SPSS-Analysis - Assignment Example

MRes Business and Management, BQ7805 Advanced Quantitative Research Advanced quantitative research articles The two research methods that we will be looking at are Partial least squares path modeling (PLS-SEM) and exploratory factor analysis (EFA). The first method, PLS-SEM, was utilized in the research article “Influences of gender and product type on online purchasing” done by Pascual-Miguel, Agudo-Peregrina, and Chaparro-Peláez (2015). The second method EFA was utilized in the research “Customer loyalty in e-commerce: an exploration of its antecedents and consequences” done by Srinivasana, Andersona, and Ponnavolu (2002). Based on the factors being analyzed and the sample sizes for each of the two research papers, the data analysis methods needs to be effective to provide a comprehensive report on the findings. These two articles cover all the key data analysis requirements. Furthermore, the data analysis section for both of these has been comprehensively outlined and explained to ensure the reader understanding the steps that were taken in order to generate results. An inspection on both articles shows that each article followed the systematic process required within the chosen technique. Pascual-Miguel, Félix J., Ángel F. Agudo-Peregrina, and Julián Chaparro-Peláez (2015). "Influences of gender and product type on online purchasing." Journal of Business Research 68.7: 1550-1556. Partial least squares path modeling (PLS-SEM) is among the favorite methods often utilized in business research. According to Hair, Sarstedt, Pieper, and Ringle (2012), PLS-SEM has gained significant popularity within the business research discipline over the last two decades. Several reasons have been given for this increased interest. The most consistent reasons being its resourcefulness in non-normal data, working with formative measures, accuracy in small sample sizes, and for studies that are focusing on prediction (Hair, Sarstedt, Pieper, & Ringle 2012). In the research by Pascual-Miguel, Agudo-Peregrina, and Chaparro-Peláez (2015), the method was chosen because of its ability to analyze complex models and for predictive purposes. With gender being an essential characteristic of the research, the authors took this into consideration and split the sample to carry out a multi-group analysis. The Partial least squares path modeling (PLS-SEM) model is carried out in two-stage algorithm with the first stage consisting of five steps. In stage one an “iterative estimation of latent constructs” is carried out which includes the four steps that build upon approximation and estimation of latent construct scores and proxies for structural models (Hair, Ringle, & Sarstedt 2011). The four steps involve “Step 1: outer approximation of latent construct scores; step 2: estimation of proxies for structural model; step 3: inner approximation of latent constructs; and Step 4: estimation for proxies for coefficients.” (Hair, Ringle, & Sarstedt 2011) In stage two, the least square method is utilized to estimate the final coefficients by analyzing each partial regression model. To carry out a research analysis using the Partial least squares path modeling, researchers have to submit all their partial regression models through the iterative procedures outlined by the PLS-SEM algorithm (Hair, Ringle, & Sarstedt 2011). The article by Pascual-Miguel, Agudo-Peregrina, and Chaparro-Peláez (2015), utilizes the PLS-SEM method in its analysis. In the presentation of their results, the authors explain a systematic approach they used to arrive at the results of the study. The approach is consistent with the two-stage standard procedure required in PLS-SEM method. For the first stage they carry out an empirical, construct reliability, convergent, and discriminant validity analyzes (Pascual-Miguel, Agudo-Peregrina, and Chaparro-Peláez 2015). For the second stage, they carry out a multigroup analysis for each of the four partial regression models. Following these observations, it is clear that the authors do not deviate or skip any of the stages required in PLS-SEM method. Looking at the presentation of the results and explanations provided in the Pascual-Miguel, Agudo-Peregrina, and Chaparro-Peláez artcicle (2015), it is evident that all the necessary steps were covered. Both stages were covered and the authors clearly explain what was done in each stage in order to generate the desired results. The only difference with the main steps of the method is that in stage one, the authors have to adjust the four steps so that they can fit the current data. None of the steps or stages is given more space than it deserves as compared to the others. Srinivasan, S. S., Anderson, R., & Ponnavolu, K (2002). Customer loyalty in e-commerce: an exploration of its antecedents and consequences. Journal of retailing, 78(1). 41-50. Exploratory factor analysis is another widely used statistical method used in the social sciences, hence making it a good choice for analyzing data on individual and group behaviors. Despite its popular use, EFA analysis comprises of complex procedures that require a good understanding of the method in order to produce effective results during a research procedure (Anna and Osborne 2005). However, one of the advantages leading to many researches sorting after this method is its ability to work with a large set of variables (Norris & Lecavalier 2010). When carrying out Exploratory factor analysis (EFA), research face a hurdle of deciding on the several options they have with the method because of its sequential and linear approach. To overcome this seemingly complex procedure involved with exploratory factor analysis (EFA) method, a decision protocol was developed for researchers to use in deciding the best option. The five decision steps start with a look at the data and deciding whether it is suitable for EFA. The second step analyzes the factors that will be extracted in order to guide the next step in the process. The third step is closely followed the second step as it assists in the determination of factor extraction. After the factors have been extracted, Step four involves selecting the best rational method to be used. Finally, Step five involves interpretation that refers to the process of grouping variables according to the selected factors and giving them a theme or name (Brett, Brown, & Onsman 2012). In the presentation of their results, the authors explain a systematic approach they used to arrive at the results of the study. The approach is consistent with the two-stage standard procedure required in PLS-SEM method. For the first stage they carry out an empirical, construct reliability, convergent, and discriminant validity analyzes (Pascual-Miguel, Agudo-Peregrina, and Chaparro-Peláez For the second stage, they carry out a multigroup analysis for each of the four partial regression models. Following these observations, it is clear that the authors do not deviate or skip any of the stages required in PLS-SEM method. In the second article, Srinivasana, Andersona, and Ponnavolu (2002) comprehensively follow the necessary steps required in the EFA method. They start off by analyzing the exploratory data set in order to test if the scale items loaded as expected, which is consistent with EFA method procedure (Srinivasana, Andersona, & Ponnavolu 2002). The author also cover steps two and three by “testing their refined scale items for reliability and unidimensionality.” For step four, a linearized logarithmic transformation scale using ordinary least square method was chosen and step five was captured under the 8Cs categorization (Srinivasana, Andersona, & Ponnavolu 2002). In essence, all the necessary step required with the EFA method were covered in the study. Srinivasana, Andersona, and Ponnavolu (2002) cover all the analysis steps. Through their work flow and presentation they ensure that none of the key steps are unexplained in their results presentation. Although, some of the steps are modified to fit the research data, none of the steps is skipped or forgone for the other. The steps are also given equal space as necessary with none of them being omitted. With the least space being offered for the results presentation, the authors still manage to explain all the five steps in equal detail. Conclusion After review the analysis and results of both articles, it is conclusive that the researchers were keen on the expected data analysis section of their chosen methods. Srinivasana, Andersona, and Ponnavolu (2002) did not deviate from the necessary procedures for carrying out an EFA analyses and their results are clear and easily understood by the audience. As for Pascual-Miguel, Agudo-Peregrina, and Chaparro-Peláez (2015), they had to modify a section of the data analyses for it to fit their collected data and the end results were satisfactory. References Anna, Costello B., and Jason W. Osborne (2005). "Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assesment." Research and Evaluation 10.7: 1-9. Brett William, Ted Brown, and Andrys Onsman (2012). "Exploratory factor analysis: A five- step guide for novices." Australasian Journal of Paramedicine 8.3: 1. Hair, Joe F., Christian M. Ringle, and Marko Sarstedt (2011). "PLS-SEM: Indeed a silver bullet." The Journal of Marketing Theory and Practice 19.2: 139-152 Hair, Joseph F., et al (2012). "The use of partial least squares structural equation modeling in strategic management research: a review of past practices and recommendations for future applications." Long range planning 45.5: 320-340. Norris, Megan, and Luc Lecavalier (2010). "Evaluating the use of exploratory factor analysis in developmental disability psychological research." Journal of autism and developmental disorders 40.1: 8-20. Pascual-Miguel, Félix J., Ángel F. Agudo-Peregrina, and Julián Chaparro-Peláez (2015. "Influences of gender and product type on online purchasing." Journal of Business Research 68.7: 1550-1556. Srinivasan, S. S., Anderson, R., & Ponnavolu, K (2002). Customer loyalty in e-commerce: an exploration of its antecedents and consequences. Journal of retailing, 78(1): 41-50. Read More