research technology consulting simo goshev alex storer steve worthington ista zahn...

27
Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn [email protected] http://rtc.iq.harvard.edu

Upload: jaliyah-ramsbottom

Post on 15-Jan-2016

225 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Research Technology ConsultingSimo GoshevAlex StorerSteve WorthingtonIsta Zahn

[email protected]

http://rtc.iq.harvard.edu

Page 2: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Consulting Goals

Data analysis support and programming services

Research project planning and guidance selecting appropriate technology for research projects

Facilitating appropriate organization, storage and sharing of data

Training on the use of both established software packages and emerging tools

Page 3: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Scope

Free!

Support the entire social science community

Consults measured in hours rather than weeks or months

Currently doing outreach to departments, student groups and centers

Drop-ins on Fridays at 1pm in the training lab, Appointments, Help Tickets and casual chats in K306

Page 4: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Simo GoshevSteven

WorthingtonAlex Storer Ista Zahn

EconomicsBiological

AnthropologyNeuroscience Psychology

ScopeWho We Are

Page 5: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Alex Storer

BS,BA - UC BerkeleyElectrical Engineering & Computer

Science, Cognitive Science

PhD – Boston UniversityCognitive & Neural Systems

Analysis:Machine LearningSignal ProcessingSurface Based TechniquesSimulationOptimization

Tools:Matlab, R, PythonEmacs, LaTeX, Linux

Page 6: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Steve Worthington

BA / MS – Durham, UKAnthropology & Archeology

PhD – NYUBiological Anthropology

Analysis:Linear models (OLS, GLS, PLS, etc.)Resampling (permutation, bootstrap)Ordination (PCA, LDA, CVA, etc.)

Tools:Mainly RSome SAS, SPSSAquamacs

Page 7: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Choropleth Maps

Pop

Soda

Coke Other

Page 8: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Map Workflow

Merge by

spatial ID

Attribute data

Map

Spatial data

State, county, or zipcode level data on any topic.

Shapefiles, ArcGIS files, etc.

e.g., state or county name.

Page 9: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

ICPSR example workflow

Long data

frame

ICPSR Attribute data

CreateMap

using R

USA Spatial data

ICPSR county level data 2003-2005

Unemployment rate, Crime rate, Federal spending etc.

Page 10: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Cleaning / Merging data

Long data

frame

Attribute data

CreateMap

using R

Spatial data

ICPSR county level data 2003-2005

Unemployment rate, Crime rate, Federal spending etc.

Page 11: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Cleaning / Merging data

Long data

frame

Attribute data

CreateMap

using R

Spatial data

ICPSR county level data 2003-2005

Unemployment rate, Crime rate, Federal spending etc.

Page 12: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Unemployment Rate 2005 (annual average estimate)

Page 13: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Unemployment Rate 2005 (annual average estimate)

1.9%

20.9%

Page 14: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Unemployment Rate 2005 (annual average estimate)

1.9%

20.9%

Page 15: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Unemployment Rate 2005 (annual average estimate)

1.9%

20.9%

Page 16: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Unemployment Rate 2005 (annual average estimate)

1.9%

20.9%

Page 17: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Crime Rate 2004 (annual average per 100K people)

0 13700

Page 18: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Simo Goshev

BA – Sofia, BulgariaApplied Econometrics

MS – McMaster UniversityStatistics

PhD – McMaster UniversityEconomics

Analysis:Econometrics Applied MicroeconometricsPanel DataApplied statistics

Tools:Mainly StataSome R

Page 19: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Help with econometrics and/or Stata

How do I bootstrap estimates from survey data?

Stata does not support weighting for the type of graph I want to create. Can you help?

I want to study the factors underpinning scholar migration. What identification strategies would you suggest?

I have two datasets and I want to merge them. However, there are not unique case ID’s in either. What would you recommend?

I have a large and highly sensitive dataset stored in a SQL database. Can you help me access the database directly from within Stata?

Page 20: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Help with computation/estimation

I am trying to estimate a model but for some reason the routine fails. Could you have a look at my script ?

I am working with a large dataset and my machine is giving up on me. Do you have any suggestions?

Which estimation method is best for…?

Page 21: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Re-normalized Multinomial Logit

Common normalization in multinomial logit

Patron:

• Doctoral Student, Department of Economics

Goal:

• Estimate multinomial logit model with non-standard normalization

Page 22: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

But what if different normalization is needed?

Page 23: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Ista Zahn

BS – University of OregonPsychology

PhD (ABD) – University of RochesterSocial Psychology

Analysis:RegressionMixed ModelsScale Development

Tools:R, Stata, SAS, SPSSEmacs, LaTeX, Linux

Page 24: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Workshops(schedule at http://rtc.iq.harvard.edu)

Page 25: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Other Workshops

Page 26: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

IQSS Services

T H E I N S T I T U T E F O RQuantitative Social Science

a t H a r v a r d U n i v e r s i t y

Research Computing Environment

Page 27: Research Technology Consulting Simo Goshev Alex Storer Steve Worthington Ista Zahn support@help.hmdc.harvard.edu

Contact Us!

[email protected]

http://rtc.iq.harvard.edu/

CGIS-Knafel, Room K306

Fridays afternoons, K018

Twitter: @iqssrtc