The classical Orthogonal Regression analysis relies heavily on the normality assumption. However, sometimes we might be uncertain of the underlying distribution of our dataset or the sample size might be small, which would cause an inaccurate inference on the parameter if the data is not normally distributed. This leads to the main objective of this paper which is to examine alternative methods to the parametric OR analysis which do not rely on the normality assumption. In this paper, the nonparametric jackknife and bootstrap resampling techniques were applied to assess the bias, standard errors and confidence intervals for the parameters of the model. We studied on the method of delete-one jackknife and bootstrapping the observations and made comparisons between the two methods as well. Under bootstrapping, three methods were considered to construct the confidence intervals which include percentile interval, bias-corrected (BC) interval and bias-corrected and accelerated (BCa) interval. Based on the results, it was found that the bootstrap estimators were closer to the values of classical OR analysis compared to jackknifed estimators. Besides, the jackknife estimates of bias and standard errors were slightly larger than that of bootstrap. Furthermore, we also found that the confidence intervals for the parameters constructed from jackknife have longer lengths and closer to that of OR. This showed that jackknife performed better in constructing confidence interval than the bootstrap.
