Social Network Analysis of Farmers after the Private Cooperatives’ “Intervention” in a Rural Area of China—A Case Study of the XiangX Cooperative in Shandong Province

Round 1

Reviewer 1 Report

Overall, I think this research has merit because of the important topic of changing agricultural systems globally, as well as the importance of social networks in farmers’ livelihoods and the well-being of societies generally. However, there are several important issues that need to be address before the paper is of a high enough standard for peer-reviewed academic publication:

1. Most critical: The research design is a quantitative cross-sectional methodology, yet the authors treat the data as if it is cause-effect. This is not a longitudinal study nor is it qualitative—if this is not true, then the authors need to heavily revise the methods. If this is true, the entire paper requires re-framing to remove qualitative inferences and interpretations.

2. The logic of the paper from literature review to theoretical framework to methods to findings is weak. Too many different concepts are introduced without critical summary (or basic definitions, in many cases). There is little coherence from one section to the next. The authors should focus on the use of SNA to operationalize the “structure of grade” theory and then revise the literature review to critically summarize the state of the research laying the foundation for this decision—and then build from that to organize the findings and discussion.

3. The writing is not academically rigorous. There is speculative interpretation throughout. The results sections should present the analyzed data and evaluate the results--not draw conclusions based on assumptions and information not included in the research design (collected as data). Note: it is impossible to explain WHY something is happening based on quantitative SNA data such as in this study...quant only gives us HOW or WHAT (qualitative data provides WHY--but this is not a qualitative study). This is why SNA is problematic. Here is one critical paper that explain why we need to go beyond quantitative social network measures:

Diehl, J.A. and M. Bose.  (2023).  A sustainable livelihoods approach to measuring mobilization of resources through social networks among vulnerable populations: A case study of Delhi farmers. Social Science and Humanities.

The authors might find some useful references in the citation lists of these papers too. I am not suggesting they change the research design—only that they need to acknowledge the serious and real limitations of using a quantitative SNA approach. The concept of “trust” or any other subjective phenomenon cannot be captured using SNA. Which is fine, but they should focus on what they HAVE measured and interpret only that data deeply.

4. Authors should acknowledge strengths and weaknesses of the study design and data—including reliability and validity, generalizability, and other conceptual validity or bias issues.

Other comments are noted in the attached file.

The topic of this paper is important and timely; however, it is just as important that the data are subordinated in the larger discourse and that the authors do not over-interpret analysis.

Comments for author File: Comments.pdf

English language is adequate--could be copy-edited again to improve grammar/syntax. A bigger issue is to reduce jargon and use language accessible to a broad academic audience. Terms should have definitions! Concepts should be described!

Author Response

Thank you very much for taking the time to review this manuscript. We have benefited a lot from your comments.  The following is our response to the summary questions of your response, in addition to your annotations in the PDF, we have also made a point by point reply.  Detailed changes can be found in the newly uploaded manuscript.

1. Most critical: The research design is a quantitative cross-sectional methodology, yet the authors treat the data as if it is cause-effect. This is not a longitudinal study nor is it qualitative—if this is not true, then the authors need to heavily revise the methods. If this is true, the entire paper requires re-framing to remove qualitative inferences and interpretations.

Response: Thank you. We appreciate your insight. This research is a qualitative study of smallholders’ social network based on fieldwork in a case area. The method of social network analysis is suitable for our research because of its convenience of network visualization and qualification. We agree that inferential explanations should not be excessive when using this method. According to your comment, we have re-checked all of the text and removed inferential interpretations and subject in.

2. The logic of the paper from literature review to theoretical framework to methods to findings is weak. Too many different concepts are introduced without critical summary (or basic definitions, in many cases). There is little coherence from one section to the next. The authors should focus on the use of SNA to operationalize the “structure of grade” theory and then revise the literature review to critically summarize the state of the research laying the foundation for this decision—and then build from that to organize the findings and discussion.

Response: Thank you for your suggestion. We have made massive revision to the logic of the paper, linking background introduction, literature review, theoretical framework, result analysis and discussion. The definition of social network theory is introduced clearly in the methodology section, the SNA and structural hole are introduced as the branches of social network theory in detail in lines 124-139, 151-156, and 176-180. We enhanced the logic from the theory to the hypothesis, to make the theoretical framework clearer. In the result section, we linked the results with our theoretical hypothesis in line 340-344, 361-366, 408-412 444-446. In the discussion section, we enhanced our research implications with the theoretical framework by rewriting the whole text. In addition, we redraw figure 2, 4, 5, and added figure 1, 8, A1-A5. We believe that the revised manuscript is easier for readers to understand the logic of this paper.

3. The writing is not academically rigorous. There is speculative interpretation throughout. The results sections should present the analyzed data and evaluate the results--not draw conclusions based on assumptions and information not included in the research design (collected as data). Note: it is impossible to explain WHY something is happening based on quantitative SNA data such as in this study...quant only gives us HOW or WHAT (qualitative data provides WHY--but this is not a qualitative study). This is why SNA is problematic. Here is one critical paper that explain why we need to go beyond quantitative social network measures:

Diehl, J.A. and M. Bose. (2023). A sustainable livelihoods approach to measuring mobilization of resources through social networks among vulnerable populations: A case study of Delhi farmers. Social Science and Humanities.

The authors might find some useful references in the citation lists of these papers too. I am not suggesting they change the research design—only that they need to acknowledge the serious and real limitations of using a quantitative SNA approach. The concept of “trust” or any other subjective phenomenon cannot be captured using SNA. Which is fine, but they should focus on what they HAVE measured and interpret only that data deeply.

Response: Thank you for your deep thinking. We agree that social network analysis should focus on "what" rather than "why". After reading the literature you recommended, and widely checked relative studies, we have made a lot revisions in the revised manuscript. In the result section, we highlighted the characteristics of farmers' social network, and removed all  subjective interpretations of the results.

4. Authors should acknowledge strengths and weaknesses of the study design and data—including reliability and validity, generalizability, and other conceptual validity or bias issues. Other comments are noted in the attached file.

Response: Thank you. We have done validity and reliability tests, but they are not mentioned in the original manuscript. In the revised manuscript, we added the description of reliability and validity work in lines 239-241. Please see our reply to your comments for details.

5. The topic of this paper is important and timely; however, it is just as important that the data are subordinated in the larger discourse and that the authors do not over-interpret analysis.

Response: We appreciate your insight. The data in this study is indeed small in scale, and it is a "small study in a large context". In response to your comments, we have substantially deleted over-interpret analysis of the results (Although some of these explanations were field facts and not "extrapolations," we removed them as well). Limited by some difficulties to conduct field research during the epidemic, we did a small case study under the social network theory. In the last paragraph of the conclusion section, we have added the shortcomings and research prospects.

Author Response File: Author Response.pdf

Reviewer 2 Report

General opinion:

The topic is important and exciting, and the rationale of the reseach, the importance of cooperation and its impacts on social networks is well explained. The theoretical background is generally good, though the references should be expanded to include basic references of the SNA methodology. The hypotheses are reasonable, though the authors fail to show a clear connection between the hypotheses and the applied methodology. Therefore the results and the conclusions should be better explained, establishing a closer connection between the quantitative results and the conclusions derived. The Discussion and the Conclusion sections should be improved, too. There are some minor language errors, and the quality of some figures should be considerably improved. Therefore I suggest that the manuscript should go through major revisions before publication.

See details below.

Abstract:

The abstract is generally good. In line 22 the expression „didn’t showed” should be „did not show”.

Introduction:

line 71-72: Please give justification for the following sentence: „…chieving an overall victory in the battle against poverty.” What does it mean, and what data support this statement?

line 79-81: Please give justification for the following sentence: „Such cooperatives have made an important contribution to small farmer agricultural production, transforming it from poverty allevi-

ation to enrichment [21].”

Two minor language usage errors

·        In lines 86, 87: „researches… focuses”  is not correct, either you use „research….. focuses”, or „researchers ….. focus”.

·        In lines 123  „showed” should be „shows”. The same error is found in line 124. Please correct these.

Theoretical analysis and hypothesis: (a better subtitle would be „Theoretical foundation and hypotheses”.

line 126: „…Structure of Grade” (Fig. 1) Proposed by…” should be „ …Structure of Grade” (Fig. 1) proposed by…”.

Figure 1 should be enlarged and sharpened for better readability of the text.

line 156: instead of „compare” you should write „compared”.

line 159: instead of „exist” you should write „existence”.

Study area overview:

At line 206 you should give a better explanation for the choice of the cooperative for the analysis. To support your choice some background data are needed: how many cooperatives exist in China? How many in the Shandong province? What is the average number of members in these cooperatives? What is the proportion of the 50+ age group among cooperative members? And among all farmers in China?

Figure 2 should be a bit larger and sharper for better visibility.

After line 227 some descriptive statistics should be presented about the 114 respondents, regarding their age, land use, neighborhood relationships, and all the main aspects they were asked about.

Paragraph from line 229 to 235: When explaining the concepts of social network analysis, you should include some basic references – as the methodology goes back tot he 1990ies, and several excellent books have been published about the concepts and the methods that you will also apply in your research.

In Table 1 the xij values should be specified more precisely. E.g. when you define the neighbor values, you write:  „If smallholder i and smallholder j are neighbors of each other's house sites, then 𝑥𝑖𝑗=1, 𝑥𝑗𝑖=1 in the matrix.”, you should add: otherwise xij=0.

Similarly, the explanation should be more precise for the other 5 aspects, too.

lines 244-246: the sentence is a repetition of that of Table 1,  under the explanation of the labor.sharing relationship, so it is redundant here.

line 247-250: Please be more precise with the definition of the matrix for the social network, e.g if xij=1 in kinship, and xij = 1 in labor but xij=0 in the other 4 aspects, then xij=2 in the entire social network. (or should it be xij=1 in the social network?, please give a precise explanation).

In this paragraph the notations should be clearer, too. The xij values for the kinship matrix could be denoted by xij(K), for the labor-sharing matrix xij(L), and for the entire social network xij(SN), or something similar.

line 256: „…the more connections the members have [53].” should rather be „…the more connections the members have, on average [53].”

line 261: What does a „redundant relationship” mean here? Please explain.

line 267: What does a „unidirectional” relation mean here? You have not mentioned anything about directed networks yet.

After line 275 you should explain what the theoretical minimum and maximum values are for ND, NE, NH and NC, because later you refer to computed values as „low” and „high”, so we should know what low or high mean for each inticator.

Lines 276 and 295: the subheadings/subtitles are not necessary, they should be omitted.

Line 287: „proximity centrality” should be „closeness centrality”.

Line 293: „intermediation centrality” should be „betweenness centrality”.

After line 303: After defining ND, NE, NH and NC and the three centrality measures and QAP, you should explain how these measures are related to your research questions and how they will be used to prove your hypotheses.

Results:

line 309: Please add a reference tot he Ucinet software (a website or some descriptive publication).

L310-312: The sentence is not clear, please explain your meaning: what do you mean by „two along ’corridors’”?

Paragraph from line 310-321: Please explain your meanings in a more detailed way, referring to what is seen in Figure 3. Explain the lines, the nodes, the notations briefly. Nodes like Z10, Z15, cannot be seen in the figure, as the notations are too small. How can we identify the two villages in the figure? When you mention Z10 and Z15 specifically, you should give a detailed description of their characteristics, otherwise, their mentioning is not meaningful.

Figure 3 should be enlarged to make the legends readable.

In lines 325 to 335 you evaluate the ND, NH, NC and NE values as „low” or „high” – but we should know what range these values could theoretically cover – theoretical minimum and maximum values should be given for your dataset. You say that NE values are „low”, but actually, Table 2 shows that NE values are the highest in the table!.

In the paragraphs from line 325 to 357 you only describe the values, but their socioeconomic meaning, or the  interpretation  of NE, ND, NC, NH is missing!

Line 357: The H1 (first hypothesis) is not justified: you claim that linkages weakened, but to prove this you should have a „before” and an „after” situation, which you do not present. Please explain in more detail why you think that H1 is proven here.

Figure 4 is not necessary, - or at its size its significance cannot be seen.

Section 5.2:

The nodes mentioned in this section are not visible, because Figure 5 is too small. Besides, The centrality values presented in Fig 5 should also be presented in a table that could be presented in Appendix A, if not in the main text. Such a table  is needed to support H3 (hypothesis 3)

Figure 5 is too small, is should be enlarged, to make the values and the legends readable.

line 412: Please explain why you need all the three types of centraility and how these types are related to your H3 hypothesis.

Discussion:

In Discussions your results should be explained and compared to previous research findings – either to those that analysed the farm cooperation using SNA methods, or those that assess the impacts of cooperation on the strength  f socila network relationships among farms, or the between-village and within-village connections, or the existence of „elites” in such networks. These comparisons are missing.

However, in this section the three subheadings discuss ideas that should be placed in the section Conclusions, as they refer to the consequences and implications of your results.  So I suggest that these ideas should be summarised not here, but in Conclusions, in a somewhat shorter form.

For this reason Figure 8 seems to be redundant here, as it is not strictly related to your findings.

Conclusions:

The contents of this section should be placed in the Discussion section, and besides summarising your results, you should also compare your findings to previous research results that  apply similar methodology, and/or address similar research questions.

Conclusions should also include a paragraph about the limitations of your work, such as choice of locations, size of the sample, age group of the respondents, etc..

 With these corrections, major revision is recommended.

There are minor mistakes about singular-plural usage, and some typos. See details in the general comments.

Author Response

Thank you very much for taking the time to review this manuscript. Your deep thinking has provided us with a lot of help. You can find responses to each of your comments in the documents we uploaded. We have also made corresponding changes in the latest uploaded manuscript.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

First, impressive improvement given the short time to revise the manuscript. It is much more clear and better organized. I offer several high-level comments to improve the style, tone, and academic language of the paper:

1.        Manuscript could use a comprehensive copy-edit to improve grammar and syntax.

2.        Introduction can be edited to be more succinct.

3.        Discussion: add first paragraph with high level summary of the main research objective and state which hypotheses were supported or not supported. This helps the reader make sense of the detailed discussion sections. Need to link findings for all three hypotheses back to the literature—how does this research confirm/ complement/ contradict previous research? Note: hypotheses cannot be “proven” only supported—this is social science where we can only calculate probabilities not proofs. I still think the discussion requires some revision. Here is a guideline for organizing it:

1. Begin by stating if the hypothesis is confirmed
2. Link results (show relationships) (don’t repeat findings)
3. Relate results to the literature
4. List implications
5. Claim significance
6. Question the findings
7. Note limitations
8. Suggest future research
3. Conclusion: some of this can move to the discussion section. Authors should focus on high level significance and contribution of this study to the future of enhancing sustainability of agriculture.

See comments above.

Author Response

First, impressive improvement given the short time to revise the manuscript. It is much more clear and better organized. I offer several high-level comments to improve the style, tone, and academic language of the paper:

1. Manuscript could use a comprehensive copy-edit to improve grammar and syntax.

Response: Thank you. We have used the language editing service of MDPI before the submission, and in the revised version, we have asked a native English colleague to help us check the language again.

2. Introduction can be edited to be more succinct.

Response: Thank you. We condensed the introduction by cutting out much superfluous expression. . The introduction section has been reduced from 1,123 characters to 926 characters in the revised manuscript.

3. Discussion: add first paragraph with high level summary of the main research objective and state which hypotheses were supported or not supported. This helps the reader make sense of the detailed discussion sections. Need to link findings for all three hypotheses back to the literature—how does this research confirm/ complement/ contradict previous research? Note: hypotheses cannot be “proven” only supported—this is social science where we can only calculate probabilities not proofs. I still think the discussion requires some revision. Here is a guideline for organizing it:

Begin by stating if the hypothesis is confirmed

Link results (show relationships) (don’t repeat findings)

Relate results to the literature

List implications

Claim significance

Question the findings

Note limitations

Suggest future research

Response: Thank you for your comment. Your suggestion have helped us well in revising the discussion. We re-adjusted the structure of the discussion, and linked the hypothesis, the reults, the relevant literature, the implications and research outlook. Meanwhile, we replace the word “prove” in the hypothesis with “support”.

4. Conclusion: some of this can move to the discussion section. Authors should focus on high level significance and contribution of this study to the future of enhancing sustainability of agriculture.

Response: Thank you. According to another reviewer' opinion, we kept a short summary of the discussion in the conclusion section. In the revised conclusion, we add the significance of this study.

Reviewer 2 Report

The authors made an extensive improvement of the manuscript, including structure, content and quality of the illustrations. They have properly addressed my comments and criticisms, and the solution they chose is acceptable for me. Therefore I consider the paper suitable for publication.

Author Response

The authors made an extensive improvement of the manuscript, including structure, content and quality of the illustrations. They have properly addressed my comments and criticisms, and the solution they chose is acceptable for me. Therefore I consider the paper suitable for publication.

Response: Thank you. We appreciate your previous comments, which help us a lot in improving the quality of the paper. Based on the comments of another reviewer, we have revised the introduction, discussion and conclusion, which you can see in the newly revised uploaded.