دانلود رایگان مقاله حسابداری برای وزن نمونه گیری در مدل سازی مسیر حداقل مربعات جزئی

Abstract

Applications of partial least squares (PLS) path modeling usually focus on survey responses in management, social science, and market research studies, with researchers using their collected samples to estimate population parameters. For this purpose, the sample must represent the population. However, population members are often not equally likely to be included in the sample, which indicates that sampling units have different probabilities of being selected. Hence, sampling (post-stratification) weights should be used to obtain consistent estimates when estimating population parameters. We discuss alterations to the basic PLS path modeling algorithm to consider sampling weights in order to achieve better average population estimates in situations where researchers have a set of appropriate weights. We illustrate the effectiveness and usefulness of the approach with simulations and an empirical example of a job attitude model, using data from Ireland.

6. Implications and future research directions

This study proposes a new modified version of the original PLS path modeling approach, namely the WPLS algorithm that incorporates sampling weights. It shows the new approach's appropriateness with an illustrative example and simulated data. The results show that the new modified version takes the specified weights correctly into account. In addition, this algorithm provides better average population model parameter estimates than the basic PLS algorithm when sampling weights are available. In particular, correcting the estimates for deviations in the sampling procedure provides less biased results that are closer to the population parameters. If researchers are interested in inference to the population, they should ensure that they correct the sampling deviations of their data set, as well as ensure that they use sampling procedures that allow them to draw these conclusions. The empirical examples' results also show the importance of applying sampling weights in model estimations. For example, applying the sampling weights available in the ISSP Work Orientations 2005 to a simple job attitude model shows that drawing conclusions could be misleading when weights are not included in the model estimation. In particular, the results show that although applying the weighting does not alter the measurement model evaluation's results, the structural model results are substantially different in the weighted and unweighted models. Not only was a significant path in the nonweighted model found to be nonsignificant in the weighted model, but the magnitude of the path coefficients may also change substantially. This deviation can have consequences for the theoretical and managerial implications drawn from the analysis.