2014
Observational and Quasi-Experimental Research Methods
Description: Observational studies draw inferences about the effect of a treatment on subjects, where the assignment of subjects into groups is observed rather than manipulated by the investigator. Quasi-experimental designs are similar to experimental designs in that there is a specific intervention for the “treated”group, but lack the randomization of subjects to treatment and control groups. This workshop will focus on these designs and the inherent trade-offs between internal and external validity. We will use actual data sets to explore examples of selection bias and confounding, and how they are problematic. We will then devote the majority of this session to demonstrating the methods used to address these issues with a focus on multivariate modeling and propensity score analysis.
Link to slides: Here
Instrument Selection and Development
Description: This workshop will address the challenges of instrument selection and development for geriatric palliative care research. We will review currently available instruments for core areas of geriatric palliative care research(e.g., physical and psychological symptoms, communication, caregiver outcomes,healthcare utilization), describe the populations in whom they were validated,and identify limitations in their use. We will address core issues underlying instrument development and focus particularly on issues of respondent burden,responses from participants with impaired cognition, and proxy respondents.Trade offs between using current imperfect measures and the long and complex process of developing new instruments will be considered, together with the need for rigor in patient-reported outcome instrument development. Participants will leave the session with a toolkit of unified core measures to use in their ongoing research.
Link to slides: Here
Use of Secondary Data Relevant to Geriatric Palliative Care
Description: This workshop will discuss the opportunities and challenges in using existing data. Numerous studies have generated rich data sets that researchers can use to address questions related to geriatric palliative care. These data sets give researchers efficient opportunities to develop and test hypotheses and write papers and grants - yet a variety of scientific and practical issues attend the use of these data sets, often impairing their productive and accurate application. This session will focus on answering the following key questions:What makes a data set is a good fit for your research? How do you learn about a data set? How do you structure a paper or grant based upon existing data? How do you decide if the available measures (often not the ideal) are good enough? How do you navigate issues such as access to data, costs, and common analytic pitfalls? Finally, it will review data sets particularly relevant to geriatric palliative care (e.g., Health and Retirement Survey, Medicare claims, Dartmouth Atlas, Long Term Care Minimum Data Set, National Home and Hospice Care Survey).