Chapter 4

Chapter 4.b.

1. script.ENchap4.b.R

script.ENchap4.b.R can be downloaded in UTF-8 format from:

Download the database

Download Aspiration_Int_UK.RData in UTF-8 format data from the Internet and save it to some directory:

2. Loading Xplortext package

3. Loading the database

4. Building the LT

5. Correspondence analysis applied to the LT

First CA, neither contextual variables nor segments are teken in to account

Word representation: only the metakeys

Repeated segment visualization: process chain

Correspondence analysis computing the segments coordinates

Second CA, equal to the former but segments are taken into account

Document representation

 

Segment representation

Only the second plot is reproduced

The first one is used to select those segments to visualize in the second plot

In order to take into account contextual variables, TextData is runned again

Education and Age_Education categories representation. Trajectories

Gender and Gender_Education categories representation. Trajectories

 

Test values on the position of the represented categories

Answer clustering

Selection how many axes to keep from reading the metawords/metakeys issued form the former CA

6 axes have to be kept. CA has to be runned again, to keep only these 6 axes

Lexical description of cluster 7: characteristic words, paragons and specific answers

Association between the partition and the qualitative contextual variables: cluster 7

Association between the partition and the qualitative contextual variables: cluster 7

For comparison, clustering the answers in the global space