Statistical analysis
For proper statistical analyses, a Windows-based SPSS
24.0 statistical analysis program was used (SPSS Inc., USA) .
We examined variables via visual (histograms, probability plots) and
analytical methods (Shapiro-Wilk’s and Kolmogorov-Smirnov test) to
determine whether they were normally distributed or not. Variables
specified as mean±standard deviation (X±SD), the mean difference between
groups, 95% confidence interval (95%CI), median (minimum-maximum
(min-max)), U value, frequency (n), and percentage (%).
Student t-test , Mann-Whitney U test, and Chi-square test were
used to compare normally distributed, undistributed, and categorical
variables. Pearson and Spearman’s tests were conducted to show
relationships between normally and non-normally distributed and/or
ordinal variables. The level of significance was as p≤0.05. For the
multivariate analysis, the possible factors identified with previous
analyses were further entered into the logistic regression analysis to
determine independent predictors of study outcomes. Hosmer-Lemeshow
goodness of fit statistics was for evaluating model fit. A %5 type-1
error level was accepted to infer statistical significance. The
diagnostic values of AAWT, HWT, FWT, and TATT measures in predicting
labor prolongation, arrest, and cesarean delivery were examined by ROC
curve analysis. When a significant cut-off value was observed, the
sensitivity, specificity, positive and negative predictive values were
presented. While evaluating the area under the curve, a %5 type-1 error
level was used to accept a statistically significant test variable’s
predictive value.