Advantages of Nonlinear over Segmentation Analyses in Path Models

Advantages of Nonlinear over Segmentation Analyses in Path Models

Ned Kock (Division of International Business and Technology Studies, Texas A&M International University, Laredo, TX, USA)
Copyright: © 2016 |Pages: 6
DOI: 10.4018/IJeC.2016100101
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

The recent availability of software tools for nonlinear path analyses, such as WarpPLS, enables e-collaboration researchers to take nonlinearity into consideration when estimating coefficients of association among linked variables. Nonlinear path analyses can be applied to models with or without latent variables, and provide advantages over data segmentation analyses, including those employing finite mixture segmentation techniques (a.k.a. FIMIX). The latter assume that data can be successfully segmented into subsamples, which are then analyzed with linear algorithms. Nonlinear analyses employing WarpPLS also allow for the identification of linear segments mirroring underlying nonlinear relationships, but without the need to generate subsamples. The author demonstrates the advantages of nonlinear over data segmentation analyses.
Article Preview

Illustrative Model And Data

Our illustrative model contains only two variables, e-collaboration technology use (ETU) and new product quality (NPQ), and one causal link: ETU → NPQ. E-collaboration technology use (ETU) measures the extent to which a web-based e-collaboration tool has been used by teams of employees of a multinational company. The fictitious multinational company was assumed to develop and sell consumer products. Each team developed a new product, such as a new brand of toothpaste, whose market success was measured through the variable new product quality (NPQ). Both variables were measured through single indicators.

We employed the Monte Carlo method (Robert & Casella, 2005) to create sample data based on this model. The sample size was 250; meaning 250 rows of data, with each row referring to a new product development team. In addition to the two variables ETU and NPQ, we also created a numeric column and a column of text labels referring to three countries. The data sample was created based on the assumption that it was comprised of three subsamples, each coming from a country where the multinational company conducted operations. This is illustrated in Figure 1, where each of the data points represents a new product development team.

Figure 1.

Underlying nonlinear relationship and country-specific patterns

Data Segmentation Results

Table 1 shows the linear path coefficients and corresponding P values, for each of the three countries. The “View or change data modification settings” option in WarpPLS 5.0 allows users to run their analyses with subsamples defined by a range restriction variable, which we chose to be our numeric column referring to each of the countries by a number: 1 for Country1, 2 for Country2, and 3 for Country3. Using this option, we were able to conduct linear analyses for each separate country without having to use different datasets in WarpPLS for each of the countries.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 13: 4 Issues (2017): 2 Released, 2 Forthcoming
Volume 12: 4 Issues (2016)
Volume 11: 4 Issues (2015)
Volume 10: 4 Issues (2014)
Volume 9: 4 Issues (2013)
Volume 8: 4 Issues (2012)
Volume 7: 4 Issues (2011)
Volume 6: 4 Issues (2010)
Volume 5: 4 Issues (2009)
Volume 4: 4 Issues (2008)
Volume 3: 4 Issues (2007)
Volume 2: 4 Issues (2006)
Volume 1: 4 Issues (2005)
View Complete Journal Contents Listing