Skip to main content
Apply

Vice President for Research

Open Main MenuClose Main Menu

Symposium Proceedings

 

Note: Presentations are grouped by the student’s area of research (based on the faculty mentor’s academic department), not the student’s academic major.

 

Poster Session D: 

1:30 - 2:30 pm

Chemical Engineering; Chemistry; Civil and Environmental Engineering; Computer Science; Electrical and Computer Engineering; Engineering Technology; Fire Protection and Safety Engineering Technology; Geology; Industrial Engineering and Management; and Mathematics (40 posters)

 

 

Presentations:

 

D-01     Samantha Stobbe

Research Collaborators:  Ravi Teja Tirumala and Sundaram Bhardwaj Ramakrishna

Research Presentation Title:  Visible Light-Driven C-C Coupling Reaction of Terminal Alkynes at Atmospheric Temperature and Pressure Reaction Conditions using Hybrid Cu2O-Pd Nanostructures

Faculty Research Mentor:  Marimuthu Andiappan, Chemical Engineering

 

The field of nanocatalysis has been largely driven by plasmonic metal nanostructures (PMNs). These metals exhibit plasmonic Mie resonance driven by localized surface plasmonic resonance (LSPR). The materials that largely fall into this category are Au, Ag, Cu, and Al. While significant progress has been made in these materials for a wide range of applications such as photovoltaics, photocatalysis, etc. due to their strong enteric-field enhancement with light interaction. However, PMNs suffer from heat losses, a low lifetime of charge carriers, and scalability issues for high-volume manufacturing. Here, we present an alternative to PMNs with dielectric Mie resonators nanoparticles (MRNPs), which are medium-to-high refractive indexed metal-oxide nanostructures such as Cu2O, CuO, α-Fe2O3, CeO2 TiO2, etc. All these materials exhibit dielectric Mie resonance resulting in both electric and magnetic field enhancements almost of equal magnitude. In this work, we chose Copper (Cu)-based MRNPs namely Cu2O for pharmaceutically relevant coupling reactions. The choice of Cu2O is due to previous work done by Andiappan and coworkers on the same.1 In this work, we hybridize MRNPs (HMRNPs) with catalytically active materials such as palladium (Pd) to form Cu2O-Pd. To test the efficacy of these HMRNPs we used an oxidative homocoupling of phenylacetylene (PA) as the probe reaction. We see a higher rate of enhancement under light compared to dark conditions.2 This idea can be extended to a wide range of industrially important reactions such as CO2 reduction, ammonia synthesis, chemicals, and fuels.

 

D-02     Kaitlin Ashcraft

Research Collaborators:  Mayokun Ayodele and Spencer Pitre

Research Presentation Title:  Chain propagation in the Minisci reaction of Heteroarenes

Faculty Research Mentor:  Spencer Pitre, Chemistry

 

The main goal of this research study is to discredit the use of light on/off experiments to consider if the overall mechanism of photoredox reactions involves a chain propagating component. The common misconception is that in a light on/off experiment, if no conversion is observed during the light off period, there is no chain propagation involved in the mechanism. However, this method does not provide any insight into possible chain reactions, because chain reactions are usually terminated within milliseconds to seconds after the light source has been turned off. The usual method often uses minute timescales that are too slow to provide any insight into chain reactions, since these occur in shorter timescales. Our new method uses an updated “rotating sector method” that leverages modern LED technologies to demonstrate how intermittent light on a shorter timescale of microseconds to seconds can show if a chain reaction occurs. This includes those relevant to typical chain propagating lifetimes, particularly the average lifetime of a chain sequence. In this work, the use of our intermittent illumination method to study chain reactions occurring in the Minisci reaction of heteroarenes will be presented.

 

D-03     Alexandria Bias

Research Collaborators:  Christopher Fennell

Research Presentation Title:  Dynamics of a Spherical Harmonic Water Model

Faculty Research Mentor:  Christopher Fennell, Chemistry

 

A single-point model for molecules is a benefit to computational research because it will allow higher efficiency in simulations and calculations. Using spherical harmonics allows a single-point model to have directionality, which is important in appropriately modeling certain molecules. By targeting only a few variables such as hydrogen bond strength and size, spherical harmonic water approximates the intermolecular interactions found through experimental data very well. In this work, we study both translational and orientational dynamical behavior of an extensive library of prospective models for single point water to evaluate how well these dynamical properties work for selection of optimal model candidates. We find that single point models are generally more extensively networked than multipoint models, evidenced by larger longitudinal and transverse orientational relaxation times, this despite similar translational diffusion coefficients. These results indicate that relaxed orientational structuring of spherical harmonic-based potentials could lead to more accurate future computational modeling of liquids.

 

D-04     David Bruce

Research Collaborators:  James Moulton and Barry Lavine

Research Presentation Title:  Kinetic and Thermodynamic Characterization of Functionalized N-isopropylacrylamide Polymers for Optical pH Sensing

Faculty Research Mentor:  Barry Lavine, Chemistry

 

Copolymers of N-isopropylacrylamide (NIPA) that swell and shrink in response to pH were prepared by dispersion polymerization and cast into hydrogel membranes made by mixing the pH sensitive polymer particles with aqueous polyvinyl alcohol (PVA) solutions followed by crosslinking with glutaric dialdehyde.  Large changes in the turbidity of the PVA membrane were observed as the pH of the buffer solution in contact with the membrane was increased or decreased.  Changes in the turbidity of the membrane were monitored using a Cary 6000 UV–visible absorbance spectrometer. The buffer capacity, pH, and temperature of the solutions in contact with the membrane were varied systematically to provide a complete pH profile of each copolymer. The swelling and shrinking cycles of many of the NIPA copolymers investigated were reversible.  Both the degree of swelling and the apparent pKa of the NIPA based polymers varied with temperature.  A unique aspect of this study was the investigation of the response of the NIPA based polymers to changes in the pH of the solution in contact with the membrane at low buffer capacities (0.5 mM). The rapid response rate and the reversibility of polymer swelling even at low buffer capacities are crucial to ensure the successful coupling of these materials to an optical fiber.  Several application areas for these NIPA based polymers are envisioned, e.g., monitoring the rising acidity levels in the ocean due to water that has become enriched in carbon dioxide, thereby putting shell building organisms at risk by reducing the amount of CO32- available in the water.

 

D-05     Zane Calvert

Research Collaborators:  Subruta Pal

Research Presentation Title:  Cobaloxime Complexes as Photocatalytic Radical Initiators in Giese-type Reactions

Faculty Research Mentor:  Spencer Pitre, Chemistry

 

Cobaloxime complexes and Vitamin B12 have been shown to be effective radical initiators of carbon radicals for use in organic reactions. This is because of the super-nucleophilic properties of the central cobalt (Co) atom of the complex when in the Co(I) oxidation state. However, despite their similar nucleophilic properties, cobaloximes suffer from low catalytic activity compared to Vitamin B12 as radical initiators in organic reactions. In this work, the use of cobaloxime for Giese-type reactions, and the effect different ligands have on the catalytic activity of the cobaloxime complex will be presented. The expectation holds that more basic ligands will act as better electron donating groups to the Co atom; thus, increasing the nucleophilicity of the complex. As hypothesized, the efficiency of the cobalt complexes as radical initiators in Giese-type reactions increases as the basicity of the ligands increases.

 

D-06     Rachel Crittell

Research Collaborators:  Margaux Lavenue, Scott Hutchinson, Rehema Nakiwala and Jeanne Bolliger

Research Presentation Title:  Access to Tricyclic Heteroarenes by an Iodine-Promoted Cyclization Reaction

Faculty Research Mentor:  Jeanne Bolliger, Chemistry

 

For many years, investigations on nitrogenous heterocycles have been an essential part of organic chemistry due to their widespread use in medications. Our research examines alternative ways to prepare nitrogen-containing heteroarenes while minimizing the formation of unwanted by-products. Many of our target molecules have structural similarities with biologically active compounds that are commonly used for numerous conditions, such as cancer, anxiety, and fungal infections. By optimizing a preliminary procedure from the Bolliger lab, we were able to improve the isolated yields for the iodine-promoted cyclization step. In this presentation, we will discuss the results of our cyclization reaction and discuss possible reaction mechanisms. Ultimately, we hope our research will pave a pathway to novel species with biologically relevant functional groups.

 

D-07     Georgia Eastham

Research Collaborators:  Shivangi Kharbanda, Osaid Alkhamayseh, Jimmie Weaver

Research Presentation Title:  Adaptation of strain-loadable alkenes for biological uses

Faculty Research Mentor:  Jimmie Weaver, Chemistry

 

Bioorthogonal chemistry allows further exploration of the intricacy of life through the development of highly selective reactions that occur reliably, even in a complex biological environment. Past work in the Weaver group developed a reaction system that utilizes an iridium-based photocatalyst, a cycloalkene, and blue light to transform light energy into molecular strain energy within the cycloalkene. The strained cycloalkene can then conjugate with an azide via a “click” reaction. This reaction is highly selective, fast, and only occurs in the presence of blue light, priming it for use for biological tagging and probing. My objective is to modify the structure of the cycloalkene such that it can be readily synthesized, conjugated to biological tools easily, and maintain ideal physiochemical properties to be used in biological systems. Specifically, I aim to synthesize cycloalkenes that are water soluble and have flexible handles that allow for loading of the alkene with the functionality that can be used to accomplish various biological goals. I am synthesizing a new alkene that will improve upon the original alkene. We hypothesize that a CH2 of the cycloalkene can be replaced with an N–H, and that doing so would increase its water solubility and provide a convenient location (the N–H) for easily functionalized with various bio-probes. Both alkenes and azides have been synthesized that have “handle” locations that will increase versatility. Continued investigation is expected to expand our knowledge of this photochemical reaction and refine the technology such that it can be used in diverse applications, even clinically.

 

D-08     Keely Fielding

Research Collaborators:  Smita Mohanty, Shine Ayyappan, Pratikshya Paudel, Patrick Combs and Viswanath Nukala

Research Presentation Title:  Structure of Ligand-Free Pheromone Binding Protein 2 in Ostrinia furnacalis

Faculty Research Mentor:  Smita Mohanty, Chemistry

 

Ostrinia furnacalis, commonly known as the Asian corn borer, is an agricultural pest prevalent in Southeast Asia. Female corn borers use pheromones as a mate attractant, as pheromones are recognized by pheromone-binding protein (PBP) in the antennae of male corn borers. PBP binds pheromone at high pH (pH at or above 6.0) and releases pheromone at acidic pH (pH below 5.0). PBP is found in the sensillar lymph of the antennae, where the pH is above 6. Once it binds pheromone, it transports the pheromone to the olfactory receptor neuron, around which the pH is acidic, inducing a conformational change within the PBP that releases the ligand. Past studies have shown that the mechanism of pheromone release in Ostrininia furnacalis pheromone-binding protein 2 (OfurPBP2) contrasts with release mechanisms in other moth PBPs. In most moth PBPs, there are conserved amino acid sequences for two biological gates responsible for ligand binding and release: the C-terminus gate and the histidine gate. For Ostrinia furnacalis, these conserved sequences have substitutions in the residues, which affects the binding and release of the ligand. By obtaining the structure and mechanism of binding and release from Asian corn borer’s PBP, we could reverse-engineer a pheromone mimic that could bind to and disable PBP. In our study, we are using nuclear magnetic resonance (NMR) data for OfurPBP2 to calculate the structure of free PBP.

 

D-09     Sarah Foy

Research Collaborators:  Rajendra Maharjan

Research Presentation Title:  Ice Crystal Growth in Solutions

Faculty Research Mentor:  Chris Fennell, Chemistry

 

Salt is spread on roads in the winter to melt ice because solutes, like salt ions, are well-known to depress the freezing point of water. When ice does grow from salty water, experiments show ice to grow more rapidly. In this study, we explore this phenomenon at the molecular level using computational simulations. We construct crystal growth experiments with aqueous hydrochloric acid and sodium chloride solutions with concentrations ranging from 0 to 1 M and at temperatures ranging from 220 to 260 K. We observe extensive formation of crystal lattice defects due to ion inclusion, and the presence of these defects was highly correlated to the measured rate of ice growth. These findings indicate that ice crystal growth might similarly be influenced by nonpolar solutes and larger hydrocarbons.

 

D-10     Jacob Freeman

Research Collaborators:  Murthy Lakshmi Narashimha

Research Presentation Title:  Emergency bulk warming with Iron oxidation

Faculty Research Mentor:  Nicholas Materer, Chemistry

 

Blood and plasma can be in great demand in emergency situations. Ensuring that fluids being injected do not cause more damage is incredibly important. This project was to make a bulk warming system that can prepare a large quantity of fluid for injection quickly in high-stress emergency situations. It is done using several different factors of chemistry and engineering. The reaction being formed is the oxidation of iron to generate a massive amount of energy quickly. The only activation used for the reaction is water and salt as a catalyst. By the use of Phase Change Material (PCM) and an expanded graphite brick, we are able to better control the duration and intensity of the heat transferred from the source to the payload. The PCM has a set range in which it changes states and helps with the gradual release of energy into the payload. The expanded graphite is important to increase the thermal conductivity of the brick to be more efficient. Making a reliable way to provide heated life-saving fluids in a crucial time is essential and it will be used in the military and emergency services for a reliable heat source.

 

D-11     Noah Holt

Research Collaborators:  Barry Lavine

Research Presentation Title:  Electrochemistry of Tricyclic Heteroarenes

Faculty Research Mentor:  Jeanne Bolliger, Chemistry

 

The Bolliger group has prepared a new class of tricyclic aromatic compounds. In preliminary screening these compounds displayed unusual electrochemical behaviors which warranted a more in-depth examination. This investigation examines the fundamental properties and the effects of various substitutions using a variety of experimental and simulation techniques.

 

D-12     Hope Koch

Research Collaborators:  Haoran Zhong and Barry Lavine

Research Presentation Title:  Infrared Imaging Microscopy Applied to the Forensic Examination of Automotive Paints

Faculty Research Mentor:  Barry Lavine, Chemistry

 

In the forensic examination of automotive paint, each layer of paint is typically hand sectioned and analyzed individually using infrared (IR) spectroscopy. However, sampling too close to the boundary between adjacent layers may pose a problem as it can produce an IR spectrum that is a mixture of two layers. Not having a “pure” IR spectrum of each layer prevents a meaningful comparison between each paint layer or in the situation of searching an automotive database prevents a forensic paint examiner from developing an accurate hit list of potential suspects. These two problems can be addressed by collecting concatenated IR data from all paint layers in a single analysis by scanning across the cross sectioned layers of an automotive paint sample using a FTIR imaging microscope. Decatenation of the IR imaging data is achieved through multivariate curve resolution. Comparing the reconstructed IR spectrum of each layer for a paint sample against the IR spectral library of an in-house automotive paint database containing the paint sample in question demonstrates that it is possible to identify the correct line and model of the vehicle from these reconstructed IR spectra.  This approach to IR analysis of automotive paint, not only saves time by eliminating the need to analyze each layer separately, but also ensures that the final spectrum of each layer is “pure” and not a mixture of two distinct paint layers.

 

D-13     Mary McKee

Research Collaborators:  Spencer Pitre and Prasadi Gallage

Research Presentation Title:  Visible light photochemistry of 1,4-dihydropyridine anions in dehalogenation and detosylation reactions

Faculty Research Mentor:  Spencer Pitre, Chemistry

 

1,4 –Dihydropyridines (DHP) are a group of compounds that are distinguished by a pyridine ring with a hydrogen atom attached to the nitrogen atom and to the C-4 position. In visible light meditated reactions these DHPs can absorb light and generate excited states that can trigger various chemical reactions. In presence of a suitable base, DHPs can undergo deprotonation to form the corresponding anion, resulting in a substantial redshift in the absorption. In this work DHP anion is easily generated in-situ using Cs2CO3 in acetonitrile at rt and it serves as a strong visible light absorbing reagent and we are evaluating the competency of the DHP anion as a photoactive reductant. Using this strategy, we have developed a method for the reductive dehalogenation of aryl chlorides and have extended the reaction conditions for detosylation reactions. The effectiveness of this method will be explained with the reactions scope as well as the mechanistic experiments.

 

D-14     Julia Murphy

Research Collaborators:  Charles Weinert, Thad Stancil, Vanessa Fortney and Laura Levescy

Research Presentation Title:  Germylamines as Versatile Amidation Reagents for Acid Fluorides

Faculty Research Mentor:  Scott Weinert, Chemistry

 

The formation of amide bonds, which are essential in the production of pharmaceuticals and peptides, is being investigated in this study by utilizing germylamines to form amides from acid fluorides. This process obviates the need for a costly metal catalyst, which makes using these reactions economically and practically favorable. Previous studies have confirmed the efficacy of converting benzoyl fluorides into amides using germylamines and this particular study analyzes the scope and limitations of this reaction by adding an electron withdrawing nitro group first in the para and then in the sterically-hindering ortho position on the benzoyl fluoride. Results indicate that the nitro substituents in both the para and ortho position do not obstruct the amidation reaction as shown by the clean formation of the amide product. 

 

D-15     Megan Padgett

Research Collaborators:  Mayokun Ayodele

Research Presentation Title:  The Effects of Modifying Substituents on Hydroquinone Catalysts for Halogen Bonding Interactions

Faculty Research Mentor:  Spencer Pitre, Chemistry

 

When there is evidence of a net attractive interaction between a nucleophilic region in one chemical entity and an electrophilic region connected to a halogen atom in another, the interaction is known as a halogen bond. The aim of this work is to investigate the effects of substituent modifications on the hydroquinone catalysts have on the halogen bonding enthalpies with alkyl iodides. In initial work from our lab, the hydroquinone catalyst used had tert-butyl group that was interfering with the halogen bond interaction. To preserve time and resources, our group is employing computational chemistry to discover a catalyst better suited for this interaction. Using WebMo, multiple catalyst structures were optimized along with the CF_3 I molecule. Subsequently, we studied the interaction of the CF_3 I molecule with each catalyst to determine the halogen bonding enthalpies. The results of these computations which will help identify which catalysts are more desirable as synthetic targets will be presented.

 

D-16     Justin Rein

Research Collaborators:  Nima Noei

Research Presentation Title:  Synthesis and Characterization of Tridentate Quinoline-based NCO Ligands

Faculty Research Mentor:  Nima Noei, Chemistry

 

The first ligand was introduced in 1991 by Mark J. Burk. He was depleting them for a hydrogenation reaction. Since then, ligands have assisted chemistry and medical practices. A ligand is an ion or molecule that binds to a central atom to form a complex configuration.  The synthesis of Tridente Quinoline-based NCO Ligands is a five-step process. Usually resulting in a central atom composed of Rhodium, Palladium, and Platinum. These materials are expensive, rarely available, and inefficient. However, replacing the core with a copper center will provide a more affordable option that is readily available.

 

D-17     Bailey Robertson

Research Collaborators:  Spencer Pitre and Tarannum Tasnim

Research Presentation Title:  Hydroquinones as Halogen-Bonding Catalysts for Radical Perfluroalkylation Reactions

Faculty Research Mentor:  Spencer Pitre, Chemistry

 

Developing a better understanding of how hydroquinones serve to catalytically activate alkyl halides through halogen-bonding interaction to form visible-light absorbing charge-transfer complexes (CTC) holds immense value. In this presentation, a CTC that occurs through halogen-bonding interactions will be highlighted. A CTC is the generation of a new ground state aggregate by the association of a electron-rich compound and an electron-poor compound in solution. The CTC can absorb in the visible light region (400-700 nm), while individual substrates are unable to. Therefore, the CTC provides an alternative strategy for performing visible-light-mediated reactions. In the Pitre Lab, we are working to develop catalytic systems for charge-transfer photochemistry. In this work, charge-transfer photochemistry that includes the use of hydroquinones (HQs) as catalysts that activate alkyl bromides towards visible-light irradiation through halogen-bonding interactions is presented. The main goal of our project is to develop a system with iodide salts that undergo a substitution reaction on the alkyl bromide to make an alkyl iodide in situ, since bromides are less reactive than iodides towards halogen bonding. Later works will focus on expanding the scope of these reactions and work to determine the optimal conditions.

 

D-18     Moriah Thompson

Research Collaborators:  Tiwalola Ogunleye

Research Presentation Title:  Sample Preparation of TCRα for Studies by EPR    

Faculty Research Mentor:  Gabriel Cook, Chemistry

 

T-Cell Receptor Protein (TCR) is a single-pass transmembrane protein found in T-cells is involved in the signaling pathway in immune response.  TCR binds antigens, activating the T-cell through signal transduction.  The two chains of the receptor, alpha and beta, connect through a disulfide bond.  While this is an essential protein for the immune system, the function and structure of TCR is not fully understood because the hydrophobic transmembrane region makes it difficult to study in typical sample conditions.  Hydrophobic proteins like this receptor must be incorporated into detergent or lipid environments to run experiments that are commonly used to determine their function and structure.  Our lab is developing methods to incorporate this hydrophobic protein into samples so that we can use Electron Resonance Spectroscopy (EPR) to study these properties.  For this work, we are concentrating on expressing and purifying the alpha chain (TCRα).  To run EPR on TCRα, we must attach a nitroxide group to the protein.  TCRα was mutated at four positions, in four separate sequences, to replace a wild-type amino acid with a Cysteine residue.  These Cysteines will form a thiol linkage with the nitroxide group.  This presentation will explain the methods used to express and purify one of these TCRα mutants, Alanine 18 to Cysteine (A18C), for the preparation of samples that can be measured by EPR.

 

D-19     Kirsten Albert

Research Collaborators:  Mary Foltz and Jehan Shwiyyat

Research Presentation Title:  Dissolved nitrous oxide from groundwater wells in Oklahoma

Faculty Research Mentor:  Mary Foltz, Civil and Environmental Engineering

 

Understanding how nitrogen is transported and transformed will enable management that decreases negative consequences of excess nitrogen (e.g., climate change, eutrophication). Although nitrate is often measured in groundwater, the amount of nitrous oxide emissions from groundwater is largely underreported. Nitrous oxide, a potent greenhouse gas, can be produced through biological reduction of nitrate in water or soil. The purpose of this research is to quantify indirect nitrous oxide emissions from groundwater through well sampling and measurement of dissolved nitrous oxide. For our research, groundwater monitoring wells in Oklahoma with differing nitrate concentrations were identified and data is currently being collected to better understand the relationship between nitrate and nitrous oxide in these wells. We are using a peristaltic pump, gas chromatography, and syringes to collect water, which are then brought to equilibrium in the lab and analyzed for nitrous oxide. After sampling and analysis, the relationship between dissolved nitrate and nitrous oxide will be assessed using correlation statistical tests. We hypothesize that the dissolved nitrous oxide in groundwater samples will vary based on the concentration of dissolved nitrate.

 

D-20     <Withdrawn>

 

D-21     Shelby Maggard

Research Collaborators:  Lizzie Long

Research Presentation Title:  Nitrous oxide emission estimates using different field chamber designs and measurement methods

Faculty Research Mentor:  Mary Foltz, Civil and Environmental Engineering

 

Soils can produce and release greenhouse gases like nitrous oxide to the air and contribute to climate change. Manual static chambers are an effective and affordable option for capturing gas emissions for subsequent analysis by gas chromatography in the lab. Automated field chamber and analyzer combos are a more expensive option with real-time gas flux measurement in the field without requiring sample transfer and analysis in the lab. There is little research on how measurements differ between manual and automated chamber methods, so we compared nitrous oxide emission estimates from different chamber dimensions and methods (manual vs. automated). Our initial field campaign using manual chambers found that chamber dimension did not significantly influence nitrous oxide flux estimates. The manual vs. automated chamber comparison field campaign is ongoing, but we expect automated chambers to provide better estimates of gas fluxes because of real-time data reporting. The quantitative data related to chamber design and emission measurement method will support future work through improved greenhouse gas emission measurement methods.

 

D-22     Carly Noone

Research Collaborators:  Sreemala Das Majumder

Research Presentation Title:  Estimating life cycle greenhouse gas emissions from constructed wetlands

Faculty Research Mentor:  Mary Foltz, Civil and Environmental Engineering

 

Constructed wetlands (CW) are built to mimic the natural benefits of wetlands and are able to decontaminate wastewaters through natural processes. These natural processes may consume and/or produce greenhouse gases (GHG), such as nitrous oxide (N2O), carbon dioxide (CO2), and methane (CH4). Furthermore, GHG can be emitted through the construction and operation phases of a CW. In this study, we quantify life cycle GHG emissions from CW using data from published literature and Simapro software to pinpoint the largest contributors to GHG emissions through the lifespan of a CW. Aspects being studied in Simapro include materials used in construction of CWs, transportation of materials, and energy consumption. Our literature review also considers the influence of design choices on GHG emissions.

 

D-23     Daisy Rosas

Research Collaborators:  Wyatt Barrier and Jaime Schussler

Research Presentation Title:  Evaluating Effects of Droughts and Wildfire through Small-Scale Rainfall Simulation

Faculty Research Mentor:  Jaime Schussler, Civil and Environmental Engineering

 

Regionally, Oklahoma has experienced prolonged and intense droughts, which causes depletion of soil moisture and loss of vegetation. As the soil moisture and vegetation are depleted, the soil surface becomes hard and compact. Under these conditions, soils with high clay content are prone to crusting, and act more like an impervious surface when rainfall occurs. Several runoff methods, including the Soil Conservation Service Curve Number (SCS CN) and Rational Method, use static values to account for infiltration or runoff, and the changing climate and local seasonal parameters are not considered. Researchers at Oklahoma State University hypothesize that infiltration and runoff patterns may be altered after extended drought. To collect local conditions, researchers at Oklahoma State University have diligently gathered daily precipitation and soil moisture data from every Mesonet station within the state of Oklahoma. The data collection encompasses Mesonet measurements from the preceding five-year period spanning from 2017 to 2022. The stations will be divided and analyzed according to the nine distinct eco-regions in the state. This information will be used to design a rainfall simulation regimen to quantify the impact of drought on hydrologic soil characteristics. Ultimately, such an approach would aid in identifying trends and facilitate more accurate stormwater runoff predictions and mitigation. After the data analysis, we will collect benchmark soil samples from field sites and cure them according to simulate the historical soil moisture data. The cured soils will then be subjected to rainfall through small-scale rainfall simulation. The small-scale rainfall simulator will be constructed with apparatuses to capture infiltration and runoff volumes, which under rainfall simulation, will provide researchers with data on how droughted soils behave during subsequent rainfall intensities and durations. By running these tests, researchers hope to find a calibrated curve number and runoff coefficient for Oklahoma soils under various drought conditions for more accurate runoff estimations using the SCS CN or rational method.

 

D-24     Reece Harrel

Research Presentation Title:  Artificial Intelligence in Healthcare: What to consider

Faculty Research Mentor:  Mayfield, Computer Science

 

Computers have become an everyday part of our lives. One field on the forefront of computer science research is artificial intelligence. Although some people may have fears of artificial intelligence, the technology can be beneficial. It has the ability to enhance or replace human capabilities in many areas. Artificial intelligence, by definition, is any machine processing simulation of human intelligence. As research begins to produce more accessible and powerful technology that utilizes artificial intelligence, careful deliberations must be given to its implementations. Specifically in the field of healthcare, artificial intelligence has the ability to revolutionize the field. But what considerations must be given before turning over our healthcare decisions to artificial intelligence? The purpose of this research is to evaluate the current studies  on artificial intelligence in healthcare and explore the ethical dilemmas surrounding its applications.

 

D-25    John Doudican

Research Collaborators:  Charles Bunting

Research Presentation Title:  Behavior of electromagnetic waves in reverberation chambers

Faculty Research Mentor:  Charles Bunting, Electrical and Computer Engineering

 

The opportunity to conduct undergraduate research in the college of engineering is not an opportunity that comes to many students. For those who earn the opportunity, they can develop skills to further advance technology. Dr. Bunting’s research involves the behavior of electromagnetic waves in a reverberation chamber. In this research, I would be given the opportunity to learn about the behavior of these waves in the chamber. For instance, one of the potential projects I could participate in would involve his areas of interest, like RF antennas. In addition to conducting research on electromagnetic topics, I will also present my research to my fellow researchers as well as my classmates. Receiving this research opportunity will allow me to further my career in this field and to better understand the world around me.

 

D-26     Luella Hollis

Research Collaborators:  Weihua Sheng

Research Presentation Title:  Enabling Sound-Based Human Activity Monitoring for Home Service Robots

Faculty Research Mentor:  Weihua Sheng, Electrical and Computer Engineering

As technology continues to advance, applications for it also continue to grow. The market for service robots will continue to grow as a result. A home service robot will help to mediate between medical professionals and patients while away from the clinic or hospital. This home service robot is designed to be able to monitor patient health, keeping track of them throughout their environment. The information obtained from this person can be sent to healthcare personnel, who can determine the proper course of action given the circumstances. The methods with which this research is intended to be conducted involve the use of sound-collecting devices as well as visual devices to better monitor the patient's behavior. This research also involves creating a digital twin of a house or apartment. By creating a digital twin of a home, the home service robot can understand and interpret its surroundings to help the patient further by being able to navigate through its surroundings. This digital twin will be constructed through the use of Revit and Unity3D. Revit will be used to build the house, while Unity3D will be used to be able to view it in 3D with an Oculus headset. This digital twin would create a way in which an environment could be coded into the robot. Problems that may arise include conversions of a model from one software to another, since there is no guaranteed way to import and export between Revit and Unity3D. Refining the efficiency with which the robot can gather and interpret this data is one of the main goals of this research, allowing it to collect better data to help the patient in every possible way. A digital twin would further the collection and interpretation of a patient’s health-related data. The intended result of this research includes the robot being fully functional with the ability to gather patient data, while also maintaining an awareness of its surroundings. The research is worth conducting to elevate the quality of life for those needing extra help with daily activities.

 

D-27     Christian Moser

Research Collaborators:  John Hu

Research Presentation Title:  Python codes to implement fastpower SCA on deep learning FPGA accelerators

Faculty Research Mentor:  John Hu, Electrical and Computer Engineering

 

Artificial Intelligence (AI) and Machine Learning (ML) models are increasingly deployed on handheld devices like Apple Siri and iPhone Face ID. However, the degree to which we can secure AI/ML limits the domain in which we can use AI and ML in security-critical applications, such as autonomous driving. Portable devices have hardware-related vulnerabilities that could compromise the AI/ML integrity during field application. One such vulnerability is the microelectronics’ susceptibility to side-channel attacks (SCA). By monitoring hardware’s power or electromagnetic (EM) emanation, an adversary can infer AI/ML’s inputs [1], which may breach user privacy in medical applications. SCA may also extract the entire neural network (NN) model [2,3], which enables adversaries to design more potent adversarial examples. This undergraduate research project intends to develop Python codes to implement fastpower SCA on deep learning FPGA accelerators. Field Programmable Logic Arrays (FPGA) are widely used to accelerate AI/ML in portable devices. However, different FPGA implementations may exhibit different SCA resiliency and hence require generic and quantitative benchmark. Furthermore, such a common benchmark is essential to guide the design of effective hardware countermeasures and defense.

 

D-28     Aaron Rosen

Research Collaborators:  John Hu

Research Presentation Title:  Toward Adversarial Robustness of Deep Learning Accelerators

Faculty Research Mentor:  John Hu, Electrical and Computer Engineering

 

Neural networks (NN) provide state-of-the-art performance on many Machine Learning (ML) tasks, such as facial recognition and object detection. However, NN is vulnerable to adversarial examples. Google famously demonstrated that their GoogleNet could manipulate a "panda" into a "ribbon" even if human eyes cannot see any difference. Other researchers showed that carefully placed stickers on Stop signs could mislead a NN classifier to recognize it as a speed limit. Crafting novel adversarial examples and robust defense mechanisms has become one of the hottest areas in cyber security, electrical engineering, and computer science. This undergraduate research project intends to develop Python codes to evaluate the adversarial robustness of deep learning FPGA accelerators. Field Programmable Logic Arrays (FPGA) are often used to accelerate AI/ML in portable devices. However, different FPGA implementations may exhibit various adversarial robustness, which calls for design-specific benchmarking. Furthermore, such a benchmarking framework is essential to guide the design of hardware-based countermeasures and defense. The baseline models will be multi-layer perceptron (MLP) and convolutional neural networks (CNN) with weights and biases restricted to +1 and -1. The binarization reduces the computation and memory storage requirement on FPGA. Specifically, the vector multiplication can be replaced by bit-wise XNOR and Pop count [3] to utilize Lookup Table (LUT)-based logic instead of DSP slices. Step 1: Binary MLP and CNN training. Step 2: FPGA deployment on Xilinx-7 series. Step 3: Generate adversarial examples reusing Python codes from AdverTorch [4] and ART [5] Toolbox and adjust them to the MLP/CNN Xilinx-7 FPGA implementation. We will focus on Projected Gradient Descent (PGD) [6] and C&W [7] method to generate adversarial examples because they are among the strongest best-effort empirical attacks to date. Step 4: Plot clean and adversarial accuracy vs. perturbation magnitude ().

 

D-29     James Boudreaux

Research Collaborators:  Mazin Mustafa and Dorian Cathey

Research Presentation Title:  Design and Fabrication of a Model Submarine Hull for EMC Measurements Relating to Anti-Detection Architecture

Faculty Research Mentor:  Charles Bunting, Electrical and Computer Engineering

Objectives: To utilize scale models of varying submarine hulls, with similar electromagnetic reverberation characteristics determined by reverberation chamber theory, as seen in real world hulls. Measuring internally excited electric fields on the models, as well as the effect of different types of absorbers on the electric field levels, to prevent “sub” detection through the means of magnetic anomaly detectors or radar while protecting sensitive equipment.

Research Purpose: Modern Submarines utilize several electronic components to navigate and monitor under sea activity. These components, such as long-range communication systems, pressure monitors, sonar sensors, and on board power generation can result in strong electromagnetic fields that would alert adversaries to the presence of the submarine. These processes take place in irregularly shaped hulls designed to be hydrodynamic with complex electrical grids and varying signal generation. Because of these conditions, the strength of both internal and external fields generated by the submarine must be found using a statistical field description. To achieve this, scale models of the United States’ Virginia Class (SSN-774) Nuclear Submarine will be fabricated with realistic reverberation characteristics to that of the steel tear shaped hull. To ensure this, the model will be fitted with a mechanical tuner redistributing the resonant energy from a series of signals ranging from 50 MHz to 8 GHz. The shielding effectiveness and quality factor of the electric field within the model hull, based on various aperture and absorber tests, will be measured to determine the accuracy of the model as a representation of the Virginia Class Hull. Upon the completion of an effective model, the measurement probes will be used to map the resonant field levels within the hull, and the RF emitted externally. During this phase of testing, the effects of dampening materials such as absorber blankets on the established fields will be determined and mapped. The goal of the absorption material is to redistribute the energy created by the electric fields away from detectable signals while protecting sensitive instruments and components.

 

D-30     Emory Meursing

Research Collaborators:  Jordan Fogg, Zeyu Deng, Long Huang, Huaxia Wang and Chen Wang

Research Presentation Title:  Remote Robot Control with Low-cost Robotic Arms and Human Motions

Faculty Research Mentor:  Huaxia Wang, Engineering Technology

 

Geographically separated people are now connected by smart devices and networks to enjoy remote human interactions. However, current online interactions are still confined in a virtual space. Extending the pure virtual interactions to the physical world requires multidisciplinary research efforts, including sensing, robot control, networking, and kinematics mapping. This paper introduces a remote motion controlled robotic arm framework by integrating these techniques, which allows a user to control a far-end robotic arm simply by hand motions. In the meanwhile, the robotic arm follows the user’s hand to perform tasks and sends back its live states to the user in video stream. Furthermore, we explore using cheap robotic arms and off-the-shelf motion capture devices to facilitate the wide use of the platform in people’s daily life. No professional knowledge is required from the user. Moreover, we implement a testbed that connects two US states for the remote-control study. We investigate different types of latency that affect the user’s remote-control experience and conduct comparative studies. Results show that the current commercial motion capture device, low-cost robotic arms and networks are already available to provide physically augmented remote human interactions.

 

D-31     Michael Kubicki

Research Collaborators:  Haejun Park

Research Presentation Title:  A Method for Quantifying Exit Usage

Faculty Research Mentor:  Haejun Park, Fire Protection and Safety Engineering Technology

 

Building and life safety codes such as those produced by the International Code Council and the National Fire Protection Association have specific requirements for the minimum width of exits and the number of exits needed to evacuate the occupants of the building. However, in these requirements, there is an assumption made that the occupants will distribute themselves proportionally throughout the exits. This is known not to be the case, in many instances there has been an observable, imbalanced use of a building’s exits. During an evacuation from a building, the use of particular exits may vary significantly resulting in some becoming crowded while others are rarely used, which may cause inefficient egress resulting in harm during an emergency evacuation. In this study, a new concept of quantifying this varying exit usage and subsequently categorizing them into five different performance groups is proposed. This method is expected to allow us to compare the performance of exits in the same building as well as exits between separate buildings. To quantify the exit usage, the actual numbers of occupants using specific exits for a certain period are needed as well as the design occupant load and design number of occupants using the specific exits. The former values are obtained by actually counting occupants and the latter values are determined by exit design criteria, i.e., calculations found in building and life safety codes. Seven exits are selected from two buildings and their performance is compared quantitatively as a proof of concept. These methods will allow all interested parties to understand the usage of each exit to further enhance future and current designs to improve exit usage distribution, which is especially critical in performance-based fire safety design. The new methods can also be tied into future research regarding the human factors going into exit and route selection.

 

D-32     Mason Martinez

Research Collaborators:  Bryceston Smith and Todd Halihan

Research Presentation Title:  Accuracy and Precision In Characterization Of Aquifer Geothermal Profiles

Faculty Research Mentor:  Todd Halihan, Geology

 

Geothermal profiles of aquifers are valuable for evaluating flow processes by modeling existing data or determining deviations from anticipated changes in temperature with depth. In aquifers with small changes in temperature, this is difficult as many environmental sampling tools are not precise temperature sampling tools. Determining accurate vertical geothermal profiles requires an understanding of the accuracy and precision of temperature measuring devices. In the Arbuckle-Simpson aquifer of central Oklahoma, temperatures vary by less than a degree Celsius across spatial and temporal ranges in springs and wells. We utilize well logging, transducer, temperature/level/conductivity probes, and precision platinum thermometers to evaluate the CAMARR (City of Ada Managed Aquifer Recharge Research) Site near Fittstown, OK. The results show small, but significant changes resulting in an inverted geothermal profile, with colder temperatures of the aquifer in depths of 300 meters or greater.

 

D-33     Olivia Fulkerson

Research Collaborators:  Erik Inman, Akash Deep and Hadi Noori

Research Presentation Title:  Accelerating Design and Additive Manufacturing of Polymer Composites for Aerospace Applications

Faculty Research Mentor:  Srikanthan Ramesh, Industrial Engineering and Management

 

Polymer matrix composites (PMCs) combine the beneficial properties of a polymer and a reinforcement agent to achieve structural and functional properties that cannot be achieved by one of the components alone. Their exceptional mechanical performance and lightweight characterize them. Fittingly, PMCs are sought-after building blocks for commercial products with a high strength-weight ratio. Additive manufacturing (aka 3D printing) (AM) is a layer-by-layer manufacturing method that creates unprecedented opportunities for safer and more efficient production of functional parts with composite materials. However, the selection of appropriate process parameters in AM is non-trivial, as it directly influences the defect formation and mechanical performance of the printed part. Currently, the optimization of process parameters revolves around a trial-and-error approach that can increase the lead time to produce tailored parts, thereby increasing production costs. This project aims to develop and validate an optimization framework to accelerate the design and production of fiber-reinforced parts (FRPs). The goal is to construct a Bayesian framework to accelerate the design and manufacture of carbon fiber-reinforced polymer (CFRP) parts with a high strength-to-weight ratio. The framework will model a black-box function with minimal experimental data. The goal will be achieved by completing three tasks: (1) random experimentation, (2) model building and optimization of the acquisition function, and (3) data-driven experimentation. At this stage of our study, we have characterized the effect of three crucial process parameters, (i) infill density, (ii) printing speed, and (iii) layer height on the fracture energy of the 3D printed CFRPs. Future work will focus on building the Bayesian optimization model and conducting data-guided experiments to rapidly identify optimal process parameters for maximizing the strength-to-weight ratio of the 3D-printed CFRPs.

 

D-34     Erik Inman

Research Presentation Title:  Sensitivity of Mechanical Strength to Surface Defects in Additively Manufactured Polymer Composites

Faculty Research Mentor:  Srikanthan Ramesh, Industrial Engineering and Management

 

Fused Filament Fabrication (FFF) is one of the most accessible forms of thermoplastic-based additive manufacturing. A three-axis motion system deposits layers of plastic to produce a 3D part. It has become increasingly popular in many industries for its ability to produce complex geometries while minimizing material usage. Manufacturing industries have been able to take advantage of this in recent years. While this has reduced material waste, the incidence of manufacturing defects is still a concern. The ability to characterize a defect’s impact on a part’s mechanical strength is foundational to quality control in this regard. Observation of an ongoing FFF study indicated that mechanical properties of 3D printed parts with defects are likely to exhibit a nonlinear sensitivity to fracture. In other words, there seems to be no functional difference between defective and non-defective parts in strength for extremely low infill density parts and extremely high infill density parts, while there is a proportional higher sensitivity to defects in midrange infill densities. Defects are inherent, and in a quality control setting, it is paramount to establish allowable variations in the process. When strength is key in an additively manufactured part, predicting the effect of a defect is essential. Currently, there is no established research into characterizing the sensitivity to surface defects of additively manufactured parts. This phase of investigation aimed to understand how defect parameters such as shape and size reduce the strength of tensile specimens. The design of the experiment was as follows: tensile test specimens were produced on a FFF printer with a fixed infill pattern, while the infill density was varied between 5% and 100%. A defect-free set was tested to establish baseline strength values. Sample sets of these were then produced with artificial defects that mimic observed part defects. The defects varied in shape (circle and square) between sample sets and each shape varied in size among sample sets. Samples were loaded in tension, and force and displacement data were collected to analyze fracture energy. It was postulated that there is a range of infill densities that are strong but have minimal sensitivity to defects.

 

D-35     Jacob O'Hara

Research Collaborators:  Yuxuan Li and Ziyang Zhang

Research Presentation Title:  Data-driven Diabetic Retinopathy (DR) Prediction with the Assistance of a Score-based Diffusion Model

Faculty Research Mentor:  Chenang Liu, Industrial Engineering and Management

 

Diabetic Retinopathy (DR) is a common microvascular diabetes complication in clinics, which may result in vision loss due to the damage of blood vessels in the back of the eye. Currently, it is also the leading cause of blindness among American adults. Thus, with the available electronic health records (EHR) data, establishing an effective classification model to accurately predict the risk of DR can significantly benefit the diabetic patients who have not been affected by DR yet. However, one of the biggest challenges is that the collected EHR data are usually imbalanced, as in general the number of patients who are diagnosed positive for DR is much less than the negative. For instance, the imbalanced ratio of positive to negative patients may be less than 1:10, and then it will result in significant training bias and reduce the accuracy of the classifiers. Therefore, to handle the data imbalanced issue, this paper aims to develop a data augmentation approach with the integration of machine learning models based on the EHR data collected from diabetic patients’ primary care visits. This paper leverages a score-based diffusion model to generate high-quality EHR samples for DR patients and improve classification performance. The classification results show a 7% improvement in terms of the F1 score, which demonstrates that the proposed diffusion model based EHR data augmentation approach is promising Key words: Data augmentation, diabetic retinopathy (DR) prediction, diffusion model, electronic health records (EHR), imbalanced classification.

 

D-36     Kent Slater

Research Collaborators:  Chenang Liu, Yuxuan Li and Yongwei Shan

Research Presentation Title:  A Generative Adversarial Network (GAN)-Assisted Data Quality Monitoring Approach for Out-of-Distribution Detection of High Dimensional Data

Faculty Research Mentor:  Chenang Liu, Industrial Engineering and Management

 

Data quality monitoring plays a critical role in various real-world engineering system inspection problems. Anomalous or invalid inspection data commonly exist due to computer/human recording errors, sensor faults, etc. Thus, an efficient tool to detect data anomalies is critically needed. However, it is still challenging due to high dimensionality, unknown underlying distribution, insufficient sample size, and high level of noise. To address these challenges, an effective approach that can learn the underlying distribution of normal data with anomaly detection rules was developed. In this approach, the Generative Adversarial Network (GAN) was employed to identify the underlying distribution of normal data and filter out noise. After using the trained GAN to generate points of the learned distribution, a -nearest neighbor based approach is used to define the anomaly detection rules. In the proposed approach, the normal records are used to train the GAN and establish the control rule. Specifically, after training the GAN using the normal records, the pairwise distances over all the GAN-generated data points are calculated, and the -nearest neighbors for every single data point are accordingly determined. Then, the average distance from each single data point to its -nearest neighbors is calculated as the statistics to indicate the data quality and establish a control chart. When a new record comes in, its similarity to the GAN-generated distribution can be evaluated by the established control chart to identify whether the new record is anomalous or not.

 

D-37     Brent Bertaux

Research Collaborators:  Jaxon Castillo

Research Presentation Title:  Discrete Differential Geometry in Virtual Reality

Faculty Research Mentor:  Henry Segerman, Mathematics

 

Is it possible to turn a sphere inside out? This is the type of question that is addressed in discrete differential geometry, which focuses on the polygons seen in computer graphics. While the eversion of a sphere has been demonstrated through video animation, what if we could experience this transformation with our own hands? By creating a virtual reality simulation that puts the user in control of discrete geometric objects, one can get a better understanding of the complex structure of polygons. To make this possible, all we need is the proper coding that can detect and inhibit the user from crossing illegal bounds. While current 2D versions of this code exist, there yet exists a 3D simulation in virtual reality. Creating this experience will provide an improved visual representation of discrete geometric objects, which can assist mathematicians in improving the understanding of discrete differential geometry. The simulation is being produced with Unity programming on a standard Vive headset, and will allow users to interact with 3D shapes from every possible angle.

 

D-38     Jensen Bridges

Research Collaborators:  Emily Armstrong, Zachariah Kline, Jacob Antici, Kian Greene and Zoe Friedman

Research Presentation Title:  A Comparison of Dimensionality Reduction Techniques for Prediction of Plutonium in Nitric Acid Concentrations

Faculty Research Mentor:  Mihhail Berezovski, Mathematics

 

This project is a collaboration with the Pacific Northwest National Laboratory, a section of the Department of Energy. The overarching goal within this project is to optimize the prediction of plutonium given the measurements of the nitric acid solution it is contained in. The problem at hand is to determine if non-negative matrix factorization is a better prediction model than the current method of prediction used by the Department of Energy, principal component analysis. This is needed in order to monitor the nuclear fuel cycle and ensure that the amount of radioactive by-products can be estimated. While principal component analysis is quite accurate, the manner of the results does not allow for much explanation outside of the mathematical terms used to analyze it. The comparison between principal component analysis and non-negative matrix factorization will not only be of accuracy of prediction, but also results that can be understood in the context of the scientific problem at hand. The conclusion reached was that while non-negative matrix factorization yields contextually significant results, the accuracy of principal component analysis proves to be superior enough to justify continuing with the current prediction method.

 

D-39     Taylor Johnson

Research Collaborators:  John Weaver

Research Presentation Title:  Reflective Abstraction as a Mechanism for Developing Pedagogical Content Knowledge

Faculty Research Mentor:  Michael Tallman, Mathematics

 

Reflective Abstraction as a Mechanism for Developing Pedagogical Content Knowledge

A mathematics teacher’s instruction is informed by their pedagogical knowledge (teaching theories and tactics) and content knowledge (ability to solve and understand math problems). A successful teacher must utilize both of these knowledge bases to support students’ conceptual learning of mathematics. Last year, we analyzed interviews previously conducted by my research mentors to examine how a pre-service secondary mathematics educator’s own conceptualization for rate of change, a foundational mathematics topic, progressed. In addition to a series of interviews to assess the participant’s individual understanding of rate of change, the participant reflected on the same set of recorded lessons of her teaching rate of change in the classroom pre-intervention interviews and post-intervention interviews. The participant noted moments of high-quality instruction and room for improvement in each lesson, however these noted moments changed after her series of interviews. Prior to the interviews asking the participant to delve into her own capabilities of interpreting rate of change, her assessment of her teaching was focused on the pedagogical tactics implemented in her classroom and their effectiveness. However, after the interviews, she took into account whether her teaching was actually helping her students make the connections they needed in order to understand the concept. We hope to distinguish and emphasize the differences in her reflection before and after working through interviews that challenged to use reflective abstraction and blend her pedagogical and mathematical content knowledge. The goal of this journal submission is to build off of our previous one discussing the transition between a well-known tiered model of mental conceptualizations of rate of change previously reported in literature in order to aid students’ critical thinking skills and teachers’ abilities to support students’ learning of this foundational concept. This journal article will illustrate the benefit to teachers implementing the practice of reflective abstraction when it comes to their own teaching. By being conscientious of the relationships within pedagogical content knowledge, teachers can aid in students’ understanding in their class and the students’ future mathematics courses as well.

 

D-40     Hadlee Shields

Research Collaborators:  Allison Dorko

Research Presentation Title:  Preliminary Findings Regarding Student Learning from Lecture and Homework

Faculty Research Mentor:  Allison Dorko, Mathematics

 

Lectures and homework provide opportunities for students to learn mathematics. Dorko and Cook (under review; 2022) propose attending to learning across the two settings is crucial to improving student learning in both settings. To that end, our study addresses the questions (1) What did two developmental mathematics students learn in lecture, and what did they learn from homework? and (2) How can we explain why they learn particular things in a particular context? The data collected spanned eight 75-minute lectures, three written homework assignments, six online homework assignments, and one exam from a developmental math class at a large US university. This constituted the material from the class period following the first exam to the class period directly prior to the second exam, and the exam itself. The content included creating graphs on a graphing calculator, using the graphs to solve inequality and optimization problems, solving linear equations by hand, and using slope to find horizontal or vertical changes in right triangle problems. The data corpus includes videos of the lectures, videos of each student doing the homework, and an interview with each student about their graded exam. In the homework sessions, the interviewer asked students to “think-aloud”, asked where they had learned particular ideas, showed students clips from the lecture video, and asked what they learned from that portion of the lecture. In the post-exam interview, the interviewer and student discussed where the student felt they had learned the material tested in each question. Data analysis is ongoing, following a thematic analysis method (Braun & Clark, 2020). Preliminary findings indicate students learned (a) how to identify independent and dependent quantities; how to use the calculator to make graphs, (b) how to interpret graphs, (c) how solving inequalities differs from solving equations, and how to do both on the calculator; (d) how to optimize functions on the calculator. Students commented that getting answers wrong in the online homework helped them understand how to solve inequalities and optimization problems.

Back To Top
SVG directory not found.
MENUCLOSE