5. Sample and statistical results
One hundred companies listed in NYSE were selected randomly using the data base of
the Wall Street Journal for the period 1999-2003, the available complete data was for
only 90 companies. Table II presents classification of the companies according to their
activities and the mean and the standard deviation for the variables used in the model.
The research design is structured primarily on the basis of calculating three
different measures of the quality of earnings on the industry level and on the company
level. The analysis is directed at testing whether there is consistency among the three
measures for one industry or one company in order to have strong evidence about
whether the quality of earnings is low or high. The quality of earnings will be marked
as questionable if there is no consistency among the three measures. In this case, the
quality of earnings measure needs more analysis and investigations and may be in
some cases different techniques to confirm whether it is high or low.
Table III presents the results of the empirical study for the industry level.
As shown in Table III, there is consistency among the three measures of the quality
of earning for the full sample, the banks, insurance, and investment industry and for
the technology industry. For the manufacturing companies and the mining, oil and gas
companies the quality of earnings is questionable and cannot be assessed based on
these three measures.
Table IV presents the empirical study results for the company level.
As shown in Table IV, the manufacturing industry has 13 companies (42 percent)
with high quality of earnings, one company with low quality of earnings, and 17
companies (55 percent) their quality of earnings measure is questionable and cannot be
assessed based on this model and needs further investigation and analysis.