Corrigendum

• Date: 01 June 2020
• Published date: 01 June 2020
• DOI: http://doi.org/10.1111/emre.12403

Corrigendum

In Werneret al. (2019), we regret to advise two errors that

occurred in reporting our findings. First, the correct χ

2

value for our alternative model 2 in Table 3 (page 751)

should be 1279.18. Hence, the corrected Table 3 reads:

Consequently, the results of the χ

2

difference tests

reported on page 749 in the section “Structural model”

in the paragraph “Alternative model testing”also need

correcting. Thus, the respective corrected paragraph on

page 749 reads as follows:

“We co nd uc te d χ

2

difference tests for all models. The

model with no control variables (Δχ

2

= 3230.52, Δdf =

1365, p<0.01) and the model including only significant

controls (Δχ

2

=2759.11,Δdf =1146,p<.01) both had a

significantly better fit than the hypothesized model with

all control variables.”

Second, on page 749 in the section “Structural model”

in the paragraph “Alternative model testing”, the model

with significant controls was incorrectly reported as

showing a better fit than the model without controls.

However, the model without controls had the better fit.

The reasoning for using the model including only

significant controls to test the hypotheses lies in the

theoretical relevance of the controls.

Thus, the respective corrected paragraph on page 749

reads as follows:

“The significant χ

2

difference test contrasting the models

without controlsand with significant controls showedthat

the model without controls had the better fit (Δ

χ

2

=471.41,Δdf =219,p<0.01). Nevertheless, the

model including only significant controls (χ

2

= 1279.18,

df = 695, RMSEA = 0.05, CFI = 0.92, IFI = 0.92) was

used to test the hypotheses due to the strong theoretical

rational for assuming a relationship between the included

controls and the study’s variables (Bernerth and

Aguinis, 2016) and since the model had an adequate fit.

Following the recommendations of Becker et al.(2016),

we also tested the model without controls as a robustness

check, which showed no changes in the pattern and

significance levels of the results”.

To allow for maximum transparency, we would be

willing to provide the raw data and our calculations to

whomever requests them.

The authors would like to apologize for the errors.

References

Becker, T. E., G. Atinc, J. A. Breaugh, K. D.Carlson, J. R.

Edwards and P. E. Spector, 20 16, “Statistical control in

correlation studies: 10 essential recommendations for

organizational researchers”.Journal of Organizational

Behavior,37:157–167.

Bernerth, J. B. and H. Aguinis, 2016, “A critical review and

best-practice recommendations for control variable usage”.

Personnel Psychology,69:229–283.

Werner, A., T.Rabl, and H. Best, 2019, “Managers’corruption

prevention efforts in small and medium-sized enterprises: An

exploration of determinants”.European Management Review,

16:741–759. https://doi.org/10.1111/emre.12165

Tab le 3 Structuralmodels

Model χ

2

df χ

2

/df RMSEA CFI IFI

Hypothesizedmodel 4038.29 1841 2.19 0.06 0.76 0.77

Alternative model 1:

No controlsmodel

807.77 476 1 .70 0 .05 0.96 0.96

Alternative model 2:

Only significant

controls

1279.18 695 1.84 0.05 0.92 0.92

N= 339; RMSEA = root mean square error of approximation;

CFI = comparative fit index; IFI= incremental fit index.

European Management Review, Vol. 17, 593, (2020)

DOI: 10.1111/emre.12403

© 2020 European Academy of Management