Alright, thank you for going on my poster. This is research that we are going to be presenting at the American Psychological Association, APA. The team is great practice for me to present the date. The title of this research is concerned and successfully learn at home or any environment predicts math performance. This work was done by me. Remember that? I'm a student at University of our second year PhD student. Dr. Allison's angle ascii to that University of California, Davis. So you only see them McGill University Dr. carried out here at University of Delaware. Dr. Tim my Rutherford, who is also here at University of Delaware. This research was a part of the workshop de facto Rutherford put on over the summer, last summer. And it was all about data mining and playing with this large datasets such as that I'm going to talk a little bit about. So a little bit of introduction. So we know that students benefit from engaging in math at home, particular when caregivers are all. But we know that they benefit from engaging with math at home. So with remote learning, the pandemic and eyes of the merits of remote learning are in question. We don't know if kids are actually learning from this. We're hoping there. So we need more research to know best is actually happening. We do know that there are differences in performance based on motivation. So a student who was more motivated, it's going to perform a little bit better than this, isn't motivated. And we predict that this might moderate relationship for how they play at home or in the lab. So more motivated student might do better at home. I'll ask, motivated student might do better or at school. And it's important. All of this data comes from before the pandemic. We do have information about their motivation and we have information about their play environment, whether at home or in school. So pre-pandemic data might be very insightful for us. Especially these make decisions concerning whether or not we should move kits online, who should go online? Things like bath. So a couple of the questions that we asked, we asked him in particular, the students by hand by Ram, predict their performance in math game. And then we ask those student motivation moderate this relationship. So these questions can be seen in this model that I drew down here. So here we can see environment, but they're not, they play at home or in school, predicts performance, whether or not they pass or they fail level. And we want to see of motivation affects this relationship. And here we define motivation as expectancy, value and importance. Desire measures. I don't remember what year it is by Jacqueline Act was defined to expectancy value theory. Some of you may have heard it. This is the theory that we're basing our motivation contracts off of our dataset. So I guess that Dr. Rutherford works with a educational technology company called spatial temporal map. They've already game data and motivation that surveys from 3 thousand nurses, seventh graders, they brought it all AMR, we add a third betters, fifth graders, fourth graders, kindergartners. We had a whole slew, but we haven't narrow it down a little bit. So we chose fifth graders in particular. After cleaning bird, the missing data and a couple of inclusion criteria like students who didn't play at all I home or studio and play all at school, which I don't they actually happened. We ended up with 1600s, 66 fifth graders. And the missing data, some students might not have done the motivation survey. And we need the motivation for this. So that's how we know the population. So here on the left you can see I to emoji. And this is how students responded to motivation questions. So here it's asking. I'm going to zoom in, how well do you think you'll do on math this year? And they have it on a Likert scales so they can go strongly disagree. To strongly agree. There's five points, so there's neutral, 0.1 of them. Like cool things about this is if they pick strongly disagree with this, meter gets really sad. If they've been strongly agree, it's a super-duper, happy. That's a fundamental elements to make the motivation survey one more easier to understand for their kids might struggle with reading, but also just fun. So on the right you can see a picture of the math game. What's important is there is no tax scheme. So they're trying to teach math skills to students using just visual aids. It's really cool. It's a lot of fine. If you want to see examples of it, I'm worth the watch me show you. So what's important is these for theaters were nested. That was nested data. They were nasa, but then 146 teachers and they average 155.7 levels of worry here. So why does that matter? Well, if a student is with a teacher who's like play a lot at home, and then another student's eye, you can't play at home at all. We need to control for that. So knowing that, that it is nested, It's going to help us with our methods later on. So a little bit about our methods. So we use that brings up a logistic regression or hierarchical an AR model. With level pleases our first level, students as our second level and teachers, that's a third tuple. And we wanted to predict whether or not a student pass the level on their environment, whether they played at home or in school. Their motivation, so expectancy, value and importance. And then we also controlled for dimeric perhaps variables like ethnicity, English-language learners status, special education status, and gender. And then we also tested for our second question. We tested a second model. We added motivation as a moderator and not just a protector. So yeah, a little bit of early results. I'm going to zoom in so you can see it's graphs a little bit easier. So the motivation and environment interaction did not improve model fit. That wasn't significant, so it doesn't moderate relationship. But we did find some really cool stuff. So on the right here you can see I'm going to explain the graphs lot about the red color is when they felt the bluish green colors when they pass level. So this first one that is the environment on the left, his home and Underwriters Lab. Now, these are proportions. So here we can see that students are playing at home. 0, 0, 0, I less than students were playing in the lab. But we can also see here on the pass and fail spectra or scale, that students are passing more than they're failing at home or in the lab, it's about an even split. So we actually found a significant finding here and our logistic hair on your model. So students are more likely to pass a level when they're playing at home. Here we can see the expectancy motivation variable. And as their motivation goes out, their likelihood of passing goes up to. So really love here, really higher. But what's important to notice here and kind of a limitation of our study is clearly we had a ceiling effect. Students really to know rate low on the meditation surveys and HCI importance as well. But presumably we had a 1 times 6, presumably somewhere. And those lower numbers, there's maybe a 100, maybe a couple of 100. So the significance is so relevant. So students with higher math expectancy or mercury to pass, students with higher math importance are also, are, are, are less likely to pass. So here, this is where I start to go hazy on light the interpretation, because these are really small numbers. But we did get a significant finding. So it is important to talk about. So here you can kind of see the fill rates are a lot higher for these low important students. So overall, while we know is that these students were more likely to pass level, it's at home them in school. The students being fifth graders, that point estimate. We don't know entirely if the teachers right, burying, encouraging a plane levels at home. It could be the case that teacher said, I only pay levels you've already played at home. But it also varies between all 146 teachers. But presumably with our model, we did capture some of that. What's important, and I mentioned this a little bit and then traction is potentially caregiver involvement may improve their math learning at home. So if a student is struggling, they might build, ask mom or dad like, Hey, can you help me out? Whereas at school, they may not have that all the time. A teacher has 20 gates, so they're a little bit stretched. So, but one of the important things to consider here, and we didn't find any demographic significance. But if caregiver involvement is promoting math learning, we do have to remember that not every student has access to a caregiver that can give them help as needed. And then finally, an important alternative consideration is that school environments may just be distracting and maybe hinder performance. There's my cat. So students might be distracted in school, which might make them feel a little bit more. And that's about it. So thank you for coming to my closer. Sorry, I might play with them. And you guys have a good one.
Delaware Day Poster Session: Raymond Patt, Allison Zengilowski, So Yeon Lee, Kerry Duck and Teomara Rutherford
From Pauline Himics February 28, 2022
7 plays
7
0 comments
0
You unliked the media.