Brown Bag 1: 10/1/14 Joan Newman and Georgia Brooke
Ongoing Research On Effects of Environmental Pollution: An Application of Propensity Score Analysis
At our first brown bag of Fall 2014 we had a 2 speakers: Dr. Joan Newman and Dr. Georgia Brooke speak about their research. Dr. Brooke is a recent graduate from our doctoral program and she used a dataset from an ongoing research project with Dr. Newman for her dissertation. This brown bag provided an overview of the larger research project and how Dr. Brooke used Propensity Score Analysis to the dataset for her dissertation.
The project started in 1995 investigating the effects of PCB pollution on the Akwesasne Native American community. PCBs are man made chemicals that were banned in the 1970s but prior to that were legally dumped into rivers like the Mohawk and Hudson rivers and remain persistent in the environment today. Native American communities were exposed to these PCBs as a population that utilized the Mohawk river and ate fish as a dietary staple. PCBs have previously been found to have negative heath outcomes and community concerns arose about human health, children’s development, and the disruption of culture and lifestyle. Newman et al., (2006) investigated the effects of PCBs and cognitive functioning of Mohawk adolescents and found that PCB levels were negatively related to delayed recall, long term retrieval and comprehension knowledge.
For her dissertation, Dr. Brooke used the dataset to investigate the relationship between breastfeeding and cognition. There have been inconsistent findings in the literature on breastfeeding – a majority of studies report a cognitive benefit associated with breastfeeding while others report no benefit. For this particular population of Native Americans, this relationship was even more interesting to investigate because of PCB contamination in their environment. The question driving the research was: Should women exposed to PCBs through contaminated fish consumption breastfeed their infants? Georgia then showed how Propensity Score Analysis was a useful approach for answering the research question. The previous literature on breastfeeding and cognition could be criticized because of selection bias: breastfed babies may perform better cognitively because of their mothers, not because of breastfeeding itself. PSA addresses the selection bias common to observational designs. Step 1. Generate a propensity score – who is mostly likely to breastfeed and who is unlikely to breastfeed based on certain characteristics (e.g. maternal education, IQ, SES, smoking habits). Step 2. Match treatment and control cases with similar propensity scores, or group similar propensity scores into clusters. Step 3: compare matched cases or clusters on outcome of interest (e.g. cognitive scores). Her results showed different effects for low PSA group (least likely to breastfeed), moderate PSA group, and high PSA groups (most likely to breastfeed), meaning cognitive benefits of breastfeeding could vary by population (which explains the inconsistencies in the literature). The populations that seem to benefit the most from breastfeeding are those most disadvantaged (low birth weight, low SES, cigarette smoke).
Brown Bag 2: 10/29/14 Kim Colvin - Intro to Generalizability Theory
Evaluation of Measures of Resilience and Vulnerability in Post-Earthquake Haiti: Using Generalizability Theory on the Road to a Validation Study
At this brown bag we got a chance to hear our new faculty member Dr. Kim Colvin speak about Generalizability or “G” theory. She provided a short overview of the theory to introduce us to main concepts and terms followed by an application of G-theory to a study she recently worked on involving subjects in post-earthquake Haiti.
G-theory is an extension of classical test theory where an observed score equals true ability plus error. According to CTT, though, there’s only one interpretation of error, meaning all potential sources of error are lumped into 1 error term which can be limiting in field research where conditions of measurement vary. G-theory then gives researchers the opportunity to parse out and estimate different sources of error and leads to better estimates of reliability coefficients. Researchers can estimate what proportion of the total variance is due to factors that often vary in assessment such as setting, time, items, and raters. G theory is particularly useful for assessing the reliability of performance assessments – which is exactly what Kim Colvin did in her research with subjects in post-earthquake Haiti.
The goal of the post-earthquake Haiti study was to validate the House-Tree-Person test, a projective drawing test, as a way to measure resiliency and vulnerability in people affected by the 2010 earthquake in Haiti. The test was culturally adapted for Haiti and was used to identify children who positively adapted or were negatively affected by the earthquake. HTP test data was collected from 77 children ages 4-15 years old in 2012 and 2013 in different locations (e.g. a town not widely affected by the earthquake, relocation camps with varying amounts of aid/resources, and an orphanage). 3 additional assessments were used to collect data in 2013. The goal was to find an assessment that could be easy to administer and score in order to quickly aid relief workers after a traumatic event. The data collected would be used to evaluate where improvements can be made in order to collect large datasets to better inform validation study of the HTP test.
An analysis of the raters to determine differences in scoring due to gender and age showed that there were no differences between genders but there was an age difference for attribute (vulnerable vs. resilient). An analysis based on location showed that the biggest difference was due to attribute and it was most notable for the orphanage location. No variability was due to raters – this is encouraging because it means new raters can be trained and results won’t be affected.
After the presentation, we discussed why the HTP was used instead of other more reliable tests: it is nonverbal, easy to give, easy to score.
Student Panel Discussion: The Ins and Outs of Attending Professional Conferences
Presenters: Fei Chen, Julio McLaughlin, Stella Li, Tom Robertson
Moderator: Angela Lui
Every semester we try to do a student panel uniquely aimed at addressing concerns/issues/topics of being a graduate student in our program. Previously we’ve had panels on the dissertation process, and the first year of teaching experience. This semester we put together a student panel discussion on attending professional research conferences like AERA, APA, AEA, APS and other acronyms. It was a very interactive panel where both students attending and students presenting discussed their ideas and experiences. We talked about some of the purposes/benefits of attending conferences (e.g. networking with other researchers, professors, universities; keeping up to date with current research; traveling; additions to your CV). As a graduate student, however, conferences are also hard to attend because of the expenses involved. Our panel provided students with advice on how to navigate this aspect – how to search for different funding sources, how to use money wisely and attend best conferences for graduate students and how to save on printing/presentation materials. More and more conferences are now offering webinar options, and special sessions and rates for graduate students. Other tips: wear comfortable shoes, plan ahead especially if there are multiple sessions, and the overall perception of conferences is that people are more friendly and approachable/less intimidating at regional conferences compared to the big national conferences.
Brown Bag 3: 11/24/14 Dr. Dai
Cradles for Talent: Specialized Selective High Schools
For our last brown bag of the Fall 2014 semester, we decided to try an evening meeting and attendance was overwhelmingly positive! At this Brown Bag Dr. Dai discussed gifted education, talent development, and specialized selective high schools like Bronx Science, Stuyvesant, Brooklyn Tech, High School of American Studies and Brooklyn Latin School in NYC. First, he discussed the background and the psychological and educational theoretical underpinnings of gifted education. For example: talent development vs. creativity development, formal vs. informal learning, production (prescribed education) vs. client based (choices, designed for what students wants) education systems, and learning content vs. developing a way of thinking. Then we got a closer look into how specialized selective high schools work in terms of their institutional leadership, admission policies, curriculum structure and requirements, and infrastructure supports (i.e. teaching capacity, network of support outside of the school, technology). We discussed the importance of schools having a rigorous curriculum with post AP courses and research oriented courses with a wide range of options and choices. Providing “threshold experiences” such as internships outside of the school, engaging in project based learning, and choosing a major much like you do in college leads to a more open and motivating school culture where students are treated as responsible individuals capable of thinking for themselves and making good decisions. They have good ownership of their learning, strive for the best, and achieve personal goals. It is especially positive for adolescents who are trying to find their niche and explore their identity and new horizons. However, we also discussed some of the problems with these types of schools – particularly the admissions process which is a topic of debate among the schools and administrators. Currently, there is only 1 criteria for admission to specialized selective high schools: scores on the Specialized High Schools Admissions Test or SHSAT. Some people say that is not enough to base admission on and other factors should be considered. Other critics say it leads to racially unbalanced schools, with an overpopulation of Asian students and admissions criteria should be reformed so that the schools’ populations more accurately reflect the area’s demographics. Another issue that was discussed was the amount of stress experienced by students who attend specialized selective high schools and how counseling programs and social and emotional supports are needed not just during high school but afterwards as well when they transition to college (Big Fish Little Pond Effect). It was particularly interesting to hear from some of the students in attendance who went to specialized selective high schools in NYC who could offer their opinions on these issues and discuss their experiences.
On October 11, 2014 we did our annual apple picking and picnic event at Indian Ladder Farms and Thacher Park.
We also went out for a night on the town at the end of the semester to celebrate people passing comprehensive exams and to get together one last time before winter break. So it's a little late but HAPPY HOLIDAYS from the Ed Psych family and friends! :)