This thesis develops the theory of failure prediction for UK oprotruction companies. A questionnaire was devised and included 17 questions related to failure for both an "at risk" group classified as vulnerable, i.e.
those scoring negatively by the Z-score model and a positively scoring
The existence of managerial factors related to failure was
investigated in the questionnaire using a multiple choice method. Both
groups proved adequate for ccmparison purposes, and were therefore
included in the A-score model.
The A-score for a company is obtained by adding the weight of all
factors and errors together, and a cut-off value determined. The model
was statistically verified by the t-test method at 1% significant
level and further examined by the Willcoxon (Rank Sum) tests null
hypothesis rejected at 5% level of significance.
An attempt was also made to relate A-score and Z-score values,
unfortunately statistical analysis indicated only 67.7%
intercorrelation between A-score and Z-score i.e. not very strong.
However, the Z-score value of zero corresponded to an A-score cut-off
value of about 50, these being critical values in both modes.
Finally, trend analysis was shown to be a suitable extra check in
objective evaluation of company performance, and an improved method of
systematically appraising contractors was produced. However, the
developed models should only be used as part of an overall asse4mment
of company stability. Any predictions should be interpreted with
caution as the models require further testing on a broader range of
companies. It is also important to appreciate that the use of such
models to exclude ocmpanies from tender lists could accelerate or even
A Doctoral Thesis. Submitted in partial fulfillment of the requirements for the award of Doctor of Philosophy of Loughborough University.