New Directions in the Study of Community Elites Laumann

1
New Directions in the Study of Community Elites
Laumann Pappi

• The study gauges the influence and status of
  members of a small Germany community.
• The study shows the data along two axis
• Integrative Centrality
• Sector Differentiation

2
New Directions in the Study of Community Elites
Laumann Pappi

• Integrative Centrality persons playing key
  coordinating roles in a given structure will tend
  to be located in the central region of their
  space. Those on the periphery are of declining
  importance.
• Sector Differentiation dividing of the space
  into relatively homogeneous regions radiating
  from the center including personnel who typically
  occupy key positions in the same institutional
  sector or share common concerns.

3
The effect of Spatial Arrangement on Judgments
and Errors in Interpreting Graphs.McGrath, et
al.

• Factors Influencing Perception of Graphical
  Images
• Proximity to the center of spatial arrangement
  impacts perception of prominence.
• Positioning between clusters of nodes in spatial
  arrangement impacts perception of bridging.
• Spatial clustering of groups of nodes in spatial
  arrangment impacts perception of grouping.

4
The effect of Spatial Arrangement on Judgments
and Errors in Interpreting Graphs.McGrath, et
al.

• Keys to the Best Spatial Arrangement
• Highlights the characteristics of the network.
• Highlights prominence and bridging.
• Clearly displaying group structure.
• Currently, there is no single arrangement method.

5
The analysis and interpretation of multivariate
data for social scientists. Bartholomew, Dave

• Multidimensional Scaling (MDS) is one of several
  multivariate techniques that aim to reveal the
  structure of a data set by plotting points in one
  or two dimensions.

6
The analysis and interpretation of multivariate
data for social scientists. Bartholomew, Dave

• Classical MDS The distances used on the graph
  would be the same as those used in the original
  data matrix. This form of scaling uses the lowest
  number of dimensions as possible.
• Ordinal MDS Looks at the value of the data
  matrix, and its relation to the distances between
  other object pairs. This deals with putting all
  the data in the same rank order as the original
  data matrix.

7
The analysis and interpretation of multivariate
data for social scientists. Bartholomew, Dave

• Interpreting Visual MDS Solutions
• The configuration can be reflected without
  changing the inter-point distances.
• The inter-point distance are not affected if we
  change the origin by adding or subtracting a
  constant from the row or column coordinates.
• The set of points can be rotated without
  affecting the inter-point distances.

8
The analysis and interpretation of multivariate
data for social scientists. Bartholomew, Dave

• A good fit is measured using the sum of squares
  equation. The closer to zero the stress value,
  the better fid the MDS solution is.

9
The analysis and interpretation of multivariate
data for social scientists. Bartholomew, Dave

• Dimensions As the number of dimensions
  increases, the stress decreases but there is a
  trade-off between improving fit and reducing the
  interpretability of the solution.
• Stress is assessed using Kruskals Type I stress
  test

10
The analysis and interpretation of multivariate
data for social scientists. Bartholomew, Dave

• Basic Steps of MDS
• Standardize variables.
• Compute distances.
• Fitted distances will be proportional to actual
  distances, and then are graphed.