Skip to main content
SpringerLink
Log in

Why Should PLS-SEM Be Used Rather Than Regression? Evidence from the Capital Structure Perspective

  • Chapter
  • First Online:
Partial Least Squares Structural Equation Modeling

Part of the book series: International Series in Operations Research & Management Science ((ISOR,volume 267))

Abstract

This study examines capital structure determinants using a simultaneous causal model with interaction effects between manifest and latent variables. Partial Least Squares (PLS) is an approach to Structural Equation Models (SEM) that allows researchers to analyse the relationships simultaneously. It is interesting to compare and contrast this approach in analysing mediation relationships with the regression analysis. In addition to statistical data, logical arguments are presented supported by two case studies from PLS-SEM and regression models. We find that the choice between regression and PLS-SEM matters even with the simplest scenarios per item for constructs. This study’s originality is the provision of new comparative analyses of PLS-SEM versus regression analysis in the context of capital structure determinants. The “indirect” and “mediate” macro syntax normal theory of the Sobel test, and the bootstrapping techniques are compared with PLS-SEM. We find that the PLS-SEM analysis provides less contradictory results than regression analysis in terms of detecting mediation effects.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution
Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 169.00
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 219.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info
Hardcover Book
USD 219.99
Price excludes VAT (USA)
  • Durable hardcover edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    See Iacobucci et al. (2007) that report the mediation analysis using Baron and Kenny’s (1986) approach in Journal of Consumer Psychology (JCP), Journal of Consumer Research (JCR), Journal of Marketing Research (JMR) and Journal of Marketing (JM).

  2. 2.

    Mackinnon et al. (2002, 2004) included these methods to examine the performance on their type error rates and power.

  3. 3.

    The mediation bootstrap t-statistic is computed from macro syntax in SPSS , the bootstrap test provides summary results of percentile-based and bias-correlated bootstrap confidence intervals. From a statistics point of view, interval estimates of a popular parameter estimate the reliability . The underlying confidence interval contains the two-sided confidence limits from a confidence interval in which their one-sided counterparts are denoted as lower or upper confidence bounds.

  4. 4.

    The outcome coefficient and standard error estimated by bootstrapping is used in computing the Sobel statistical value.

  5. 5.

    The regression result will show of total debt ratio (TADR), followed by debt to total capital of book (TDTC-BV) and market value (TDTC-MV), long term debt to capital of book (LTDTC-BV) and market value (LTDTC-MV) and short term debt to capital of book (STDTC-BV) and market value (STDTC-MV), respectively.

  6. 6.

    c’, that is “c’=c−ab” defines the difference between the total effect (path c) and indirect effect of X on Y via M (path ab).

  7. 7.

    This can be seen in both summary results from the bootstrap technique with percentile-based, bias-correlated (BC) bootstrap confidence intervals (CI) or from the normal theory of causal effects in Table 6.8 of Appendix 6.1

  8. 8.

    The Goodman Test is a robustness test that is also commonly used to test mediation effects, but it is not as popular as the Sobel Test .

  9. 9.

    The test statistics from the Sobel and Goodman Tests were generated from the macro documented at http://www.comm.ohio-state.edu/ahayes/sobel.htm.

  10. 10.

    See Appendix 6.1 in Fig. 6.14 for complete results.

  11. 11.

    Simplest data (i.e., one item for variable X, one item for variable M, one item for variable Y).

  12. 12.

    X = firm size (total assets), M = TADR, Y = ROE.

  13. 13.

    When the model diagram shows that at least paths “a” and “b” are significant, there is strong support that the mediation exists. The significance of the mediation can be confirmed with the t-statistics test.

  14. 14.

    The collinearity statistics matrix (VIF) also estimates other relationships such as capital structure determinants ; they have consistent results.

  15. 15.

    AVE is the parameter to measure convergent validity that estimates the “degree to which two measures of the same concept are correlated (Hair et al. 2013)”. Specifically, it measures the degree of multiple items in the same construct. The cut-off value for good convergent validity in AVE is set as 0.5 or above. That score of 0.5 means 50% of the measurement variance is accounted for (Fornell and Larcker 1981; Hair et al. 2013). Composite reliability (CR) is assessed by means of all the indicators assigned that have strong mutual correlations to the same construct. So, CR is used to check how well all indicators are connected with the construct. The acceptable threshold cut-off value for CR is 0.7 or above. But cut-off values of 0.5 and 0.6 have also been acceptable (Bagozzi and Yi 1988; Hair et al. 2013).

  16. 16.

    The regression assumptions are: (i) the relationship of Xs and Y is linear, (ii) “non-stochastics” of Xs and no exact linear relationship exists between any two Xs (i.e., Xs is known with certainty which is the Xs are measured without errors), (iii) the value of error term are expected value of zero, having a constant variance for all observations, uncorrelated across observations and normally distributed (see the Gauss-Markov theorem).

  17. 17.

    The case of assume the optimal exists is when developer of PLS-SEM started to asserted such as, Wold (1982, p. 28) says “scarifying optimally”, “the gains of sacrificing optimally are considerable”.. “overall optimization property”… “reliability ”.. etc.

References

  • Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173.

    Article  Google Scholar 

  • Bennett, M., & Donnelly, R. (1993). The determinants of capital structure: Some UK evidence. The British Accounting Review, 25(1), 43–59.

    Article  Google Scholar 

  • Bevan, A. A., & Danbolt, J. (2002). Capital structure and its determinants in the UK—a decompositional analysis. Applied Financial Economics, 12(3), 159–170.

    Article  Google Scholar 

  • Cepeda, G., Nitzl, C., & Roldán, J. L. (2017). Mediation analyses in partial least squares structural equation modeling: Guidelines and empirical examples. In H. Latan & R. Noonan (Eds.), Partial least squares path modeling: Basic concepts, mhodological issues and applications (pp. 173–195). Cham: Springer International.

    Google Scholar 

  • Chang, C., Lee, A. C., & Lee, C. F. (2009). Determinants of capital structure choice: A structural equation modeling approach. The Quarterly Review of Economics and Finance, 49(2), 197–213.

    Article  Google Scholar 

  • Chiarella, C., Pham, T. M., Sim, A. B., & Tan, M. M. L. (1991). Determinants of corporate capital structure: Australian evidence. In G. S. Rhee & R. P. Chan (Eds.), Pasific basin capital markets research (Vol. 3, pp. 139–158). Amsterdam: North Holland.

    Google Scholar 

  • Chin, W. W. (2010). How to write up and report PLS-SEM analyses. In V. E. Vinzi, W. W. Chin, J. Henseler, & H. Wang (Eds.), Handbook of partial least squares: Concepts, methods and applications in marketing and related fields. Berlin: Springer.

    Google Scholar 

  • Goodhue, D. L., Lewis, W., & Thompson, R. (2012). Does PLS have advantages for small sample size or non-normal data? MIS Quarterly, 36(3), 981–1001.

    Google Scholar 

  • Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis (7th ed.). Upper Saddle River: Prentice-Hall.

    Google Scholar 

  • Hair, J. F., Sarstedt, M., Ringle, C. M., & Mena, J. A. (2012). An assessment of the use of partial least squares structural equation modeling in marketing research. Journal of Academy of Marketing Research, 40(3), 414–433.

    Article  Google Scholar 

  • Hair, J. F., Hult, G. T. M., Ringle, C., & Sarstedt, M. (2013). A primer on partial least squares structural equation modeling (PLS-SEM). Thousand Oaks: Sage.

    Google Scholar 

  • Hair, J. F., Hollingsworth, C. L., Randolph, A. B., & Chong, A. Y. L. (2017). An updated and expanded assessment of PLS-SEM in information systems research. Industrial Management & Data Systems, 117(3), 442–458.

    Article  Google Scholar 

  • Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2017). A primer on partial least squares structural equation modeling (PLS-SEM) (2nd ed.). Thousand Oaks: Sage.

    Google Scholar 

  • Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76(4), 408–420.

    Article  Google Scholar 

  • Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York: Guilford Press.

    Google Scholar 

  • Henseler, J. (2010). On the convergence of the partial least squares path modeling algorithm. Computational Statistics, 25(1), 107–120.

    Article  Google Scholar 

  • Henseler, J., Hubona, G., & Ray, P. A. (2017). Partial least squares path modeling: Updated guidelines. In H. Latan & R. Noonan (Eds.), Partial least squares path modeling: Basic concepts, methodological issues and applications (pp. 19–39). Cham: Springer International.

    Chapter  Google Scholar 

  • Iacobucci, D., & Duhachek, A. (2003). Advancing alpha: Measuring reliability with confidence. Journal of Consumer Psychology, 13(4), 478–487.

    Article  Google Scholar 

  • Iacobucci, D., Saldanha, N., & Deng, X. (2007). A meditation on mediation: Evidence that structural equations models perform better than regressions. Journal of Consumer Psychology, 17(2), 139–153.

    Article  Google Scholar 

  • Jairo, I. (2009). The use of structural equation modelling (SEM) in capital structure empirical analysis. KCA Journal of Business Management, 1(1), 11–35.

    Article  Google Scholar 

  • Jannoo, Z., Yap, B., Auchoybur, N., & Lazim, M. (2014). The effect of nonnormality on CB-SEM and PLS-SEM path estimates. International Journal of Mathematical, Computational, Physical and Quantum Engineering, 8(2), 285–291.

    Google Scholar 

  • Jose, P. E. (2013). Doing statistical mediation and moderation. New York: Guilford Press.

    Google Scholar 

  • Judd, C. M., & Kenny, D. A. (1981). Process analysis estimating mediation in treatment evaluations. Evaluation Review, 5(5), 602–619.

    Article  Google Scholar 

  • Latan, H. (2015). Applied statistical data analysis with IBM SPSS. Bandung: Alfabeta.

    Google Scholar 

  • Latan, H., & Noonan, R. (Eds.). (2017). Partial least squares path modeling: Basic concepts, methodological issues and applications. Cham: Springer International.

    Google Scholar 

  • Lee, L., Petter, S., Fayard, D., & Robinson, S. (2011). On the use of partial least squares path modeling in accounting research. International Journal of Accounting Information Systems, 12(4), 305–328.

    Article  Google Scholar 

  • Lin, F.-J. (2008). Solving multicollinearity in the process of fitting regression model using the nested estimate procedure. Quality & Quantity, 42(3), 417–426.

    Article  Google Scholar 

  • Lohmöller, J.-B. (1989). Latent variable path modeling with partial least squares. Berlin: Springer.

    Book  Google Scholar 

  • MacKinnon, D. P. (2000). Contrasts in multiple mediator models. In J. S. Rose & L. Chassin (Eds.), Multivariate applications in substance use research: New methods for new questions. Mahwah: Erlbaum.

    Google Scholar 

  • MacKinnon, D. P., Warsi, G., & Dwyer, J. H. (1995). A simulation study of mediated effect measures. Multivariate Behavioral Research, 30(1), 41–62.

    Article  Google Scholar 

  • MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological Methods, 7(1), 83.

    Article  Google Scholar 

  • MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the indirect effect: Distribution of the product and resampling methods. Multivariate Behavioral Research, 39(1), 99–128.

    Article  Google Scholar 

  • Marcoulides, G. A., & Saunders, C. (2006). Editor’s comments: PLS: a silver bullet? Mis Quarterly, 30(2), iii–iix.

    Article  Google Scholar 

  • Marsh, P. (1982). The choice between equity and debt: An empirical study. The Journal of Finance, 37(1), 121–144.

    Article  Google Scholar 

  • Neter, John, Wasserman, W., & Kutner, M. H. (1990). Applied Linear Regression Models. Homewood, Illinois: Richard D. Irwin, Inc.

    Google Scholar 

  • Nitzl, C., Roldán, J. L., & Cepeda Carrión, G. (2016). Mediation analysis in partial least squares path modeling: Helping researchers discuss more sophisticated models. Industrial Management & Data Systems, 119(9), 1849–1864.

    Article  Google Scholar 

  • Noonan, R., & Wold, H. (1986). Partial least squares path analysis. In T. Husen & T. N. Postlethwaite (Eds.), The international encyclopedia of education (Vol. 7, pp. 3769–3775). Oxford: Pergamon Press.

    Google Scholar 

  • Ozkan, A. (2001). Determinants of capital structure and adjustment to long run target: Evidence from UK company panel data. Journal of Business Finance & Accounting, 28(1-2), 175–198.

    Article  Google Scholar 

  • Preacher, K. J. (2013). Calculation for the Sobel test: An interactive calculation tool for mediation tests. Retrieved January 2, 2013, from http://quantpsy.org/sobel/sobel.htm.

  • Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36(4), 717–731.

    Article  Google Scholar 

  • Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40(3), 879–891.

    Article  Google Scholar 

  • Rigdon, E. E. (2013). Partial least squares path modeling. In G. R. Hancock & R. O. Mueller (Eds.), Structural equation modeling: A second course (2nd ed., pp. 81–116). Greenwich: Information Age.

    Google Scholar 

  • Rönkkö, M., McIntosh, C. N., & Antonakis, J. (2015). On the adoption of partial least squares in psychological research: Caveat emptor. Personality and Individual Differences, 87(1), 76–84.

    Article  Google Scholar 

  • Sarstedt, M., Ringle, C. M., & Hair, J. F. (2017). Partial least squares structural equation modeling. In C. Homburg, M. Klarmann, & A. Vomberg (Eds.), Handbook of market research (pp. 1–40). Heidelberg: Springer International.

    Google Scholar 

  • Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychological Methods, 7(4), 422.

    Article  Google Scholar 

  • Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13, 290–312.

    Article  Google Scholar 

  • Sobel, M. E. (1986). Some new results on indirect effects and their standard errors in covariance structure models. Sociological Methodology, 16, 159–186.

    Article  Google Scholar 

  • Titman, S., & Wessels, R. (1988). The determinants of capital structure choice. The Journal of Finance, 43(1), 1–19.

    Article  Google Scholar 

  • Vanderweele, T. J. (2015). Explanation in causal inference: Methods for mediation and interaction. Oxford: Oxford University Press.

    Google Scholar 

  • Wold, H. (1982). Soft modelling: The basic design and some extensions. In K. G. Joreskog & H. O. A. Wold (Eds.), System under indirect observation (Vol. 2). Amsterdam: North-Holland.

    Google Scholar 

  • Yang, C. C., Lee, C. F., Gu, Y. X., & Lee, Y. W. (2010). Co-determination of capital structure and stock returns—A LISREL approach: An empirical test of Taiwan stock markets. The Quarterly Review of Economics and Finance, 50(2), 222–233.

    Article  Google Scholar 

  • Zhao, X., Lynch, J. G., & Chen, Q. (2010). Reconsidering Baron and Kenny: Myths and truths about mediation analysis. Journal of Consumer Research, 37(2), 197–206.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Faculty of Economics and Muamalat, Universiti Sains Islam Malaysia, Bandar Baru Nilai, Negeri Sembilan, Malaysia

    Nur Ainna Ramli

  2. Department of Accounting, STIE Bank BPD Jateng, Semarang, Indonesia

    Hengky Latan

  3. Department of Accounting, Petra Christian University, Surabaya, Indonesia

    Hengky Latan

  4. Department of Economics and Finance College of Business and Law, University of Canterbury, Christchurch, New Zealand

    Gilbert V. Nartea

Authors
  1. Nur Ainna Ramli
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Hengky Latan
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Gilbert V. Nartea
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Hengky Latan .

Editor information

Editors and Affiliations

  1. School of Business, University of Queensland, St Lucia, Queensland, Australia

    Necmi K. Avkiran

  2. Institute of HRM, Hamburg University of Technology, Hamburg, Germany

    Christian M. Ringle

Appendix 6.1

Appendix 6.1

6.1.1 6.4.1. Study 1

Table 6.8 Maco syntax matrix regression procedures
Full size table

Direct effects of mediators on DV (b paths)

Total effect of IV on DV (c path)

M to Y

Coef. Β

Std.error

t-value

p-value

X to Y

Coef. Β

Std. error

t-value

p-value

TADR

−56.6422

10.5804

−5.3535

0.0000

Firm size (log total assets )

−2.73

1.544

−1.769

0.0770

TDTC BV

−0.782

0.1108

−7.0552

0.0000

Firm size (log total sales)

−2.099

1.471

−1.427

0.1540

TDTC MV

−10.7824

7.3952

−1.458

0.1449

Economic growth (GDP)

4.051

0.623

6.502

0.0000

LTDTC BV

154.9192

11.9324

12.9831

0.0000

Economic growth (GDI)

0.711

0.38

1.869

0.0620

LTDTC MV

−6.7315

5.8122

−1.1582

0.2468

     

STDTC BV

−29.1996

11.5452

−2.5291

0.0115

     

STDTC MV

−10.7524

6.8313

−1.574

0.1155

     
 

Mediation effects

Percentage of the total effect

Bias corrected confidence intervals

Bootstrap results for mediation effects

Sobel test

Goodman test

 

Lower

Upper

 

Sizeta → TADR → ROE

−5.0957

−5.0895

66.3305

−5.269

−0.5492

Support for mediation

Sizesales → TADR → ROE

−5.1318

−5.1247

76.3888

−4.6141

−0.5123

Support for mediation

GDP → TADR → ROE

4.0932

4.066

6.1302

0.0684

0.6876

Support for mediation

GDI → TADR → ROE

−0.1708

−0.1681

−0.5716

−0.073

328

No support for mediation

Sizeta → TDTC BV → ROE

−6.8492

−6.8453

128.3982

−8.2226

−1.3988

Support for mediation

Sizesales → TDTC BV → ROE

−6.9112

−6.9067

152.5059

−7.3196

−1.1653

Support for mediation

GDP → TDTC BV → ROE

5.0416

5.0173

8.8805

0.1124

1.0321

Support for mediation

GDI → TDTC BV → ROE

−1.9174

−1.9009

−8.7543

−0.2512

−0.013

Support for mediation

Sizeta → TDTC MV → ROE

−1.2634

−1.2543

7.6537

−0.3303

−0.1514

Support for mediation

Sizesales → TDTC MV → ROE

−1.3315

−1.3214

9.4334

−0.297

−0.1337

Support for mediation

GDP → TDTC MV → ROE

1.2529

1.1873

0.6479

0.0103

0.053

Support for mediation

GDI → TDTC MV → ROE

0.224

0.1907

0.1793

−0.0121

0.0162

No support for mediation

Sizeta → LTDTC BV ->ROE

11.811

11.8049

−233.5834

−2.4789

24.4326

No support for mediation

Sizesales → LTDTC BV → ROE

11.1461

11.1384

−235.9986

−1.989

20.1296

No support for mediation

GDP → LTDTC BV → ROE

−5.3272

−5.3129

−12.5172

−1.8063

0.2173

No support for mediation

GDI → LTDTC BV → ROE

1.9135

1.9068

13.6538

−0.0315

0.7993

No support for mediation

Sizeta → LTDTC MV → ROE

−0.8669

−0.843

2.9747

−0.2293

−0.0116

Support for mediation

Sizesales → LTDTC MV → ROE

−0.7303

−0.6759

2.0349

−0.1603

−0.0054

Support for mediation

GDP → LTDTC MV → ROE

0.9142

0.8161

0.3187

0.0000

0.0451

Support for mediation

GDI → LTDTC MV → ROE

1.0405

0.9376

1.2344

0.0000

0.0418

Support for mediation

Sizeta → STDTC BV → ROE

−2.2382

−2.2338

25.7008

−4.4546

2.4637

No support for mediation

Sizesales → STDTC BV → ROE

−2.3695

−2.3658

35.9862

−5.0232

2.4132

No support for mediation

GDP → STDTC BV → ROE

2.1745

2.1524

2.6783

−0.3848

0.8552

No support for mediation

GDI → STDTC BV → ROE

−2.0551

−1.9905

−4.5947

−0.2356

0.0958

No support for mediation

Sizeta → STDTC MV ->ROE

−0.8695

−0.8622

7.6794

−0.3662

−0.1322

Support for mediation

Sizesales → STDTC MV ->ROE

−1.1065

−1.0988

10.5401

−0.3558

−0.1413

Support for mediation

GDP → STDTC MV → ROE

1.3019

1.2606

0.9183

0.0165

0.0666

Support for mediation

GDI → STDTC MV → ROE

1.2411

1.1711

2.3517

0.0033

0.0424

Support for mediation

Fig. 6.14
figure 14

Single equation model

Full size image

Rights and permissions

Reprints and permissions

Copyright information

© 2018 Springer International Publishing AG

About this chapter

Check for updates. Verify currency and authenticity via CrossMark

Cite this chapter

Ramli, N.A., Latan, H., Nartea, G.V. (2018). Why Should PLS-SEM Be Used Rather Than Regression? Evidence from the Capital Structure Perspective. In: Avkiran, N., Ringle, C. (eds) Partial Least Squares Structural Equation Modeling. International Series in Operations Research & Management Science, vol 267. Springer, Cham. https://doi.org/10.1007/978-3-319-71691-6_6

Download citation

Publish with us

Policies and ethics

Search

Navigation