Hartford Area SAS User Group Announcement and Directions

[Charles Patridge]

HASUG MEETING ANNOUNCEMENT AND DIRECTIONS

Our next meeting is on: Thursday, February 19, 2015
At
UCONN Medical Center
Cell and Genome Building
400 Farmington Ave
Farmington, CT 06030
9:00 am to noon
Registration will be open until noon on Wednesday, February 18th. 
Refreshments to be served & time for conversation, beginning at 8:30AM.

---

Our Topics and Speakers will be:

Other People’s Macros

Terry Murphy, Yale
Use of Other People’s Macros

• Advantages and Pitfalls
• Group-Based Trajectory Analysis

Proc Traj (free downloadable SAS macro)

• Functional Ability Before Death
• Functional Ability Before/After Fall-Related Hosp.
• Functional Ability Before/After ICU
• Conclusions

Dr. Terry Murphy is an Assistant Professor in the Geriatics Section of the Department of Internal Medicine of the Yale University School of Medicine. He also a senior biostatistician at its Program on Aging and a member of the Biostatistics Core of Yale’s Claude D. Pepper Older Americans Independence Center. Complementing his training in engineering and statistics, Dr. Murphy has been working on aging-related studies for nine years. His research interests involve the modeling of complex datasets that typically involve missing data, longitudinal trends, spatial clustering, and multiply correlated outcomes, especially within the Bayesian context. For over six years he co-led a Bayesian study group and has published methods papers that demonstrate the advantages of Bayesian approaches in spatial applications and longitudinal modeling.

---

One step closer to personalized medicine statistical modeling: Recent advances in testing causal mediation in SAS

Emil Coman, UCHC
This is a presentation of some recently developed SAS code for testing ‘causal mediation’. It reviews first statistical options for estimating classical BK (Baron-Kenny) indirect effects, presents the new ‘causal mediation’ directions and proposed software implementations (currently available in Mplus, SAS, Stata, R and SPSS), and then particularly illustrates the SAS options for continuous mediator and outcome, for simplicity. A brief, necessary review will be shown of the Potential Outcomes (PO)/Counter-Factual (CF) view of mediation, including alternative definitions for direct and indirect effects. SAS options for classical BK testing as simple regression-based models are briefly presented, then more detail on options for causal mediation like mcmed and process (Hayes), mediation (Valeri & VanderWeele), and the recent ‘unification’ perspective of Tyler VanderWeele which proposes a decomposition of effects into controlled direct effect, ‘pure’ indirect (BK), ‘pure’ interaction effect (reference interaction), a combination of them – mediated interaction, as well the ‘percent eliminated’ effect. Some implications for the practice of group-based and personalized modeling of effects will be provided.

Dr. Emil Coman is a health communication researcher with particular interest in causal modeling of complex processes. He applies his modeling tools nowadays to issues of patient centeredness of healthcare, comparative effectiveness of interventions and treatments, disentangling causal mechanism leading to desired change and lack thereof, and a variety of health disparities research questions. He has explored models of dynamic mediation, based on true (or latent) change analysis of dynamic coupling between concomitant and successive changes in two or more longitudinal health outcomes, is promoting multiple-group models of change as alternative to current tests of indirect effects, and is exploring what causal modeling can provide as evidence for the benefit of individual patients.

---

Directions to Next Meeting Place on February 19, 2015    UCHC - http://www.uchc.edu/directions/offcampus.html#cellandgenome

UCONN Medical Center
Cell and Genome Building
400 Farmington Ave
Farmington, CT 06030

Please go to HASUG.org and register for the upcoming HASUG meeting.  Be sure to invite your colleagues and get them to register for future HASUG meetings  by registering their email address.

In addition, we are always looking for new hosting facilities as well as volunteers to make presentations to our membership.  And, if you are so inclined, HASUG steering committee could use some new "blood" and new "thoughts" for future meetings - consider volunteering to be a member of the SC - we are the oldest SAS local user group on the planet earth!

Look forward to seeing you at our upcoming meeting on February 19, 2015 as well as future meetings, too!

THE HASUG Steering Committee

[Charles Patridge]

For those of you planning on attending the HASUG Meeting on Feb 19, 2015, please park in the area designated as "Area 2".

If you have not registered as yet, you have until tomorrow, Wednesday, Feb 18th to register.

The HASUG Steering Committee

 

--
You received this message because you are subscribed to the Google Groups "Hartford Area SAS User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hartford-area-sas-us...@googlegroups.com.
To post to this group, send email to hartford-area-...@googlegroups.com.
Visit this group at http://groups.google.com/group/hartford-area-sas-user-group.
For more options, visit https://groups.google.com/d/optout.