How to Build Pride in the Workplace?

Round 1

Reviewer 1 Report

Dear Authors.

Excellent statistic work!!!

Yet, "Right conclusions based on wrong assumptions"?

I'd like to look at Your questionnaire (81 questions) because I can not see the definition of "organizational pride" in Your paper, what, in my opinion should be the pillar of conducted research.

Without defining as above, the whole concept (although very valuable) is biased.

Author Response

Dear Reviewer,

Thank you very much for your contributes. Please check our comments below, and all the changes implemented in the manuscript.

All the best.
The authors

Dear Authors.

Excellent statistic work!!!

Yet, "Right conclusions based on wrong assumptions"?

I'd like to look at Your questionnaire (81 questions) because I can not see the definition of "organizational pride" in Your paper, what, in my opinion should be the pillar of conducted research.

Without defining as above, the whole concept (although very valuable) is biased.

A: Thank you for your comments, we changed the literature review in order to better define Organizational Pride and all the concepts that we proposed to be related to it. Regarding the questions that characterize Organizational Pride they are all included in the table 2 and were gathered from previous studies.

 

 

Reviewer 2 Report

The text presented is of high academic quality. However, there are some points that could be improved:

1. The theoretical framework needs to be expanded a little more, focusing on the issue of organizational pride. Too much is devoted to the topic of job satisfaction. The theoretical framework is strong but short.
2. It needs to take care of the order of its ideas. Points are made with references from years ago that lack a logical order. For example, he points out something from 2020, and then points out something from 2017, as if it were a newer contribution. He needs to take care of the order in which he arranges the texts and references.
3. The theoretical framework is so short that it refers to notions such as social responsibility or citizen behavior without explaining it.
4. Previous research works are mentioned, paying attention to what they contribute but not to what was missing and what the present study does consider.
5. As it points out, the sample is too small, and focuses only on one type of company. Unfortunately, the fact that the study is based on a business company, but that the studies in which it is argued and previous results were in industrial companies makes the research in general a little questionable. It must be considered that there are cultural aspects, training and socioeconomic and educational level that may cause differences between what this text discovers and what could happen in industrial companies.  This is the point that I consider to be the most delicate. Perhaps it would be advisable to extend the research and include a company from another line of business to strengthen the results and give reliability to the study.

6. In order to remedy the above, the author could expand the sample to include a company from another line of business, or expand the theoretical framework and look for research carried out in business companies or in the line of business chosen as a sample.

7. For clarity, it is suggested to include hypotheses and research questions. These should be related and proven in the discussion of results.

Author Response

Dear Reviewer,

Thank you very much for your contributes. Please check our comments below, and all the changes implemented in the manuscript.

All the best.
The authors

 

The text presented is of high academic quality. However, there are some points that could be improved:

1. The theoretical framework needs to be expanded a little more, focusing on the issue of organizational pride. Too much is devoted to the topic of job satisfaction. The theoretical framework is strong but short.
   2. It needs to take care of the order of its ideas. Points are made with references from years ago that lack a logical order. For example, he points out something from 2020, and then points out something from 2017, as if it were a newer contribution. He needs to take care of the order in which he arranges the texts and references.
   3. The theoretical framework is so short that it refers to notions such as social responsibility or citizen behavior without explaining it.
   4. Previous research works are mentioned, paying attention to what they contribute but not to what was missing and what the present study does consider.
   5. As it points out, the sample is too small, and focuses only on one type of company. Unfortunately, the fact that the study is based on a business company, but that the studies in which it is argued and previous results were in industrial companies makes the research in general a little questionable. It must be considered that there are cultural aspects, training and socioeconomic and educational level that may cause differences between what this text discovers and what could happen in industrial companies.  This is the point that I consider to be the most delicate. Perhaps it would be advisable to extend the research and include a company from another line of business to strengthen the results and give reliability to the study.
2. In order to remedy the above, the author could expand the sample to include a company from another line of business, or expand the theoretical framework and look for research carried out in business companies or in the line of business chosen as a sample.
3. For clarity, it is suggested to include hypotheses and research questions. These should be related and proven in the discussion of results.

A: Thank you for pointing out these issues. We agree with them and have re-written the literature review in order to address them. We have re-organized the ideas so their main focus is Organizational Pride, they are presented in the correct order and all relevant concepts introduced were also better defined. Most studies present only a very segmented analysis and that is their main flaw. Furthermore, even though we looked for some more references there was none available in the same context. We emphasize this is the limitations of the study as well as the sample size that we could not address at the moment. We have also added all the hypotheses in order to facilitate the understanding of the model and they are all addressed in the discussion.

Reviewer 3 Report

Thank you for the opportunity to revise this manuscript. Although I recommend major revisions, I truly believe that the paper has good potential. I am sure that after careful revisions, it will qualify for publication.

But at this point, here are my comments and suggestions.

Line 32: the beginning of the sentence needs revisions. The same applies to Lines 36, 37.  In fact, I found many sentences throughout the paper that are incomplete or seem to be combinations of different sentences merged together.

Lines 51-70: there is a vast literature on organizational commitment, including famous theoretical models. What the authors do as a theoretical background does not capture any of these models and does not clarify the model adopted by the authors. The same applies to the next paragraph that explains the role of pride in organizational commitment.

In my opinion, the Lit review section has to be rewritten to clearly explain what the exact theoretical underpinning of this research is, what is the model followed by the authors, what are the hypotheses, and what is their support in the existing literature. As it is now, this section is a sequence of information related to the topic, but not the core information on the matter. 

The research model to be estimated in the section devoted to data analysis must be grounded in extant literature. As it is now, it seems to be derived from intuition and a sort of statistical engineering. This is somehow evident in lines 109-112. Besides, the sentence on line 109 does not really make sense. 

Lines 113 - 117: a better description of the method is required, along with why this method is preferred over other methods. The fact that there are complex relationships is not an argument: mediation and moderation analysis can be performed in multiple regression or cov-based SEM. Also, the authors ought to mention the software used to analyze the data. 

I am not sure that I understand why the authors need an OLS model and a PLS model. A PLS model capturing the complex relationships is enough. The models presented in Table 1 are problematic anyway: the R squared is so high that it most likely indicates some structural issues. 

Lines 127-128: the authors say that "Although all models obtained very high R2, indicating that those aspects can explain 127 most of the variance of the dependent variable, this analysis does not take into account 128 the interactions between all the aspects."

It is not only that this very high R2 is problematic, but the argument that these models do not take into consideration interactions (and this is why you need PLS-SEM) is weak. Why don't you add interaction terms to your multiple regression model? The problem doesn't come from interaction terms but the models as such. There is no clear theoretical background in support of these models. There is no check for potential multicollinearity. The models result from a forward feature selection when they should have been driven by existing literature. 

Table 2: no need to use more than 3 decimal places.

Also, Table 2: do the items included in your measurement come from previously validated instruments? Are the dimensions presented in Table 2 known in the literature? How can you be sure that all latent constructs are unidimensional? For instance, Recognition is measured using 7 items. How can you know that these items load in only one latent construct and not two? 

Also Table 2: the authors mentioned that they drop items with loadings less than 0.7 (which is the right thing to do). This suggests that your scales contained more items than reported: do the original scales split into specific subscales and the authors ignored this aspect?

Line 105: the authors say that the survey comprises 81 questions: where these questions come from? Are they known instruments? Where they previously used? What is the structure of each instrument? Do they support formative or reflective measurement?

Figure 1 is very nice, but the model is plagued by serious endogeneity and circular relationships. I recommend a careful analysis of the existing literature and the estimation of a decent model. This one is seriously flawed. 

To conclude, the topic of the paper is interesting, but the manuscript has to be rewritten. First, the authors must ground their work in the right literature, as this will help them identify the correct model to estimate. The methodology must be carefully revised. Please consider my comments above as the starting point to improve your work.

Author Response

Dear Reviewer,

Thank you very much for your contributes. Please check our comments below, and all the changes implemented in the manuscript.

All the best.
The authors

 

Thank you for the opportunity to revise this manuscript. Although I recommend major revisions, I truly believe that the paper has good potential. I am sure that after careful revisions, it will qualify for publication.

But at this point, here are my comments and suggestions.

Line 32: the beginning of the sentence needs revisions. The same applies to Lines 36, 37.  In fact, I found many sentences throughout the paper that are incomplete or seem to be combinations of different sentences merged together.

Lines 51-70: there is a vast literature on organizational commitment, including famous theoretical models. What the authors do as a theoretical background does not capture any of these models and does not clarify the model adopted by the authors. The same applies to the next paragraph that explains the role of pride in organizational commitment.

In my opinion, the Lit review section has to be rewritten to clearly explain what the exact theoretical underpinning of this research is, what is the model followed by the authors, what are the hypotheses, and what is their support in the existing literature. As it is now, this section is a sequence of information related to the topic, but not the core information on the matter. 

The research model to be estimated in the section devoted to data analysis must be grounded in extant literature. As it is now, it seems to be derived from intuition and a sort of statistical engineering. This is somehow evident in lines 109-112. Besides, the sentence on line 109 does not really make sense. 

Lines 113 - 117: a better description of the method is required, along with why this method is preferred over other methods. The fact that there are complex relationships is not an argument: mediation and moderation analysis can be performed in multiple regression or cov-based SEM. Also, the authors ought to mention the software used to analyze the data. 

I am not sure that I understand why the authors need an OLS model and a PLS model. A PLS model capturing the complex relationships is enough. The models presented in Table 1 are problematic anyway: the R squared is so high that it most likely indicates some structural issues. 

 

 

Table 2: no need to use more than 3 decimal places.

Also, Table 2: do the items included in your measurement come from previously validated instruments? Are the dimensions presented in Table 2 known in the literature? How can you be sure that all latent constructs are unidimensional? For instance, Recognition is measured using 7 items. How can you know that these items load in only one latent construct and not two? 

Also Table 2: the authors mentioned that they drop items with loadings less than 0.7 (which is the right thing to do). This suggests that your scales contained more items than reported: do the original scales split into specific subscales and the authors ignored this aspect?

Line 105: the authors say that the survey comprises 81 questions: where these questions come from? Are they known instruments? Where they previously used? What is the structure of each instrument? Do they support formative or reflective measurement?

Figure 1 is very nice, but the model is plagued by serious endogeneity and circular relationships. I recommend a careful analysis of the existing literature and the estimation of a decent model. This one is seriously flawed. 

To conclude, the topic of the paper is interesting, but the manuscript has to be rewritten. First, the authors must ground their work in the right literature, as this will help them identify the correct model to estimate. The methodology must be carefully revised. Please consider my comments above as the starting point to improve your work.

A: Thank you for raising all these issues, we agree with them and have re-structured the whole model. First, we have re-written the literature review to support all our hypotheses in previous studies and gather a better definition of Organizational Pride and what has been done to understand it. The methodology section was also improved to better explain the models used, the OLS although flawed allowed us to understand some contradictions found in the literature. Its flaws were also emphasized in the results. The measurement instruments were better explained and were all based on previous studies. We have found a different model, based on previous studies and using validated measurements which allowed to support some of our hypotheses.

Round 2

Reviewer 1 Report

OK

Author Response

Thank you very much!

All the best.

Reviewer 2 Report

I believe that the changes made are in line with what was requested. Although unfortunately the study is still very limited, I believe that this opens the possibility for the researcher to continue with other articles that improve the results.

Author Response

Thank you very much!

All the best.

Reviewer 3 Report

Dear Authors

Your work improved significantly in the Lit review and Discussions part, and you also tried to clarify the method. However, I still have some important concerns.

1. The OLS model seems to use different measurement items than the PLS model. How can you know, then, that the model is flawed because is an OLS model, and not because you have different variables? And how can you compare the results of the models? You say that you have 81 variables, but you choose to use only a few in your OLS model. Ground this choice in extant literature, or simply drop it!
2. The very high R2 values of your OLS models may be the result of multicollinearity, as you state on line 192, but this can be tested via VIF values, and not only hypothesised. My suggestion for you is to simply give up the OLS model: you said in your answer to me that you keep this model to clarify some contradictions in the existing literature: in order to do so, you need to explain that this exact model was used before and compare your results with those results. But considering that on Lines 325-328 that some authors already used SEM in approaching this topic: why, then, going back to OLS? 
3. You said that you conducted your PLS-SEM in Python, using the pslpm package: most likely you mean the "plspm" package, a port of the R package plspm developed by Gaston Sanchez, with additional features adopted from the R package seminr. Please, make the correction.
4. Figure 1 presents your research model. I will discuss only one path, so you can understand my concern: Management support goes into Appreciation, which goes into Affective commitment, which eventually goes into Commitment. This is a typical case of endogeneity, that must be addressed. You can see what I mean, by watching this video: https://www.youtube.com/watch?v=qCvvUxR978U After addressing the endogeneity part, you may want to come back to what you say on Lines 377 - 379. 
5. In Figure 1 you have not only total effects, but also indirect effects and direct effects (total effects decomposition). All these results, including direct, indirect, and total effects, must be reported in a separate table and discussed in terms of mediators' relevance. If the authors decided to adopt a structural model, then it is appropriate to discuss the findings in terms of mechanisms that underline their relationships. 
6. Line 263: you say that "The OLS model, despite all its flaws, indicates commitment as a cause of Organizational Pride". How can you be sure that what this model indicates is not misleading, exactly because the model is flawed? 
7. Lines 175 - 176: you write that "The PLS was the SEM model chosen because the main goal 175 was to identify the key “driver” constructs". Any regression technique identifies drivers, the reason why you need PLS-SEM is not that you want to explore predictors, but that your predictors are latent constructs, your variables are not normally distributed, your sample size is small, and your research is exploratory. These are the four reasons accepted in the literature for using PLS-SEM. In addition, PLS is not an SEM model but a modeling technique. Please, make the correction.
8. Lines 162 - 166: this sounds as a weak argument to me. There are many alternatives to your OLS model, such as extracting factor scores from your organisational pride items, then performing OLS regression, or creating a composite index as the average score of the items measuring organisational pride. Why do you think that presenting several OLS models is a good choice? 

Overall, the paper improved to a significant degree, and for this reason I am sorry to have to recommend once again major revisions. But the OLS model, if the authors want to keep it, has to be grounded in previous literature as it is (why exact this model, with variables measured in exact this way).

Second, the authors ought to explain why, when they have 81 variables, they choose to create OLS models with only a few predictors and why they think that this is the right thing to do.

Third, Figure 1 shows a complex model, that suffers from endogeneity. This aspect must be corrected, or the authors ought to explain why endogeneity may not be a problem in this case.

Last, but not least, if the authors estimate a structural model, they ought to discuss the many mediating effects they captured.  

If these aspects are clarified, I will be happy to recommend acceptance. But for now, I regret to have to say that I cannot. I would also appreciate it if the authors provided item-by-item answers to my comments, not only one paragraph at the end of my letter, as they did this time. Item-to-item response to the reviewer's comments and recommendations is the norm. 

Author Response

Dear Reviewer, thank you very much for all your comments. Please check below our answers. 

All the best.

"Dear Authors

Your work improved significantly in the Lit review and Discussions part, and you also tried to clarify the method. However, I still have some important concerns.

The OLS model seems to use different measurement items than the PLS model. How can you know, then, that the model is flawed because is an OLS model, and not because you have different variables? And how can you compare the results of the models? You say that you have 81 variables, but you choose to use only a few in your OLS model. Ground this choice in extant literature, or simply drop it!

The very high R2 values of your OLS models may be the result of multicollinearity, as you state on line 192, but this can be tested via VIF values, and not only hypothesised. My suggestion for you is to simply give up the OLS model: you said in your answer to me that you keep this model to clarify some contradictions in the existing literature: in order to do so, you need to explain that this exact model was used before and compare your results with those results. But considering that on Lines 325-328 that some authors already used SEM in approaching this topic: why, then, going back to OLS?"

A: Thank you for your remark, we have dropped the OLS model and section and adjusted the Methodology, and Discussion parts accordingly.

"You said that you conducted your PLS-SEM in Python, using the pslpm package: most likely you mean the "plspm" package, a port of the R package plspm developed by Gaston Sanchez, with additional features adopted from the R package seminr. Please, make the correction."

A: Thank you for your note, we have corrected the reference of the packages used.

"Figure 1 presents your research model. I will discuss only one path, so you can understand my concern: Management support goes into Appreciation, which goes into Affective commitment, which eventually goes into Commitment. This is a typical case of endogeneity, that must be addressed. You can see what I mean, by watching this video: https://www.youtube.com/watch?v=qCvvUxR978U After addressing the endogeneity part, you may want to come back to what you say on Lines 377 - 379."

A: Thank you for your remark. The main goal of the paper was to identify among the exogenous constructs those related to Organizational Pride, and even though the model suffers from endogeneity it affects mostly the dimension of the effects of mediator constructs such as Organizational Identification or Affective Commitment (or even Pride when considered towards Commitment). We want to continue this research further, after also gathering more data, and that was identified in the limitations as an aspect to include in next stages of our research. We have also eliminated from the conclusion the statement that we analyze the relationship between all the constructs, which can only be achieved after eliminating any endogeneity issues.

"In Figure 1 you have not only total effects, but also indirect effects and direct effects (total effects decomposition). All these results, including direct, indirect, and total effects, must be reported in a separate table and discussed in terms of mediators' relevance. If the authors decided to adopt a structural model, then it is appropriate to discuss the findings in terms of mechanisms that underline their relationships."

A: Thank you for your comment, we have added the table to distinguish between direct and indirect effects.

"Line 263: you say that "The OLS model, despite all its flaws, indicates commitment as a cause of Organizational Pride". How can you be sure that what this model indicates is not misleading, exactly because the model is flawed?"

A: Thank you for your note, we have eliminated the OLS model as you suggested.

"Lines 175 - 176: you write that "The PLS was the SEM model chosen because the main goal 175 was to identify the key “driver” constructs". Any regression technique identifies drivers, the reason why you need PLS-SEM is not that you want to explore predictors, but that your predictors are latent constructs, your variables are not normally distributed, your sample size is small, and your research is exploratory. These are the four reasons accepted in the literature for using PLS-SEM. In addition, PLS is not an SEM model but a modeling technique. Please, make the correction."

A: Thank you for your remark. We had already mentioned some of those aspects, we have included the ones missing and removed the lines 175-176.

"Lines 162 - 166: this sounds as a weak argument to me. There are many alternatives to your OLS model, such as extracting factor scores from your organisational pride items, then performing OLS regression, or creating a composite index as the average score of the items measuring organisational pride. Why do you think that presenting several OLS models is a good choice?"

A: Thank you for your note, we have eliminated the OLS model as you suggested.

"Overall, the paper improved to a significant degree, and for this reason I am sorry to have to recommend once again major revisions. But the OLS model, if the authors want to keep it, has to be grounded in previous literature as it is (why exact this model, with variables measured in exact this way)."

A: Thank you for your note, we have eliminated the OLS model as you suggested.

"Second, the authors ought to explain why, when they have 81 variables, they choose to create OLS models with only a few predictors and why they think that this is the right thing to do."

A: Thank you for your comments, we have removed the OLS model as you suggested.

"Third, Figure 1 shows a complex model, that suffers from endogeneity. This aspect must be corrected, or the authors ought to explain why endogeneity may not be a problem in this case."

A: Thank you for your note. The main goal of the paper was to identify among the exogenous constructs those related to Organizational Pride, and even though the model suffers from endogeneity it affects mostly the dimension of the effects of mediator constructs such as Organizational Identification or Affective Commitment (or even Pride when considered towards Commitment). We want to continue this research further, after also gathering more data, and that was identified in the limitations as an aspect to include in next stages of our research.

"Last, but not least, if the authors estimate a structural model, they ought to discuss the many mediating effects they captured. "

A: Thank you for your remark, we have added a table distinguishing between direct and indirect effects what helped understanding mediating effects included in the discussion.

"If these aspects are clarified, I will be happy to recommend acceptance. But for now, I regret to have to say that I cannot. I would also appreciate it if the authors provided item-by-item answers to my comments, not only one paragraph at the end of my letter, as they did this time. Item-to-item response to the reviewer's comments and recommendations is the norm."

A: Thank you for your comments, we apologize for not answering item-to-item. .

Author Response File: Author Response.docx

Round 3

Reviewer 3 Report

Dear Authors,

It was a wise choice to drop the OLS model. I still have a problem with the endogeneity of your model: it may bias your results, making the current estimations misleading. Considering that the rest of the work is fine, I will leave the final decision to the Editor. 

Good luck!