The QMSS research assistantship program, initiated in 2004, provides students with hands-on learning experience and the chance to interact and be mentored by leading social scientists at Columbia University and its affiliated institutions. Students awarded the assistantship work closely with an affiliated faculty member on the faculty member's research, giving students the opportunity to apply the tools and methodologies taught in the core curriculum to current, cutting-edge work in the social sciences. Participating students often use the data and research from these projects as the foundation for their theses.
At the beginning of the academic year, interested students may submit their résumés to participating faculty members. In order to ensure a proper fit, faculty members interview and select the students they feel have the relevant interests and skills for their project. You can read about some past projects below.
Columbia Business School
The Static and Dynamic Allocation of Jobs (2014-2015)
A field experiment in Liberia which randomized job offers and assignment to positions so as to measure the social benefits of job creation and how to optimally assign workers to positions. The RA conducts data analysis using Stata (running regressions, creating figures and tables, etc.), literature reviews, and all other RA work involved in writing the paper.
Contribution of Mutual Funds’ Holdings of Credit Default Swap (CDS) Contracts in 2007-2008 to the Increased Fragility of The Financial System (2014-2015)
Using a comprehensive dataset of mutual funds’ holdings of credit default swap (CDS) contracts in 2007-2008, we analyze the motivation for and consequences of mutual funds’ participation in the CDS market. Consistent with the predictions of Rajan (2006), we document that the largest mutual funds were more likely than the smaller ones to take “hidden tail risk” by selling CDS contracts. Additionally, we show that smaller funds exhibited herding behavior in which they followed leading funds in initiating the selling or speculative buying of CDS contracts on new reference entities. Moreover, the reference entities that attracted the most selling interest from the large mutual funds are disproportionately firms that were perceived to be “too large to fail” or “too systemic to fail.” Our combined evidence indicates that the use of CDS by mutual funds heightened the correlation in financial health among the systemically important financial institutions and contributed to the increased fragility of the financial system. The research assistant retrieves the N-CSR filings by mutual funds from the SEC's website and then record the information of CDS positions. The RA can develop a thesis on mutual fund holdings and CDS using the data s/he collects.
Faculty Investigator: Wei Jiang, Graduate School of Business, Finance and Economics Division
Geisinger Medical Center Access Study (2014-2015)
The broad goals of the research project are to (1) understand the factors that are obstacles to providing the right level of care to the right patients at the right time and (2) provide guidance on capacity and management decisions to assure the more efficient and effective use of ICU care. The project involves statistical and econometric analysis of a large dataset of all inpatient visits to the Geisinger Medical Center in Pennsylvania from 2009-2014. The goal of this analysis are to examine when and why patients awaiting admission to an ICU are blocked in getting a bed (access block) and when and why patients awaiting discharge from the ICU are unable to be moved to a lower-acuity bed (reverse access block) as well as the resulting implications on clinical outcomes. The Research Assistant’s primary tasks consist of data cleaning and processing as well as statistical and econometric analysis.
Employee Ratings of Their Companies (2014-2015)
The primary project is to conduct analysis of employee ratings of their companies to determine the association between those employee ratings and company outcomes such as financial indicators of performance. The RA extracts, organizes, and conducts preliminary analysis of the data using STATA, R, and Python. This project is novel in that there is no quantitative study of employee ratings of this scale, and no other study has time series data on employee attitudes to capture how variance overtime relates to the organizations actions and success. The RA will be able to pursue analysis on cross-industry analysis, take a rare opportunity to quantify culture, and note how changes in the firm (such as getting a new CEO) affects its employees for hundreds of firms over time. The project offers the opportunity to develop new methodological skills such as study design - a skill that is difficult to master in the classroom alone.
The Performance and Fees of Wealth Managers (2013-2014)
The objective of this project is to investigate the role and behavior of the investment advisors of high net worth U.S. households, their skills, value added and fees. By exploiting a unique data set containing information on the portfolio positions and transactions of a sample of high net worth U.S. households over the period from 2005 to 2009, as well as the wealth advisors among which they have distributed the management of their wealth, we are able to investigate various questions related to both the relationship between clients and managers, and the performance and characteristics of the wealth management industry, which, due to lack of high quality data, has been so far under-studied by financial economists. The RA is involved in Stata analysis mainly focusing on panel data methods. He/she also collects information on wealth managers, news and events and possible bankruptcies, SEC proceedings and scandals. More generally, he/she will become familiar with the research method in finance and micro-finance in particular, and will participate in the discussion on how to structure the project.
Internal Conflict and Self Control (2013-2014)
This project that investigates how the experience of internal conflict influences self-control. The research team is interested in investigating how the perception of internal conflict influences individuals’ choices and decisions. The RA conducts literature review on relevant topics (self-control, self-regulation, motivation) as well as assists with data collection and data analysis.
The Stress Buffering Impact of Power (2010-11)
The project is investigating the effect of power on stress. Carney theorizes that power buffers individuals from stress. We are testing her idea that people who feel powerful will experience less stress and recover more quickly from that stress. To test this idea the investigators will run subjects through experiments in which the subjects are introduced to psychological or physical stressors. Prior to the stress manipulation, subjects will be assigned to either a low or a high power condition. Outcome measures include psychological, cognitive, emotional, physiological, and neuroendocrine markers of stress and resilience.
Stress, Risk-taking, and the Mediating Effects of Illusory Correlation (2010-11)
The objective of this assistantship is to collaborate with Dr. Richard Ronay, instructor and postdoctoral associate in the Management Division of Columbia Business School, to develop new experimental protocol in order to explore the relationship between stress and risk-taking. A mystery remains as to the exact relationship between psychological stress and risk-taking. In certain domains, stress has been shown to increase risk-taking, while in others, stress has been shown to decrease risky behaviors. One possible explanation is that stress can tax cognitive resources, thus leading to the use of biases, heuristics, and other automatized processes in place of higher, executive cognitive systems. Illusory pattern perception is one example of these heuristics and has been shown to be associated with stress, arousal, and lack of control. Thus, the goal of the proposed research is to test the possible mediating effects of illusory pattern perception on the relationship between stress and risk-taking.
The experiment will manipulate stress, with participants being randomly assigned to one of two groups: control (no stress) and stress. Participants in the stress condition will be exposed to a stressor, either the Cold Pressor Task or the Trier Social Stress Test, while participants in the control group will complete a modified task to elicit no stress. Then, both groups will complete an illusory pattern perception test and a measure of risk-taking, such as the Balloon Analogue Risk Task (BART). The measure of illusory pattern perception will be a modified version of the Implicit Association Test (IAT). One of the primary tasks of this project will be to redesign the IAT using DirectRT software to measure associations made with the stimuli presented in the BART. Once all of the measures are modified for the purposes of this experiment, an experimental protocol will be written for future use and for submission to the Columbia University Institutional Review Board.
Biculturalism and Cultural Identification as they Impact Measurement of Preferences (2010-11)
The assistant will edit survey questionnaires, recruit Korean participants, and manage the SPSS dataset for a Korean-American study of how culture biculturalism and cultural identification affect the ability to estimate the preferences of American versus Korean constituencies. Additionally, the research assistant is responsible for coding non-verbal and verbal video content for the Women's Negotiation Study.
Evaluating progress towards United Nations Millennium Development Goals from all the Millennium Villages sites (2013-2014)
The M & E team is currently developing a 2015 end-of-project evaluation methodology and framework for analyzing progress towards the United Nations Millennium Development Goals from all the Millennium Villages sites. These sites are located in 10 sub-Saharan African countries. As part of that evaluation, project will entail using different sources of data: quantitative household surveys, economic costing and qualitative process data to help measure progress towards the MDGs. For implementation projects, such as the MVP, multiple-sources of data are required to understand the implementation issues of carrying out a complex development project. The Research Assistant contributes skills related to evaluation methodology, research design and analysis, multi-source data management, analysis and writing towards summarizing MDG indicator results for each of the MV African countries. Necessary skills include: Be familiar STATA or SAS; Conduct academic literature reviews; Have basic knowledge of Microsoft Excel for aggregating financial accounting data; Compile information in coherent and organized manner; Conduct multi-variate regression models; t-tests; Ability to synthesize data and write well. Much of the learning will be done through exploring multiple datasets. There will be good mentorship and flexibility, based on level of statistical understanding.
Faculty Investigator: Uyen Kim Huynh, Center on Globalization and Sustainable Development, The Earth Institute
Unemployment Insurance Programs in Brazil (2013-2014)
The enforcement of social program eligibility is a major challenge in developing countries, where a large share of the labor force is employed informally, in jobs that escape monitoring from the government. In a context of high informality, it is widely believed today that unemployment insurance (UI) programs must lead to large distortions in workers’ and firms’ decisions. On the one hand, informal job opportunities may exacerbate the disincentives for UI beneficiaries to return to a formal job. UI beneficiaries could work informally while drawing UI benefits. On the other hand, informal job opportunities and government poor monitoring capacity may exacerbate the impact of UI on the occurrence of eligible unemployment. The costs associated with actual layoffs, temporary layoffs, or “fake” layoffs (quits disguised as layoffs or officially displaced workers staying as informal workers in the same firm) are reduced if displaced workers cumulate UI benefits and informal pay. Our project (with Gustavo Gonzaga at PUC-Rio) aims at quantifying the costs associated with both types of distortions for the case of the UI program in Brazil, and at using our results to inform policy debates on the viability of social insurance programs in developing countries. We exploit a unique combination of two large restricted access datasets covering 15 years of Brazil's recent history: a longitudinal matched employee-employer dataset covering the universe of formally employed workers and a UI registry covering the universe of UI payments. We exploit the data in several ways, including the use of quasi-experimental techniques to estimate the impact of UI extensions on program costs (difference-in-difference and regression discontinuity design). The data is currently analyzed using STATA. Help from the Research Assistant in order to transfer (part of) the data construction/analysis to software that is more efficient at handling large datasets may be valuable.
Empirical Political Economy (2010-11)
Research with Suresh had an empirical political economy theme - analyzing the interaction between political institutions and economic outcomes. In particular, we examined how political events shaped issues of public finance - or taxation. The primary project utilized forensic econometric tools, usually associated with finance, to explore how U.S. multinationals shelter taxable income and how these activities are sensitive to political events. Other projects entailed working with new micro data on political violence from the developing world, including Haiti and Zimbabwe, investigating questions of voter repression, and vulnerability in the wake of disaster.
The Quality of Growth (2008-09)
In this research project, we ask whether and how aggregate economic growth has brought about the progress of society considered as a whole, using individual income as the proxy measurement. According to the Pareto criterion, social progress can be said to have taken place when no incomes have decreased, and there has been an increase in aggregate income. On the other hand, it is widely recognized that growth involves development, a disruptive process, which － through the structural change which defines it － harms some while benefiting others, at least in the short run. If such a result does arise, does it appear immediately or with a lag? How long does it take for a rising tide to lift all boats? Does the answer depend on who is sitting in them or where they are floating? Our contribution is two-fold. First, we apply existing concepts in a novel inter-temporal context to derive conclusions about whether ‘growth delivers.' In this way, we relate the existing theoretical literature on welfare comparisons to an important applied concern. Second, we develop and demonstrate the application of new concepts which can be used to characterize and even quantify the ‘quality of growth.' Specifically, we apply our work to three empirical cases (Brazil, China and the United States) to show the application of the tools we define in diverse settings and to derive some comparative insights.
The research assistant is expected to conduct literature review on economic growth, welfare criteria, and partial ordering techniques, collect data from various sources to construct the global and regional income profiles, and carry out preliminary data analysis.
Quality Growth and Income Distribution (2007-08)
This project was similar to the more recent one--please see above for more detail.
Towards a 'More Perfect' Payments System (2007-08)
The research assistant's work is part of a larger research project to analyze how Civil War monetary-banking-fiscal policies affected both the means and mechanisms for making long-distance domestic payments, and in doing so forged a more integrated payments system, characterized by lower costs and risks in transferring funds over time. The research assistant's project continues the work from 06-07, examining the impact of how the National Banking Acts (1863-1865) reconfigured correspondent banking relations and caused the greater centralization of correspondent relations and deposits in New York City national banks, reinforcing the city's role as national money center.
Do NPO Actions Promotion of Entrepreneurship in Africa Produce Results? (2007-2008)
The main question is: What actions undertaken by NPOs, international institutions, and CSR departments of large corporations, in the area of promotion of entrepreneurship in Africa, deliver measurable and significant results? The research assistant will do web research, as well as research on donated material, of NPOs and institutions who claim to provide aid to entrepreneurs in an attempt to analyze actual provisions of such aid. In the later stages of the research, the research assistant will design surveys and econometric tests to analyze these claims.
Vouchers and Private School Entry (2006-07)
This project addresses two questions regarding Chile’s educational market. First, what effect does private school entry have on the sorting of students across public and private schools within local markets? For example, does the entry of a private school induce “cream-skimming,” under which the highest-income or ability students exit public schools? Second, what effect does private school entry have on the net outcomes of students within local markets? Do average outcomes rise because of the salutary effects of competition, or do they fall because sorting has deleterious effects on some students?
Socio-Economic Outcomes of Unwanted Children (2006-07)
The research project explores the relationship between the availability of fertility control and social and economic outcomes of children in developing countries. The unique history of abortion legislation in Romania - a "special case" in the field of demography and reproductive behavior - provides an excellent opportunity to study the effects of access to abortion and contraceptives on child outcomes. Using a number of new and unique datasets, mostly from the Ministry of Education in Romania, this analysis will examine in detail the school performance of a cohort of children born around the time of a second change in reproductive health legislation.
Towards a "More Perfect" Payments Union: The Civil War as a Second Revolution (2006-07)
The research assistant's work is part of a larger research project to analyze how Civil War monetary-banking-fiscal policies affected both the means and mechanisms for making long-distance domestic payments, and in doing so forged a more integrated payments system, characterized by lower costs and risks intransferring funds over time. This particular project examines the impact of the National Banking Acts (1863-65), which reconfigured the correspondent relations and deposits in New York City national banks, reinforcing the city's role as a national monetary center. The research assistant 's analysis focuses on New York State banks and examines how the adoption of the NBAs caused many banks to shift their reserve agents-deposits from Albany to New York.
Export Responses to Real Exchange Rate Fluctuations (2006-07)
This research project studies how responsive export flows are to changes in real exchange rates and if this response differs with the size of the real exchange rate movement, the margin of production, and the type of exported good. This evidence can shed light on the theoretical literature that tries to disentangle the main response channels and mechanisms activated during real depreciations.
Faculty investigator: Mariana Colcelli, Department of Economics, Barnard College
What are the Effects of Loans from Microcredit Organizations on Borrowing Households (2005-06)
This project focused on the effects loans provided by microcredit organizations have on borrowing households. Using a methodology that accounts for selection bias, this project examined the impact of credit offered through the type of contract most popular among microcredit organizations, the group contract.
What are the Drivers of Economic Performance? (2005-06)
This project re-examined the long-term underlying drivers of economic performance in a large cross section of countries. Seeking to develop concepts distinct from the commonly used notion of average growth, the project examined episodes of economic failure and success, and how these relate to economic reforms as well as geographic/ecological factors, conflict incidence, and quality of institutions.
Political Economy of Local School Expenditures (2004-05)
This project examined questions related to the political economy of local public school expenditures in the United States. States and districts vary by direct versus representative voting and local versus state control in how school expenditure levels are determined. Reback's empirical work investigated whether there are important interactions between these political institutions and other factors, such as demographic trends, that influence public school expenditures.
The Work and Social Organization of Chinese Miners from 1848-1910 (2014-2015)
Analysis of the first mass migration of Chinese labor beyond SE Asia and the gold rushes of which they were a part as constitutive of the economic integration of the Pacific world, racial politics in the various white settler colonies, and the evolution of a global anti-Chinese racial discourse. Using empirical research in previously unused or underused sources, as well as research on discourses of racial exclusion of Chinese gold miners and the "Chinese Question" in the US (California), Australia (Victoria) and South Africa (Transvaal). The Research Assistant would compile data from 4-5 counties in California that are drawn from IPUMS and the software program is STATA/SPSS. There are also data that needs to be identified in the federal census of industry (also California). Finally, the entire data set needs to be compiled into tables.
Normalizing the Socialist Good Life: Consumption, Consumerism and Political Legitimacy in post-1968 Czechoslovakia (2008-09)
The full chapter will be concerned with analyzing two sociological surveys that were requested by the Communist Party’s Central Committee, prepared according to international norms, and cleaned and filtered by professional sociologists, but never analyzed. What we will do is compare the published data from the first large-scale investigation of Czechoslovak society, undertaken in 1967 on over 13,000 subjects with the hitherto unused data from the similar study of over 17,000 subjects undertaken in 1978. Then we will compare the data from the 1978 study with that of the also unused data from a similar study of over 15,000 subjects in 1984. The questionnaires for these two surveys run to over 200 questions each, and cover a wide variety of social indicators, including asking specifically about the ownership of various kinds of consumer goods and consumption habits. The data will illustrate two important points. First, during the period of “High Normalization” (1968-1977) when the regime was trumpeting consumption and material comfort as an intrinsic part of socialism and the socialist way of life, the material comforts available to the population greatly expanded in several ways. Then, the data from 1984 will show along the same axes that this material expansion ceased as Czechoslovakia slipped into “Stagnant Normalization” (1977-1989).
Faculty Investigator: Bradley Abrams, Department of History
Normalizing the Socialist Good Life: Consumption, Consumerism and Political Legitimate Czechoslovakia after the Prague Spring (2006-07)
This project provided a descriptive analysis of the two sociological surveys that were requested by the Communist Party’s Central Committee, prepared according to international norms, and cleaned and filtered by professional sociologists, but never analyzed. The questionnaires for these two surveys run to over 200 questions each, and cover a wide variety of social indicators, including asking specifically about the ownership of various kinds of consumer goods and consumption habits. The data will illustrate two important points. This project prepared the faculty investigator for the next year's project which completed the chapter of focus.
Faculty Investigator: Bradley Abrams, Department of History
Mailman School of Public Health
MAC AIDS Food and Nutrition Study (2014-2015)
An ongoing, longitudinal survey persons living with HIV/AIDS (PLWHA) in New York City and its northern suburban counties. The goals of the research are to provide policy relevant data on the need for health and social services, client experiences with services, and outcomes of services for health, mental health, and quality of life. The RA assists in the preparation of reports, fact sheets, and manuscripts on subjects ranging from effective food and nutrition services for PLWHA, to food insecurity among PLWHA; creats analytical data sets; cleans data; analyses of data using Stata; and conducts literature reviews.
Faculty Investigator: Angela Aidala, Mailman School of Public Health, Sociomedical Sciences
PREVSIM: A Primary Prevention Microsimulator for Health Policymakers (2014-2015)
The goals of this study are to (1) systematically review and synthesize the prevention science literature to identify high yield primary prevention efforts to prevent leading causes of morbidity and mortality in the United States, and (2) to further synthesize and translate this knowledge for easy consumption by policymakers. We here aim to a) develop a population representative microsimulation tool for each of the 10 administrative regions of the Department of Health and Human Services to faithfully estimate life course morbidity and mortality to obesity, diabetes mellitus, cardiac disease, cerebrovascular disease, and cancer across population subgroups and overall; b) to compare various primary prevention packages within each region and across the United States to maximize efficacy, cost-efficiency, and equity of interventions; and c) to create an online, user-friendly, freely available tool that policymakers can use to understand the implications of various primary prevention efforts for health policy. The RA conducts a systematic review of the peer-reviewed and gray literature regarding the effectiveness, cost-efficiency, and equity implications of primary prevention interventions; regression-based data analyses of large, national datasets (NHANES, NHIS) using SAS and/or STATA toward the construction of the PREVSIM microsimulation model; helps program the PREVSIM microsimulation model in Python; and prepares manuscripts for peer-reviewed publication.
Influenza Outbreak Prediction: Applying Data Assimilation Methodologies to Make Skillful Forecasts of an Inherently Chaotic, Nonlinear System (2014-2015)
In order to understand how influenza forecasts are used to inform decisions by lay individuals, government officials and public health workers, the research assistant will be working with me and researchers at the Center for Research on Environmental Decisions to quantify the effects through survey design and quantitative analysis. These findings will be used to improve communication of infectious disease forecasts. The research assistant is able to explore how the forecasts we produce inform decisions and result in feedbacks on behavior and infectious disease outcomes. He or she has the opportunity to work with my group and our dynamic modelers and statisticians, as well as the researchers at the Center for Research on Environmental Decisions. These interactions provide a wealth of opportunities to study the interface between natural and social science systems, and to approach these problems using qualitative and quantitative approaches.
Sexual Networks and HIV Infection in Malawi (2010-11)
Whereas sexual relationships among low-risk individuals account for the majority of HIV infection in sub-Saharan Africa, limited knowledge exists about the structure and characteristics of sexual networks among the general population in sub-Saharan Africa. The research assistant will assist in network analysis of data collected via a roster of potential sexual partners and a sexual network survey conducted in region of interest.
Social Dynamics of Sexual Initiation: A Complex Systems Approach (2009-10)
The key feature of this model is that the long-term dynamics of the prevalence of sexual initiation are driven by 1) the rate of change within each cohort and 2) the feedback of social influence from higher position cohorts to lower position cohorts, both of which are highly influenced by the social network topology between and within cohorts. Prof. Orr had already developed a prototype agent-based model of this process, but it lacked the social network component. In this model, there is clear evidence that, when the diffusion rate is relatively high within a single cohort, this can contribute to a rapid increase in the prevalence rate within a position. However, this project needed expansions to make it into a publishable study of excellent quality, which the Research assistant assisted with.
The research assistant’s tasks included learning how to program an agent-based model and conducting relevant literature searches. The modeling environment implemented was Netlogo, and the researcher provided the research assistant with training in the program.
Can Small Class Sizes in Childhood Reduce Health Disparities in Adulthood? (2009-10)
This project explored whether a proven education intervention, small class size in grades kindergarten through 3, reduces health disparities by income and race in adulthood. Reduced class size in childhood has been shown to improve educational outcomes, such as high school graduation. Improved high school graduation rates, in turn, are widely believed to lead to increases in earnings, health, and to reduce premature adult mortality. Specific research questions explored included: What is the relationship between education in childhood and health, disability, and earnings in adulthood? Through which mechanisms does education work to produce health? What are some of the specific causes of adult death that childhood education might avert? This study used data from Project STAR (Student Teacher Achievement Ratio), a 1985 multi-center randomized trial in Tennessee of 11,601 students in kindergarten-3rd grade across 46 school districts.
The researchers linked data from subjects enrolled in Project STAR to Social Security Administration (SSA) earnings and disability data, providing them with information on each subject’s earnings, disability claims, and vital status of each of the 23 years since subjects were first enrolled in the Project STAR study. This data was then used to examine the causal relationship between reduced class sizes in childhood and disparities in earnings, disability, and mortality in adulthood. As the Social Security Administration has strict privacy policies that limit the use of its data, it was necessary to write SAS code and then send it to them for execution.
The research assistant worked closely with the researchers, Peter Muennig and Ty Wilde, generating SAS code for submission to the Social Security Administration and was given the opportunity to participate as an author on the paper by reviewing the literature and providing extensive editing.
Can Physicians Improve Physician Performance? (2009-10)
Together with Dr. Lawrence Casalino of Cornell University, Dr. Wilde studied the impact of feedback on physician performance to determine if physicians improve after they’re told how well, or badly, they’re doing. They used a uniquely rich and valuable data source from the New York Business Group on Health that tracked the performance of 15,000 physicians in the New York area from 2005 to 2008. Because of a quirk in how these reports were distributed, they deployed an innovative statistical method called regression discontinuity analysis. This method has recently proved to be a powerful tool in empirical economics, but has been little used in health services research. They also stratified the sample to evaluate the results for a key group of physicians – those who work in poorer neighborhoods. The burden of many chronic illnesses is particularly heavy in low-income and minority populations, so physicians serving those communities should be a special focus in efforts to implement preventive-care improvements. Given the importance of chronic care to any future health reform efforts, it is essential to understand how feedback affects physician performance.
The research assistant focused on cleaning the raw data, using STATA to infile the data, create variables, label variable names, check for missing values, and find logical inconsistencies within the data, and assisting in the analysis of the data by tabulating variables, running regressions, and creating tables and charts for presentation. Additionally, he aided in writing grant applications.
Agent Based Modeling of Social Norms (2009-10)
This project involved construction of a computational model of the spread and emergence of social norms in an adolescent social network. In our model, agents randomly distributed across a social network change their behavior based on the actions of others closest to them. Based on the updating of actions, we show that complex systems of social norms are self-organizing, as well as self-perpetuating. We hope to improve on our modeling, making the decision processes of the agents more realistic, and testing the emergence of social norm behavior against school social network data. The goal of our research is to identify entry points for public health policy.
Advanced Methods in Understanding the Relationship among Mood, Drug Use Triggers, and Health among Gay Men in New York City (2009-10)
The assistant worked on three inter-related tasks conducted in collaboration with Dr. Wilson. The first was analyzing data from a longitudinal study that uses a weekly, Internet-based drug use diary methodology to examine relationships among substance use and mental/physical well-being among 200 HIV-positive substance-using men who have sex with men (MSM) in New York City. The research assistant conducted advanced statistical analyses including multi-level modeling (i.e., HLM/random effects regression analysis), latent growth curve modeling, and multi-variable panel plotting. The second task involved participation in the collection and analysis of pilot data from 20 MSM to examine the feasibility of using a daily phone-based diary to describe relationships among mood and drug use triggers among MSM. Both studies involved the use of innovative methodologies (i.e., Internet- and phone-based weekly or daily diaries), which allow for rigorous statistical analyses that can provide enhanced descriptions of the nature of the causal relationships among key variables. The final task involved active participation is the write-up of an R01 grant proposal, including helping to shape the methodological approach, that aimed to explore health, mood, and drug use triggers among HIV-positive and HIV-negative MSM in New York City.
HIV Risk Behaviors and Rapid Testing Among Washington Heights, NY Residents (2014-2015)
The Research Assistant would help analyze data collected for a Washington Heights/Inwood Informatics Infrastructure for Community-Centered Comparative Effectiveness Research (WICER)- affiliated study. This study examines characteristics of approximately 200 individuals who have a greater likelihood of accepting an offer to receive rapid HIV testing and post-test counseling. Data from this study will also be merged with the larger WICER database (N=5868) to examine research questions related to a variety of health outcomes. The Research Assistant assists in developing and implementing data analytical plans to generate reports and disseminate research findings. The RA is expected to conduct basic analysis such as frequency distributions, and more advanced analyses such as logistic and multiple regressions. Familiarity with SPSS and/or SAS is necessary. These programs are used for all data analysis. The RA will be given co-authorship on any manuscripts published in peer-reviewed articles and/or submitted to professional conferences and meetings.
The causes and political consequences of mass and subgroup changes in public opinion concerning domestic and foreign policy issues in the United States, including especially the most recent period of partisan polarization in American politics (2014-2015)
The research assistant assists in identifying a large number of policy issues to study, as well as assists in updating data related to long- and short-term trends for issues for which partisan and ideological conflict have been great in recent years. This work heavily uses the Roper Center for Public Opinion Research's data archive as well as other online sources. This is in a sense a "big data" project since we will need to assemble a great deal of mass and subgroup opinion data, as well as learn the most efficient way to search and assemble the data from the online holdings, which have benefitted from some recent software advances at the Roper Center for example. For the 2014-2015 year, I would like to assemble data for 10-15 possible papers that track fully opinion trends on 10-15 different issues and could lead to submission of papers to "The Polls--Trends" section of the journal Public Opinion Quarterly.
The Political Economy of Energy Use in Rural India (2014-2015)
The goal of the project is to understand the evolution of rural energy use and the policies behind it.The research assistant manages and prepare the dataset on rural energy use with more than 250 villages and 5,000 households, conducts descriptive and explanatory analysis in Stata and/or R using econometric and statistical methods, and conducts regression diagnostics and problem solving.
Testing game theoretical models of negotiation and committee bargaining in dynamic environments (2013-2014)
The Research Assistant will work on a project that plans to employ laboratory experiments to test game theoretical models of negotiation and committee bargaining in dynamic environments. This broader research project includes two related experimental studies. In the first study, the PI designs and conducts experiments that look at the consequences of (private and public) communication on bargaining processes conducted by standing committees (that is, committees that can reconsider their initial decision at any point in time). In the second study, the PI designs and conducts experiments that look at the consequences of different institutional arrangements (deliberation by unanimity, deliberation by majority rule, or oligarchic decision by an elite) on the financing of public goods and the accumulation of public debt. The Research Assistant helps the PI in designing the experiments, testing the software, recruiting the subjects, administering the experiment and analyzing the data. The quantitative skills this project requires are the ability to use STATA (or, alternatively, R) and Excel to analyze and visualize the data collected in the lab. Regarding the data collection process, I plan to train my Research Assistant to design and administer a laboratory experiment and I do not expect him/her to have any earlier knowledge or skill in these tasks. Basic knowledge of game theory is preferred.
Context, Strategic Choice, and the Design of International Agreements (2010-2011)
This project engages a paradox of rational choice institutionalism in international relations (IR) theory. Our theory-testing agenda arises out of a simple observation: negotiations over the creation, or modification, of agreements take place within case-specific networks of prior legal commitments (‘institutional context’), together with past histories of interactions (‘social context’). We hypothesize generally that variation in institutional and social context is an important determinant of beliefs about the nature and severity of the problems states face in overcoming barriers to collective action. Such beliefs, in turn, affect actor preferences over what formal mechanisms (e.g. ex ante limitations on duration, provisions for monitoring, exit, termination, and dispute resolution) are needed to reach agreement, together with possible tradeoffs among them.
To test a set of more detailed hypotheses about the expected relationship between legal/institutional embeddedness and the design of agreements, we have constructed an original dataset of bilateral agreements, state participants, and membership in an international intergovernmental organization. The agreements were selected randomly using the United Nations Treaty Series from among the roughly 34,000 bilateral treaties deposited at the United Nations as of 2002. Also included, however, is data on state membership in multilateral intergovernmental organizations from the Correlates of War-2 (COW-2) International Organizations Dataset Version 2.0.
The research assistant will work with an appropriate social network analysis tool for mapping and analyzing institutional relationships among states and help to visualize and analyze connections among state dyads in the bilateral data and within the organizations in the COW-2 data.
Faculty Investigators: Tonya Putnam, Department of Political Science; Mark Copelovitch, Political Science at University of Wisconsin - Madison
County Level Election Returns, Policy Salience, and Candidate Characteristics (2010-11)
This project works with faculty from Harvard and the University of Chicago to examine the development of electoral competition in U.S. politics over the course of the 20th Century. A large part of this project concerns developing methods for gathering historical information and transforming that information into a format that can be analyzed using quantitative methods. We will be creating datasets that can be used to answer questions such as: “Why do we have a large incumbency advantage in U.S. elections today when this advantage did not exist prior to the 1960s?,” “Why did U.S. states adopt mandatory primary elections for all statewide and federal offices during the 20th century, and what has been the impact of this institution on U.S. politics?,” and “Why did U.S. states adopt civil service reforms and what impact did this have on electoral competition?”
The research assistant is involved in several aspects of the project. One of the primary tasks will be to assist with data collection. We are currently working on creating three major databases: County Level Election Returns, Policy Salience, and Candidate Characteristics.
Should Parties Move to the Right or the Left to Get Votes? (2010-11)
This project is replicating a recent QMSS thesis project that asked “Should the Democrats move to the left on economic policy?” using 2004 National Election Study data, using National Election Study data from other years to get a time series of estimates. It will then look for similar survey questions in other recent polls to get some independent verification of the determined estimates.
Cross-national Election-related Violence and Violent Overthrow Attempts (2009-10)
This project involves three projects to answer the larger question. First, the research assistant will assist in cleaning GIS cadastral data from colonial Kenya. Next, the student will collect cross-national data on election-related violence and violent overthrow attempts. Finally, the student may assist in cleaning data from a survey on property rights in Kenya.
Political Geography: Explaining Trade Protection in the U.S. (2006-07)
This work draws on research from the previous year's project, which showed that political representatives are encouraged to provide assistance in the form of tariffs to declining industries, because they are credibly assured that additional output in those industries will remain in their district, using existing plants and facilities. This research project applies that argument to trade protection in the United States. The methodological interest of the project requires merging geographic data on industry location and industry characteristics with political geography data on Congressional districts.
Do Peacekeepers Keep Peace? (2005-06)
The project here was the initiating phase for a book written by Fortna: Does Peacekeeping Work? : Shaping Belligerents' Choices after Civil War. The investigators are interested in historical changes in war initiation and termination, where civil wars see an increase in declarations post-1945 while inter-state wars appear to decline in formalization of violence during this period. The project would involve the compilation of data that exists with regard to these questions, including lists of relevant variables, etc. The research assistant will compile data not only on the post-1945 period, but previous periods of war, and brief summaries of the data for each civil war.
Political Consequences of Social Inequality (2005-06)
This project examined the political consequences of social inequality, particularly its potential role as a source of social stress. Using vast panel data sets from the UK and Germany containing individual level information on income and consumption, as well as on political attitudes and life satisfaction, this research explored the political consequences of the rise in income inequality of the past 20-30 years.
Under What Conditions Do States Die? (2004-05)
This project asked a fundamental question for international relations: under what conditions do states die? Fazal argues that geography plays a key role in determining state survival and death. Using the technology of Geographic Information Systems (GIS), this project collected and analyzed historical data on terrain roughness for states in the international system over the past 200 years.
Faculty Investigator: Tanisha Fazal, Department of Political Science, University of Notre Dame
Brain Mechanisms of Conscious Visual Perception (2008-09)
Placebo Effects and the Expectation of Pain (2007-08)
The development of noninvasive techniques for studying human brain activity has convinced many scientists that the time is ripe for major breakthroughs in understanding the neural bases of cognition, emotion, and social behavior. Indeed, it can be argued that understanding brain function using human neuroimaging is currently a central locus of knowledge integration in the sciences, with major contributions from psychology, biology, the physical sciences, and engineering. Neuroimaging, and functional magnetic resonance imaging (fMRI) in particular, is unique among neuroscience methodologies in that it provides hundreds or thousands of simultaneous measurements of activity throughout the brain in each participant. However, most commonly used analyses do not make use of this information: data are usually analyzed one brain region at a time- ignoring information on the functional interconnectivity of brain regions and its relation to psychological states- or modeled as distributed ‘sets' of coactivated regions in multivariate analyses. Few existing methods allow for inferences on functional pathway strength, and none provide a way to locate brain regions that form part of a functional pathway in a whole-brain analysis. Therefore, there is a need for analysis methods that can identify and test the integrity of functional pathways and their links to context and behavior.
Exploring Stress as a Social Psychology Experiment (2007-08)
A major reason for the dearth of connectivity analyses is that fMRI data are multilevel: the first level models connectivity among regions over time for a single subject, and the second level models the variability in these patterns of activity across subjects. Multilevel connectivity models that would be appropriate have not been fully developed, though some promising inroads have been made ([10-13]).
The multilevel mediation/moderation (M3) framework proposed here is designed to accomplish these goals and allow researchers to localize and test functional pathways in humans. The M3 path analysis approach is a very simple, three-variable form of structural equation modeling (SEM) that has been extended to the multilevel setting , making it feasible to treat linkages (i.e., connectivity between regions) as random effects. As developed for fMRI, the M3 framework will allow functional pathway building in a hybrid exploratory (search) / confirmatory (inference) fashion in regions of interest (ROIs) or whole-brain fMRI. The M3 approach combines two concepts that have been enormously successful for modeling multivariate relationships in the social sciences: the concept of mediation/moderation in path analysis, and the concept of mixed-effects (or hierarchical) models.
Quantitative Methods in the Social Sciences
Pilot Feasibility Study: Creating a Sample of Teachers (2010-11)
The Farkas Duffett Research Group (FDR) and the Quantitative Methods in the Social Science Program (QMSS) will work together on a pilot data project during Fall 2010, developing a technique to generate a list of teachers and their contact information, drawn from a small sample of schools. The goal of the project is to determine the feasibility of generating a list of teachers – complete with subject area, phone, email addresses, instructional grade, school name and address – using data from school websites and through telephone contact with schools. In addition to feasibility, we are also trying to develop estimates of how long it will take to complete the tasks of selecting a subsample of viable schools and collecting information on said schools.
World Newspaper Project: The Future of the News (2010-11)
Working with the World Association of Newspapers (WAN), the research assistant conducted extensive analysis of data that the WAN had collected in the 2010 Future and Change Study. The Study was conducted by the Shaping the Future of the Newspaper project of the World Association of Newspapers and News Publishers' (WAN-IFRA), in partnership with the Norwegian School of Management (BI) and the University of Central Lancashire in the United Kingdom. Conducted in nine languages, the survey aimed to reach top executives from management, commercial, and editorial sectors to understand how each branch views the changes that must happen to manage change and flourish in the new media environment.
Assessing the Relationship between Media, Freedom, and Democracy (2010-11)
The Media Map project will fill the gap in current research, regarding the most effective use of media resources, by analyzing the empirical correlations between data on the media and its enabling environment on the one hand, and data on development on the other. It will also closely describe how specific media and communications interventions have affected development outcomes, focusing on the impact of media assistance in the development areas of democracy and governance, economic growth and poverty reduction, human rights, and gender equality and health.
A two-year collaboration between Internews and governance, development, and media economists at the World Bank Institute and the Brookings Institution, the project is composed of three phases. In the first phase, we collect and examine existing data on media and global development. The second phase, simultaneous to the first, will investigate the evolution of how donors evaluate the impact of media development interventions over the past two decades by reviewing available studies and conducting a series of interviews. The third phase uses five country-level case studies to analyze the most effective investments in media development and their outcomes from interventions over the last 20 years. Due to the extended nature of this project it will hold several research assistants.
Evaluating Data Sources in Creating a Census of Physicians (2010-11)
Working as part of the Built Environment and Health Project, this research focused on compiling spatial databases to support research on celiac disease, behavioral health risk prevalence, and NYC restaurant inspections. Data from the U.S. census, ReferenceUSA, Center for Disease Control, and New York City’s Department of Health were harvested, cleansed, and geoprocessed to build data sets that will be used as input to future statistical analytics. The primary goal of this assistantship was to utilize geographic information systems and specifically how to process data through ArcGIS.
Faculty Investigator: Institute for Social and Economic Research and Policy
Aesthetic Categories in the Iberian Worlds: Networks, Productions, and Inventions of Early Modernity (2009-10)
The Constitution and the Courts (2009-10)
The research assistant worked with faculty at the Columbia Law School to analyze data from a collaborative national study on the Constitution and the Courts.
Evaluating the Penny Harvest Program (2009-10)
The first project under CREED's umbrella will be its work with Common Cents Penny Harvest Program which is already occurring. CREED has already published work on the first wave of this evaluation under QMSS/ISERP and is continuing data collection, evaluation of collection materials and processes, and grant-seeking to expand collection to more Penny Harvest sites.
Assistant responsibilities will be related to the work with Penny Harvest. The assistant will have several tasks. First, the assistant will assist with the administration of surveys to grade-school students participating in the Penny Harvest. Also, the assistant will be primarily responsible for transferring survey data into analytical formats and preparing the analysis of data. The data entry will take the assistant through the process of survey research from transcribing survey data into analytical format to coding and cleaning data. Additionally, theassistant will participate in the continuing development of the instrument itself resulting from and contributing to the coding and cleaning of existing data. Additionally, the assistant may assist Center staff in developing existing literature repositories for future work in this area.
Evaluating Job Training for the Formerly Incarcerated (2009-10)
Working with data provided by a New York City-based non-profit, this project examined the effectiveness of a comprehensive employment services program designed to assist people returning to New York City from stays in prison and jail. The program is structured to help former inmates, immediately after their release, to find rigorous pre-employment training, paid transitional work, job placement services, and long-term retention services to help participants learn to work, obtain employment, and stay employed. This study was conducted to evaluate the impact of post-employment services on employment retention among program participants.
Examining the Effects of School Quality on Later Life Cognitive Outcomes (2008-09)
This effort is part of a larger study to examine whether and to what extent schooling - particularly the amount of schooling and school quality - influences a set of cognitive outcomes in later life. The researchassistant worked with various archival sources to obtain historical data on aspects of school conditions and quality throughout the United States in the first half of the twentieth century, then linked these data to cognitive assessments of a group of elderly residents of Upper Manhattan.
Built Environment and Health (2008-09)
Faculty Investigator: Katheryn Neckerman, Institute for Social and Economic Research and Policy
Collecting New Accounts Related to the Rockefeller Drug Laws (2008-09)
Related to the previous year's project examining the effects of implementing a new, much harsher legal framework for combating drugs, this project sought to better understand the social conditions that led to the creation and implementation of the Rockefeller Drug Laws in the early 1970s. The research assistant examined data from press archives to assess the public mood toward narcotics and various strategies for treating offenders and addicts, with an eye toward identifying both social trends in attitudinal change and whether any key significant events helped to change public sentiment.
Rockefeller Drug Laws Case Study (2007-08)
Recent research on the rise of the current incarceration regime in the United States has examined the effects of political realignments and changes in sentencing policies. In this analysis, we extend this research by examining the differential impact of what we label “higher order” policies, such as the creation of sentencing laws, with the implementation of such policies at the grass roots level. Using the implementation of New York State’s Rockefeller Drug Laws as a case study, we examine how the development and implementation of these laws contributed to a rise in the incarcerated population. Our preliminary analysis indicates that while the passage of the Rockefeller Laws was a necessary condition for a rise in the incarcerated population, it was not sufficient. Moreover, our preliminary analysis also shows evidence of the importance of predicate or second felony offender laws on the growth of the prison population.
Examining Patterns of Immigration and Police Activity in New York City (2006-07)
Faculty Investigator: Aarti Shahani, Revson Fellow Institute for Social and Economics Research and Policy
School of International and Public Affairs
Oversight Mechanisms in the Peruvian government’s anti-corruption agency (the Contraloría General de la República) (2014-2015)
Our ultimate goal is to test different oversight mechanisms for curbing corruption in public works. As a point of clarification, the term “public works” refers to infrastructure projects that are financed and constructed by the government. These projects may include the paving of roads, the laying of water pipes, the equipping of public hospitals, and so on. The research assistant helps collect, organize, and analyze data, including sensitive government information; is responsible for non-administrative tasks that help move the project forward; finds and summarizes relevant readings, analyzes census data, and copyedits draft papers. The research assistant acquires first-hand experience conducting field experiments, learn about dealing with the logistical and public relations challenges of launching collaborative studies with government agencies, and engage the following crucial research questions:
1. How do we ensure that unelected officials are accountable to the public interest?
2. What are effective strategies for fighting corruption?
PoLAR, a 5-year project funded by the National Science Foundation as part of the Climate Change Education Partnership (2014-2015)
This project uses a variety of innovative methods to educate the public and specific segments about climate change, and it has a significant research component. The RA works with several sorts of data from one component, Future Coast, an internet-based game activity. Participants are invited to compose voicemails from the future, reporting on conditions in the future. The voicemails are uploaded to a site, where participants can listen to them and link multiple voicemails into storylines. The RA conducts multivariate analysis of the voicemails, which have been coded on a number of thematic and expressive variables; conducts analysis of the links between voicemails from users on the website, as well as analyzing the traffic on the website and on social media (Facebook, Twitter); and produces visualizations of the data. SPSS, SAS and STATA would be useful. Familiarity with R also a necessary for this project.
Interviews/Group Discussions about Glacier Retreat (2014-2015)
The RA supports the analysis of data on climate change perception and adaptation. These data come from mountain communities located near glaciers that are rapidly shrinking. The data include focus group transcripts and minutes of community meetings from sites in the Italian Alps, Peruvian Andes and western US. The focus groups responded to open-ended questions about the glaciers and changes in the communities, while the community meetings were self-organized, addressing the issues raised by community members themselves. We are interested in the socio-spatial scale (how large is the group for which the speakers are concerned) and the temporal scales (how far in the past and future they look) that the speakers use when discussing their communities. I seek to understand the connections between: individual socio-economic and demographic data, socio-spatial and temporal scales, the forms of change community members are most concerned about, their perceptions of the change (in terms of uncertainty, severity, reversibility), and the possibility of responding to the change. Desired project outcomes include (1) differences in perception between and within the three sites;(2) differences in perceptions between focus groups and community meetings, since they correspond to different social settings; these settings, in turn, could influence the perceptions of change. Proficiency with R, multivariate statistics and large data sets are all needed.
The Social-Scientific Analysis of Language (2014-2015)
The research team is using large text corpora of political, legal, and contractual language. We plan to use the set of judicial decisions on the Federal Circuit Courts of Appeal to implement a LASSO estimator for the effect of legal variation. Additionally, we are using similar natural language processing techniques to examine CEO contract language in the vast collection of regulatory filings at the SEC's EDGAR database. Finally, we are building a large corpus of collective bargaining agreements with the help of Cornell's Kheel Center to analyze the relationship of labor union contract clauses with firm outcomes. The project has already yielded a number of new datasets, on CEO pay, labor union contracts, judicial opinions, and Congressional language. The scope for further analysis of these datasets is quite large, depending on the student's skills and interests. As just one example, a student might look at the effect of the TARP caps on CEO pay on the language of CEO contracts. The RA conducts Programming in Python, R, and Stata. In particular, Python is used to clean corpora and convert them to vectors of text features (such as N-grams). R is used to run many sophisticated text-processing algorithms, and Stata is used to conduct final analysis.
Judgment and Decision Making and Public Policy (2010-11)
The project applies Judgment and Decision Making research to public policy. There is a status quo bias such that people, all else equal, tend to prefer things the way they are (i.e., the status quo). Because of this, people are often resistant to change. However, if a change is made, people slowly adapt and the new state eventually becomes the status quo. At some point in time, this new status quo is preferred. The project is investigating this process as it applies to changes in public policy, such as the 2003 NYC smoking ban. The questions it asks are: Do people eventually come around to changes in public policy (e.g., Do they eventually approve of / prefer no smoking in bars/restaurants)? If so, how long does it take for public opinion to come around? Is it a gradual, linear process?
The research assistant helped with various facets of the process, including data collection (using LexisNexis to search for relevant articles), data coding (using Excel and NVivo to code for pro and con statements and LIWC and LSA to code for other statements), data analysis (using Excel, SPSS, and R to analyze the data), and writing up the results for publication.
Shaping the Future of the Newspaper (2010-2011)
The World Association of Newspapers and News Publishers' (WAN-IFRA) Shaping the Future of the Newspaper project produces the annual World Newspaper Future & Change Study, which captures details of publishers' financial challenges around the world, as well as their specific strategies to rebuild their businesses for the future - all with new products, cost-cutting tactics, and organizational and employee development. Newspapers are learning how to adapt to and grow in this new ecosystem, where outside forces, such as technological advances, the economic climate, and reader consumption habits dictate the new building blocks that will enable media houses to survive and flourish. These crashing waves of change have left in their wake both devastation and a mindset of renewal for many of the world's publishers, according to the study's results. In some parts of the world, the economic downturn, coupled with debt challenges and advertising revenue shortfalls, have left many newspaper companies shifting gears into new, high potential revenue strategies. The 2011 Study, sent to thousands of media executives and editors in chief worldwide in ten languages, will be conducted from November 2010 to April 2011. The QMSS research assistant is contributing to the analysis of the collected data.
Journalism Training on Business and Economic Reporting in Sub-Saharan Africa (2009-10)
This study is focused on determining the effectiveness of journalism training programs in sub-Saharan Africa and whether such programs improve coverage of economics/business in Africa. The media in much of sub-Saharan Africa is severely constrained by several factors: lack of resources, government pressure, the influence of media ownership and the declining quality of secondary education and professional education. In many countries, newspapers are unable to perform the role of watchdog or effectively educate the public in part because of the difficulties faced by the journalists in their employ. Into the breach has stepped a plethora of foreign organizations, including NGOs and professional media groups, which provide journalism training. The final research paper will provide an evaluation and impact analysis of these training programs, and one of objectives of this research project is to provide recommendations for future training.
The responsibilities of the research assistant included: helping design the survey questionnaire and the sampling and survey procedures, using software such as Atlas TI or NVivo to analyze the results of our survey, employing statistical analysis techniques, and helping determine appropriate ways to quantify certain qualitative data.
Long-Term Water Management in Response to Climate Change (2008-09)
This project is part of a multi-disciplinary effort to develop analytical tools to guide long-term water management and policy choices that are responsive and adaptive to the effects of climate change. In many rapidly growing regions of the world – where we face increasing demands for water supplies that are increasingly variable, both spatially and temporally – such management and policy tools are essential. The project developed statistical analysis tools that can be made available to policymakers, managers, or investors, to 1) help calculate the water supply risk profile of a given location and 2) identify, quantify, and rank, via optimization models, water management investment strategies that will mitigate such risk for a region.
As background to the project, an analysis of the key water supply and demand attributes that are critical for analyzing a region’s sensitivity to climate risk has been performed. We have also collected data for a pilot-scale analysis in the semi-arid tropics of Andhra Pradesh in India to demonstrate the appropriateness of statistical methods for the assessment of climate risk and its mitigation. Longitudinal data on ground water and surface water resources, types of water users, climate, demographic trends, and agricultural production have been compiled over the last year and are readily available. This available data is also supplemented by GIS / remote sensing data.
Faculty Investigator: Tanya Heikkila, School of Public Affairs, University of Colorado Denver
Capital Flow and Balance of Payment in Emerging Markets (2008-09)
The research aims to investigate the optimal level of international reserves for emerging economies through multivariate models and to quantify its influences on domestic absorption to sudden stops and interest rate for the country's long-term debt. The research assistant met regularly with Professor Calvo and his Ph.D. student and discussed about plausible measurements for the probability and cost of sudden stop, and also the opportunity cost of holding reserves. He also assisted Professor Calvo in applying probit model with various financial data sets of 16 emerging economies and predicting their future level of international reserves with statistical software.
The Winds of Change: The Global Development of the Wind Energy Industry and the Environmental Movement (2007-08)
This project focused on the research question: why is it that, while wind power stands out as one of the splashiest success stories in renewable energy in some regions, it has failed to reach its true global potential and has developed unevenly in different parts of world? The research assistant had two main responsibilities. The first responsibility was to help with the development of databases that have organizations as units of analysis. He had to find additional dependent variables and merge new and existing files. The second responsibility was to analyze the data using event history analysis with Cox regression models –the dependent variable being the “risk” of buying RECs, recorded as the year when the organization made the decision to buy RECs, using SAS, STATA, and/or SPSS.
Linking Soybean Yields to Weather Outcomes in Brazil (2007-08)
While there has been a lot of research on this subject, the novelty would be to look for break-points in the non-linear relationship between yields and temperature. Professor Schlenker has done this in the US already, and yields start to rapidly decline once temperatures exceed 29C, which has strong implications for climate change. The assistant 's tasks included: a) assisting in preparing the data for analysis, b) linking it to weather outcomes, and c) running regressions, beginning with a simple panel and advancing it from there.
Do Peacekeepers Keep Peace? (2005-06)
This project examines the empirical effects of international peacekeeping in civil wars. The central question of this project is whether peacekeeping is effective in keeping peace. That is, does peace last longer when international peacekeepers are deployed than when belligerents are left to their own devices to implement a cease-fire?
Faculty Investigator: Tanisha Fazal, Department of Political Science, University of Notre Dame
School of Social Work
Investigating a 10-year Dataset on Individuals with Serious and Persistent Mental Illness (2009-10)
Primary psychotic disorders, specifically schizophrenia disorders, bipolar disorders with psychotic features, and major depressive disorders with psychotic features (SD/BD/MDPF), are serious and persistent mental health problems that increasingly affect individuals, their families, and society. The Suffolk County Mental Health data set was initially collected to extend our understanding of the course of illness following initial hospitalization for a psychotic episode (SD/BD/MDPF) by following a cohort of patients with psychotic condition. Using the Suffolk County Mental Health data set, this project aims to: 1) Describe the diagnostic stability of the psychotic disorders (psychotic features, and major depressive disorder with psychotic features) across time. 2) Identify whether the association between treatment adherence and outcomes is similar for individuals with different ages of onset of psychotic disorders. 3) Examine prognostic outcomes of patients with schizophrenia disorders, bipolar disorder with psychotic features, and major depressive disorder with psychotic features. 4) Examine the extent to which stigma, SES, social factors, substance use, comorbidity, demographic characteristics, and health behaviors explain differences in treatment adherence, prognosis, and outcomes. The research assistant used SAS and SPSS 15 software for all analyses. He was invited to assist in contributing with the manuscript development and writing, which can lead to co-authorship on publications.
Clinical Trial Network/Brazil Family Health Program (2009-10)
The research assistant worked with two data sets, the Clinical Trial Network (CTN) data set, aimed to detail the CTN workforce from 150 clinics throughout the United States, and the Brazil Family Health Program (BFHP) data set, designed to study the roles of Community Health Workers and the medical staff of the BFHP in two cities. The research assistant was engaged in the conceptualization phase of one paper from each data set described above. This level of engagement provided the assistant an opportunity to learn how to pose research questions and to form hypotheses. The assistant conducted descriptive analyses, described the samples, and then worked on specific research questions, such as: what differences are there between Community Health Workers and the medical staff in two different cities in Brazil? The research assistant was also asked to build tables with results and to summarize the methods we used. Additionally, the research assistant was encouraged to develop her own project and was invited to co-author a paper.
Circumstances Leading to Parent and Child Health in Fragile Families (2004-05)
The QMSS research assistant analyzed data from the Fragile Families study, which follows a birth cohort of (mostly) unwed parents and their children over a five-year period. The study is designed to provide new information on the capabilities and relationships of unwed parents, as well as the effects of policies on family formation and child wellbeing. The research project examined narrative accounts (rather than traditional survey-based methods) of the circumstances leading to parent and child health in a subsample of the Fragile Families sample.
Parental and Child Health in Fragile Families (2004-05)
The QMSS research assistant analyzed data from the Fragile Families study, which follows a birth cohort of (mostly) unwed parents and their children over a five-year period. The study is designed to provide new information on the capabilities and relationships of unwed parents, as well as the effects of policies on family formation and child wellbeing. The research assistant focused on descriptive understandings of the data and analyzed trends found within.
New York City's Changing Ethnic Neighborhoods (2013-2015)
This project will document the changing ethnic neighborhoods in New York City. In particular, it will pay attention to the city's ethnic enclaves, their historical emergence, their cultural significance and their ongoing transformations. We will seek to identify changing neighborhoods and to explore the history and development of these places, including demographic patterns. We will also examine the culture of places (including celebrations, food, language, housing, etc.), the socio-cultural, political, and/or economic achievements/challenges, and notable individuals from these neighborhoods. The RA works on the following tasks: analysis of census data to identify ethnic neighborhoods and its demographics; historical analysis of the formation and emergence of ethnic neighborhoods; exploratory and descriptive mapping of these neighborhoods using ArcGIS; secondary research on notable traditions and individuals from the neighborhoods; final write-up and synthesis of the analyses. Strong familiarity with Stata and excellent research skills are required for this project.
Race and Ethnic Differences in College Application Patterns and Institution Choices (2014-2015)
This project that aims to examine post-secondary education decisions and choice of institutions using a series of large-scale longitudinal data in the U.S. (High School & Beyond (HS&B), National Education Longitudinal Study of 1988 (NELS 1988), Education Longitudinal Study of 2002 (ELS: 2002), and High School Longitudinal Study of 2009 (HSLS:09)). Each dataset provides a unique array of information related to this project. When pooled together, the four datasets can yield rich insights into students’ post-secondary outcomes. The main research questions include: How do students and parents construct their post-secondary institution choice set and how does it vary by race and ethnicity? How does the process of college preparation and final institution decision differ by race and ethnicity? And how do the above patterns change over the past few decades? I plan to use a series of statistical methods to address these questions, such as basic regression analysis, longitudinal data analysis, and matching techniques. The research assistant assists in reviewing the relevant literature, preparing the four data sets for analysis, and carrying out descriptive and multivariate statistical analysis (basic regression analysis, longitudinal data analysis, and matching techniques).
Network / Hierarchy Analysis of Social Elites (2010-11)
This project explores how actors navigate status. The QMSS research assistant would participate in the quantitative portion of the project involving GIS and network analysis using data from "famegame.com." The questions explored will be: a) How do you climb a hierarchy? b) how do you maintain your position? and c) how does one fall from status?. The QMSS research assistant will provide descriptive analysis on the distribution of the events geographically, what a "small world" looks like with regard to its status system, and what careers of particular actors look like within the worlds of the data.
Venture Creation in the U.S. and Germany: A Comparative Study (2010-11)
Entrepreneurship is one of the key factors of economic prosperity. Given this wide-spread belief of entrepreneurship as a means to stimulate new firm foundations, it is striking to note the limited success of such policies in Germany despite massive subsidies. The reason for this seems to be that German policies mimic those of U.S. Silicon Valley policies. Here German policy-makers ignore that entrepreneurs in Germany are exposed to different institutional constraints.
QMSS research assistants will participate in gathering data via structured telephone interviews with entrepreneurs. Research assistant will be trained for such interviews by the faculty and then independently arrange and conduct interviews with U.S. and German entrepreneurs.
Government Response to Network Failures: The Case of the Manufacturing Extension Partnership (2010-2011)
This project investigates the rationale for government intervention to foster innovation in the United States. Government intervention is typically justified as a response to market failure, and represents an effort to make markets more competitive. However, a growing body of literature argues that innovation tends to occur in "collaborative spaces" that are sheltered from market competition. The focus of the research is to examine whether there is a mismatch between the process of innovation and the rationale for innovation policy. The research examines the relative importance of public inputs to network production among small and medium-sized firms in a study of the Manufacturing Extension Partnerships (MEPs) sponsored by the National Institute of Standards and Technology. The research assistant will help the investigator to quantify the relative importance of the factors contributing to the success of collaboration projects by developing and analyzing survey and interview-based indicators of public inputs to network production among small and mid-sized firms.
China International Migration Project (2009-10)
This project is a comparative study examining the role of immigrants’ transnational ties and social networks in the host societies in the labor market attainment processes among Chinese immigrants in the U.S. and Europe. The study draws on theories of social stratification, transnationalism, social capital, and ethnic enclaves. It attempts to assess 1) the extent to which transnational ties and local social networks provide immigrants with access to capital, labor, goods, and other dimensions of support that facilitate their incorporation into the labor market in receiving societies; 2) the mechanisms through which these two forces interact in determining immigrants’ occupational attainment; 3) how the mechanisms vary for wage earners versus entrepreneurs, and for documented versus undocumented immigrants; and 4) the cross-setting similarities and differences underlaid by context-specific institutional arrangements. The data used is from “The China International Migration Project” collected by collaborators at SUNY-Albany, and it collected ethnosurveys in both the migrant-sending communities in China and the receiving regions in the U.S. and Italy, including adequate coverage of undocumented immigrants. When the research assistant participated in this project it was still at a very early stage, and therefore he primarily assisted in reviewing the relevant literature, preparing secondary data for analysis, carrying out descriptive and multivariate statistical analysis (and programming when necessary), and writing up results.
Trajectories, Tastes, and Network Structures of Elites (2009-10)
This project examined the experiences, social networks, and tastes of students who attended elite boarding schools. Using surveys, interviews, and online media to follow graduates from ten such schools, this project asks how the experiences and trajectories of wealthy students differ from those less advantaged, and how an increasingly diverse alumni differ from those whose families have held advantages for generations. The goal of the project is to better understand the key mechanisms for the production of advantage and begin to make sense of the experiences and trajectories of a diversifying elite. Graduating classes from the top ten boarding schools in the United States will be surveyed 2, 5, 7, and 10 years after graduation, in order to gather social, demographic, and experiential information as well as information on social networks. The research assistant worked to construct the survey instrument, conduct the survey, and analyze the results.
Sociology of Religion and Politics (2009-10)
Dr. Barkey utilized a research assistant to assist in contributing to research of the currently highly relevant topic of the sociology of religion. The three major Abrahamic religions have all experienced a serious renaissance, mostly in their most fanatic forms. As a result even the most adept secularization theorists have admitted that religion continues to be well and alive and even exuberant in its outreach. Furthermore we need to see that there is great variation in the resurgence of religion as an important factor in politics and public life. There is a lot of new research in the field of religion and politics and a more general sociology of religion, though there seems to be less of a gathering of analytic ways of thinking about this very crucial relation. A sociology of religion and a subfield of religion and politics needs to pay more attention to large scale macro historical and comparative questions, causal mechanisms, the various foundations of religious behavior, the role of institutions, as well as the actions of individuals embedded in the context of these institutions.
The Effects of Choice on Race, Ethnicity and Gender Inequalities in Education (2008-09)
This assistantship provided the faculty investigator with work that assisted in the production of her book: Arresting Development: Race, Place, and the End of Adolescence
The Role of Non-Profits in Social Service Provision in NYC (2004-05)
This assistantship is part of a larger study researching the role of nonprofits in providing government-funded social services in New York City. Using data from a public database of contracts for social service provision and new spatial analysis tools, Marwell focuses particularly in the inequity in distribution of agencies and funds and what factors are related to over- and under-representation.
Constraint or Opportunity: What Drives Labor Supply Among U.S. Undergraduates? (2010-11)
Today’s college enrollees are more likely to work, and work more,than those of the past. CPS data reveal that between 1970 and 2003, average labor supply among 18 to 22 year old full-time, four-year undergraduates nearly doubled, from 5.0 hours to 9.6 hours per week. An immediate concern is that this increase may reflect tightening financial constraints. Unless student employment has other benefits, students would be better off borrowing money instead of spending time working, so that they could finish college faster or better. But if students’ ability to borrow is not rising as fast as college costs, they may have little choice but to work. This in turn may delay or diminish their acquisition of human capital, thus decreasing the return on their educational investment. When broadly defined to include “fuzzy” constraints on borrowing for discretionary consumption as well as self-imposed constraints on borrowing, credit constraints appear to be driving the trend even among high-income populations.
More recent data, however, suggest that the long-term trend towards increased student employment may have slowed or even reversed since 2003, which raises questions such as: is this break in trend “real” or simply a statistical blip, perhaps resulting from changes in who decides to attend college?, and if it is real, to what extent is it driven by changes in labor market opportunities? The research assistant will work with more recent data to help develop hypotheses and design and implement multivariate regression analyses to explain the patterns of college student employment over time as well as code data.
Expanding Notions of Social Reproduction: Grandparents Educational Attainment and Grandchildren's Cognitive Skills (2008-09)
This assistantship resulted both in an MA thesis and a published paper. Inherited privilege and status remain powerful factors in the distribution of opportunity in American life. These transfers of socioeconomic resources across generations are facilitated by the links between adult educational attainment and children's cognitive skills. Our current study expands the notion of social reproduction beyond this narrow two-generation approach to investigate the links between grandparents’ educational attainment and their grandchildren's academic abilities. Using a nationally representative sample of over 13,000 children who participated in the Early Childhood Longitudinal Study—Kindergarten Cohort (ECLS-K), we find that familial advantages in human capital persist over time and that these advantages are associated with improved cognitive outcomes among later generations. Even after controlling for a wide array of socioeconomic and demographic characteristics, young children with college-educated grandparents possess stronger literacy and mathematics skills at the start of formal schooling. Propensity score approaches, which address concerns regarding the endogeneity inherent in the topic, yield similar results, suggesting the robustness of our findings.
The Influence of Social and Structural Contexts on Children's Cognitive Development and How they Fuel Social Reproduction (2007-08)
The specific tasks to be completed depend in part on the research assistant's initial skills, and the expertise they develop during the course of their QMSS program. The overarching goal would be for the student to complete a solid quantitative study for which they would have primary responsibility. This may or may not become the student’s master’s thesis. The tasks involved are those that accompany any quantitative study: situating the study within a theoretical framework; selecting appropriate analytic samples; conducting both descriptive and analytic analyses; and presenting findings in logical visual and written formats. The research assistant would also have the opportunity to present this paper—and others on which we might collaborate—at major national conferences, including the annual meetings of the American Sociological Association, the American Educational Research Association, and American Evaluation Association.
This work resulted in a published paper by the student and faculty investigator on Disparaties in Cognitive Gains within Higher Education.