Chapter 1 the study of collocations

vii) S-Nodes per T-Unit 
 
 
The essays were also analysed with respect to the number of sentence 
nodes (S-Nodes) they contained.  This was a measure of syntactic 
accumulation.  The number of underlying sentence nodes, indicated by tensed 
and untensed verbs, was calculated for each essay and then the average 
number of S-Nodes per T-Unit was estimated by dividing the number of S-
Nodes in each essay by the number of T-Units in that essay (see Long 1991, 
unpublished paper).  Inter-rater reliability on 5% of the data was also obtained 
from two other raters, and was at 98%.  Each of the raters was given the same 
sample of essays and instructions for counting the S-Nodes in each essay (see 
Appendix L).  A short training session was given by the researcher on two 
other essays.  After the measurement, the number of S-Nodes marked by the 
raters and the researcher were compared and the inter-rater reliability (98%) 
was considered sufficient for the Error-Free T-Units measurement.