GRIDSTAT Version 5 May 2009

This is the latest version of a program which runs under DOS. It can carry out a wide range of analyses of a repertory grid. This version has been revised so that the options are more sensibly structured, and new options for hierarchical representations and assessment of conflict / inconsistency have been added.

1. Looking at the Grid:. Basic Statistics & Correlations. This now also includes

assessment of extremity and midpoint ratings and

has a focussed or rearranged representation of the grid from a k-means clustering.

2. Finding Patterns of Relationships among Constructs or Elements.

(i) Principal components analysis of construct or element (invariant) correlations with a choice of orthogonal or oblique rotations

(ii) Clustering of Element or Construct Similarities with a choice of method and measure of association.

(iii) Ordinal Relationship Analysis of Constructs or Elements.

(a) Analysis by Asymmetric Coefficients (see Bell, 2004c)

(b) Partial Order Scalogram Analysis (Shye, 1985)

3. Finding Patterns of Relationships between Constructs & Elements.

(i) Singular-Value-Decomposition Representation of a Grid [the famous INGRID approach of Patrick Slater] with a choice of pre-transformations (see Bell, 2004b)

(ii) Correspondence Analysis

4. Conflict Analysis.

(i) Element-Construct Triangular Inequalities (see Bell, 2004d)

(ii) Dilemmas [Self-Ideal Matching]. (Feixas & Saul, 2004)

(iii) Sign Balanced Triads of Correlations (Bassler et al, 1992)

5. Dependency grid measures. (see Walker, Ramsay, & Bell, 1988; & Bell, 2001)

[Note: references are given in the manual]

The program reads the grid in a simple format from a text file and writes output to a user specified text file. It also now contains options to edit & import and export grids in some other formats. It is accompanied by a manual and sample input and input in a zip file. To download the zip file click here.

GRIDSCAL

This program is designed to analyse a set of grids that have common constructs or common elements (or both). It can carry out

· individual differences multidimensional scaling of the common facet (either elements or constructs) with individual weights could be calculated for the separate grids. Where both elements and constructs are common (ie equivalent to semantic differential data) the user can choose which is the basis of the scaling representation and which is the basis of the individual weights)

· Where both elements and constructs are common, a form of three-mode principal components.

· Analysis of variance breakdown of variance in the grid.

The program reads the grid in a simple format from a text file and writes output to a user specified text file. It is accompanied by a manual and sample input and input in a zip file. To download the zip file click here.