Response to Editor’s suggestions#1:

I would like to thank you for your patience. Both reviewers liked your paper, and I believe that their comments will help you improve the article. In addition to their feedback, I have the following suggestions:

General author response to Editor’s comment

Thank you very much for your most helpful comments. We are really happy to receive the comments from reviewers and additional suggestions from you. We have gone through all the comments and have responded to each of them. In case any of these responses need further modification, we would be most willing to do the same.

Comment 1

• p. 6 (Measures): You may wish to change the intervals, as in their current form, they overlap (e.g., does a student with parents' income 20000 Rs belong to interval 2 or 3?).

Authors’ response to editor’s comment

Thank you for your valuable observation. In our revised manuscript, we have changed the intervals to the following format.

1=<5000 Rupees per month

2=5001-20000

3=20001-50000

4=50001-100000

5=100001-above

Thus, a student with Rs 20000 family income will be belong to the interval 2.

Revision on Page 4,

Students reported parents’ monthly income on a five-point measure ranging from 1 to 5, where 1 =<5000 Rupees per month (which correspond to the lower class family income); 2=5001–20000 (lower middle class); 3=20001-50000 (middle class); 4=50001–100000 (upper middle class); and 5=>100001 Rupees per month (upper class) (41, 42).

Comment 2

• p. 7 (Statistical Analyses): You are using a random intercept model. Have you considered also estimating a fixed effect (dummy variable) model?

Authors’ response to editor’s comment

Thank you very much for this suggestion. Yes, we also ran a fixed effects model, i.e. a model in which schools were not considered random but fixed effects (where a coefficient is estimated for each school). Results were comparable.

However, we felt that the random effects model better corresponds to our study design.

Comment 3

• On top of p. 9, you refer to Figure 1; however, the figure is not provided.

Authors’ response to editor’s comment

Thank you very much for these specific observations. Figure 1 is dropped and we have removed reference to the figure from the revised manuscript.

Comment 4

• On the same page, the estimated Beta is 4.12, but the value in Table 5 reads 5.12.

Authors’ response to editor’s comment

Thank you very much for identifying this error. The estimated Beta is 5.12 as on Table 5. Correction has been made in the revised manuscript.

Comment 5

• It would also be useful to explain all abbreviations in table notes (and then referring to the first Table where the respective abbreviation was explained). Furthermore, you may wish to use the same number of decimal places in Tables 4 and 5. Is there a reason why you prefer two decimal places in Table 4 and three (or even four) in Table 5? Values such as 0.000 or 4.263e-05 should be replaced by <0.001.

Authors’ response to editor’s comment

Thank you for this comment. We have placed all abbreviations in Table notes. Table 5 has been modified by using 2 decimal places as in Table 4 and errors such as 0.000 or 4.263e-05 have been corrected in the revised manuscript

Comment 6

• In Table 5 you refer to the term "Gov.rank", which should probably be "Govt.class".

Authors’ response to editor’s comment

Thank you very much for pointing out this error. It has been corrected as ‘Govt. class’ in the revised manuscript.

Comment 7

• Also, please, check all your references carefully, as they appear to be incomplete (and, e.g., ref 21 also contains JEL classification).

Authors’ response to editor’s comment

Thank you very much for this comment. We have revised the reference section thoroughly and made all the corrections.

Response to Additional Journal requirements #2:

1. Please provide additional details regarding participant consent. In the ethics statement in the Methods and online submission information, please ensure that you have specified (1) whether consent was informed and (2) what type you obtained (for instance, written or verbal, and if verbal, how it was documented and witnessed). If your study included minors, state whether you obtained consent from parents or guardians. If the need for consent was waived by the ethics committee, please include this information.

Authors’ response to the requirements

We have adhered to all these requirements and obtained written informed consent from participants in the age range of 17-18 years as well as from parents of participants in the age range of 15-16 years.

2. During our internal checks, the in-house editorial staff noted that you conducted research or obtained samples in another country. Please check the relevant national regulations and laws applying to foreign researchers and state whether you obtained the required permits and approvals. Please address this in your ethics statement in both the manuscript and submission information.

Authors’ response to the requirements

Thank you for your ethics statement : "The research project was submitted to the Cantonal Ethics Committee (Basel-Stadt and Basel-Land), which positively acknowledged the study protocol and informed consent forms, but which also required the necessary permission from the respective school management trusties as well as the permission of school principals, which was then ob-tained in subsequence. "

3. Please amend your current ethics statement to confirm that your named institutional review board or ethics committee specifically approved this study.

Authors’ response to the requirements

We obtained the required permits and approvals for conducting research and obtaining our sample in India. We have revised the manuscript by specifically mentioning the Ethics committee approval.

4. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

Authors’ response to the requirements

The minimal anonymized data is uploaded as required

5. Please upload a copy of Figure 1, to which you refer in your text on page 16. If the figure is no longer to be included as part of the submission please remove all reference to it in the text.

6. Please include a separate caption for each figure in your manuscript.

Author response to reviewers’ comment

Figure 1 is dropped and we have removed reference to it from the revised manuscript. There are no figures in the revised manuscript.

Response to Reviewer #1:

I found your paper very motivating and has global approach towards issues on discrimination. A few more changes as suggested in the report will improve the quality of the paper. Wish you all the best.

This paper studies the impact of caste system on academic performance of Indian students in grades 10 and 11 who study in similar school environment. I found the topic very interesting and relevant. Academic performance in these grades are very important for Indian students and likely to affect the subjects they chose and the schools/colleges they choose for further study. The results are also very thought provoking and can have significant impact on policies directed towards backward communities in India.

General author response to reviewers’ comment

Thank you very much for your most appreciative comments as well as for providing a report to improve the quality of this paper. Thank you especially for highlighting the strengths and scope of the study. We have reworked thoroughly on the manuscript to incorporate the suggestions have pointed out. We have taken great care to adhere and respond to each of them and revisied our manuscript accordingly. If you deem our responses as not fully fitting or in case, we missed your point, we would of course be willing to address them.

Comment 1

Describe the difference between caste system and government class in a bit more detailed manner. The authors can either incorporate this description in the introduction or put it in a footnote. The authors can also incorporate history on caste system to explain why this system is so deep rooted in India even till this day. Of course, apart from history there are other factors such as education that has kept this system alive in India till this day. This will help the readers who are not well aware of this difference.

Author response to reviewers’ comment

Thank you very much for this comment. We agree that our manuscript benefits from including more information on the caste system and its relation to the governmental classes. We have made the following additions in our manuscript in response to this suggestion:

Revision in the manuscript on page 3:

Although the Indian constitution of 1950 eradicated the caste system, inequalities based on caste by birth has continued to hinder the national development (1. 2). The Indian caste system can be defined as a system of social stratification, which divides the society into groups based on its members occupations and is closely associated with Hinduism (3,4). With the aim of uplifting the disadvantaged groups, the government of India has grouped the traditional caste system into three classes, i.e., General Class (GC), Other Backward Classes (OBC) and Scheduled Castes/Scheduled Tribes (SC/ST). The major differences between caste and class are that the membership in the caste is given by birth and that caste is a closed group characterized by endogamy whereas class is an open group. Also, in the class system vertical mobility is possible, such as a person can move higher and go down, whereas in caste there is no such mobility. Finally, a given class can be distinguished from another class on the basis of economic criteria such as income, occupation whereas caste is based on religious and mythical traditions and may have hereditary and traditional occupation (5). But although these governmental classes are thought to replace castes and thus to eradicate caste-related discrimination, they still represent basically the caste system, since class assignment is based on sub-caste affiliation, as opposed to individual socioeconomic status (6, 7). Thus, in the system of the governmental class, the so-called untouchables (Dalits) are assigned to SC/ST, other socio-economically unprivileged castes, such as shudras are grouped into OBC while members of the highest caste being assigned to General or Forward class (6-8).

Comment 3

There are several examples of caste-based discrimination in India. It includes, honour killing in case of inter-caste marriages, bullying students from lower castes in higher educational institutions, untouchability, refusal to eat food cooked by people belonging to lower castes etc. Using a wide array of examples from different spheres of life in India will help readers understand the extent of discrimination and the impact on the life of people belonging to lower castes due to this discrimination.

Author response to reviewers’ comment

Thank you very much for these specific suggestions to improve the quality of our introduction. As you have rightly mentioned, caste-based discrimination is a reality in India and is manifest in various forms in different domain of life including education. We have made a mention of this in the introduction in the following line:

There are studies which show that the caste system contributes to economic inequality (11, 12). Students belonging to the backward castes are exposed to various forms of daily humiliation, exploitation and exclusion in the schools (2, 13,14).

The experience of discrimination is likely to impact students in their educational outcome. However, while planning our paper, we decided to make a mention without elaboration about the problem of caste-based discrimination in school along with other important factors. This was based on the logic that the focus of the study would be to dispassionately approach the class categories. Since the government proposed to address the problem of caste by subsuming castes under the three classes, the aim of our study was to examine whether the class division continues to reflect the problem of caste system, by looking at academic performance. We would prefer not to dilute this focus of the study by indulging in the larger issues related to caste system such as caste discrimination and injustice especially in the introduction. The results show that lower class has significantly lower academic performance, and since economical and certain psychological factors have been controlled for, the results point to the impact of social status, which is majorly linked to caste-based social identity and experiences. In this context, we discuss potential reasons related to social status that may have influenced the academic performance. Here, in the revised manuscript caste discrimination is discussed as one of the potential factors influencing academic performance. Provision of tuition is another such factor that may influence academic outcome, as you have later suggested. Our results provide the rationale for future studies to examine at length the factors related to caste-based social status that may inflence academic performance.

Revision on page 13

Factors specific to belonging to a household of backward caste must have contributed to the poorer academic performance of the students from the backward castes/classes. Caste-based social identity may be strengthened by caste-based social discrimination. For example, Nambeeshan (12) documents the discriminative experiences of the Dalit students in relation to a) access to school including facilities and resources b) participation in different spheres of school life and c) Social relations with teachers and peers in schools in the State of Rajastan. Discrimination can lead to decreased school engagement and rate of attendance as well as increased absenteeism and school drop out (12, 13). Maurya (13) examined 13 Dalit narratives and found that Dalit students experienced exclusion and humiliation from classmates, teachers and administration. An important theme that emerged in the narratives was ‘learned self-devaluation’, a tendency to devalue onself as part of a group and be resigned to the inequalities/injustices imposed by others. Thus, it was observed that Dalit students would not ask for clarifications in a classroom or express their legitimate needs because of lack of assertiveness and self-confidence (13).

Comment 4 & Comment 7

Comment 4

Explain why choose Kerala and Madhya Pradesh. This is important and interesting since Kerala has more liberal outlook to caste, religion etc. compared to Madhya Pradesh which is very traditional.

Comment 7

Since Kerala and Madhya Pradesh are culturally quite different, it will be good if the authors can control for state characteristics in their regressions. They can either use state fixed effects if possible or can use data on caste-based crimes. Such data can be found in the website of National Crime Records Bureau (India). Also, the authors can run two separate regressions for the two states and compare the results.

Author response to reviewers’ comment

Thank you very much for these important observations. As you have rightly pointed out, the two States are interesting for their differences more than similarities. In our revised manuscript we have incorporated information about the States of Kerala and Madhya Pradesh and highlighted the relevant differences.

Revision on page 7

The study was conducted on school students from the states of Kerala in South India, and Madhya Pradesh, in North India. Kerala ranks high among Indian states on social developmental and quality of life indicators, while Madhya Pradesh is close to India’s average rates (Census of India, 2011). Kerala has the highest literacy rate (94%) among Indian states, higher than Madhya Pradesh (70%) as well as the national literacy rate (74%). Besides, Kerala boasts of equal educational opportunity for male and female children as compared to the other Indian states including Madhya Pradesh where females lag behind (Census of India, 2011).

And again, on page 7

The participating schools from the State of Madhya Pradesh followed CBSE syllabus (Central Board of Secondary Education) and Kerala schools followed State Board of Education (SBE) syllabus.

And on page 8

According to NSS 2011-12 (15), Kerala and Madhya Pradesh differ regarding the proportion of the three govt. classes. Kerala has only 9.5% persons of SC-ST, 65% OBC, and 25% GC, compared to 39% SC-ST, 42% OBC, and 19% GC in Madhya Pradesh. In Madhya Pradesh, there is a lower rate of persons who complete secondary education as well as lower gap across the three categories (SC-ST = 7%, OBC = 10%, GC = 15%), as compared with Kerala (SC-ST = 15%, OBC = 18%, GC = 23%) (15). Thus, our sample had a lower representation of the lowest class (SC-ST) because their proportion is less than 10% in Kerala and only 7% of the SC-ST studied till Secondary School in Madhya Pradesh (15).

And on page 14

According to NSS 2011-12 (12), in Kerala, there is a higher rate of persons who complete secondary education as well as lower gap across the three categories (SC-ST = 14-16%, OBC = 18%, GC = 23%), as compared with Madhaya Pradesh (SC-ST = 5-8%, OBC = 10%, GC = 15%). Thus, although Kerala is relatively higher on social indicators (15, 51), the results for KL2 (Table 5) was similar to the overall sample.

However, in this study we have not sought to represent the States of Kerala and Madhya Pradesh with our sample. We have used a purposive rather than random sample. We have only examined data of 3 schools each in these States and only recruited students from English medium schools of state board in Kerala and central board in Madhya Pradesh. The design of the study is not one of comparative investigation of the two states. Hence comparative analyses are beyond the scope of the study objectives. Besides, without representative samples from the two states, comparative regression models using either state fixed effects or crime data may not yield any meaningful results.

Comment 5

In the methodology section, explain in brief the nature of assessment that students were asked to do.

Author response to reviewers’ comment

Thank you for this comment. The assessment included the measures of self-esteem and life satisfaction, self-report of family income in likert scale, and self-report of gender and age and the governmental class and academic marks/grades were gathered from school records as explained in the methodology section.

Comment 6

In the section statistical analysis, please check line 5 (in that section). I believe a little rewording and rearranging the words can make the sentence construction better.

Author response to reviewers’ comment

Thank you very much for this important observation. We have revised the manuscript and corrected the sentence construction.

Revision on page 10

A multilevel model was then set up to assess the relationship between caste affiliation school performance for all six schools combined, once again controlling for the above mentioned five covariates.

Comment 7

Were the students asked if they take private tuition and if yes, then for how many subjects? Indian students have a tendency to go for private tuition and that can impact their academic performances.

Author response to reviewers’ comment

Thank you very much for this comment. We have not sought information on availability of tuition/coaching at home or outside. Having tuition/coaching or having a family member who can help in clarifying doubt are factors associated with the student’s socio-economic status. As the reviewer has rightly observed, these factors need to be explored in future research.

Revision on page 15

Also, availability and utilization of tuition or special coaching at school, home or elsewhere need to be also explored as a factor that may influence academic performance of students in the Indian context.

Comment 7

It would have been interesting if the authors could include data on students in the arts and humanities stream. In India, there is a bias against students in humanities.

Author response to reviewers’ comment

Thank you very much for this comment as well. Yes, it is true that there are more students who go for science and commerce streams at higher secondary level and more schools at private sector which provide these streams.

Response to Reviewer #2:

The manuscript deals with the relationship between the government classes and academic performance (AP) of the students of class X and XI in two states of India. The topic is of immense importance and quite under-researched. While the paper has potential, it has not fully utilized the data that might have been available to the authors. The literature survey needs to be broadened and application of literature to the findings and discussion needs tightening. Following are my specific comments

Response to the Comment

Thank you very much for your highly helpful comments. We have taken utmost care to incorporate all your suggestions and to respond to all of observations. We have revisied our manuscript accordingly. We have modified the introduction, method, results and discussion on the basis of your comments. If you deem our responses as not fully fitting or in case, we missed your point, we would of course be willing to address them further.

Comment 1: Introduction

What are the enrolment proportions and drop-outs in secondary and senior secondary school for the government classes and their performance at the national level? See MHRD report (ESAG).

Authors’ response to reviewers’ comment

Thank you very much for the observation and for providing a relevant reference. This information is definitely useful while discussing the difference in academic performance across the governmental classes. However, we do not have data about the proportions of enrollment, drop-outs and academic performance of all the three governmental classes in the Ministry of Human Resource Developmentestimates (ESAG 2018). The report mentions theoverall data and then separate data for only SC-ST categories. There are no separate data on the Other Backward Castes (OBCs) and General Class. Based on the ESAG 2018, at Senior Secondary school, out of the 24.7 students who get enrolled, 5.9 are from the SC-ST category, which is slightly less than their nationalproportion of the SC-ST category in India (27.5%). Similarly, out of 19.4 million students who enrolled for secondary school board examinations in 2016, 4.9 million were from SC-ST. As per the 68th Round of the National Sample Survey (NSS 2011-12), the rate of educational attendance in India in the age range of 5-14years were 88-89% for SC-ST, 90% for OBC, and 93% for GC. For the age range of 15-19 years, the rate of attendance dropped and the gap between the three categories widened to 54-57% for SC-ST, 64% for OBC, and 71% for GC. Drop out rates are higher for students from the SC-ST categories (19.4% for SC and 24.7% for ST) as compared with drop out rates of all students (17.1%).More importantly, as per the MHRD report the proportion ofstudents who passed the secondary school board examinations in 2016 was lower for the SC-ST categories of students (73% for SC and 65% for ST) when compared to all students (78.7%). Though this document does not mention GC and OBC, nonetheless, this information could be indicative of a trend that students from lower classes perform poorly at secondary schools across students of Central and State boards. Similar trend is observed also in the NSS 2011-12. In separate estimates for the States, in July 2012 arelatively lower rate of persons from the lower classes completed secondary education in Madhya Pradesh (SC-ST = 7%, OBC = 10%, GC = 15%), as well as Kerala (SC-ST = 15%, OBC = 18%, GC = 23%).

Revision on page 4

According to the National Sample Survey (NSS 2011-12, 68th Round), the proportion of persons SC-ST in India is 27.5%, OBC 44% and GC 28.5% (15). Based on the yearly report of Ministry of Human Resource Development (ESAG 2018), at Senior Secondary school, out of the 24.7 students who gott enrolled, 5.9 were from the SC-ST category, which is slightly less than their national the proportion (27.5%) in India (16). Similarly, out of 19.4 million students who enrolled for secondary school board examinations in 2016, 4.9 million were from SC-ST. As per NSS 2011-12, the rate of educational attendance in India in the age range of 5-14years were 88-89% for SC-ST, 90% for OBC, and 93% for GC. For the age range of 15-19 years, the rate of attendance dropped and the gap between the three categories widened to 54-57% for SC-ST, 64% for OBC, and 71% for GC (15). Drop out rates are higher for students from the SC-ST categories (19.4% for SC and 24.7% for ST) as compared with drop out rates for all students (17.1%) in India (16).

Revision on page 5

According to the ESAG 2018 report (16), the proportion of students who passed the secondary school board examinations in 2016 was lower for the SC-ST categories of students (73% for SC and 65% for ST) when compared to all students (78.7%). Though this document does not mention GC and OBC, nonetheless, this information could be indicative of a trend that students from lower classes perform poorly at secondary schools across students of Central and State boards. Similar trend is observed also in the NSS 2011-12 (15). In separate estimates for the States, in July 2012 a relatively lower rate of persons from the lower classes completed secondary education in Madhya Pradesh (SC-ST = 7%, OBC = 10%, GC = 15%), as well as Kerala (SC-ST = 15%, OBC = 18%, GC = 23%) (15).

References added:

National Sample Survey, 68th Round (July 2011 – June 2012). Employment and Unemployment Situation among Social Groups in India. Ministry of Statistics and ProgrammeImplementation, Government of India. 2015. http://mospi.nic.in/sites/default/files/national_data_bank/pdf/NSS_68Round-563.pdf

Educational Statistics at A Glance. Ministry of Human Resource Development, Government of India. ESAG 2018. https://mhrd.gov.in/sites/upload_files/mhrd/files/statistics-new/ESAG-2018.pdf

Comment 2: Literature survey

Bourdieu’s theory of ‘cultural capital’ is useful here to explain differences in academic achievement based on socio-economic status.

Authors’ response to reviewers’ comment

Thank you very much for your valuable suggestion. We have revised the discussion in the manuscript to incorporate mention of how social and cultural capital is transferable to economic capital and hence how the caste system can disadvantage the students from lower castes economically.

Revision on page 5

Bourdieu (22) proposed an elaborate theory of social, economical, and cultural capitals. According to him possessing economic resources (economic capital) contributes to increased social connection and influence (social capital). In turn, social capital may facilitate increased educational opportunities/achievement, which is part of the cultural capital (22, 23). In the Indian context, caste affiliation along with the financial condition of the family constitutes a major part of students’ socioeconomic and cultural capital (24, 25). Caste affiliation may determine to some extent the socioeconomic status of the family (11). Also, economic status irrespective of caste/class affiliation may determine academic achievement (17, 19). Gupta (26) found evidence for a differential influence of social status and economic status on academic performance in Indian college students. While students from lower castes were more likely to perform poorly in academics, students from lower economic status were not more likely to perform poorly.

References added:

Bourdieu, P. The forms of capital. In J. Richardson (Ed.) Handbook of Theory and Research for the Sociology of Education (New York, Greenwood). 1986: 241-258.

Huang, X. Understanding Bourdieu - Cultural Capital and Habitus. Rev. Eur. Studies. 2019: 11(3), 45-49. doi: 10.5539/res.v11n3p45

Comment 3:

‘Alternately, the associated socio-economic status may determine academic achievement, irrespective of caste/class affiliation.’ Please give an example/reference.

Authors’ response to reviewers’ comment

This statement means that having higher/lower financial and social status may lead to access to better/poorer schooling and additional coaching opportunities, and lead to better/poorer academic achievement. As you have already suggested, the concept of different forms of capitals (Bourdieu, 1986) and their relations can explain this. The manuscript has been modified by providing references for this statement.

Comment 4:

The authors might benefit from the following article that deals with the relationship between caste and academic performance in higher education in India:

Namrata Gupta (2019) Intersectionality of gender and caste in academic performance: quantitative study of an elite Indian engineering institute, Gender, Technology and Development, 23:2, 165-186.https://www.tandfonline.com/doi/abs/10.1080/09718524.2019.1636568

Authors’ response to reviewers’ comment

Thank you very much for suggesting this reference to improve the literature survey. The manuscript has been modified to incorporate reference to this article. The referred study found that while students from lower castes were more likely to perform poorly in academics, students from lower economic status were not more likely to perform poorly.

Revision on page 5

Gupta (26) found evidence for a differential influence of social status and economic status on academic performance in Indian college students. While students from lower castes were more likely to perform poorly in academics, students from lower economic status were not more likely to perform poorly.

Comment 5:

Some of the references are not specific and does not lead the reader to the source. For instance: ‘India Go. Ministry of Social Justice and Empowerment, Government of India 2018, Sep’. Which document are you referring to?

Authors’ response to reviewers’ comment

Thank you very much for bringing to notice this error in referencing. In the revised manuscript, we have also gone through all references and ensured specificity as you have suggested.

The mentioned reference is corrected as:

Social Justice AR 2018-19. Ministry of Social Justice and Empowerment, Government of India 2018 http://socialjustice.nic.in/writereaddata/UploadFile/Social_Justice_AR_2018-19_English.pdf

Comment 6:

On Pg. 10: ‘Though Indian studies have examined self-esteem and life satisfaction in the context of educational outcomes, …’ What Indian studies are you referring to?

Authors’ response to reviewers’ comment

Thank you very much for the comment. In the revised manuscript we haveprovided the references for this part of the statement. The references are (15) and (29).

Comment 7: Methods

We need more information about schools. Are they government, private or mixed schools? What is their language of instruction?

Were there any difficulties for the participants in filling up the questionnaire and how were they handled? For instance, was the assessment administered in English? If yes, could the students understand the language particularly in the self-esteem and life satisfaction scales?

What is a z-score?

Authors’ response to reviewers’ comment

Thank you very much for this comment. We have revised the methods section and the suggested details regarding participants and procedure have been incorporated. Since we conducted the study only in urban/semi-urban English medium schools, the questionnaires were administered in English. Students were assisted by their respective Eglish teachers in responding to the questionnaires.

The raw scores of academic performances from different streams of schools were converted to z-scores for making them comparable.

Revised on page 7

The participating schools from the State of Madhya Pradesh followed CBSE syllabus (Central Board of Secondary Education) and Kerala schools followed State Board of Education (SBE) syllabus. The XI grade students were recruited from two different streams, i.e. science and commerce. These schools were from the private sector and English was the medium of instruction in all these schools.

Revised on page 9

The assessments were administered during school hours and in classrooms. Students were given 30 minutes to complete the assessment. The medium of assessment was English since all the participants were from English medium schools.

Comment 8a: Results

P. 14: ‘Predicted school performance values were 45.5 (±3.6), 49.1 (±3.3), and 51.4 (±3.3) for low, medium and high-class levels respectively, and were thus increasing with increasing levels of governmental class.’ Please specify if these are mean values taking all the schools together. Also, in the column on AP, there does not seem to be a linear relationship between class categories and AP, since, SC/ST average is 44.5, OBC 43.0 and GC 53.8. Please explain.

Author’s response to reviewers’ comment

We distinguish between AP mean values based on descriptive statistics (mean values across all students and schools by caste-class) and predicted mean values based on a multilevel model with caste taken as a factor with three levels (low, medium, and high). In both cases, all schools were included to compute means.

In addition (and independent of the computation of means as explained above), we set up an apriori hypothesis in which we wanted to test whether AP linearly increased with higher caste class (from SC-ST to OBC to GC). To this end we used caste-class as a linear predictor in a separate multilevel model.

Comment 8b

In Table 2, AP, SES and SWLS numbers for various categories are out of what? AP seems to be percentage, but the rest?

Authors’ response to reviewers’ comment

Thank you for making these observations. These numbers are Mean (SD) as given on the Table description on top [Table2: Descriptives: Academic performance, self-esteem and life satisfaction Mean (SD)]

Comment 9: Discussion

P. 16: Line: ‘Also, an increase of family income may not necessarily correspond to better educational outcomes for the students of low social class. In other words, the educational disadvantage that we observe in our results in the lower governmental class, may not be equated with economic disadvantage.’ Please state the reason for this.

Authors’ response to reviewers’ comment

Thank you very much for making these careful observations. That social backwardness may affect educational outcome independent of financial status is an argument that is presented in the introduction with supportive references(page5). Further, the results also point in this direction because caste affiliation contributed to academic performance when family income was controlled for. Additionally, to support this argument further, here we have added a descriptive observation that in our sample a relatively higher proportion of OBC class reported higher family income when compared with GC (This is only an additional observation and not a statistical evidence). In the revised manuscript these two sentences (‘Also, an increase of family income.... with economic disadvantage.’ have been since the argument is discussed elsewhere.

Comment 10

P. 16: ‘In a study (40) conducted in the state of Haryana in which students from a socioeconomically advanced OBC subgroup were included, the academic performance of these OBC students was found to be on par with the General Class.’

This statement appears flawed as the Yadav and Chahal, the authors of the paper cited here state in their own paper: ‘The reserved categories include OBC’s which is socially backward but economically prosperous category.’ (p.36: Yadav and Chahal 2016). Hence, Yadav and Chahal seem to have studied OBCs that are socially backward and cannot be referred to as ‘socioeconomically advanced’.

Authors’ response to reviewers’ comment

Thank you very much for carefullyidentifying this error. The statement has been corrected as “economically backward”. The phrase “socially backward families which are economically advanced” in the next sentence would further clarify it.

Comment 11

P. 14: Line: ‘In one school (KL2), academic performance was highest in the highest class (GC) and lowest in the lowest class (SC-ST) (p<.001) while in another school (KL3), academic performance also varied among classes, albeit being lowest for the medium class (OBC) (p=0.014). For the remaining four schools, no significant differences in academic performance among the three class levels were found…’

As your Table 1 shows, schools KL1 and MP2 have either a single or no SC/ST students. Hence the claim that: ‘For the remaining four schools, no significant differences in academic performance among the three class levels were found…’ does not seem pertinent. Instead it will be better to show the relationship between caste and AP for GC and OBC for all the schools. The status of the OBCs in Kerala and MP should also be discussed and related to the findings.

Authors’ response to reviewers’ comment

Thank you very much for this observation. The main focus of the study is to look at the overall sample (N=858) in the three categories. Examining of separate schools is only additional and not part of the design or objectives of the study. Since we planned to include the whole classroom while sample recruitment, the actual presence of the three categories of classes in these schools are reflected in the sample. Thus, KL1 and MP2 have either a single or no SC-ST students, and MP2has only 6 and MP3 has only 7 OBC students. GC are relatively better represented in all six schools. It may be interesting to notice that in KL2 where all three categories have sizeable representation, academic performance was highest in the highest class (GC) and lowest in the lowest class (SC-ST) (p<0.001), which is similar to our results for the overall data. As you have suggested, we have modified the text by omitting the statement ‘For the remaining schools...’ andour focusinstead will beon the three categories for all schools. Besides, as suggested, we also elaborate on the status of the govt classesin Kerala and MP.

Revised section on page 8

According to NSS 2011-12, 93% population in Kerala as compared to 72% in Madhya Pradesh are literate. The two States also differ regarding the proportion of the three govt. classes. Kerala has only 9.5% persons of SC-ST, 65% OBC, and 25% GC, compared to 39% SC-ST, 42% OBC, and 19% GC in Madhya Pradesh. In Madhya Pradesh, there is a lower rate of persons who complete secondary education as well as lower gap across the three categories (SC-ST = 7%, OBC = 10%, GC = 15%), as compared with Kerala (SC-ST = 15%, OBC = 18%, GC = 23%). Thus, our sample had a lower representation of the lowest class (SC-ST) because their proportion is less than 10% in Kerala and only 7% of the SC-ST studied till Secondary School in Madhya Pradesh.

Revised section on page 11

Only in one school (KL2), there was a sizeable representation (See Table 1) of all the three govt. classes. For the other 5 schools, there was no proportionate/adequate representation of all three categories in each school when taken separately. Thus, when school KL2 was separately examined, it was found that academic performance was highest in the highest class (GC) and lowest in the lowest class (SC-ST) (p<.001). Assuming a linear functionality between governmental classes and academic performance, we found a positive association in this school (KL2: β=4.12, SE=1.23, t=4.17, p<0.001)where the students showed increasing academic performance with increasing class levels. For the other five schools, a significant pattern of association could not be observed (Table 5).

Revised sections on page 14-15

For instance, according to NSS 2011-12 (15), one in four SC-ST households had no literate family member of age 15 years and above in rural area, as compared to one in ten households in urban area. On comparison, among OBC, 18% households in rural areas compared to 7 per cent in urban areas and among GC, 11% in rural areas compared to 3% in urban areas had no literate member of age 15 years and above (15). The persons from the higher class were more likely to live and study in urban area as compared to the lower classes. Also, family members in the higher class were much more likely to have self-employment or salaried jobs as compared to the members of the lower classes who work mostly in the primary sector (15). Thus, the role of (lack of) education, social exposure, career aspirations and achievement motivation in relation to the caste-based social identity must be examined in future studies.

According to NSS 2011-12 (15), in Kerala, there is a higher rate of persons who complete secondary education as well as lower gap across the three categories (SC-ST = 14-16%, OBC = 18%, GC = 23%), as compared with Madhaya Pradesh (SC-ST = 5-8%, OBC = 10%, GC = 15%). Thus, although Kerala is relatively higher on social indicators (15, 51), the results for KL2 (Table 5) was similar to the overall sample. Hence, the difference among the classes across the States in India as well as the influence of the specific context of the States need to be also examined for a better understanding of the association of caste affiliation and academic performance.

Comment 12

Based on the literature, the authors might like to discuss as to why there is a relationship between government classes/caste and the AP.

Authors’ response to reviewers’ comment

Thank you very much for the suggestion and our discussion has been modified to incorporate it. In this study, we have not examined the factors that might have contributed to the relationship between the classes and AP. Hence, after corroborating our results by refering to other studies, we go on to make hypothetical speculations on possible reasons for the relationshipin three paragraphs.Following are someof the reasons that we have discussed in the original manuscript:

1) The reason for our finding must be related to social status over and above the potential influence of economic status (family income)as well as some of the psychological factors (self-esteem and general wellbeing/life satisfaction) that were controlled for. In other words, reasons specific to belonging to a household of backward caste must have contributed to the poorer academic performance of the students from the backward castes/classes. We leave it to future studies to examine what aspects of caste identity and social status influence academic performance. In our revised manuscript, we include that limited social exposure and minimal aspiration/achievement motivation are likely to influence academic performance.

2) The influence of social backwardness on education may be increased by associated economic backwardness. We refer to study which found that socially backward students who were economically advanced performed well in academics on par with GC students (The manuscript has been modified to bring emphasis on this argument).

3) Yet another reason could be English language proficiency since our participants were from schools with English as medium of instruction. Students from the lower classes are much more likely to come from families with parental lack of education and lack of proficiency in English language. This is likely to influence their academic performance.

The manuscript has been modified to strengthen these arguments.

Revised section on page 14-15

The reason for our results must be linked to the influence of social status over and above the potential influence of economic status (family income) as well as some of the psychological factors (self-esteem and general wellbeing/life satisfaction) that were controlled for. Factors specific to belonging to a household of backward caste must have contributed to the poorer academic performance of the students from the backward castes/classes. Students from the backward castes are likely to be low on educational and career aspirations. This could be so because the household of the students may have less understanding of or exposure to educational and career opportunities different from their traditional caste-based social and occupational roles. Hence, the potential role of caste-based social identity and its relation to (lack of) social exposure, career aspirations and achievement motivation must be examined in future studies.

The factors that influence academic performance may be diverse and complex for all the numerous subgroups placed under the three governmental classes. Some of the social subgroups under OBC category have high economic prosperity and respectable social status in some states/districts of India. In a study (27) conducted in the state of Haryana in which students from an economically advanced OBC subgroup were included, the academic performance of these OBC students was found to be on par with the General Class. Thus, the influence of social backwardness on education may be increased by associated economic backwardness. Hence, the differential influence of governmental class for students from socially backward families which are economically advanced versus economically backward needs to be examined in the Indian context.

Another important factor that may have influenced the students’ academic performance reason could be English language proficiency since our participants were from schools with English as medium of instruction. Students from the lower classes are much more likely than their counterparts to come from families with parental lack of education and lack of proficiency in English language. This is likely to influence their academic performance.

Comment 13

‘(11) observed that students from lower socioeconomic status are likely to have a positive influence...’

Do you mean to say ‘…. likely to be positively influenced…’? Please replace (11) by the last name of the referred author.

Authors’ response to reviewers’ comment

Thank you for making this suggestion and pointing out the error. Yes. We meant to say “... likely to be positively influenced”.Also, (11) is replaced with the author’s last name.

Comment 14

‘This would imply that the gap in academic performance between students from higher and lower governmental classes are likely to widen if students from the schools of the governmental sector and from the rural background are investigated.’

Because??

Authors’ response to reviewers’ comment

We collected our data from urban areas where all the three class categories of students are likely to study together in the same schools. This co-education is likely to improve the academic performance of the lower class as observed in the referred article. On the other hand, in rural areas and especially in schools of government sector which provide free education, students from higher class are seldom represented and hence there would not be possiblity for co-education of higher and lower classes. Hence, we specualte that without the potential advantage of co-education, these students are likely to perform poorer academically, if compared with students of lower class who study with students of higher class. The text has been modified to make this argument more lucid.

Revised text on page 16

Thomson (18) observed that students from lower socioeconomic status are likely to be positively influenced when they study along with students of higher socioeconomic status and are likely to perform better academically. Thus, if studying together with students from higher governmental class is advantage for students from lower class, then it would imply that we may find a greater gap in academic performance if we include students from rural schools who are mostly from the lower classes and who are less likely to study with students of higher class.

Attachments

Attachment

Submitted filename: 3-Response to reviewers 2.docx

Dear Editors

On behalf of my fellow authors I herewith resubmit the manuscript “The Governmental Ranking of Class and Academic Performance of Indian Adolescents” PONE-D-19-31769R1 for consideration in PLOS ONE. We have revised our manuscript substantially and according to the reviewer’s sugges-tions. All authors listed in the title page have read the manuscript, attest to the validity and legitimacy, and agree to its submission to PLOS ONE. The authors report no financial interest or potential conflict of interest.

We believe that our manuscript would perfectly fit in your prestigious journal and thus we would be delighted to undergo your review process and are looking forward to your valuable response.

On behalf of the authors, yours sincerely

Roshin Kunnel John and Jens Gaab

Reviewer #1:

In page 10, in the line "A multilevel model was then set up to assess the relation-ship between caste affiliation school performance for all six schools combined, once again controlling for the above mentioned five covariates.", there is a small error. I believe the correct way to write this sentence will be: A multilevel model was then set up to assess the relationship between caste affiliation `AND' school performance for all six schools combined, once again controlling for the above mentioned five covariates.

Response: We have corrected the sentence accordingly.

Reviewer #2:

While there is a detailed discussion of results on academic performance and its relation to socio-economic status, there is no discussion on the two other measures, self-esteem (SES) and satisfaction with life (SWLS). How were these related to academic performance? Also, what were the gender differences in performance with respect to their caste?

Response: We have included the calculations and results for SES, SWLS as well as gender, family income and age in the results section and also addressed these findings in the discussion section with two added sentences.

The Discussion needs to engage with Bourdieu, that is, how far it affirms the theory of cultural/social capital.

Response: While we fully agree that our manuscript is possibly relevant to the area of sociology, we must admit that none of the authors has any backgrpund or experience in this discipline. Thus, any informed discussion of Bourdieu would appear assumptive from our side. We hope that this is not a major problem, but of course we hope that our paper and the results will be picked up by relevant disciplines, such as sociology, educational sciences and policy studies.

Line in Introduction: 'Based on the yearly report of Ministry of Human Resource Development (ESAG 2018), at Senior Secondary school, out of the 24.7 students who got enrolled, 5.9 were from the SC-ST category..." What are 24.7 and 5.9? percentages?

Response: Yes, the % was missing and we have corrected the sentence accordingly.

The line in the Discussion 'Thus, the influence of social backwardness on education may be alleviated by better economic condition and worsened by economic backwardness.' is confusing. It might either be modified or deleted.

Response: We deleted the sentence.

The last line in the Discussion section: 'then it would imply that we may find a greater gap in academic performance if we include students from rural schools who are mostly from the lower classes and who are less likely to study with students of higher class'. This line is also not clear. "include" where?

Response: We deleted the sentence.

Attachments

Attachment

Submitted filename: Response to reviewers.docx