april 8, 2010 meeter center lecture hall calvin college grand rapids, michigan 1 center for social...
TRANSCRIPT
1
April 8, 2010Meeter Center Lecture Hall
Calvin CollegeGrand Rapids, Michigan
Center for Social Research
Staff
IntroductionJim Penning, Director of CSR
Presenters Neil Carlson, Assistant DirectorChristina der Nederlanden, Research AssociateNathan Mosurinjohn, GIS TechnicianJeff Schiman, Statistical Analyst
Closing remarksEdwin Hernandez, Ph.D., RDV Corporation
3
Overview Presentation purpose About KCCS Congregational population
(2009 census data) Spatial and social distance
measures(2007 survey data)
Local generosity analysis (2007 survey data)
Future research and community impact(Edwin Hernandez)
4
Presentation Purpose Present new information from
the Kent County Congregations Study 2009 congregational census 2007 social network data 2007 generosity analysis
Demonstrate data visualization tools ArcGIS mapping NodeXL network analysis Tableau charting
Prompt discussion of your research ideas and interests
9
Data visualization rising
Low cost, relatively easy-to-use visualization tools arriving
Transition in progress from: Before: Visuals too difficult
and time consuming to create except as final presentation aids
After: Visuals an integral round-trip part of data analysis
10
Data visualization keys Data integration:
a few linked sources, many uses Rapid, interactive what-if analysis Data processing matters:
interdependence with command line and spreadsheet-style tools
Edward Tufte, dataviz guru: Recovery.gov makeover in NYT Japanese math and science
superiority Don’t condescend—show dense info Small multiples have big impact
11
ArcGIS Market leader in Geographic
Information Systems Calvin has a site license ArcGIS Server (gis.calvin.edu)
can support interactive map analysis over the web for students with little training
Dr. Jason Van Horn’s GIS course just approved by Faculty Senate for addition to the core for “visual rhetoric”
12
NodeXL Open-source, free template
add-in for Microsoft Excel 2007, developed by Microsoft staff and a coalition of academics.
Works with any network data: people and their relationships =>vertices and edges
Students can use familiar spreadsheet tool to visualize network data
13
Tableau Award-winning interactive charting
tool developed at Stanford Drag-and-drop interface, “shelf”
metaphor (Super)rows, (super)columns,
shape, color, text, size, filter, page Roots in business intelligence Public version free for general use,
available to students; Server-side: Publish interactive
visualizations to the Web
14
About KCCS Sponsored by the
Doug & Maria DeVos Foundation Goals
Serve the community, especially children in public schools
Create original research Project history
2006: partial census and list collation
2007: countywide survey 2008: Gatherings of Hope report 2009: countywide census
15
www.calvin.edu/go/kccs
16
Sample KCCS facts Just 33% of surveyed
congregations reported working with a public school (but 50% of Reformed congregations!)
583 congregations reported over 2,000 programs and over 9,000 services and activities for members and the general public.
“Replacement value” of these services exceeds $89 million a year.
17
2009 Census
Team effort by staff and students
Census managed byChristina der Nederlanden,recent Calvin grad in psychology
18
Census MethodsMission?
Update the KCCS congregation information (2007) & identify new congregations since last canvas of Kent County.
What is a congregation?
Definition: a group that meets regularly on an on-going basis, comes together primarily for worship, meets and worships at a designated place and has an official name and formal structure that conveys its purpose.
Out in the fieldTwo-student team
Census tract map
Census update forms
20
Canvassing Process
Process began in June of 2009 and ended mid-November 2009.
Able to cover all of Kent County, including areas not completed in 2006.
124 tracts covered, 24 new congregations discovered, 35 moved, 3 closed, 760 congregations confirmed.
21
Neighborhood ContextStreet condition/cleanliness, land use, greenery:
Tableau
Final char
t
27
Spatial relationships
Who is my neighbor in physical space?
Census data includes latitude and longitude of all cases
Nathan Mosurinjohn,recent Calvin graduate in GIS
28
Geographic Information Systems
Coordinate Data Collection
Analyze data
Visualize data
29
Coordinate Data Collection
Create waypoints for neighborhood context
Used MPS Atlas tool to create booklet of maps for canvassing
Data Collection
31
Data analysis
Neighborhood context rating Cronbach’s alpha scale of
several observational items (alpha = 0.67)
“Kriging” interpolated surface with statistical error estimates
Krigedsurface
of neighborho
odcontext
33
Data analysis
Neighborhood context rating
Congregation changes Newly opened or discovered Closed Moved
Over-time change
GR detail, over-time change
GR area, tradition and size
Miles to nearest neighbor by tradition
Miles to nearest neighbor of a different primary ethnicity
Miles to nearest neighbor by ethnicity
Miles to nearest neighbor of a different primary ethnicity
41
Social interactions
Survey respondents were asked to name up to 3 relationships with other congregations or organizations
We coded these to link to any congregation within Kent County
4,658 interactions, 1,589 with other congregations
Baseline measures to evaluate coalition-building efforts
Interaction counts
Interethnic interactions
Inter-traditioninteractions
All Interactions
Joint Service Projects
47
Congregational Social Networks
NodeXL Interaction Types: General
Interaction, Joint Service Project, Personal Associations, Shared Building, Shared Leader, Association
Betweenness Centrality Closeness Centrality Eigenvector Centrality Clustering Coefficient
Diagram of Joint Service Projects
Who are our neighbors?
Who are our neighbors?
Joint service projects with neighbors2 miles away or less
Network statistics by interaction type
Closeness centrality by ethnicity(axis reversed; smaller values = fewer steps to neighbors)
Closeness Centrality by Size and Tradition
54
Local charitable giving
Do we show charity to our neighbors in proportion to our capacity to do so?
KCCS data includes local vs. nonlocal domestic vs. international cash giving
Becky Haney (economics) and Jeff Schiman, recent economics grad, are working on analysis
55
Congregational GivingCongregational giving totals
All ($75,651K)
Reformed ($31,189K)
Mainline or other Protestant ($7,250K)
Other traditions ($803K)
Black ($868K)
0% 20% 40% 60% 80% 100%
22,223
3,653
9,596
1,554
1,499
5,585
337
20,883
341
970
32,303
6,430
16,435
657
3,626
5,070
86
31,790
256
158
10,305
1,406
2,531
889
1,230
4,177
73
10,171
95
10,820
2,259
2,628
295
896
4,436
308
10,598
176
Denominational International U.S. outside Kent CountyKent County
Proportion of total reported giving (amounts in $000s)
Figure 20, Page 24
Questions
Are local and non-local giving complements or substitutes?
How do local giving and total giving vary as a proportion of congregation budget?
Are total giving and local giving inferior goods, normal goods, or luxury goods?
Of Interest
Few have studied congregation level giving. We try to determine the relationship between models of individual and congregation giving.
We test the methodological approach in the current literature.
Very Preliminary ResultsIncome Elasticity
Method used: Both Heckman SelectionPseudo R-Square: 0.3295
Local and non-local giving
Data suggest complementarity. For every 10% increase in non-local giving, local giving increases by 0.68%.
Other significant and positive relationships: Income of the church: 2.6%The number of regular participants: 3.9%Urban congregation: 29%Pastor hours spent in community
participation:1.6% Joint service projects: 0.97%
60
Future Work & Resources
Data available for further study
KCCS offers analytical assistance to selected projects
Online directory Support for the Doug &
Maria DeVos Foundation’s Successful Futures Neighborhood Initiative
61
Closing Remarks
Edwin Hernandez, Ph.D.RDV Corporation
62
Questions and Comments?