didn’t actually you know check another source this doesn’t work but this is the place to learn that and if they’re not. Being encouraged to try these new platforms. They’re not going to learn how to use them responsibly as you’re kind of mentioning so I definitely thought that’s great that you’re finding a space for that and even just yeah, giving the space for that type of thing.
Angelica
Yeah, totally.
Olivia
I wanted to you know, think a little bit of that we talked about the data bias that you’re mentioning, and you know I’m sure your students as data scientists understand a little bit more about the intricacies there where you know all these models of Generative AI have been you know trained on existing data and that data is not always. I should just say is bias right? It’s biased already and so thinking at that I’m curious your thoughts on why it might be crucial to have data scientists from diverse backgrounds working in the field particularly considering all of these issues of bias.
Angelica
Yeah, and just like you mentioned Olivia, a lot of the data that we have is bias or the bias exists in the data. We don’t have right? So, we’re trying to apply some of these data science solutions or models. Groups that we don’t even have the data for, right? So, if I’m trying to do a model to predict policing and this is a very popular algorithm that came out in 2016 called the compass algorithm that was. Unfairly putting African American males behind bars at a higher rate because the algorithm wasn’t even trained on African American males to begin with so it could in decipher in an image. An African American male from actually the nighttime lighting so it had something to do with lighting as well and so these are things that we don’t want to fall victim to as data scientists and so to help with that. Maybe having a diverse member on the team that comes from the African American culture that can say hey you know our skin tones don’t show up well at night. So, if we want to use these algorithms and police cameras Can we make sure that we have different types of lighting different types of filters that we have more data of African American images so that we’re not falsely putting people behind bars right? So those are that’s an example of an algorithm that can really cause a negative societal impact all because there was bias and there wasn’t enough diversity there to recognize the bias. So having diverse teams that can really look at the data understand what’s in the data was not in the data because sometimes what we don’t capture is just as important as what we do and really go through actually labelling the data as well. And this is coming from people not just from diverse race ethnicity gender or sex but also with age right? I know that a lot of now analysis that we do with social media. We’re trying to do sentiment analysis. How do people feel about. The new iphone that’s coming out. Do they feel good? Do they feel bad, and I don’t know about you Olivia, I don’t understand a lot of the acronyms in some of the younger generations I don’t understand those emojis right? So, I may miss label sentiment as bad when it was actually good right? So even that type of diversity from all age groups is very important right? So, we want to make sure that we are examining the data sets that we have this diverse team that’s able to look at Bias coming in as well as bias coming out so that we don’t make wrong business decisions or worse negatively impact the society where we harm somebody’s health or we take somebody’s life away from them early just because an algorithm told us to do it.
Olivia
Yeah I really appreciate that and I appreciate your emphasis on it’s the data that we might not have right? if it’s just you know lacking because I think that when people think well you know we’ve done this study. We’ve done a lot of you know we’ve pulled all this data. Maybe the numbers right? We have thousands of datasets. Whatever samples, whatever it is but when the team is not thinking about the diversity of the samples and making sure that it’s ranging and getting data from different types of people or wherever it is. You’re right that, that data is not accurate. You know the quantity doesn’t matter when the quality is not quite there and I really love you bringing up the sentiment analysis, the age because you’re right I mean emotions are notoriously difficult right? We’re all working on understanding our own emotions and then when you get into this sentiment analysis here. I mean Sarcasm is notoriously also difficult to understand and so when emojis are the thing that changes the entire message. Well, we need to make sure we have people on data teams who understand that and can kind of tag things correctly. So that we do get an accurate um, understanding because as we Know so much is happening on the internet right now politically and just kind of understanding the state of American Affairs and I think it’s important that we have a grasp on that for you know to make sure that we have a society that we want to live in so I definitely appreciate you mentioning that.
Angelica
Yeah, absolutely I think as educators as well if you gave your students end of semester surveys and only the top students who got A’s in your class responded to the survey, are you really capturing the complete picture of areas of improvement. So, because the other students who may have gotten A, C, A, D or an F didn’t even respond right? So, there’s no response bias as well as response bias and how do we actually build relationships so that we can get more responses because data can help us solve so many issues out here. We just need to have the data in the first place and having diverse teams to kind of gather that data is very important.
Olivia
Right? And I won’t pick your brain too much on this. But I even think about like you know with a restaurant review, right? like isn’t is it all just 5 stars and 1 star because people who are choosing to fill out surveys are either really happy or really upset. You know we want to make sure that we’re getting data samples that are actually representative of the average experience of a whoever is being interviewed.
Angelica
Olivia, you actually use like my first example that I always tell my students when I’m talking about bias. I ask them raise your hand, how many of you have filled out a restaurant survey or a fast food receipt survey in the last three months and literally no student raises their hand.
Olivia
Really.
Angelica
Right? Because most of the time the people who fill out their surveys are either really really mad or they’re really really happy. There is no in between right? So Olivia, you’re already thinking like a data scientist. I need to have you on my team.
Olivia
Well thank you? Maybe that’s going to be my next pivot there. I’ll follow you in the pursuit of biology into data science. Well I definitely you know, thinking about all of this and I guess my next question would just be thinking about you know you have these students who are working in the data science fields and they’re actually looking at data and they’re also using AI and how do you then assess and evaluate the learning outcomes of those students and how do you think about? You know how they’re going to be prepared for this world where AI and data, I should say the data they’re producing is going to be used to train more AI, you know gender AI models. How do you actually assess and evaluate the learning outcomes of your students.
Angelica
Yeah, and I think especially recently I would say at least over the past five years if not decade with a controversy around standardized testing and assessments right? Do tests and do standardized testing really show a student’s knowledge of a subject right and it’s very controversial. Some people may say yes, some people may say no, but we all can agree that assessments aren’t the only way that we can assess our students right? So even though I don’t allow them to use AI generated content on their graded assessments, I’m still able to assess certain soft skills for instance if they do cite the source in their projects and cite other sources, I can get a feel for their integrity right? Which is really great when you want to go into the work field how and how much integrity does a candidate have. I also can assess their critical thinking. So are they actually just on projects regurgitating a code and the cult was wrong or do they have the critical thinking ability to really fact, check it and understand that hey this was really. A bias result or an inaccurate result and challenge it right outside of that I can assess their creativity. So are they using these tools to enhance their creativity. Do I see slide decks that have bells and whistles in it. Do I see where they have generated image through text that showing that that fits the business problem. Are they kind of thinking outside the box as well as peer feedback. So, my students do group projects as well as solo projects. So, us having their peers give them feedback about how great of a team player they are, how much they came on time to meetings, how approachable they were if they were able to explain technical things to a non technical audience. All of those are great soft skills in assessments that’s going to benefit them when they go out into the workplace or even college or even their next journey in life. So, I don’t just rely on graded assessments, I rely on them as a person and as a student right. Do they have that integrity, do they have that creativity, do they have that critical thinking, do people want to work with them and then that’s how I go about assessing them overall as a student.
Olivia
Yeah, well I got to say I hope some educators are listening because this is exactly where you know I think it’s this is one of the big conversations happening is how do we actually assess students and you know; however, people feel about actual tests themselves. It’s really what are we actually assessing students for and as you mentioned creativity is huge but then also how well do you work with others and they think are you actually kind of having that ethical mindset and being able to kind of be credible there and I think that just given you know ethical and being an ethical person is of benefit to in any job in any role and I think thinking about even those interview questions that your students are going to end up facing they need to be prepared and so what better way you know there’s nothing worse than a classroom where students learn all these things than they’re like wait a second I’m not ready for the world. They’re not you know. This isn’t I. My skills are half baked and I need to kind of get that next level ups that I can actually get a job and so it’s great that your learning outcomes are really kind of skills that employers out there are looking for.
Angelica
Yeah, most definitely and having that one on one feedback sessions with students that I do at least every three weeks to kind of say this is your progress but this is also your peer feedback and how you can react to that feedback. And use it to be better in the next upcoming three weeks is very beneficial because at the end of the day AI is going to be an assistive technology in a lot of these industries at least for data science roles, right? So, I can teach you the technical, but I can’t teach you how to be a great person. Right? You have to really embody I want to be a great person, I want to have integrity, I want to enhance these soft skills that’s going to put me in a great position when I go into the workplace.
Olivia
Yeah, well you know you are someone who’s in the workplace and so I’m curious how you’ve been staying updated with all these advancements in technology and AI and thinking about you know as you mentioned how to assess your students. How do you actually continually enhance your teaching methods.
Angelica
I’m good. I’m not going to lie to Olivia it seems like every day there’s like 15 new AI tools coming out like you can think of this as back as the space race when we teach students like the space race if you know the United States can get to the moon quicker than any other country right.
And so, this is what’s going on in AI right now I will say over the last month, we’ve probably have had a hundred new AI tools just come on the scene. So, it’s no possible way that you will have time to go through each tool. However, I do love these virtual conferences now I know cover it was a detriment but one of the benefits that it had to the workplace as far as professional development is concerned that a lot more industry leaders are hosting virtual conferences. So, I just attended an open data science conference that also had an on sites location in New York snorkel. I just had a conference this month data camp just had an AI conference. Microsoft just had a VS code day as well as cascade and databricks and this is all within two months that I’ve attended all these conferences.
Olivia
Wow.
Angelica
And these conferences, learning from panels that have been using AI in the workforce how to create business strategies, how to help my students what project management all of those have been free and very beneficial conferences that you can just look up right on LinkedIn. I want to find a conference about X, Y and Z and they’re going to pull up right as well as yet is awesome online educational platforms like I said I’ve always kept up to date with the new Google analytics certification that just got launched a couple of months ago which I had a hand in creating so really shout out to the Google Analytics certifications on that as well as data camp khan academy of course. All of these online educational platforms are putting out courses for prompt engineering. Generative AI one, to really give exposure to people that may not even have a tech role but they have a non tech role because we want them to become data literate at least and then eventually AI fluent. Because that is just where the industry is going to go no matter what role that you’re in so those educational platforms are always great and of course following industry leaders on social media platforms right? Use it for the benefit I’m definitely a LinkedIn person times ten so making sure that soon as somebody posts a nice little gym, I put that little gym in my pocket and make sure I share it with others as well.
Olivia
Yeah that’s great and I think you know for some educators out there LinkedIn is not always the most. There aren’t as many educators active on there. I don’t want to say a blanket statement than none are but I think that there are you know as you are, people in the industry who are on LinkedIn a bit more. There are just you find out about great events in that way and some of I’m sure some of those conferences were you know posted about there and just finding I think the right voices that are kind of aligned. You know there’s a million people out there and there’s a lot of people of great content but finding just 1 or 2 that are kind of posting content and tapped into the networks and Conferences and things that you want to attend is a great way to kind of stay abreast there as well saying that. That’s great to push for that and definitely these virtual conferences. I mean even I just attended the ISTE conference in Philadelphia and got to see just so much new educational technology and so. You know, even being there in person it couldn’t get you know see all of it and so I think just attending them as often as one can and especially the sessions with industry leaders can be super super helpful. So, it’s great to hear that and I would just love to think a little bit you know, clearly you are supporting in your own classroom students exploring AI and at all of the applications of it that are out there and in your opinion what measures can schools and educational institutions take to guarantee this variability fair availability of AI and digital learning resources to all their students.
Angelica
Yeah, I love this question because accessibility is very important, and the reality is that not everyone has access to be able to use these tools and technology right. So, it starts with us empowering our teachers. So just as I’ve mentioned yes, there are free resources out here as far as when you go online, you’re looking for AI tools. But then if you don’t have the computers in your classroom is not going to be a benefit right. So really empowering teachers through our state budgets, right? And making sure that we have what we need in the classroom to really make sure that students are not falling behind on this technology innovative. Demand that’s coming out like digital transformation and digital technologies are going to be used a lot more in college as well as in the workplace. So really, really preparing the teachers even through conferences as we just mentioned right the conference costs. How are you allowing your teachers to take their professional development days or their teacher workdays and actually learn some of the Generative AI technologies that are out here that they can incorporate in their classrooms. So, I think it starts all with preparing the teachers and then providing access.
Yes, if there is not a way where you can actually blame computers into the class then showing a presentation about generative AI, a video there coding with paper, going on a low cost field trip to one of the tech industries out here that it would be great to have you to teach students about AI. All of these are good things that we can utilize and then if we do have those tools in the classroom now how can we innovate the curriculum right, how can we add this AI knowledge, and I will just give a shot out real quickly to upshots. Upshots is an online platform for gamification, so it helps students learn any content by giving them badges and leaderboards and things of that. There’s also lesson plans dot AI as well as education copilot.com and auto classmate and all of those help teachers develop AI generated lesson plans that can increase engagement across any subject that you’re teaching. And they are honestly great AI tools that the teachers can use even if the students can’t use AI quite yet.
Olivia
Yeah, and I love that you’re pushing for that teacher education and access as well because I think just hearing speaking to you, you know you are staying up to date with all these tools and I think that you know there’s no one that their students are going to trust more than their own teacher and so when they are not sure as we said you know the internet can be an overwhelming place. There are so many tools out there but having a teacher that encourages them to just try 1 or 2 and can say you know I’ve done this I can tell you how this one works just allowing students to have a little bit of guidance with at least the first couple tools that they try I think can really help students then take that leap and the next step and actually try the tools out on their own. But I do think you’re right? It starts with the teachers. Yeah well, I think you know Angelica it’s been so lovely talking to you and that it’s going to bring us to the end of today’s episode.
Angelica
Yeah, definitely agree.
Olivia
I do really want to extend a heartfelt thank you to our incredible guest Jelly Angelica for joining us and sharing her in valuable insights on our podcast today. Her expertise and passion have truly enriched our discussion and I really hope you, our listeners have gained as much knowledge and inspiration as we have here. We encourage you to express your thoughts and share your experiences about today’s topic and social media using the #MagicEducatorInsights. Thanks for joining us. Remember to tune in next time for more engaging conversations on Educator Insights. Until then, stay curious and keep exploring bye.
Angelica
Bye.
