[Inaudible] that’s under paper, the issues that I got with it are there are a whole bunch of different effects involved in it. The high false alarm offices tend to be in the South East, where we have a larger number of fatalities because the mobile home population is a lot higher. Tornadoes got [inaudible] after dark, in the middle time and moving faster and so I am not completely sure that we have really seen it, there is a whole bunch of things that should increase your FAR, but they will also increase fatalities that are together. So I am not sure that actually just saying let’s reduce the false alarm rate, will reduce [inaudible] on fatalities. This is [inaudible] I am an anthropologist, so I am thrilled to hear you talking about culture even if we have something, the different definition of it. But I think one of your last questions was, get into the issue of what is arrow of causality, how is this related to each other and a flood for some social science methods ethnography in these offices will help you understand questions of the false alarm rate, happiness you know which involves embedding the social scientist and using a toolkit of social science methods to really you know talk to people in semi-structured interviews, I think would probably add a lot to the study, even this is a platform for asking some deeper ethnographic questions. Yeah, the problems we got to go to 50 offices and do it. Well I could probably do a sample that would maybe completely representative, but would give you insight to which direction this arrow of causality might be going, what the points of causality are. Have you read the Gary Fine? I haven’t read the whole thing to know about — Yeah, pretty amazing. I would like to see how the negative offices at the bottom you know seen to be consistent to that whole time period, I am wondering of — And the middle ones, the middle ones don’t change much either. Right, I mean you know average is average, but if you rated you were another SFA and get the bottom ones or whatever that might be a different group of offices, they were in this one. That’s right. Now over the time period that you show, those offices didn’t really show a lot of improvement or anything of that even though there have been advances in technology during that timeframe, so I guess I am kind of wondering out loud here if you have a negative score or skill and you get mad with the tension you know to try and drive those numbers up and get better going weather that has a negative feedback and reduces morale in the office and that sort of – Yeah, that is a good question. To tell the truth, my impression is that I was surprised that how little of that there is, I mean the Fine book points out you know the semi- autonomous regions and some offices operate very, and in some cases there is a reluctance to do that sort of thing. Fine talks about the Jack Kelly era and Jack Kelly you know General Jack got in there and he really raised the flag up the flagpole about this and he got some reaction, but it’s not clear to me how effective that top-down is on influencing. But there certainly is, it seemed like, as you alluded to as well, I mean the offices that have events where there are fertilities or missed tornadoes, it has a big at least temporary negative impact and maybe an overreaction kind of thing I don’t know how long that last. But the organization is not that big you know it’s 4000 people and so people kind of know everyone and dealing with problem offices seems to be a delicate issue and it’s done discretely with trying to preserve, I mean the thing that you don’t want to stress is that performance of, I mean this you could say the service is for the country on the whole, the national, it’s pretty damn good, there some offices are doing maybe really-really good and the others are doing good enough, who also gets in that how do you address in a way that that is effective in making the changes you want. I am not saying that that negative pressure is necessarily you know top-down or whatever, all the offices they know what their performance statistics are each year or whatever and the forecaster is thinking about POD and FAR when they are trying to issue a warning and that’s in the back of their head that’s probably a negative feedback. Right, yeah I think that, yeah especially after the previous event that these [inaudible] or something. I have a couple of comments that the first one is that I think you are going to find if you could measure it that this culture, and better culture means better performances also, tied to better science and better understanding, I mean do go hand in hand. And second, having been a manager in NOAA not at Weather Service office, but two different parts of NOAA, it’s hard to change the culture with our performance management system and our performance evaluation system and when I say, hard to change the culture, also hard to change the science, the education, how people reform and it’s a delicate value, if you are too heavy handed, you can hurt that and if you don’t do anything at all, you are not likely to make any change. Right. And so it isn’t necessarily easy to be this. Final comment, I served in two different parts of NOAA and I found different cultures in the management sides in the [inaudible] issues. And one part of NOAA, I didn’t think supported with our managers very well at all and the other part of NOAA did and that could be [inaudible] part of your 54%. Right. You just said you did interviews with the MICs – Top-down. Yeah. Did you do interviews with the MICs [inaudible]? To me, I wouldn’t be surprised that they would have answered the same way, at our office we do really good things, we are supportive in our office, our people are — That’s a valid criticism. I just didn’t want to go there. I am calling, if anybody you know it was politically difficult for me to do that, so I elected not to. But that’s true, it would be very interesting to see if somebody was outside the agency you know if that had been commissioned to that that would have been an interesting set of interview yeah. [Informal Talk] [Question, inaudible] The answer is no. Now I would look for, was that F-scale and the number of events that’s the only, mainly we took in account. [Question, inaudible] Early on I looked at other warning performance metrics for flash floods and severe thunderstorms and you see the signal there too but not as sharp not as crisp and I think that’s probably you know due to a variety of things probably especially with flash floods, the difficulties with verification and I guess the severe thunderstorms seem to be very routine now and it is not as not as demanding or something, but the signal for the severe thunderstorms was there, but it just wasn’t as clear cut as it was for tornado warnings. Thanks. You are welcome.