Exploration of Social Media for Observing Improper Tourist Behaviors in a National Park

Round 1

Reviewer 1 Report

This is an excelling study. However, there are some issues with the method section which are fixed by addressing some of the points below.

Please note that I also included several editing comments in the PDF which need to be fixed before the study is published.

p. 3, line 94 to 99: Please mention in this section that only public profile posts were used.

p. 3, line 100: Please mention in this section what sampling strategy you used to collect your posts?  It sounds like a purposive sampling strategy?

Please clarify what processes were used to enhance reliability and validity of findings?

p. 3, line 101: How was the text analysis used on the visual material? Please include more detail for your readers.

p. 3, line 101: Who else was involved in the data analysis (three people?) What is their expertise? Did they receive training in coding? Please provide more information.

p. 3, line 112: Please define 'qualified' in this context. Otherwise this doesn't make sense.

p. 5, line 145: Are you able to make a distinction between the different types of Likes, i.e. sad, wow, love, etc. This would add value to the findings.

P. 5, table 3: Why are these actions bundled together? Please provide a sentence to explain why the bundling was necessary.

p. 7, line 193: Your discussion focuses on 'careless actions' - what about your interpretations of all the other actions?

p. 7, line 209: But how is enhanced monitoring possible with no manpower?

Comments for author File: Comments.pdf

Author Response

The authors thank the reviewer for the valuable comments on the manuscript.  For specific responses to your comments, please see the attached response file.

Reviewer 2 Report

This paper is an exploratory study that used social media to analyze the types of improper behavior among visitors at Yushan National Park in Taiwan, the message modes that visitors use to disseminate their improper behavior, and other Facebook (FB) users’ responses to posts that involve improper behavior. Below are some suggestions intented to help the authors strengthen their manuscript.

(1) This paper is a Project Report, but does this exclude a section on Literature Review?

(2) Concerning the Methods, the readers do not know how long was necessary for collecting the data. Which period of time? Moreover, it is necessary to know more about the text analysis, including software used.

(3) Concerning the Results, I think more in depth analyses would greatly contribute to the significance of this work.

(4) Concerning the Conclusion, I would suggest rewriting this section because it is concentrated on limitations and future research.

(5) Minor comments. There are some language anomalies that need correcting, such as:

- Page 2, lines 84-85, “This park also serves as a sanctuary to 11 endangered and 51 rare bird and animal species”.

Good luck with the review

Author Response

The authors thank the reviewer for the valuable comments on the manuscript.  For specific responses to your comments, please see the attached response file.

Author Response File: Author Response.pdf

Reviewer 3 Report

I am glad to review this manuscript. This project report investigated Facebook post contents to reveal the types of improper visitor behaviors in a national park. The research idea is very interesting and timely to utilize social media data and develop implications for the national park management. The manuscript is well-written. These project outcomes may provide other researchers and practitioners insights for better understanding traveler behaviors and the preservation of the national park. Meanwhile, readers would have many questions authors need to improve this manuscript and clarify unclear statements to be published in a research journal.

1.

In the ‘Keywords,’ it seems that this study did not really conduct ‘text analysis’ with post texts. Since this study reviewed all components of a post such as photo and text, I think ‘content analysis’ would be more appropriate than using ‘text analysis.’

2.

I think that the key point of this study is to demonstrate the benefits of  ‘social media surveillance’ using this case. In the ‘Introduction,’ authors need to describe the definition, features, and process of this practice.

3.

This study collected and examined the different numbers of likes, comments, and shares of the targeted posts. However, why are the numbers of user reactions important in this study? Although the authors compared the numbers based on the types of improper behaviors, there is no justification why this study needs to compare the numbers to draw a research conclusion. To clarify the problem, the authors would need to build a research problem statement stating the necessity based on previous literature. In addition, a discussion should be improved to interpret the theoretical or managerial implications based on the comparison among the frequencies in the cross analysis.

4.

This report would need to clarify the purpose of this study. Three research questions in the ‘Introduction’ do not include the engagement indicators (i.e., likes, comments, and shares). Therefore, the authors need to specifically show the research purposes.

5.

As a practical project report, more details about the research methods should be described to provide more benefits with this project. Please consider these questions as follows:

- Data collection tool: What tools did the authors use?

- Language: What language(s) were used to search or to analyze?

- Search word: Provide exact search words used in the data collection.

- Coding procedure: How did this study develop the coding category?

- Reliability: What reliability was used? What are criteria to judge the appropriateness regarding reliability of the coding category and coding result?

6.

Regarding the data, did you collect the data from the national park community page? Or did you collect the data from individual’s pages? Since post data of individual accounts in Facebook are not visible to other without friend relationships, how can you have the access to the information if this study collected data from individual pages?

7.

According to many Facebook engagement studies, reactions like clicking likes, comments, and shares are significantly related to the number of followers (Facebook friends). Users having more Facebook friends can receive more likes or comments than users having fewer friends regardless of the topic of Facebook posts. If this study collected the post data from individual pages, this influence should be considered in the analysis.

8.

I would suggest the authors to include several examples of each improper behavior contents to better understand the types. I think the contribution of this research comes from a deeper investigation of improper behaviors of national park visitors rather than comparing the numbers of likes, comments, and shares. Therefore, I would like the authors to add the examples and interpret frequent behavior types and to infer the (expected) causes or results of the improper behaviors based on previous research articles.

9.

What are the practical implications for national park managers? Please develop more useful suggestions associated with this research topic and research methods.

I hope these comments are helpful to improve your manuscript.

Author Response

The authors thank the reviewer for the valuable comments on the manuscript.  For specific responses to your comments, please see the attached response file.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

This manuscript has been significantly improved.

Author Response

The authors thank the reviewer for the valuable comments on the manuscript. 

Reviewer 3 Report

Thank you for the efforts to improve this manuscript. Even though more information was improved, many concerns are not fully resolved yet. Please consider the additional comments to revise this manuscript.

1.

Comment 1: In the ‘Keywords,’ it seems that this study did not really conduct ‘text analysis’ with post texts. Since this study reviewed all components of a post such as photo and text, I think ‘content analysis’ would be more appropriate than using ‘text analysis.’

Response 1: ‘Text analysis’ is replaced by ‘content analysis’. Please see Abstract (page 1, lines 12 & 22) and 2.2 Studied Content (page 6, lines 128-134).

>>(2nd review comment-#1)-> Thank you for correcting the term.

2.

Comment 2: I think that the key point of this study is to demonstrate the benefits of ‘social media surveillance’ using this case. In the ‘Introduction,’ authors need to describe the definition, features, and process of this practice.

Response 2: The definition and features of social media surveillance are added. Please see pages 3-4, lines 73-75. The process of this practice is provided in this current study. Please see 2.3 Data Collection (pages 6-7, lines 137-156).

>>(2nd review comment-#2)-> I do not think that the additional sentences do not explain enough about the importance of the approach and procedure. If the sentences are not your opinions, please add citations of the resources.

3.

Comment 3: This study collected and examined the different numbers of likes, comments, and shares of the targeted posts. However, why are the numbers of user reactions important in this study? Although the authors compared the numbers based on the types of improper behaviors, there is no justification why this study needs to compare the numbers to draw a research conclusion. To clarify the problem, the authors would need to build a research problem statement stating the necessity based on previous literature. In addition, a discussion should be improved to interpret the theoretical or managerial implications based on the comparison among the frequencies in the cross analysis.

Response 3: The justification for comparing the numbers of Likes and Shares is added. Please see page 5, lines 103-109. The implications of the findings are provided. Please see pages 16-17, lines 328-340.

>>(2nd review comment-#3)->Regarding the justification, two references (Kwok and Yu, 2013; De Vries et al., 2012) are not related to the online improper behaviors. The papers are to examine the engagement patterns between a brand and social media users. As a brand does not post any improper behavioral content, this argument on your manuscript is not linked to the references.

More importantly, what is the definition of ‘intensity of support or response’ with the buttons? This concern is also related to your response #6. If it means the amount of support indicated with likes, comments, shares, how can this research control different numbers of followers? As authors mentioned, some post data came from public pages and some were collected from individual pages. To measure the amount of support only leveraged by the improper behavior content in post (which is the research design of this research), authors would need to check if the numbers of followers need to be controlled. Again, if one improper behavior post was published by a user having very many followers, the post would receive more likes, comments, or shares due to the size of followers. How do you think about the necessity of controlling the number of followers in different pages? 

4.

Comment 4: This report would need to clarify the purpose of this study. Three research questions in the ‘Introduction’ do not include the engagement indicators (i.e., likes, comments, and shares). Therefore, the authors need to specifically show the research purposes.

Response 4: The description regarding the indicators of FB users’ responses is added. Please see page 5, lines 103-109.

>>(2nd review comment-#4)->Do three response buttons have the same meaning? If those are identical, why do you need to present the tables with three buttons? If those are not identical, please explain how three buttons are distinct.

5.

Comment 5: As a practical project report, more details about the research methods should be described to provide more benefits with this project. Please consider these questions as follows: (1) Data collection tool: What tools did the authors use? (2) Language: What language(s) were used to search or to analyze? (3) Search word: Provide exact search words used in the data collection. (4) Coding procedure: How did this study develop the coding category? (5) Reliability: What reliability was used? What are criteria to judge the appropriateness regarding reliability of the coding category and coding result?

Response 5: The method of data collection is revised. Please see 2.3. Data Collection (page6-7, lines 137-156).

>>(2nd review comment-#5)->Please provide details about the coding procedure.

1)         Who are “three people”? Are they qualified to fully understand and judge the category?

2)         If authors calculated relatability, please provide the reliability in number and its threshold to confirm the appropriateness.

3)         If there were discrepancies among three coding results, how did this study resolve the differences?

6.

Comment 6: Regarding the data, did you collect the data from the national park community page? Or did you collect the data from individual’s pages? Since post data of individual accounts in Facebook are not visible to other without friend relationships, how can you have the access to the information if this study collected data from individual pages?

Response 6: This study collected data only from the public shared posts, including the national park community page and the individual accounts. The research limitation is provided. Please see page 6, line 128, and page 17, lines 341-345.

>>(2nd review comment-#6)->Although authors described the possible problem in the limitation, the research design with the data collection pages can seriously affect the research outcomes. Please see my new comment to Response #3.

7.

Comment 7: According to many Facebook engagement studies, reactions like clicking likes, comments, and shares are significantly related to the number of followers (Facebook friends). Users having more Facebook friends can receive more likes or comments than users having fewer friends regardless of the topic of Facebook posts. If this study collected the post data from individual pages, this influence should be considered in the analysis.

Response 7: This study did not examine the factors that might influence post response such as the number of followers (FB friends). The reviewer’s concern is addressed in future studies. Please see page 17, lines 346-349.

>>(2nd review comment-#7)->Although authors described the possible problem in the limitation, the research design with the data collection pages can seriously affect the research outcomes. Please see my comment to Response #3.

8.

Comment 8:  improper behavior contents to better understand the types. I think the contribution of this research comes from a deeper investigation of improper behaviors of national park visitors rather than comparing the numbers of likes, comments, and shares. Therefore, I would like the authors to add the examples and interpret frequent behavior types and to infer the (expected) causes or results of the improper behaviors based on previous research articles.

Response 8: The examples of types of improper behavior are provided. Please see page 7, lines 165-173 and page 13, lines 240-244. The expected results are added, page 13, lines 244-246.

>>(2nd review comment-#8)->I think that displaying real post examples in Results or Discussion would be very useful to preset the patterns of improper behaviors. Why do not authors add a new table with the Facebook post examples of improper behaviors?

9.

Comment 9: What are the practical implications for national park managers? Please develop more useful suggestions associated with this research topic and research methods.

Response 9: The practical implications are provided. Please see pages 14-15, lines 287-326, and pages 16-17, lines 328-340.

>>(2nd review comment-#9)-> Well-improved.

10.

>>(Additional comment) In Table 1, the total number of frequencies in ‘number of comments’ is not 62. Please check the frequency.

11.

>>(2nd review comment-#10) In line 101-104, the sentences (“These responses signify … of misconduct in real life”) need to be revised. It is not clear to understand.

Please proofread the manuscript.

I hope the additional comments are useful to revise your manuscript.

Author Response

The authors thank the reviewer for the valuable comments on the manuscript. For the responses, please see the attached response file.

Author Response File: Author Response.pdf