 So if you have any questions specific to this, you are encouraged to ask your question now. Thank you, yes. So just I will, can you tell me a little bit more about what's ethical and what's unethical and what you want to intervene because I understand that we wouldn't want to ask parents to not spend time ready to their children. But like, if that's too much to ask them to do that. But isn't it, we really deal with policy to invent new aspirants, don't rate to their children, and in this way we would be trying to reduce the gap between two groups of families in something that's very easy to meet, but maybe not unfair. So it's more difficult. I think this has been a very difficult issue for public policy makers because where does the state begin and where do parents become responsible, right? And I think the upshot of Adam Swift's point is that, well, the state should not, you know, intrude into the very private domain of what defines them, right? Therefore, you know, if there are really parents who are conscious and who are dedicated and devoted in terms of doing these little things, including when opening the door, actually talking to the child. And if any inequality emerges because of such branding behavior, there should be authority, and we should not be calling parents. Well, I am very much giving advice because it's good that parents do that. But because we know it's good, and we know it's good for kids. And as a state, we care about the education of kids or kids. Wouldn't we want to say to the parents who don't do it, or maybe you should do it but it's good for the kids. Maybe they don't know it's good for the kids. And we can send the violence to us to defend it. And I'm not sure if it's true that it's something that the parents don't need to say. Well, yes. I mean, that's exactly why you are going into the discussion of instrumental value of parents versus philosophers who are kind of interested in the intrinsic value of them and respecting that without any return, right? That we care about our loving mother. We would love to have a father who would really return to these regards or whether this will make me become an ISI. Right? So, but I think you are making a very good point because when we look into the ECD literature, there's already a lot of intervention doing exactly that. Our literature and Chico's presentation today was quite striking that before children would go to school in the very first four months, we can clearly see cognitive gap emerging by parental education background and a lot of those educated parents were actually responsible, engaged, and very intimate parents. There is this very fascinating book written in the 90s about a four million watts gap that children raised in watching parents where they are listening to fewer words versus children from professional white-collar families being exposed to higher world count and that was creating this early learning advantage. So, of course, you know, that creates very big dilemma for the state and many would take their position. I'm just presenting this competing philosophical viewpoint, you know, which is out there. And of course, it has created an ethical dilemma. Right? Because as states, we want the chorus of which. I think their response would be that, yes, we want a very good education system where regardless of children's background, there should be corrective measure, but we should not really get into the private space of family. So you can have a lot of those remedial policies at school for everyone regardless of their background. It's difficult for them because of the family. They ask them, how can we support people? Correct. They are correlated. So they are correlated by the setting of point that you also mentioned that is the composition exercise. They are distinctive in nature, not causal. And so I think the more going forward, the idea would be to revisit them with better data and establish whether the pattern is something we see in different contexts and in different studies. But of course, there is some correlation, but the correlation is nowhere close to perfect or we don't have high co-miniative problem in that sense. I mean, it's great, but at least I can't interact with everybody. Depends on the list of the location of the pattern of the model. So the internet in the region that we talk about is embedded in this term of assets and the inclusion of it. So it's very important that that is medical and the same kind of course. It can be separated in my perspective. Okay, so in our model, we have separate control for mental background, including educational background. Right? So in Chico's presentation today, he will only focus on educational background. So we are saying that, well, the actual is there and we consider that as not acceptable source of inequality. But when we have a measure which is independent of that, correlated. But it is there in the data. Does the data find a stronger correlation with measures which are in addition to panel background, but proxying for intimate familial interaction? And the answer is yes. So if your concern is true that is highly correlated, then this additional measure of intimate familial background wouldn't have much predictive power. So that is all I can add. But you know, it's a fair point that that is there in the data, that correlation is there. But what we find is that the measure of intimate familial interaction prompts the contribution of parallel education. Question one is that how do you measure the parents' efforts exactly that they need? And the second is that is it possible that different, let's say they are siblings and like for the different kids, the parents' effort could be different based on its agenda? Right. Thank you. I rushed too fast. So we have several measures. One is time spent by mother. Aiding the child for education matters at home in school closure. In minutes, the same by father. And then time spent by mother on childcare. So this is independent of educational activities. So these are three measures. They are again in time inputs. And we even collected independent measure of this from the student asking how much does your mother spend time on you while also asking mother. So we can even check for measure returns and all that. So that's where it goes. But of course, you know, we could also collect more data on bedtime stories and then, you know, other type of activities like conversation over a meal and all that. But, you know, if we go by the evidence, adding those perhaps would just increase our predictive power. Yeah. And because siblings is... Oh, so siblings, that's a good point. Right. Now, the ambulance has kind of reached the replacement level stage now so that for most families in one school age group, typically there's one child. So we have an average family of two children. And so ours is focused on private and secondary school children. So there are siblings, but very in small cases, siblings are on the same education level. So therefore, you know, they are really competing in the same category, but it's a fair point that something we don't have to do. Any other question? But I think that I contribute with the literature like with I don't know, broad aspect like showing that it's important to be exposed to like diversity not only from rich and poor students, but also like sex diversity lets to more progressive like gender abuse like in general. Yeah. And fortunately, I don't have here like single sex schools so to check like if sex schools are even worse than like mixed sex schools. So I cannot like make that policy implication like straight forward, but I think that yeah, it's like an implicit message that it's important to be exposed to a more diverse environment and more in general. Okay. Yes. My question has to do precisely with the title of Mala Narrative Facts. In particular, I was wondering if you have any idea of the support of the independent variable. I think that maybe if the independent variable is somehow bounded, the fact that we are finding is only representative for the bound of the support. Something like for instance, if the female composition of the classroom goes up to 75% of the person will make that and then you get a positive effect, but if the female composition goes over 75% then you get no effect or something of the other kind. So I was wondering if you have anything about that? I was checking if I had a viewer here like to check of the support of the variable. So I don't have single sex schools nor like, but I have a wide like support for the share of female peers in the classroom. It's not concentrated on from 40 to 60, but I have like a wide range of percentage of female peers. And as I said, like further research on this will be to analyze non-linearities and check whether the changes are the defects are explained from students moving from I don't know 45% of female peers to 55% or concentrated like moving from the 20 to the 80% of female peers in the classroom. But unfortunately I don't have the figure here, but we can talk about that later. Okay, so as a chair I'll exercise my monopoly power and ask you the last question. So just reflecting on your results and it seemed that what you claim as the care effect could be simply reflecting the fact that typically when school authorities reassigned the students they would sort students based on disciplinary behavior. And the literature shows that boys are more disruptive than girls. So it could essentially be that well larger female shares means that you have more well-behaved students in the classroom and I don't know in your context whether you have got a smaller class sizes for disruptive students there by default the female share would be low. There is actually an ice-paper and DJ on this. So if that happens then your results would be simply reflecting the fact that what you claim as female share larger female share is actually exposure to less disruptive students. Yeah, that's a good point. And fortunately I don't have in the data like disruptive behavior like reported in the data. But as I said the procedure of assignment of the students into classes is made by the head of the school and it seeks to preserve some homogeneity in a broad set of characteristics. So like behavior is one of those. If they have like previous information of misbehavior of all the students they are going to arrange classes to preserve like homogeneity on group composition within the same school because the head of the school have that information. But then I checked that following like two different approaches to check if the actual like sex composition follows as good as random distribution. And I found like supportive evidence showing that in fact the assignment of the students to classes is as good as random. So the reason why it raises is because in your actual sample there is a voice deficit, you have more girls and boys. Yeah, on average. On average. Yeah. And so because in many other countries there is missing voice phenomena so there are less interested in education and whether when you have that in the population how do you interpret your result given that well disruptive behavior or lack of discipline and motivation is actually making boys participate in education less so that you have more girls in general. Yeah, but in that case I included the previous repetition variable individually and also like a peer, the share of peers with the previous great repetition as control variables. So to like control for ability or misbehavior I have that variable and I included that at individual level and also at peer effect level. So I mean it's not perfect but it's like a proxy of ability and misbehavior. Thank you so much, Martina. It's a very nice presentation. Thank you so much. We're running a little bit behind schedule still. I hope you can stay there for five minutes and share your constructive comments, suggestions. Yeah. Yeah, no, no. So that is not possible that we can link the student with the teachers. I mean that would be really nice because then we can check the identity of both the teachers and the student but that's not possible. Yeah. So just a quick question. So I understand that the village dominance that's given based on historical development but there is nothing that stops marginalized family from moving away across locations. The migration story. So have you looked at that turn? Yeah. The creation of settlement in a village because going by your logic the equilibrium outcome is for every student cause to actually move into villages where they will have more people similar in terms of social values. Yes. What is the background? Is there a higher level of migration? Okay. So I think one thing that we have in our, this thing is that more than 95, so around 94 or 95% of the population has been staying in their village of residents for 50 years or more. So this kind of signals that there is a lower caste wise. Yeah. So this is something that, yeah, we can do. Yeah. Yeah, that we haven't checked. So the point is that your main punchline is a teacher effect and I understand that well if the teacher is more like me, I'm benefiting but then within Indian context that well if you think of teacher mobility, teachers, high quality teachers would like to go to village dominated by the student caste. Right? So therefore there is negative sorting based on teacher quality while there is positive sorting based on identity that I get. So this is where I think his question is very important that well you know if you could do maybe using another dataset to kind of match the teacher's identity and then have some measure of quality. Yeah. To kind of check which is beneath his background to horses. Yeah. Yeah. Yeah. I think that's interesting. Yeah. This is what we tried, even wanted to try but this data does not allow us. Yeah. We can look for some alternative data set where we can match this, the identity and even the location. Yeah. Yeah. That's a good. Thank you. You got one last question. So we know that in India, the North-South divide that in southern part of India your supply becomes weaker because in the region there is less of a discrimination versus northern part of India. Yes. So if you were to slice your results, do you see this being stable or does it flip? No. So this is something that we haven't done. So Sophia, but I think, yeah, this is something that we can look for the heterogeneity across the North and South. Yeah. Thank you. Thank you so much. Thank you. Thank you.