HyeJin Hwang
To the Classroom: Episode 11
May 2, 2023
Jennifer Serravallo:
Today I welcome Dr. HyeJin Hwang to talk about her recently published study on the bidirectional relationship of knowledge and comprehension in monolingual and multilingual readers. In this conversation we'll talk about different types of knowledge, ways to support knowledge building the use of informational and narrative texts in the classroom, and frameworks that can help us think about comprehension instruction. Later. I'm joined by my colleague Gina Dignon, to talk about practical takeaways. This is the To the Classroom podcast. And I'm your host, Jennifer Serravallo. Welcome. HyeJin, thank you so much for joining me.
HyeJin Hwang:
Thanks so much for having me today.
Jennifer Serravallo:
So I think you wanted to start by introducing the co-authors of the paper that we're going to be talking about today.
HyeJin Hwang:
Yeah. I enjoyed conducting this study with Christian McMaster who is a professor and chair at the Department of Educational Psychology at the University of Minnesota and Panya Kendall at the same department at the same university.
Jennifer Serravallo:
Wonderful. So let's start by talking about different types of knowledge that are important. When thinking about reading comprehension. You write a little bit about domain knowledge, general knowledge, topic knowledge, and just so everybody's clear that's listening, can you just define them and give an example?
HyeJin Hwang:
Well, yes, reading researchers have operationalized prior knowledge in varying ways, but in our study topic, knowledge was defined as reader's knowledge about the topic of text being read. So for example, when readers are reading about text about dragonfly, then reader's topic knowledge is knowledge about dragonfly and domain knowledge is broader broad knowledge related to a field of study as opposed to a specific topic. And domain knowledge can be characterized as prior knoweldge on multiple topics in one field. For example, it could be science or social studies in K12 settings and general or general academic knowledge is the broadest type of prior knowledge. It can be defined as prior knowledge related to multiple domains or fields. And general knowledge can be assessed by asking students information about multiple topics across different domains.
Jennifer Serravallo:
Thank you. I think that's so helpful to understand the differences or the different kinds of knowledge for teachers who are interested in building knowledge and supporting students knowledge building. You also write about some different models that are important to know. When we're thinking about comprehension, I think a lot of people are familiar with Simple View of Reading and Scarborough's Reading Rope, but these are more comprehension based models. So one of them is the Kintch's Construction Integration Model (it's a bit of a mouthful) of text comprehension. Can you explain what it is and what role domain knowledge plays in that model?
HyeJin Hwang:
Great question. So Kintch's Construction Integration Model is to explain processes of reading comprehension. The model shows why prior knowledge including domain knowledge is so important for reading completion. And it says there are two phases of reading comprehension and the first phase is construction phase. In this phase, construction phase is where readers need to literally understand what version propositions in text means and prior knowledge can facilitate the construction phase. For example, let's say there are two sentences, dolphins have lungs, they rise to the surface to breathe. Those readers who have prior knowledge that dolphins live in the ocean, right? If they know this knowledge, they can extract meaning from the two sentences because the readers already know that they, in the second sentence, they rise to the surface to breathe refers to dolphins so they can easily construct meaning. And now second phase it's called integration phase.
In the integration phase, readers need to integrate the extracted meaning from texts with their prior knowledge to store the expected meaning in our long-term memory. And this process is critical for readers to learn from texts and gain new knowledge. Back to the example, the dolphin example, readers who have prior knowledge that many animals need oxygen to survive and the air has oxygen in it, are likely to be successful in integrating prior knowledge with the extracted meaning. And when they are more likely to make a good inference when they are asked a question like "what happens to dolphins, they cannot come to the ocean surface?" They can make an inference that dolphins will not survive because they cannot have oxygen in the air. They have lungs, so they need oxygen in the air.
Jennifer Serravallo:
When you give an example like that, it just shows how complex and nuanced, right?
HyeJin Hwang:
Yeah.
Jennifer Serravallo:
...and kind of elegant what's going on in readers brains to help them understand attacks. Thank you so much for that. You also talked a little bit about another model called the Lattice Model. What's that? How does it differ?
HyeJin Hwang:
Yeah, yeah, you're correct. In addition to Kintch's model, the model was informed by Connor's Lattice model, and the Lattice model describes the reciprocal relationship between prior knowledge and reading and it provides a broader picture of reading development by including a wide range of factors to explain reading development. Specifically, it explains that the reciprocal relation would be influenced by all the literacy factors such as decoding and language factors like vocabulary and cognitive factors like cognitive flexibility.
Jennifer Serravallo:
So another model, Simple View of Reading,--decoding times language comprehension, which you just talked a little bit about language comprehension--decoding times language comprehension equals reading comprehension. The conclusion that some people draw when we look at that model is that we need to teach kids to decode the words and we need to teach kids to develop knowledge by reading aloud to them and then they're going to be able to comprehend the text they read themselves. Now with your findings in this recent study in mind, what would you say to that interpretation of comprehension or that interpretation of the Simple View of Reading?
HyeJin Hwang:
Just to remind you, the study was based on Kintch's Construction Integration Model. Yep. And Connor's Lattice model rather than the Simple View of Reading because our study is more aligned with the two models. But when students decode words well then they can read more fluently, they're more likely to read text fluently. And also when they have strong domain knowledge, it can facilitate language comprehension and then it is more likely that they will improve reading comprehension. I think it's reasonable prediction to some extent, but you are likely to seek differences among students. Let's suppose that there are students who have similar decoding skills and similar domain knowledge, but their reading compression can be different because some students are better at using--using--what they know while other students are not good at using what they know. Or it is also possible that some student are more actively engaging in reading while others are less likely. So what I'm trying to say is that other factors like reading strategies, inference making and reading motivation matter too in addition to decoding, domain knowledge, and language proficiency.
Jennifer Serravallo:
I totally agree with you and I know that you, like me, love and respect the Active View of Reading Model as a more, again, complex nuanced model to explain some of the different processes that are going on in the reading brain, especially as it relates to comprehension. So let's get into your study a little bit more. In your study you examined directionality and magnitude of relation between domain knowledge and reading separately for bilingual, multilingual and monolingual students, bilingual or multilingual and monolingual students. Can you talk about why it was important to look at those language differences? Because I think it was such an important, but I don't see it in every study, right? Important but kind of unique contribution. So why was it important to look at language differences?
HyeJin Hwang:
Well, it was important to look at the difference in the relation between domain knowledge and reading between multilingual students and monolingual students because there were two possibilities. So the first possibility was that some scholars have argued that the relation would be stronger in monolingual students than multilingual students because monolingual students often have more exposure to English and therefore they are more likely to have better English language proficiency. And this can facilitate using domain knowledge to make sense of text. And alternatively the second possibility. Some other scholars have argued that the second possibility is that the relation would be stronger in multilingual students than monolingual students because domain knowledge can play a compensating role for multilingual students' English language proficiency. And it turned out that our finding was that there was no difference in the reciprocal relation between domain knowledge and meeting between multilingual and monolingual students. So that means the magnitude of the positive relation between prior knowledge with later reading was similar between the two at the same time. Also, the magnitude of positive relation between prior reading with later domain knowledge was also similar between the two.
Jennifer Serravallo:
Thank you. That's a really helpful explanation. So let's get a little bit more into some other implications and ideas for strategy instruction from your study. What if a reader doesn't have a lot of knowledge related to the text that they're reading? Can they learn from the text? And if so, what do they need to be able to be ready to do? So I'm approaching a text. I don't know anything about dolphins. I don't know where they live. I've never heard about dolphins before. I'm reading this book, how do I learn from it?
HyeJin Hwang:
Well, what answer for your first question, can they learn from text would be, it depends. So we talked about the importance of having strong domain knowledge in reading development, right? Suppose students have strong science domain knowledge and they know multiple science topics. For example, they know many different science topics like animals, like birds, and insects, and they also know the water cycle, food chains, and weather. However, they have not learned specifically about rainforest ecosystem and they are going to read about a text about rainforest ecosystem. And for these students, they can use their domain knowledge about multiple science topics to understand text about rainforest ecosystem even though it is the first time they read about this topic. So for example, they know that they are, there are different types of weather and different, they also know that different places have different weather and animals and then they can leverage their knowledge during meeting to learn about weather in rainforest and how animals adapt to the rainforest ecosystem to survive and protect their babies.
Having strong domain knowledge can help students to understand text and learn from text even though they don't have topic knowledge specific to a text. However, that means for those students who don't know well about a topic of text and they have weak domain knowledge, learning from text would be more difficult than those students who have strong domain knowledge but who just do not know well about a topic specific to a text. And to answer your question for your second question, skilled readers would use reading strategies to comprehend text and learn from text, especially when they notice that they do not know well about a topic of text they are going to read. So students need to be taught to think about things they already know that are related to a topic of text. Also, they can ask questions they would like to know before or during reading and attempt to answer the question questions during reading.
What seems to be particularly important in this context is monitoring comprehension because students are reading about a topic they did not know well. So students need to pay attention to whether they understand what they are reading and they found that they did not understand well. Then for example, they can reread or they can ask questions to others and the skilled readers would use reading strategies to make sense of text. But I also like to remind you of the reciprocal model in the paper. So according to the reciprocal model becoming skilled readers are predicted by prior knowledge. So both reading strategies and domain knowledge matter.
Jennifer Serravallo:
You talked a little bit about strategies, what other strategies might be important even outside of that one example to support the activation of knowledge before I'm reading, to connect relevant knowledge while I'm reading, to learn from the knowledge in the text during and after the reading is done. What other strategies do you feel like are important when we're thinking about knowledge building?
HyeJin Hwang:
Skilled readers can learn better from text and they're more likely to use reading strategies as, for example, they could monitor their thinking processes and they're more likely to summarize what they're reading and also good readers think about what their goal of reading is. They actively think about what they know and use their knowledge to comprehend text and reading strategies. It's not a question that my study can directly answer, but skilled readers, for many skilled readers knowledge, knowledge activation can occur automatically or effortlessly, but not all students can access relevant prior knowledge automatically or effortlessly. So we need to support students to activate what they know to make sense of text.
Jennifer Serravallo:
Yeah, that makes sense. So let's talk about some implications for practice from these findings. So do we think from what you found teachers should try to frontload information before students read? So give them some background domain knowledge, topic specific knowledge, what have you, before they read, should we create text sets that are conceptually related? Should we do both of those things or something else?
HyeJin Hwang:
The major finding is that domain knowledge and reading appear to be mutually each other from all the grades and throughout the elementary years in multilingual students, in monolingual students. Then the implication is that students need instructional support for developing both domain knowledge in reading from the beginning of schooling. Currently instructional time for science and social studies is often much less than the instructional time for English Language Arts. So more science and social studies instruction needs to be provided to elementary students. And another way to support both domain knowledge and reading is to integrate content area and English Language Arts instruction. In my other study with Sonya Cabell and Rachel Joiner at Florida State University, I investigated the overall effect of integrated literacy and content area instruction on comprehension, vocabulary, and content knowledge by synthesizing visuals from different relevant research. We found that the integrated instruction can support students' comprehension as well as we found that students learn better vocabulary words and content information taught during the intervention.
And we also reviewed characteristics of integrated literacy and content area instruction. We found that in integrated literacy in content area instruction, students engage in reading and writing for the purpose of learning more information about the natural and social words. Students practice using reading strategies in the context of learning more information and teachers select texts that are connected to one another around science and social studies topics and also vocabulary words are told in relation to the content being told. So when students are learning about plants, vocabulary words told for example include roots, stem leaves, pollen and also we found some studies used concept map to show how different words and ideas are connected to one another.
Jennifer Serravallo:
Yeah, so I'm thinking there's not a wall between ELA time and science time and social studies time that although we're learning content in science and social studies, we're also bringing strategies that we teach kids during ELA time, like asking questions, checking for understanding into those times and probably think about ways to bring content into the English Language Arts classroom as well. Last question, much of your research focuses on domain knowledge related to science and social studies, but I'm wondering if you studied or are aware of studies that look at the kind of knowledge that's most helpful when children read fictional narratives. So beyond, for example, being able to visualize a setting because they're familiar with the topic knowledge of that setting, like a story that's set in the zoo. And I've been to a zoo before so I know what a zoo is like so I can visualize the zoo. More what I mean: is there knowledge about people, the kinds of things I need to know about people that I bring to reading a story or the kinds of things I know about story structure and how stories tend to go and that kind of knowledge helps me to read a fictional narrative. Can you talk a little bit about that?
HyeJin Hwang:
Yeah, and also I have to start my answer with this is also not a question that the study can directly answer, but comprehending texts involve different kind of knowledge. So what leads to have general knowledge about narratives such as characters, plots and themes to successfully comprehend narratives, text and genre knowledge should be explicitly taught to students. At the same time, domain knoweldge still plays an important role in campaigning narrative text. For example, the setting of narratives is a mountain and the narrator explains different plants and rocks where in the mountain where the characters are then skilled readers would use prior science knowledge about plants and rocks to visualize the setting and make sense of the text. My 2019 study published in Learning and Individual Differences, also demonstrated that science domain knowledge can predict comprehension of narrative texts as well as informational text in fourth grade multilingual and monolingual studnets.
Jennifer Serravallo:
Thank you. Well, I'm going to link to the studies you've mentioned in the show notes so people can check them out and read them. I think they're very reader friendly. As a teacher myself, I was able to understand them. So thank you for writing in a way that's easy to comprehend. And thank you so much for your scholarship and for joining me in this conversation today. I've learned a lot. Thank you.
HyeJin Hwang:
Thank you.
Jennifer Serravallo:
Okay. And now I'm joined by my colleague Gina, thank you so much for joining and listening into that conversation. What thoughts do you have about what we could practically do right away in the classroom?
Gina Dignon:
Well, my initial thought, and you asked it at that last question was around the genre knowledge. And I was thinking about what's the ratio of narrative versus non-fiction texts that would be ideal.
Jennifer Serravallo:
I don't want to speak for her, but I think she would say, yes, we need to be including more informational texts. We need to spend more time--that message was pretty clear--in the content areas and we need to be integrating, she said, right, integrating content area with literacy instructions. So to me that says, yeah, we need to be doing more informational text reading.
Gina Dignon:
Well, also what you're saying makes me think about the factors that she referred to when sort of responding to your question about the simple view of reading. So those factors about reading strategies, reading motivation, and there was one other one that how those play a role so that two students could have the same decoding and language comprehension ability, but their reading comprehension would differ for how--oh, I know she said how actively they're reading and I'm just curious about, I feel like these are all the things teachers know and observe in classrooms, but those aren't the things that are considered when testing kids on their reading comprehension. And I just wonder about that because that also made me think about the other aspect that cognitive flexibility that she was talking about. And I think, correct me if I'm wrong, that meant the ability for kids to, or any reader to integrate their knowledge of what they're reading and kind of attach it to whatever knowledge they have they have, how she defined the different kinds of background knowledge. Is it the domain, is it the topic, is it the general academic? So I feel those are all the things that are kind of like the glue that helps kids either really understand, really visualize, or deepen their comprehension as they read. But it's just sort of elusive to measure.
Jennifer Serravallo:
It's, it's elusive to measure. And I think it's also could be really challenging. I'm just thinking as a teacher planning, just those two sentences she talked about with the dolphins where you really sit and think about all the things you need to know to really comprehend those two sentences. On the sentence level, the vocabulary level, that related knowledge level, it's complex. And how do you plan to support the knowledge kids need?
Gina Dignon:
It made me think about, I think there was one line in the study about how kids are receiving the knowledge. Do you know, are they actually reading the text? Are they watching a video? Are they hearing the text? And I feel like those are all implications for how she referred to instructional supports. Those are all modalities that I've seen in classrooms where you can help facilitate some of the knowledge, the background knowledge that kids need or the content knowledge for social studies and science.
Jennifer Serravallo:
Just going off of what she was talking about and the findings in her study, it seems like it would be not the best use of time to just stick a text in a child's hands when they have no general or topic knowledge that's going to help set them up to be successful in that text. And so maybe part of the role of things like videos or hands on experiences or field trips or things that, different modalities to help build knowledge as a beginner. And then as you've progressed through a unit, you start reading more broadly around that topic or topics related to the general knowledge that you're building That seems to align, I think, with what she was saying.
Gina Dignon:
And all the while you could be teaching and using strategies, right? Absolutely. No matter what the input is, right?
Jennifer Serravallo:
Yeah. Right. Viewing a text. Let's talk about what's this part mostly about? Let's talk about what are the key details we learned from this video. Absolutely. It doesn't need to be that reading strategies are only for reading. I think that's
Gina Dignon:
What I'm saying, sort of the type of reading. And I kind of feel like there's so much, so many digital resources that kids are accessing, whether we know they are or not, it's teaching them how to use these strategies can help them in general, not just to gain any knowledge about a topic or a content area. Because if you think about it, kids are consuming so much video, you know what I mean, whether we know they are or not. So teaching the strategies I feel like can help them just help them think about what they're consuming in general. But yeah, I mean this reciprocal relationship she was talking about, I feel like is really important. When you mentioned the text selection for students, not just whole class text selection, but what they're choosing to read, because I just feel like there's a lot out there about should kids be reading texts that they self-select or how much time should that be? But this seems to support wide reading of lots of different kinds of texts. Do you agree with that?
Jennifer Serravallo:
Well, kind of. I feel like it supports wide reading, but within some parameters. So for example, if I'm in a study right now on let's just use ocean animals or different, since she was talking a lot about animals, maybe I have books available for kids to read at the end of the study where they can read about ocean animals, dolphins and whales and tide pools and whatever. And then I also have maybe some other books about land animals because they can apply their knowledge of life cycles or ecosystems or the way that animals have to survive in a particular habitat and how they adapt to those habitats. Cause they learn those things about ocean animals, they can apply them to other studies of animals. But maybe this is not the time to suggest kids go read about the solar system. Right? Right. Because that's such a leap. So I'm widely reading in things that are in some way related. And I love that idea about concept mapping, and I was thinking as part of planning maybe what are all the things that relate in some way to what our study is? Because I think if we're just too focused in studies, then we don't build enough knowledge across a child's experience in school.
If there's a way to connect things a little bit more tangentially yet still guide kids toward texts where they're going to be successful. So that's my "kind of" answer. I feel like it's kind of in the middle. Yeah. Yes. Broadly read, but not too focused and not too broad kind of in the middle.
Gina Dignon:
Yeah, that makes sense. And I did note about the concept mapping because I feel like that has come up in other people's work, the role of concept mapping with ELL learners and with general that best practices in vocabulary learning. So I thought that that was interesting as well. But I am just curious, what question I still have is this idea of reading widely narrative text. So if you're in a genre unit on, I don't know, reading some kind of narrative where does, cause I do feel like there is valuable academic knowledge related to fiction and narrative, but then I don't know if that really aligns with this particular work, this particular study.
But again, and then it just goes back to my initial question about the balance of narrative texts. Because I think there's a lot of this idea of integrating content into ELA is more like I do think I need to preserve some narrative writing because when kids move up the grades, what they're expected to do when they're reading a piece of narrative writing is pretty complex. So they need practice in doing that because they're then asked to analyze text in middle and high school and do literary analysis. And I don't know, I do see how having the other types of domain knowledge, the broad domain knowledge can help with that analysis. But also there is academic knowledge just around literature.
Jennifer Serravallo:
Absolutely. I mean, yeah, there's the genre knowledge, what's different between a mystery and a historical fiction text and how should I set myself up to read? But then think about what it takes for a reader to be able to interpret symbols. There's knowledge around the kinds of symbols of motifs that show up in text and strategies we could teach kids to be able to figure out what they are. Or think about the knowledge around-- I was talking about people, there's different kinds of people, different traits, different motivations, different feelings. The more you're savvy about people in life, the better you're going to be able to interpret the characters and stories. So I think maybe I need to look at a different researcher who looks at narrative specifically and explores some of those questions. But I thought her answer at least around the role of structure, genre knowledge, right there is definitely research support for the importance of those types of knowledge. And then therefore, I would say we need to devote some time to it school to help kids to develop that understanding.
Gina Dignon:
Yeah. No, I agree. Yeah.
Jennifer Serravallo:
All right. Lots of things to think about
Gina Dignon:
Yea, there's a lot to think about for sure.