openrepgrid and friends

74
OpenRepGrid & Friends Tools for the Analysis of Repertory Grid Data ICPCP Conference, Sydney, July 21, 2013 Mark Heckmann University of Bremen, Germany

Upload: mark-heckmann

Post on 11-Nov-2014

461 views

Category:

Education


3 download

DESCRIPTION

OpenRepGrid and Friends - Presentation held at the 20th International Conference for Personal Construct Psychology, Sydney, Australia, July 2013 If you have any questions about OpenRepGrid visit the OpenRepGrid Google group under http://groups.google.com/group/openrepgrid

TRANSCRIPT

Page 1: OpenRepGrid and Friends

OpenRepGrid & Friends Tools for the Analysis of Repertory Grid Data

ICPCP Conference, Sydney, July 21, 2013

Mark Heckmann University of Bremen, Germany

Page 2: OpenRepGrid and Friends

PART I •  Motivation for OpenRepGrid •  Why R? •  The OpenRepGrid project

PART II •  Getting started with R •  Getting started with OpenRepGrid •  Creating analysis reports •  How to add new features

PART III •  Graphical User Interfaces for OpenRepGrid •  Preview: A grid administration software

Page 3: OpenRepGrid and Friends

Part I

Page 4: OpenRepGrid and Friends
Page 5: OpenRepGrid and Friends
Page 6: OpenRepGrid and Friends

General requirements for (grid) software

•  Durability

•  Availability

•  Functionality

•  Compatibility

•  Documentation

•  Openness

•  Extensibility

•  Usability

Page 7: OpenRepGrid and Friends

Current    situa+on  

 

Page 8: OpenRepGrid and Friends

No grid software offers all methods

of analysis that have been

devised in the literature

Source: http://callcentersindia.files.wordpress.com/2010/06/selection-process.png

Page 9: OpenRepGrid and Friends

1

THE STRUCTURAL QUADRANTS METHOD:A NEW APPROACH TO THE ASSESSMENT OF CONSTRUCT

SYSTEM COMPLEXITY VIA THE REPERTORY GRID

JOSEP GALLIFA and LUIS BOTELLA

Department of Psychology, Ramon Llull University, Barcelona, Spain

This article presents a new mathematical method (the Structural Quadrants Method)for the assessment of construct system complexity via the repertory grid. The Struc-tural Quadrants Method (SQM) is presented step by step, and its sensitivity togrid structure is illustrated by applying it to five case studies. A validational studydemonstrating the discriminative power of the SQM and comparing it to tradi-tional measures is included. Results indicate that the SQM discriminates between agroup of 11 experts and one of 11 novices in terms of grid complexity as expected(i.e., detecting high degrees of differentiation and integration in the experts’ gridsand low degrees of differentiation and integration in the novices’ grids). The dis-criminative power of the SQM is unparalleled by the traditional measures of gridstructure compared in this article. The article ends with a section on the distinctiveadvantages of the SQM and some suggest ions for future research.

The term cognitive complexity was initially introduced by Bieri (1955),and defined as “the capacity to construe social behavior in a multidi-mensional way” (Bieri et al., 1966, p. 185). As discussed by G. J. Neimeyer(1992, p. 94), such a definition of construct system complexity empha-sizes differentiation, understood as “the relative number of differentdimensions of judgment used by a person.” In line with this notion ofcomplexity as differentiation, traditional measures of construct systemcomplexity have relied on an attempt to assess the degree of overlapor redundancy in the use of constructs, with higher degrees of overlapreflecting lower levels of differentiation and complexity (G. J. Neimeyer,1992).

A theoretically coherent notion of construct system complexity can-not solely rely on the system’s level of differentiation but must also

Received 7 May 1997; accepted 24 July 1998.The authors would like to thank Robert A. Neimeyer and two anonymous review-

ers for their insightful comments on earlier drafts.Address correspondence to Luis Botella, Department of Psychology, Ramon Llull

University, Cister 24–34, 08022-Barce lona, Spain. E-mail: [email protected]

Journal of Constructivist Psychology, 13:1–26, 2000Copyright ã 2000 Taylor & Francis1072-0537/00 $12.00 + .00

Page 10: OpenRepGrid and Friends

Not extensible by user

No framework to support experimental types of analysis

Page 11: OpenRepGrid and Friends

The output of most programs does not easily

lend itself to subsequent

computation

Source: http://www.jacksofscience.com/wp-content/uploads/2008/05/computing_occurs.jpg

Page 12: OpenRepGrid and Friends

Development by a single person or a

small group of researchers, each with

their own special competencies in a

certain field of research and often

limited time resources. (Fromm, 2009)

Page 13: OpenRepGrid and Friends

Source: http://www.kirchen-und-kapellen.de/images/content/Community.jpg

No joint community effort

to improve a program:

Development and documentation is delegated to the

software providers

Page 14: OpenRepGrid and Friends

Source:  h1p://evolu+on.berkeley.edu/evosite/misconceps/images/misconcep+ons_beavers2.gif  

Discontinued development

once its initiators have moved on to

other fields or retired

“the programs that are currently available have a short shelf-life in that they are unlikely to be updated once their creators have retired”

(Fransella, Bannister & Bell, 2004, p. 38)

Page 15: OpenRepGrid and Friends

What to do? •  Software design that enables users

to contribute to its development •  Set up infra- and communication

structures to foster participation in a joint development

Go  Open  Source!  

Page 16: OpenRepGrid and Friends

Why ?

Page 17: OpenRepGrid and Friends

•  Because I am familiar with it J

•  is the lingua franca in the field of

statistics

•  runs on all major platforms

•  is managable by non-programmers

•  becomes increasingly popular within

the social sciences

Page 18: OpenRepGrid and Friends

Mu

en

ch

en

, R. A

. (20

12).

Po

pu

larit

y o

f da

ta a

na

lysis

so

ftw

are

. htt

p:/

/r4

sta

ts.c

om

/art

icle

s/p

op

ula

rity/

Page 19: OpenRepGrid and Friends

•  provides an underlying framework for

further computations

•  provides a wealth of functionality

•  extensible (GUI, 3D etc.)

Suitable for Open Source grid software development

Page 20: OpenRepGrid and Friends

OpenRepGrid project

OpenRepGrid R package OpenRepGrid Wiki

Page 21: OpenRepGrid and Friends

User entry points

•  Give feedback: bug reports, suggestions ...

•  Supply method descriptions to trigger development

•  Provide R code snippets

•  Generate tutorials and case studies

•  Join the R package development team

Page 22: OpenRepGrid and Friends

Development process

Adding graphical user interface(s)

Create wide range of analysis methods

Documentation in Wiki

2011 2013

Phase 2 Phase 1

Researchers with basic R skills Anyone

Page 23: OpenRepGrid and Friends

Part II

Page 24: OpenRepGrid and Friends

Getting Started with R

Page 25: OpenRepGrid and Friends

Setup

+  www.r-project.org www.rstudio.org

Page 26: OpenRepGrid and Friends

30  

Page 27: OpenRepGrid and Friends

31  

Using R as a calculator

4 + 4 ! ! ! ! !# Addition!4 - 1 ! ! ! !# Subtraction!!2 + 2 * 3! ! !# standard calculating rules !(2-1) * (2+2) ! !# standard use of brackets!

Page 28: OpenRepGrid and Friends

32  

Variable assignments

a <- 2 + 2 ! ! !# assignment operator!a ! ! ! ! !# variable value is 4!!a <- 100 ! ! # change variable value!a * 10 ! ! ! ! !# use the variable !

Page 29: OpenRepGrid and Friends

33  

Objects

Objects (Variables) names may contain numbers, letters, underscores or dots. The name must not contain a function (+, - *, / etc.).

object1 <- 10!

object.1 <- “hallo”!

Page 30: OpenRepGrid and Friends

34  

Functions

Function (lat. performance, execution) Every function uses round brackets    rnorm(4)!!!!!

Page 31: OpenRepGrid and Friends

36  

“Nurse, get on the internet, go to SURGERY.COM, scroll down and click on the ,‘Are you totally

lost?’ icon.“

Page 32: OpenRepGrid and Friends

37  

Objects

Opening help files

?rnorm!

!Commands inside the function’s round brackets are called arguments. A function may take zero to many arguments.

rnorm(4, mean=10)!

!

Page 33: OpenRepGrid and Friends

Generate 10 random numbers with mean 100

and a standard deviation of 15.

You have 2 minutes

Your turn

Page 34: OpenRepGrid and Friends

Getting started with

OpenRepGrid

www.openrepgrid.org

Page 35: OpenRepGrid and Friends
Page 36: OpenRepGrid and Friends

Web Repository with 3.500+ packages

OpenRepGrid

Load package when starting R

Page 37: OpenRepGrid and Friends
Page 38: OpenRepGrid and Friends
Page 39: OpenRepGrid and Friends

44  

OpenRepGrid is shipped with several sample grid data sets from the literature. Type in the name of a data set and press enter

boeker!

!

!

8 - martin ! george - 7 | 9 - elizabeth ! karl - 6 | | | 10 - therapist ! kurt - 5 | | | | | 11 - irene ! father - 4 | | | | | | | 12 - childhood self ! mother - 3 | | | | | | | | | 13 - self before illness ! ideal self - 2 | | | | | | | | | | | 14 - self with delusion! self - 1 | | | | | | | | | | | | | 15 - self as dreamer ! | | | | | | | | | | | | | | | ! balanced (1) 1 4 2 2 3 5 2 5 4 2 6 2 2 3 3 (1) get along wit! isolated (2) 3 6 3 5 5 4 5 4 5 4 4 4 2 2 3 (2) sociable !ely integrated (3) 2 2 2 3 5 3 2 3 2 3 3 4 4 5 3 (3) excluded ! discursive (4) 4 1 3 1 2 4 2 3 3 2 3 3 3 5 4 (4) passive ! open minded (5) 2 1 2 1 2 4 4 2 4 2 6 3 2 2 3 (5) indifferent ! dreamy (6) 4 5 3 5 4 5 4 5 4 4 6 3 3 3 2 (6) dispassionate!cally oriented (7) 2 1 3 2 3 3 3 2 2 3 2 3 3 3 3 (7) depressed ! playful (8) 4 5 4 3 4 3 2 3 4 4 5 3 2 4 3 (8) serious !ocially minded (9) 2 1 3 2 4 5 4 1 3 2 6 3 3 3 3 (9) selfish ! quarrelsome (10) 5 5 5 5 5 2 5 2 4 4 1 6 5 5 5 (10) peaceful ! artistic (11) 5 1 2 4 3 5 3 2 4 3 3 4 4 4 4 (11) technical ! scientific (12) 2 1 5 3 4 4 5 3 4 1 6 4 2 3 3 (12) emotional ! introvert (13) 4 5 4 6 5 3 5 3 5 2 5 2 2 2 3 (13) extrovert ! wanderlust (14) 1 1 4 2 4 5 2 5 5 3 6 1 1 2 1 (14) home oriente!

Page 40: OpenRepGrid and Friends

For easier handling we save the boeker dataset in the object x!

x <- boeker!

!

!

!

Page 41: OpenRepGrid and Friends

Importing a dataset from a .txt file!!!x <- importTxt() # interactive dialog!!# sample file path on windows!x <- importTxt(“C:mark/grid_1.txt") !!# sample file path on MacOs / Unix!x <- importTxt(“~/mark/grid_1.txt") !

!

!

!

!!

Page 42: OpenRepGrid and Friends

Subsetting a grid using the squared brackets operator, e.g. extracting rows 1 to 5

x[1:5, ]!

!

!

!

8 - martin ! george - 7 | 9 - elizabeth ! karl - 6 | | | 10 - therapist ! kurt - 5 | | | | | 11 - irene ! father - 4 | | | | | | | 12 - childhood self ! mother - 3 | | | | | | | | | 13 - self before illness ! ideal self - 2 | | | | | | | | | | | 14 - self with delusion! self - 1 | | | | | | | | | | | | | 15 - self as dreamer ! | | | | | | | | | | | | | | | ! balanced (1) 1 4 2 2 3 5 2 5 4 2 6 2 2 3 3 (1) get along wit! isolated (2) 3 6 3 5 5 4 5 4 5 4 4 4 2 2 3 (2) sociable !ely integrated (3) 2 2 2 3 5 3 2 3 2 3 3 4 4 5 3 (3) excluded ! discursive (4) 4 1 3 1 2 4 2 3 3 2 3 3 3 5 4 (4) passive ! open minded (5) 2 1 2 1 2 4 4 2 4 2 6 3 2 2 3 (5) indifferent !

Page 43: OpenRepGrid and Friends

Creating a bertin plot

bertin(x)!

!!

Page 44: OpenRepGrid and Friends

Opening the documentation

?bertin!

!!

Page 45: OpenRepGrid and Friends

Changing the look by modifying the arguments

bertin(x, colors=c("white","darkblue"))!

!

Page 46: OpenRepGrid and Friends

Creating biplots of elements and constructs

biplot2d(x)!

!

Page 47: OpenRepGrid and Friends

Creating composite plots of elements and constructs

biplot3d(x)!

!

Page 48: OpenRepGrid and Friends

Detect implicative dilemmas using correlations

indexDilemma(x, self=1, ideal=2)!

###################!Implicative Dilemma!###################!!Dilemmatic Self-Ideal Construct Pairs!#####################################!!By A Priori Criteria:!! RexSI Rtot Constructs!1 0.52 0.63 1 balanced - get along * <==> 5 open minde - indifferen!2 0.60 0.64 1 balanced - get along * <==> 6 dreamy - dispassion!3 -0.35 -0.54 1 balanced - get along * <==> 7 practicall - depressed!4  0.35 0.42 1 balanced - get along * <==> 8 playful – serious!

RexSI = Correlations excluding Self & ideal!Rtot = Correlations including Self & ideal!Rtot was used as criterion!* = discrepant construct pair!

Page 49: OpenRepGrid and Friends

Generate a cluster dendrogram for the

boeker grid. You have 2 minutes

Your turn

Page 50: OpenRepGrid and Friends

Writing things down

Page 51: OpenRepGrid and Friends

out now!

Page 52: OpenRepGrid and Friends

Press  to  generate    document  

Insert  tags  

Page 53: OpenRepGrid and Friends

Using R markdown

•  Markdown is a lightweight language to structure a document

•  R code chunks can be directly included •  Include R code between the following

tags

```{r}

# Some R Code here

```

Page 54: OpenRepGrid and Friends
Page 55: OpenRepGrid and Friends

Generate a report with inter-construct correlations for the boeker dataset.

You have 2 minutes

Your turn

Page 56: OpenRepGrid and Friends

Adding new features

Page 57: OpenRepGrid and Friends
Page 58: OpenRepGrid and Friends

Create a function that counts the number of

midpoint ratings

Task

Page 59: OpenRepGrid and Friends

Writing new functions

f <- function(){!!# some R commands!

}!f()!!!f <- function(x){!!x!

}!f(12)!! !

Page 60: OpenRepGrid and Friends
Page 61: OpenRepGrid and Friends

!midpoints <- function(x) { !!scores <- getRatingLayer(x) !!midpoint <- getScaleMidpoint(x) !!sum(scores == midpoint)!

}!!midpoints(bell2010)!

The Midpoint Function

Page 62: OpenRepGrid and Friends

Part III

Page 63: OpenRepGrid and Friends

Graphical User Interfaces

Page 64: OpenRepGrid and Friends

OpenRepGrid  -­‐  the  workhorse  

Web  based     Local  

Building GUIs on top of OpenRepGrid

PC / Server

GUI type

OnAir  Other  special  purpose  implementa+ons  ?  

Page 65: OpenRepGrid and Friends

www.onair.openrepgrid.org

Page 66: OpenRepGrid and Friends
Page 67: OpenRepGrid and Friends

gridhub: computer-guided grid adminstration

Preview  

Page 68: OpenRepGrid and Friends

gridhub developers

Matthias Kaulartz (programming) Mark Heckmann (concept)

Page 69: OpenRepGrid and Friends
Page 70: OpenRepGrid and Friends
Page 71: OpenRepGrid and Friends
Page 72: OpenRepGrid and Friends

Want to participate?

You have suggestions or ideas?

Just get in touch!

Page 73: OpenRepGrid and Friends

Thanks!

Page 74: OpenRepGrid and Friends

Literature

•  Fransella, F., Bell, R. C., & Bannister, D. (2004). A manual for repertory grid technique (2nd ed.). Chichester: John Wiley & Sons.

•  Fromm, M. (2009). Grid Software. European Personal Construct Association. Retrieved July 25, 2010, from http://www.epca-net.org/repgrid/software.

•  Heckmann, M. (2011). OpenRepGrid - An R package for the analysis of repertory grids (Unpublished diploma thesis). University of Bremen, Bremen, Germany.

•  Muenchen, R. A. (2011). The Popularity of Data Analysis Software. Retrieved March 28, 2011, from http://sites.google.com/site/r4statistics/popularity.

•  R Development Core Team. (2011). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. ISBN 3-900051-07-0, URL http://www.R-project.org/.