36th Cordoba Guitar Festival: Spectator Analysis Using Structural Equation Modelling (SEM)

Round  1

Reviewer 1 Report

This research aims to understand spectator preferences, satisfaction and loyalty for the Cordoba Guitar Festival through SEM. The manuscript would benefit from the following revisions.

Introduction- It is not clear from the introduction why the music festival is being studied or what the benefit will be of conducting this research. The hypotheses are introduced with no context as to how they were developed or if they were informed by previous studies. I suggest either moving the literature section on previous studies to the introduction before the hypotheses or moving the hypotheses to the beginning of the materials and methods section. While hypotheses 1-3 seem to fit well with the logic of the model, hypothesis-4 does not seem well-connected and needs further explanation. What is economic impact? How is it measured?  

Literature review- The style used in the literature review is awkward. Please rewrite the sentences so as not to begin sentences with references. I suggest writing statements about previous research in your own words and then adding the references at the end of the sentence. The literature review section on music festivals needs to be expanded. (also see my previous comment about perhaps putting it in the introduction). The literature on music festivals influences the hypotheses and needs to be addressed in detail. Why are preferences important? How is satisfaction linked to loyalty? What are the economic impacts of music festivals? The literature review would also benefit from some sort of conclusion that synthesizes what previous studies covered and transitions into the hypotheses of the current study.

Materials and methods- You need to explicitly link the selection of variables to the literature. Right now it is not clear how these variables were selected or why.

Results- Results are discussed and given with no mention of the P values. Please address the significance of the P values, particularly those above 0.05.

Discussion/conclusion- The discussion and conclusions need more narrative and explanation of the hypotheses and what it means that they are confirmed. What are the implications of the model being a good fit? The fourth hypothesis is not supported through the model. Why? (Again, this goes back to my comment that this hypothesis was not well-formulated as part of this model and has not been well-defined or justified in the manuscript). Finally, what does this research contribute back to the literature on music festivals? What are the managerial implications of this research?  

Author Response

Thank you for your comments and suggestions. In what follows, we provide a point-by-point response to all the issues raised.

Introduction

a.       New paragraphs and references have been included to support why the music festival is being studied and the benefits of the research.

b.       The hypotheses are explained one by one and have been moved to the end of the literature review.

c.        The reason for the weak correlation of the fourth hypothesis with the other hypotheses may be due to the formulation of the survey questions. In this regard, the questions focused on the spectators’ individual direct expenditure for the concert and accommodation, and in a very global way, on indirect expenses such as transport, drinks, food, souvenirs, etc.

d.       In strict monetary terms, the planned average cost of attending a concert is 33.6 euros, and there are no significant differences when related to spectators’ musical preferences. However, the results do not allow us to determine that the more heterogeneous the tastes or musical preferences of the public, the greater the economic impact of the festival.

Literature review

a.       We have cited the central ideas of the previous research, mentioning the author's surname, and then inserted the bibliographic reference number. We have also added more references related to music festivals in the literature review. For the first three hypotheses we have included references to the literature on musical festivals.

b.       The study of musical preferences provides an indicator of the importance spectators give to the different programmed concerts. In this sense, there is a first dimension, which can be called ‘electro-acoustic’, which reveals preferences for concerts featuring electric guitars and/or electro-acoustic guitars. A second dimension shows greater preferences for Spanish guitar concerts. Thus, the presence of both dimensions makes it possible to show that the spectators attending the festival have different tastes or musical preferences in relation to the guitar as a musical instrument. Moreover, each segment of spectators is generally considered as a group with the same interests, identical motivations, similar perspectives and similar socio-demographic characteristics.

c.        Spectator satisfaction with the festival can be a key element to determine future behaviour (loyalty) towards the festival. The spectators reported a high level of satisfaction with their experience in the festival, with attending to the musical preferences of the public being the most significant aspect. The results reveal that the spectators’ perception of the festival differs significantly, and that audiences whose musical preferences are more heterogeneous have a much more positive image of the festival.

d.       In relation to the festival’s economic impact, please see responses c and d above related to the Introduction.

e.        The hypotheses related to the analysis of the public are inferred from previous studies. In the Conclusions section we state that the economic impact must be studied in greater depth in subsequent investigations. However, we want to point out that the main objective of this research is to demonstrate that the statistical model used is appropriate, that is, it explains three of the hypotheses very well, and to a lesser extent, one of them. This isa  normal situation in any economic scenario subjected to a statistical test.

Materials and methods

The goodness-of-fit test allows us to accept the proposed model, as the likelihood-ratio statistics is χ² = 27.79. Given that the model has 55 estimable and identified parameters and the covariance structure includes 78 distinct sample moments, the p-value is 0.224, and, thus, the proposed model is clearly accepted. In other words, the p-value is p = P[χ² ≥ 27.79] = 0.224, which is greater than any reasonable level of significance (α  = 0.05, 0.10, even 0.20). Moreover, the null hypothesis is clearly accepted, thus validating the proposed model.

Discussion/conclusion

a.       New paragraphs have been included in the Discussion and Conclusions sections with a detailed explanation of the hypotheses and the good fit of the model.

b.    The fact that the hypothesis on the festival’s economic impact does not show a very good fit could be due to the formulation of the survey questions. The questions focus only on direct expenses such as the admission price to concerts and accommodation. Nonetheless, as mentioned above, a statistical model that fits perfectly in a real economic scenario, such as the analysis of spectators at a music festival, should be considered a rare case at the very least. An ideal scenario would be one where the optimal value of the likelihood ratio statistic is equal to zero, which would mean that the covariance structure between the variables fits perfectly to the data.

Reviewer 2 Report

Dear authors,

 Please take into consideration the following aspects:

·        Lines 12 – 16: The abstract is too short. Please mention briefly the main results you achieved

·        Lines 31 – 36: I do not find appropriate to present the hypothesis in the overview section. You should explain how you developed these hypotheses, what was the scientific background you used to formulate them and what is the relationship with the literature.

·        Line 46: Instead of writing “According to [4]”, I suggest you say “According to Saayman& Saayman”

·        Line 48: you affirm that many authors have studied music festivals, but you don’t give additional details. Moreover, you are using the following references [5], [6], [7], [8], [9] and [10] only in this line, without making any additional comment on them or giving additional details. For example, why these papers are relevant to your research?

·        Line 49: I wouldn’t start a paragraph using a citation as you did here. I would use the same solution as in the case of my previous suggestion for line 46. You have many other places in the paper where I recommend you make similar changes

·        You should extend the literature review focusing on festivals’experience, attendees’satisfaction, and loyalty. You focused on the SEM literature dealing with this aspects, but it will be good to see other approaches on this topics also.

·        Lines 94 – 99: you applied a questionnaire, but you didn’t say anything about the way you implemented the survey: who applied the questionnaire, it was a paper-based or online, when it was applied etc. Also, how many and what type of questions you applied?

·        Lines 110 – 114: you define what is an exogenous variable, but I think anyone knows that. I will find more valuable to explain why you choose that five exogenous variables and how you measured them.

·        Lines 119 – 120: How did you choose the seven endogenous variables? Did you take them from similar studies or from the literature? Did you develop them?

·        Lines 138-139: this is the initial model which includes all the potential relationships? Looking at Table 1 and 3 some of these relations are not statistically relevant since p values are too high, but during the lines 162-171 you present them as they are

·        Line 155: in Table 1 you use Q65, Q64 … without explaining what they are

·        Lines 195-196: P-value in 0.224, much higher than 0.05, but besides that, you consider the model fits well the observed data. Statistically significant should be only if P< 0.05

·        Lines 232 – 236: there is no clear evidence why the H1, H2, and H3 hypothesis are validated. What is the influence of musical preference on festival experience? How is festival experience measured through satisfaction? What is the influence of satisfaction on loyalty? You didn’t explain any of these aspects!

·        Lines 250 – 255: you explain that the fourth hypothesis is not entirely supported, and you give only a statistical explanation. What is the economic explanation of this result?

·        Overall, the biggest weakness of this paper is the lack of the economic approach. You focused too much on the statistical part (SEM) and, as a result, the main conclusions you achieved have a very low relevance.

Author Response

Thank you for your comments and suggestions. In what follows, we provide a point-by-point response to all the issues raised

·        Lines 12 – 16:  Two paragraphs have been added to the abstract and the main results of the statistical model have been included.

·        Lines 31 – 36: The hypotheses are explained one by one and have been moved to the end of the literature review.

·        Line 46: We have cited the author(s) by name and have inserted the bibliographic reference number.

·        Line 48:  We have cited the central ideas of the previous research and for three hypotheses we have included references to prior studies on musical festivals.

·        Line 49:  In all cases, we have cited the author(s) by name and have inserted the bibliographic reference number.

·        You should extend the literature review ...: We have extended the literature review related to music festivals.

·        Lines 94 – 99: In section 2.1. four paragraphs have been introduced to address this observation.

·        Lines 110 – 114: In the analysis of the audience attending a musical festival, the musical styles offered are considered as independent variables (exogenous) because the musical programming does not usually vary.

·        Lines 119 – 120:  What does vary are “Performance of latest album tracks”, “Variety of tracks performed”, etc., which are selected as dependent (endogenous) variables. We developed these criteria specifically for this research based on statistical concepts.

·        Lines 138-139: Indeed, in these analyses it can happen that some relationships are not statistically relevant. But what is relevant is that the estimated model and the p-value support the goodness-of-fit of the overall test, which determine if the data are congruent with the proposed model. The optimal value of the likelihood ratio statistics is zero, thus indicating that the covariance structure between the variables fit the data perfectly (which is our objective: to propose a model congruent with the observational data).

·        Line 155: The first column of Table 1 shows the question numbers of the survey included in the SPSS database. Thus, for example, Q65 refers to question number 6, item 5, "Rock"; Q64, question 6, item 4 "Jazz". This information has been included in the manuscript. In this regard, the corresponding SPSS page editor is presented below.

·        Lines 195-196: The asymptotic sampling distribution in our test is χ² (with 23 DF), as there are 78 distinct sample moments and 55 parameters in the model. The model is identified and the chi-square statistic is χ ² = 27.79, thus the p-value is p = P[χ² ≥ 27.79] = 0.224, which is higher than any reasonable level of significance (α  = 0.05, 0.10, even 0.20). Moreover, the null hypothesis is clearly accepted, thus validating the proposed model.

·        Lines 232 – 236: First, from a conceptual perspective, the study of musical preferences provides an indicator of the importance that the public attributes to the different programmed concerts. In this sense, there is a first dimension, which can be called ‘electro-acoustic’, which reveals preferences towards concerts featuring electric guitars and/or electro-acoustic guitars. A second dimension shows greater preferences towards Spanish guitar concerts. Thus, the presence of both dimensions makes it possible to show that the spectators attending the festival have different tastes or musical preferences in relation to the guitar as a musical instrument. Moreover, each segment of spectators is generally considered as a group with the same interests, identical motivations, similar perspectives and similar socio-demographic characteristics. Likewise, spectator satisfaction with the festival can be a key element to determine future behaviour (loyalty) towards the festival. The spectators reported a high level of satisfaction with their experience in the festival, with attending to the musical preferences of the public being the most significant aspect. The results reveal that the spectators’ perception of the festival differs significantly, and that audiences whose musical preferences are more heterogeneous have a much more positive image of the festival.

Second, in statistical terms, in observing the results in the regression tables, variances and covariances, for example, for your question "What is the influence of musical preference on festival experience?", it should be noted that the results for "Behavioural loyalty" related to Q1 "Attendance to previous festivals" is 5.07. But, as mentioned above, if some incongruent estimated results are given, it is normal in this type of analysis. If absolutely all the results were perfect, it would not be normal.

·        Lines 250 – 255: The reason for the weak correlation of the fourth hypothesis with the other hypotheses may be due to the formulation of the survey questions. In this regard, the questions focused on the spectators’ individual direct expenditure for the concert and accommodation, and in a very global way, on indirect expenses such as transport, drinks, food, souvenirs, etc.

·        Overall, the biggest weakness of this paper is the lack of the economic approach...: We have deleted the epigrah 2.2. "Types of variables in SEM". Anyway, the main objective of this research is to analyse spectators attending a musical festival using SEM. The analysis of the economic aspects of the festival constitutes, here, a merely marginal aspect.

Reviewer 3 Report

The reviewed article is an interesting study concerning spectators’ responses to 36th Cordoba Guitar Festival – one of the most important cultural events in Spain. It analyses dependencies among the musical experiences, preferences and the satisfaction level of the festival attendees; attitudinal and behavioural loyalty, satisfaction and spectators’ preferences as well as the festival’s economic impact on the city. As it follows from the references, the issue has not been a very common subject of scientific publications.

To verify the four formulated hypotheses, the Authors applied a popularised structural equation modeling (SEM). A considerable part of the article has been devoted to discussing SEM and hypotheses as well as the presentation of the results of statistical analysis. Certainly, that only reveals a thorough knowledge in the field of research methodology and their research reliability.

Significant doubts, though, are in the theory, in formulated hypotheses and study conclusions.

Firstly, the overview of past academic achievement in the area of research on cultural events, such as festivals, have only been outlined, pointing out the researchers ever dealing with the issues. There should be far more detailed information on the past outcomes.

Secondly, in the theoretical section there is no conceptualisation and operationalisation process of categories, that are closely related to the study presented in the article. From the research problem point of view, the explanation of the following notions: preferences, attitudinal and behavioural loyalty, satisfaction and economic impact is much more relevant than the overview of festival-related articles.

Thirdly, it is recommended to explain the formulation of hypotheses which, unconventionally, have been included in the introduction to the article. Most commonly, hypotheses are formulated once the source query is done and before the full description of methodology is stated. There are two paths to formulate hypothesis: deductive (from already existing theory referring to the research problem) and inductive (formulated on observations and the analysis of economic practice). The latter has been used in that article.

Fourthly, in reference to conclusions, they are definitely insufficient. Presented conclusions are not directly drawn from the research (apart from the one: „the results of this study indicate a good fit of the model”). In the summary, it would be recommended to clearly indicate the scientific input as well as the practical one of the conducted study.

The paper does not have a sustainability perspective.

Author Response

Thank you for your comments and suggestions. In what follows, we provide a point-by-point response to all the issues raised.

- Firstly, the overview of past academic achievement in the area of research...:

New paragraphs and references supporting why we have studied music festivals have been included.

- Secondly, in the theoretical section there is no conceptualisation and...:

In section 2.1. four paragraphs have been included to address this observation. The hypotheses related to the analysis of the public are inferred from previous studies. In this sense, the study of musical preferences provides an indicator of the importance that the public gives to the different programmed concerts. More specifically, there is a first dimension, which can be called ‘electro-acoustic’, which reveals preferences for concerts featuring electric guitars and/or electro-acoustic guitars. A second dimension shows greater preferences for Spanish guitar concerts. Thus, the presence of both dimensions makes it possible to show that the spectators attending the festival have different tastes or musical preferences in relation to the guitar as a musical instrument. Moreover, each segment of spectators is generally considered as a group with the same interests, identical motivations, similar perspectives and similar socio-demographic characteristics.. Likewise, spectator satisfaction with the festival can be a key element to determine the future behaviour (loyalty) towards that festival. The spectators reported a high level of satisfaction with their experience in the festival, with attending to the musical preferences of the public being the most significant aspect. The results reveal that the spectators’ perception of the festival differs significantly, and that audiences whose musical preferences are more heterogeneous have a much more positive image of the festival.

- Thirdly, it is recommended to explain the formulation of hypotheses ...:

Following your recommendation, the hypotheses have been explained more clearly and moved to the end of the literature review.

- Fourthly, in reference to conclusions, they are definitely insufficient...:

The conclusions include new paragraphs related to the detailed explanation of the hypotheses and the good fit of the model.

- The paper does not have a sustainability perspective.

We believe that the methodology is applicable to the future sustainability of the festival as it suggests other programming lines that could attract more spectators with different musical preferences and thus increase the festival’s economic impact on the city as a whole. As we mentioned, the average expenditure per spectator is currently quite low. However, as we have stated in the manuscript, this aspect needs to be studied in greater depth in future research from a strictly economic approach rather than a statistical analysis of spectators.

Round  2

Reviewer 1 Report

The authors have done an adequate job of addressing my comments/concerns regarding this paper and I find it to be much improved from the original.

Author Response

We sincerely appreciate your recommendation for our article.

Sincerely,

Vicente Fruet

Reviewer 2 Report

Dear authors,

You improved a lot the previous version fo the paper. But, there still are some aspects that you should work on:

- The abstract is still too short. The purpose of the study is clear, but the main results are not. You have to say a few aspects about these hypotheses and the main conclusions you achieved when you studied them.

- regarding the statistical part, even if you adjusted the information provided, the information is not clear enough. You have 4 research hypothesis, you explained them, you used the literature for that, but you didn't mention what are the null (H0) and the alternative (H1) hypothesis for each of them

- when you discuss about the model, what are the null and the alternative hypothesis? You said that the null hypothesis "is clearly accepted". If you cannot reject the null hypothesis, it doesn't mean that the null hypothesis is clearly accepted! 

- You told me that: "the main objective of this research is to analyse spectators attending a musical festival using SEM. The analysis of the economic aspects of the festival constitutes, here, a merely marginal aspect."

When you establish a research objective, you can use different tools to achieve it, you do not have to mention that SEM will be your option. Moreover, when you analyze a socio-economic dimension of a festival, it is impossible to analyze the festival without taking into consideration these aspects. So, from my point of view, you have to update the results and discussions part by including additional explanations in order to relate the statistical approach to the economic dimension (exactly what you did in the literature review section).

Author Response

1. The abstract is still too short. The purpose of the study is clear, but the main results are not. You have to say a few aspects about these hypotheses and the main conclusions you achieved when you studied them.

Answers of the authors:

Thank you for your comments and suggestions. In what follows, we provide a point-by-point response to all the issues raised.

Regarding the first observation, new paragraphs related to the hypothesis have been added to the Abstract, including the adjustment of the results, the economic impact and the main conclusions drawn.

 2. Regarding the statistical part, even if you adjusted the information provided, the information is not clear enough. You have 4 research hypothesis, you explained them, you used the literature for that, but you didn't mention what are the null (H0) and the alternative (H1) hypothesis for each of them

- when you discuss about the model, what are the null and the alternative hypothesis? You said that the null hypothesis "is clearly accepted". If you cannot reject the null hypothesis, it doesn't mean that the null hypothesis is clearly accepted!

Answers of the authors:

Following the four hypotheses, we have added an explanation about the null and alternative hypotheses of the subject that concerns us. However, we have not been more explicit about these statistical concepts so as not to dwell too much on the text.

3. You told me that: "the main objective of this research is to analyse spectators attending a musical festival using SEM. The analysis of the economic aspects of the festival constitutes, here, a merely marginal aspect."

When you establish a research objective, you can use different tools to achieve it, you do not have to mention that SEM will be your option. Moreover, when you analyze a socio-economic dimension of a festival, it is impossible to analyze the festival without taking into consideration these aspects. So, from my point of view, you have to update the results and discussions part by including additional explanations in order to relate the statistical approach to the economic dimension (exactly what you did in the literature review section).

Answers of the authors:

We have added three paragraphs to the Results section and one to the Conclusions, which are related to the analysis of the spectator and the economic impact of the Festival.