Furthermore, analysis of variance (ANOVA) and a response table
were used to determine the optimized levels of the GMAW process
parameters [20]. This was accomplished by separating the total
variability of the S/N ratios (which was the sum of the squared
deviations from the total mean of the S/N ratio) into contributions
from each parameter and its error. The percentage contribution
was used to evaluate the effect of the parameter change on the
quality. In addition, the calculated test statistic, F-ratio [20]
(named after Fisher who developed the method), was used to
determine which welding parameters had a significant effect on
the quality. When the calculated F-ratio was greater than the critical
F, it would indicate which factor had a significant effect on the
quality.