15:13:42 Okay, I have to record it to the computer I messed up my zoom, and I am having a serious problems because I signed up. I needed for something and I use the different.
15:13:56 I don't know, I changed some setting, and now I'm kicked out of my zoom account, rather zoom account.
15:14:03 I have no clue at the moment. How to come back, but if you're if you're logging in, you have to use these or login.
15:14:13 It's far as something at the beginning, there is kind of this with with with username and something.
15:14:19 There are three icons. Below that, and one of them is like key with Isa or something, which is the connection via an organization. So you click on it and then you write Rutgers dot.
15:14:34 Zoom dot something and and then you will use your regular login to Rutgers ID systems.
15:14:42 Okay, you will have to like, like, go off the will have to log off and close the zoom and open to some again.
15:14:52 Okay, I'll try. Thank you.
15:14:57 Assuming like understood what you just said, but I think I did
15:15:04 know to chat store.
15:15:07 So we will show you what we have and then Vojtech will tell you a few things that we kind of came up with.
15:15:19 Okay, do you see that.
15:15:22 Yep.
15:15:29 Okay, so the few things that are
15:15:35 kind of new. So, you know, for a long time now we've been trying to also come up with the names of the codes.
15:15:46 So one decision was that every prob problems will have the initial two letters or one two or three letters will be the name, the reference to the Y category.
15:16:05 So, we have, as you can see, we long time ago we decided to have just two types, individual health problems so we have age psychological health and I age physical health, and then unfortunately there's the residual category which is hard to do, you know,
15:16:29 we cannot detect from the text whether it is psychological or physical. So, it is general and yeah we take.
15:16:41 Yeah. So, For example, we have this code.
15:16:46 Business closures and closures which I think are basically about the same. And we're wondering whether they should remain on the the economic problems category or y, and Zed category of problems, or should they for example creates.
15:17:06 It's under, like, move it to Kobe because they basically referred to Kobe, and I had this idea that, well, business closures, there are much broader category than just Kobe related to corporate caused business coaches, but if we for example have a code
15:17:22 and Kovats say CV, the business closures. Then we have basically all we can have two different closures in the system.
15:17:37 It's just like the advantage of having this prefix attached to them.
15:17:42 So we can use business closes that caused by pandemic and love them and whatever, and like other business coaches.
15:17:53 So we take one thing Richard's question is very important and we actually, we talked about it but we never finished, we say that the coders can restart the work and you restarted but which codes, they should be using because you know when you're coding
15:18:12 you're using only x codes. Right, the system will aggregate them automatically into why codes that back back end will be of course instructed to do that.
15:18:23 So, should they should you guys be coding using the new access or old exes, doesn't it doesn't matter. I mean you access it I know it basically doesn't matter because eventually old all access will be converted into new access but
15:18:42 in order to avoid confusion. Well, you know, one thing is the, I'm glad, Richard that you raised that because you know one thing.
15:18:50 The new exes are simpler.
15:18:54 Right. And this, this is the perfect example.
15:19:06 I don't know what it is like 20, right.
15:19:09 So, actually, 25. so
15:19:17 I don't know if it's a good instruction, but the instruction would would be, or is suggested. It doesn't matter, but preferably start using new coach.
15:19:36 Because we will we will come in this will be a bit of cleaning, but you know, Eventually, everything will be converted into new exits.
15:19:46 So, I don't know what you think about this, but at this point.
15:19:53 Vojtech when you started restarted coding What are you using new exes or I called exes, all that says all this. Okay.
15:20:03 This one comes to coding. So Richard done.
15:20:06 Now we're just gonna, we're just now coding in English.
15:20:10 Yes.
15:20:11 Well there was one of the decisions that we could only in English.
15:20:16 Okay, which will, you know, again, eventually, it really doesn't doesn't seem to make any difference because we will have the ability to sort out into three language groups anyway.
15:20:31 So, the English codes will be impossible will be easier to for everybody to see what's going on in every in each of the three language communities and if someone of course wants to begin.
15:20:42 Then they will know what they're doing in their language right and the more kind of ethnographic in depth look into something, of course will be in the, in the original language, a more practical question then do we, so we, if we already have something
15:21:00 that we coded using the old system.
15:21:03 So what happens there do we remove all the old codes and no no no no, we will do it in the backend stuff that's already been corrupted we could just leave, and we just do that.
15:21:16 Exactly. And that's why the same logic is behind this answer, it doesn't matter because we're, technically, no you can start coding in the new using new exes as much as possible you will save us some time, because we will have less than work to do to
15:21:35 convert.
15:21:37 But Wojtek tells me he started doing that already, that it goes very fast.
15:21:57 Let me just say something, it won't save any time it will take exactly the same amount of time, because it doesn't matter whether you for example code for addiction or psychological health.
15:21:54 Either way, addiction will be merged into the psychological health but I would really want to make it super clear with Alberto I mean, we have like green light for because basically will annihilate those codes.
15:22:14 When you merge them into those new contemporaries are when we are going to make them merge them into those new, new codes.
15:22:31 At this point in time, it absolutely doesn't matter which goes to us, because again it doesn't matter whether there are like just, I don't know how many annotations are there for the diction doesn't matter how many annotations does addiction have
15:22:38 either way it's going to get merged. Hopefully, into psychological health.
15:22:45 But if you start using those new x codes which do not exist yet. In this in the background, you'll create them in the back end and for some time, until we get this green light from Alberto will proliferate codes.
15:23:01 And they will be kind of like detached from the rest of the already coded material at least four songs, what do we need Alberta's green light for, For example, I would like to.
15:23:15 I'd like for him to tell us that those codes are no longer used in any ongoing projects, or that nobody is actually trying to use.
15:23:29 Add riders platform to kind of investigate different.
15:23:35 Many I see that, yes, all this is plaguing the European community or something right right right right.
15:23:43 because if the code is for example created but Amelia and then appears in three other projects and suddenly we will change this code for all three projects.
15:23:52 Yeah. Yeah, so either that he finds a way along with his team to kind of
15:24:02 create for some kind of like a sandbox environment, which we're going to do with the colors whatever we please, and create our own code book, kind of from scratch, that will apply only to borrow more texts, or, or we need to kind of make sure that we're
15:24:18 not going to meddle with the results collected, that are being collected by by anybody else, using the tools and the materials, collect on the platform.
15:24:32 Get it.
15:24:33 Yeah, so can you, can you please make a note of it because we really need to start moving faster so maybe. For now, we can write to Alberto as soon as we can, preferably today.
15:24:45 Yeah, but it's not going to change anything when it comes to a workload, we can safely use the already existing Xcode, I mean those are those that are okay,
15:24:58 And since problems is the, is a very special category in which we are suggesting kind of like such extensive emergent mergers and changes that are going to heavily affect what what is the in the backhand.
15:25:12 We just need to make sure we're going to be absolutely positive that it's it's allowed and that will not working.
15:25:21 results.
15:25:22 So I'm making.
15:25:26 I'm so.
15:25:29 The second thing that we wanted to tell you is that the way it will be written.
15:25:36 Ultimately, is that the first time, the new code appears.
15:25:43 They will be more extended name of the code or a bit of explanation in comments, but it will not be repeated.
15:25:52 When the, it is cold, you know appears the second, third and so on time.
15:25:59 And the next thing we want to tell you is you know this is the category that is very carefully vetted already by a lot of people and, you know, we'll see the name so thank you very much.
15:26:11 A, but there are still places and particularly. I'll show it to you.
15:26:20 At the end, we are creating the following situation we've, we've gone through all the codes now in problems and needs.
15:26:28 So here you have that, that what we call the tail. And in that day we have a lot of questions so your names appear here for your decisions.
15:26:40 Sometimes we say, those are suggestions right. We basically reduced the words that we use in suggestions to out for all this information merge, or there's information that he was already moved.
15:26:56 And sometimes change and the, the, the, usually, then, then always, when necessary, there's an explanation.
15:27:05 And then there are names here and there. And this if you could please go through that there's not much of it, and react or suggest, or agree with the proposed change.
15:27:19 And that goes for a while so most of it is outs, but sometimes as you can see our other suggestions. And then here we are beginning to collect outs from the
15:27:38 part of the spreadsheet that is already kind of finished right or is being just polished now and we are now going one by one. Bye.
15:27:57 Bye. Bye. One more time and we are taking out to the bottom of this spreadsheet.
15:27:59 To make it easier for everybody to quickly, have a look at it but. So, so far, you know, we've just started doing this. You can see this.
15:28:11 All of those in this case are suggestions out.
15:28:21 A okay I hope this is clear right so the key thing is your names here in comments and suggestions in the column D and see if it makes sense.
15:28:34 Here's the part of the system that we are still playing with a little bit, trying to figure out new wise or a Greg's, and he.
15:28:50 The it's going relatively well, you know, we were the. The fourth thing that we want to tell you is that we are coming up with a short names for the codes, because we're trying to see okay, you know, when, when is it visualized if the name of the code
15:29:09 is three words with spaces.
15:29:12 It will be a much more messy, but each time we you know we propose the name we, we, again, the first time, it appears there's a brief, brief explanation and comments.
15:29:25 And those names are compromises between our first attempt were those were just like three letters.
15:29:32 And then you remember you told us that this is inscrutable because then you have no clue on it so it's sort of a compromise with few letters trying to suggest what it is.
15:29:47 So for example, labor issues in security right so that's pretty clear.
15:29:51 La work shortage, we may want to come up with maybe even more compressed somewhat more compressed, so it is just one word.
15:30:03 A in the health system. So, it is HS.
15:30:14 It should be HS here for example I'm looking at line.
15:30:16 55. Right, so it's not AC but HS access, right, and age as a tentative there's, you know, but because it seems in your cars coding, there was a lot of about alternative medicine.
15:30:30 So, things like that.
15:30:36 And
15:30:36 that's basically that's that's what it is. We've been moving stuff to problems. One more time, looking carefully. In other categories. According to the rule that problems from other categories and we want to, those are sort of our dependent variables
15:30:59 right so we want to have it all as clean as possible. But we also pay a lot of attention to how it is organized a just another all as you recall, is. But once we do it right, the system will then be doing it for hours for us.
15:31:18 There should be no annotations in, in wise, remember some time ago we made his decision so it is clean, why is just apparent that the all the annotations come from the children from access, and the.
15:31:39 Okay, I guess that's all about the, the emerging code book right yeah maybe just one more thing like.
15:31:50 Since we kind of just to try to shorten them.
15:31:55 Some of them became quite cryptic so if you find a better name, more telling name that is short and the requirements of having this prefix, then you're welcome to suggest that that that there was one encoding I think about landlord abuses, which, when
15:32:16 you read it out like doesn't tell you what it originally was.
15:32:21 But generally, Yeah, yeah. So, we welcome your input.
15:32:27 So there's that.
15:32:29 There is one more thing, during our last or the last part one meeting, we considered again splitting the interviews into even smaller chunks than they are kind of like doing it more arbitrarily by putting them in the form of some kind of thematic units.
15:32:55 And I think we have a good argument against it because it opens up the whole area.
15:33:04 Well,
15:33:08 we can be quite easily accused of of our bitterness and I think it would it would totally mess up our decoder reliability, if we, if we just decided I mean, who's going to do it, or again to do it as a team or games do it individually.
15:33:25 Generally I think there are a lot of problems with doing that, making it work and yeah I think we have a little incentive to actually go for that.
15:33:38 We also try to think again about the differences between putting a whole interview in one post on the platform against versus
15:33:55 placing it in a format that we have agreed on which is like a question response in one post.
15:34:03 And the argument for adopting the latter would be that, for example, if you have an interview. That is quite long.
15:34:12 And it is recorded that's just one post we have just one post for analysis. And, and the other form, like when I looked at what money I did one interview can be presented as a thread with 60 or 80, different posts which allows us to get much more granular
15:34:37 about the data, information presented that we're able to use any given code like multiple times, because those posts are quite numerous. So, yeah, I think we have a pretty good rationale for sticking to this form this way of formatting.
15:34:59 Yeah so yeah we have we we've tried to think through and looking at the examples that those two dilemmas one is, you know, I know we all decided but we wanted to look at it one more time whether a given interviews should be treated as one post or as a
15:35:17 number of posts right.
15:35:20 And then, if you remember those of you who are there, our conversation with Alberto Alberto pointed out that it.
15:35:41 We know each method has its pluses and minuses but he, if we divide the interview into several posts. Then we are going to be more precise in catching actual meaningful edges, right co occurrences.
15:35:52 So that was the I know we made a decision already some time ago but we just wanted to revisit one more time, particularly we have to write it up in the code book.
15:36:02 But the second decision was as Voytek said even more important. And this is for your consideration. It was Danny suggested that we look at the merging if the, you know, the first question is about, I don't know, healthcare, and then they talk about something
15:36:23 else and then.
15:36:34 And then your suggestion was very obviously, the legitimate when we merge those two.
15:36:41 So we started actually we went to the interviews, you know money as text and we looked at the possible situations, and obviously we got to the point but very thick fog well those two should be much, and I thought well no they should in which immediately
15:36:59 indicates that there is a danger of arbitrariness.
15:37:03 And I guess we've we have a lot of resources, a lot of time I don't know and multi coders situation that we could do for more serious.
15:37:15 Reliability checks, maybe we could try that. But we don't have time.
15:37:19 And it's not the investment of time I don't, we think, I mean we, that's what we suggest to us is not worth it.
15:37:27 And it is not going to perhaps as much as we could see, because it's very preliminary.
15:37:34 It's not going to happen too often.
15:37:37 So, you know, maybe even I mean you can figure out the justification that if, in the course of a conversation or interview they come back to a certain topic, or it's this topic kind of re emerges in the interview, maybe it indicates that this is important.
15:37:56 So, you know, it will generate more edges for a given issue, because again, remember we Alberto showed us and we discuss that.
15:38:11 If we treat an interview, as a single unit, without dividing it.
15:38:19 For example, the, I don't know why but we've talked a lot about grandmothers well because they are in some interviews very important you know this generational gap.
15:38:27 So, a grand grandmother, the word grandmother appears in an interview say 10 times right.
15:38:35 So, if, if we code, the interview, as one unit. So we have 10 times run mother or grandmother should be coded only ones.
15:38:46 And if it is repeated in the text, we don't, you don't impose the code on it anymore. Right.
15:38:53 And that would be that is sort of undercounting, whereas assigning the code grandmother 10 times will generate tremendous inflation because then every edge will be multiplied by 10 with grandmother and they will be everything that appeared in the post
15:39:13 right.
15:39:14 But now imagine the situation that you have 10.
15:39:19 Then you have 40 units, right, posts that interviews divided into 40 units in 10, in 10 of them, the word grandmother appears, and in each of them.
15:39:33 It generates just a few edges.
15:39:37 I don't know grandmother and the dog right appears on the ones in one unit and then so we have another one edge grandmother dog, and so on so it is sort of as much as we can see that either generates the number of edges that is in between the first solution
15:39:55 and the second solution so not to excessive, not too small, but also more meaningful. It seems so that that's roughly, you know, I guess, we are telling you this because we are already beginning think you know how to write this into our code books, which
15:40:13 of course after we've written new, you will, it will give you a we'll have. We'll wait for your comments but that's that's no, that's, that's the one of the things that we discussed again on the basis on looking at actual text is then you.
15:40:36 I understand this this decision, it makes sense but I'm. For example, I'm not that familiar with it with the interviews but sometimes you speak about a topic, or someone in in more questions.
15:40:51 So, my knee elbow, I don't know or or it will explode explode some topic in like more questions.
15:40:58 But for example when they when the people answer they will say like them for her, like the Migrant Mother. So then you need, then we need to kind of cold it is her or them, or that that problem when they said that problem.
15:41:14 We have to take the code from the person or the problem to the other, to the other coast right. And I presented as always you're hitting the nail on the head.
15:41:27 Precisely, really tells Danny, because I spent some time thinking. Let me share my screen maybe.
15:41:45 That's exactly it you're absolutely right. So we actually played with the codes exactly because of that issue.
15:41:44 Okay, so I have to get out of here.
15:41:51 Okay. Unfortunately the post is in Polish but I'll try to kind of.
15:42:04 Well everybody speaks Polish Richard by now I think is fluent looks at what. So, here we have an interview with Susanna written in this preliminary say way method as one post, and obviously this causes the problems that he already mentioned, but I'm here
15:42:21 we for example, because it touches upon what what you said, Yeah, like manual says, you said about your parents and your grandfather, grandparents, they're being been like, oh God was the word
15:42:40 involves being involved in tangled in political changes, changes the religion is a safe place for them and so forth. So, for example, in such a case, when when the interview is kind of relating back to what was said before and is not present in the interview
15:42:58 itself, I think it is okay for us to code the interviewers question, and probably only then, I mean generally we should not code, the interface questions,
15:43:11 then hitting closer to what she said.
15:43:19 She says, like, yeah, yeah.
15:43:21 Indeed, what Lauren justice says is their opinions or they they kind of like match perfectly us. And so we have this, there's here. Right, Yeah, and we cannot just go like when you to know because again, let me show you the different coding
15:43:43 format.
15:43:51 So, when I haven't presented in this phone.
15:43:55 And this response. We don't know who those, they are right. So, while closing this, we should put are kinda like substitute this day, or there there's with parents and grandparents, which we have a code for, namely older generations, because otherwise
15:44:17 without that. In this short fragment which is a post itself, which, which means to us it is our unit of analysis, we won't have any subject the object.
15:44:27 Right.
15:44:28 We need to kind of, it's a different way of coding because again coming back to the older version, we will have those grandparents and parents, appearing in the text I mean coded for writing this whole interview.
15:44:42 This whole fragment that the system is going to analyze for us, whereas in this one.
15:44:48 If I codes, like all the generations here. They want to be here and there will be there will be no
15:44:55 no edges between our older generations and what he says about Lauren justice and their propaganda or like, whatever.
15:45:03 So, we need to make sure, while using this method of data record or storage that we recognize what these
15:45:17 pronouns or whatever they are, they are referred referring to, if you have this legendary Bob.
15:45:23 But we need to kind of bear that in mind, to kind of make it meaningful, which we can quite easily see, if I switch to the coding view which.
15:45:38 But you can't code money right because she is in a in individual post on. So her questions are individuals so if you call them, it does not make much sense does it.
15:45:52 Unless Unless like in this example, she's referring to kind of some conversation that that is not recorded in writing in the system, but that kind of presents us with some information that the respondent provided.
15:46:07 So for you we have like this. Empathy learning justice part indoctrination but we don't don't have this connection with the elderly. So what I should do in here is to kind of know some older generations, or whatever.
15:46:30 All the people. Right, I need to place this code here to kind of make it wholesome, but particularly that you know what happens here is something very interesting because she says, only right vape.
15:46:37 So we have to put the code on you, which is the older generation.
15:46:42 Because she says, only have they have those views, but only bacha only grandmother can sometimes empathize with me right so now there's a different information but despite of this monolithic, they having some views close to peace, Bob is the only one
15:47:01 who can is kind of softer and can empathize with the younger generation right so if we don't impose those codes then all of this kind of disappears.
15:47:14 We cannot actually kind of reflect that in our codes, because it will add only one older people code in this abstract Oh, there is like each, and only and Bobby Cha, it's like three times, actually, but it will be only one there.
15:47:31 Yes, exactly. Very important point, just one once once per unit of analysis, I think that's what we decided with Alberto, and the unit of analysis now is pretty small because this is just a chunk it.
15:47:46 I mean, basically, an answer to a question.
15:47:50 And what.
15:48:03 Sorry. No, go ahead.
15:47:54 And what if they kind of like in their in their answer or like very kind of.
15:48:00 It's something that it's very like important for them or very kind of like radical like in some amount. Would it make sense to code it like twice the same code.
15:48:09 So kind of to arbitrarily kind of strengthen the links. Yeah, but you have to burn in mind that if you for example decide to do it twice in a given piece of text.
15:48:24 Those the edges between those codes will multiply. Yeah.
15:48:30 That's what that was telling suggests that this was something really important.
15:48:34 And the multiplication, would.
15:48:39 Yeah, express that important.
15:48:40 I don't know what do you guys think in your car you had your hand up earlier I'm sorry, let me I just address that issue one one more time. Wait, wait, wait, wait, wait because your cards.
15:48:54 Okay, okay, okay, okay. Yeah, um, because for example if we have, say, I don't know, the elderly, poverty, Lauren justice political turmoil propaganda and rhetoric, you will not only strengthen the edge between the things that you kind of want to strengthen
15:49:16 but you will strengthen all edges.
15:49:18 So it comes at a cost.
15:49:21 You may choose for example to choose to codes that you think are particularly important and kind of use them twice, which will result in those codes having like double the number of edges which because they will multiply, but it will also kind of affects
15:49:38 all the other codes used in this post, you are not we are not able using the tools that we have to kind of transcend any one single connection, but by using a code code, and they've given code, more than once, we are strengthening or all those other connections,
15:49:55 which I think doesn't matter much if if the post is short, but if it gets longer and more codes, we are kind of in a way, warping the results.
15:50:11 Like the neck.
15:50:14 I mean, sorry. Yeah, because you moved on my screen. Then it was on top and you are the bottom and suddenly everything switched.
15:50:22 Sure.
15:50:24 I think that, yeah.
15:50:26 This all becomes a little bit of a problem course that we've been dealing with the cut, you know, personalizing or splitting up the interviews into more units of analysis.
15:50:41 Obviously, what you just presented is making up for the negative effect that it has that we're losing kind of like the big picture edges right which sometimes, and more often than not, have been irrelevant, but sometimes they are relevant and you need
15:50:56 to bring them in back by recording something that the bigger than the previous answer. So that that puts a little bit more stress on the consistency of the colder who should be able to kind of like follow the main narratives, and infuse them in the individual
15:51:14 units of analysis, when he deems that to be relevant right because that that also kind of like, again, it's the basic problem of qualitative coding that I realized there is during the time we've been dealing with our project right but how much semantic
15:51:32 do we want to remain an explicit and how much latency coding do we want to do that again.
15:51:40 I think could be largely solved by the new code book, and that's kind of like the first idea that I have is that I would really like to start coding the new material with the new access, and reason for that is that they were basically tailor made to also
15:51:56 adjust to the needs that we as coders might have, because we already know that there have been like, somehow massive topic shift during the time as we transfer from the first version of the court post the second core post which is basically the interviews.
15:52:13 And also it will be much easier to cold. And, yeah, there's the issue that actually you all three kinda like talked about, which is the fact that this is a unique chance to kind of like own our own codes for our own projects right because if we continue
15:52:30 using the old codes for the new corpus. We are as with except like proliferated thoroughly through reading the problem of other projects using our codes.
15:52:47 And then the merging situation and splitting the cones along the lines of the projects will just kind of like grow a little bit.
15:52:50 But that's that's that's more or less what has been talked about before. I just wanted to say that technically speaking, the fact that we are now going to be using posts, individual replies under a post as the unit of analysis is technically better because
15:53:11 it still allows us to analyze the whole threat. As we can just basically click a button in the metadata projection.
15:53:21 And we can still do the first thing, just with the more specialized coding granularized coding and I think that's great because that's already giving us two data sets, to compare in the final analysis and see how consistent.
15:53:36 The two methods are. And I think that this kind of like good reflection on the whole methodology, switch that we experience which is obviously you are nodding so you know that already I was just thinking out loud because I like it.
15:53:51 And no, it's great you know it's so important that in those conversations we kind of share our thoughts because sometimes we think about something and you know we were not sure.
15:54:02 And then you guys the particularly rewarding is when when you guys say something that we were kind of trying to figure out and you have the same or or very often really the same intuitions are the same ideas like previously standing and now you're so
15:54:19 that reinforces.
15:54:20 I guess the sense that, despite of the enormous amount of time we're spending on it that it makes sense. Yeah, I, I mean, the first thing that came to my mind was like, oh, it will be so awesome to be able to just take the new excess and just like, put
15:54:29 them in the system on block, so that we would have to hold dictionary, at our hand, but because we probably cannot do that.
15:54:46 I don't think so forth, our rights we can generate and void cold with no annotation.
15:54:53 That means that we will try to coat with.
15:54:59 Maybe if we want to do that we have to try to close with the open G docs and that's that's basically, that's basically it.
15:55:07 Which is not a big problem possibly since we can like know the old code. But, one more thing, and that's that's what's like the still in my mind which is what we touched upon, as well as the fact that we are not ended this is this is again a methodological
15:55:25 kind of like how to put it fork.
15:55:29 Because now we came to the point when we are going to be coding.
15:55:35 So latency that we don't really expect the text that's says the thing, appearing in the post were recording it because we out of the context know that it's important to create the edge.
15:55:46 So we are starting to cold and notation, independently, in a way, if I translate again. The saying what I've seen what you will have been presenting SS was the example with when you which was not that extreme we still have the pronoun, but sometimes we
15:56:04 might not get really a semantic unit, that would be a pointer to the code from the previous reply, but we still will bring it in, because of the context because otherwise the edge would be missing and the information would make sense.
15:56:27 That is some possibly mythological problem because our coding, that I mean technically is.
15:56:41 It's like language dependent right we still need the materialized language in order to create the annotation and grade to code. And it is a matter of again just I think our internal discipline as coolers.
15:56:49 text, you know like that one word like cold only only and rely on the fact that that will represent the people of older age, and why I'm saying it because this I was thinking, like, Wow, this is great because actually this kind of like an item ization
15:57:08 of the whole discourse into the individual reflects that overcomes largely the problem that I've been talking about also last time about the software that I had in my mind which could can like really closer information in terms we will be coding one whole
15:57:25 thread, as a analytical unit, but there is the kind of like a place to maybe use this because I can really do the thing that I can discover a then went to codes are coding the same text that's that's something that I can do her precisely and very safely
15:57:49 And it might be useful in terms of larger longer responses because also from the check interviews. Sometimes one response to one question, can have two pages, easily, it is still a very, very long piece of text.
15:58:04 However, the just just my last thought it is excellent and by this way we can actually completely sort out the interviewer, if we want to, or bring the interviewer back in the picture, if we want to again it's just a matter of click, and the meta data,
15:58:18 if we, because we have the speaker code right so this that's that's manageable. That's great. I think this works really really nicely.
15:58:27 That's it, that's that's all the things that came up to my mind during listening to presenting.
15:58:33 As always, great job. Thank you.
15:58:36 If I may very quickly then, you know, Vojtech and I talked about the same thing, and we very soon hopefully we will do the experiment you know code one interview in, bit by bit, and then that is the whole and ask Alberta to visualize both and see what
15:58:57 if we if there's a difference you know what's the meaning of this defense and so on. So, there's one thing that But the second thing let me please remind you, and this is really super super important.
15:59:09 The problem of new x is is really only the problem of one category, which is problems, and this we can do. Yeah, really, we can finish it very quickly.
15:59:21 In, in other categories, predominantly I don't know, almost 90 over 90% of codes, x codes are the same as old ones that the only the work that we were doing it in other categories was the work of signing access to a new newly invented categorization of
15:59:49 And that's that's it so that, you know, whatever the problems are the biggest category but you know whatever it is 20% of codes. So 80% of codes, it's fine.
16:00:03 So, this conversation makes it very clear that we have to resolve that because we can resolve that issue very quickly. We can be done with assigning new access, very quickly.
16:00:16 And yes, you're right. And that's what we wanted to hear that, that it would be better to called using the new system, but again in the new system.
16:00:27 The problem of this kind is basically only the problem of problems, because I just as you were talking I look quickly at two other spreadsheets and this is the same old x is just grouped differently, and the different wise, and that's not our problem
16:00:44 And that's not our problem because the system will be doing this for us.
16:00:49 So, yeah, we need to get this information from Alberto very quickly.
16:00:55 By take what you were saying that, you know, if we change it arbitrarily will change it for every project that is on the platform and.
16:01:05 Okay, anyway so those two things very important I wanted to share. And now, Danny again then voted in the order of hands up.
16:01:16 I think voting was first, because I'm the second.
16:01:20 Okay, but there is an hand.
16:01:23 Okay, but maybe. All right. Yeah.
16:01:27 What it is ok so now I'm the first one.
16:01:33 Well, I'm in makeup up with your answer is probably solved. So I was thinking about like coding with a new code is better, definitely.
16:01:48 My question or my note would be that after this project. Well, did the riders want to leave all the data there with the coding, because they want to have this kind of collective, whatever.
16:01:57 So if after our project we kind of delete the the hierarchy.
16:02:09 Would it kind of solve the issue for them that we kind of assign some codes to our hierarchy and then when we delete the hierarchy there won't be any or are there any hierarchies already created in other projects.
16:02:19 So if we can have.
16:02:27 If we kind of take some code from some other higher ed, our hierarchy.
16:02:30 No. Yeah, I don't know, I doubt it. I think we are clearly on the path to have sovereign project, which will have very little overlap with other projects, but as far as I can tell what the media did in the first place when she created those categories
16:02:50 during the code marathon coding marathon I don't know what it what it was but she did it globally for the whole system.
16:02:58 I mean I saw no signs of there being any other hierarchies used independently.
16:03:06 If you create a hierarchy for any given x i think it sticks globally for the, for the whole system, which kind of makes me hopeful that maybe we will get this green light because if such a
16:03:21 over overcompensating intervention say was, was given a gringo as like creating those hierarchies for all the codes in the in the backend. I think it may be possible for us to kind of like change that.
16:03:36 And like you said, answering your question for problems, we will be changing codes, we will be deleting access, in a way, I mean the nominally will be deleting them we will be keeping those annotations that we used those codes to call them for, but will
16:03:57 be creating new codes and merge all the access into those new code. So, this will be kind of probably reversible.
16:04:05 Maybe they have some, some ways.
16:04:16 Some time stamps or whatever.
16:04:12 They can probably mirror, like, at one point that can like kind of preserve the whole system and then maybe like move it. Yeah. Maybe that's why that's why this is another question for our Berta lead maybe before we do that, although we started already
16:04:27 changing something so anyways, minimal. At this point, but, you know, let me share the screen one more time quickly to say one more thing.
16:04:37 Now we have this new functionality that has been introduced me Can I share my screen first for just Okay, I'm coming back to here.
16:04:51 So again, like when you, when you look at, to the right of any given post the system tells you what codes were you then, like, again, I'm still talking about this pronoun.
16:05:05 It should kind of create a pretty understandable and coherent story so we have those older the elderly people who are kind of like religious, they sometimes demonise other groups and there is this big role played by the Laura Justice Party, and there
16:05:24 is blown ization and sense of thread among them and sense of safety which comes which stems from religiousness which is not coded. Sadly, but there should be.
16:05:35 And those things became clickable.
16:05:39 Oh, Yeah,
16:05:45 jumping with joy. Yeah, I know.
16:05:49 to kind of help with those messed up, interviews, and it can prevent some situations in the future. Hopefully I don't know I haven't checked how to act with those messed up annotations, but I think it kind of makes it much more usable now.
16:06:10 So I just wanted to add that.
16:06:16 Thank you.
16:06:18 Okay. Then you could send you new hand up.
16:06:20 Yeah, it's this much, but but the other topic, the deadlines, the time plan.
16:06:28 Yeah.
16:06:34 Well, I'm not ready to talk about deadlines I, I have the sense of enormous time pressure that's but I like, you know, we need to the way I you know this point.
16:06:57 The Voytek and I need to finish problems.
16:06:57 You know with with everything you know with everything else going on.
16:07:00 We are under enormous time pressures but we are going to talk now and we will finish the problems.
16:07:09 I don't know, maybe what is it about the June will be the that, like, kind of, we will be presenting something on this on this conference. So before June.
16:07:22 Ideally we should present something as we should have something.
16:07:26 So the coding needs to be done until let's say like, I don't know, and have a prayer, approximately.
16:07:36 money are collecting the interviews, until until April also or it should be done already you know until September.
16:07:57 Okay. Basically, you know, so we are we will help you.
16:08:03 I think that.
16:08:04 I mean, I think I am assuming, and I'm pretty sure that's been great to go that way you know once we have this sorted out and the code already that the whole thing.
16:08:16 The will accelerate and the you know right now the problem is the bottleneck is on the coding side, I think, pretty soon, the bottleneck will be on the generation on the side of generation, new material.
16:08:35 Hopefully not. But, you know, I sometimes really worried that you know this has been like, what, two years or so and we will have 50 interviews or something that's not particularly impressive I'm a break.
16:08:50 So hopefully we will have more.
16:08:53 By the end of the summer say.
16:08:57 But we'll see the, the material is pretty rich.
16:09:04 We'll see what happens. But you're right, we need to have this. Yeah, we will, we will talk about it and we will share with you that the schedule.
16:09:14 Because I know there's some other things also going on it's not it's an important question. Obviously, there was a question for Richard.
16:09:25 I think Richard told me that he has a new PhD student that is doing some interviews, instead of john, and she will be maybe she's from fatigue Yes, yes, Sabine Sabina or some Yeah, that'd be assaulted, that we probably some more interviews coming sooner.
16:09:43 My Monia and Utica are doing. Keep going, you know, and you've got me you've seen the email from eat garbage you will have some interesting stuff from Australia from the miners, which is would be very fun, there will be really great
16:10:04 Vojtech.
16:10:08 And then
16:10:15 also I just wanted to say that we have like 12 German interviews so far, and 30 something interviews in Polish and check.
16:10:20 I'm not counting the, the minus yet.
16:10:23 And yes, and the tog refer started working in March 2021. I think it's Yeah, not that much, at least as far as I know, it's not what young was counting on the, but I think for preliminary analysis that we may have to present in June, may suffice but probably
16:10:51 there will be more, I mean hopefully that will be more.
16:11:04 Well, let's assume that I don't know by the end of March, maybe we'll we'll have 40, that would be okay and then we'll have, you know, but as soon as coding, restart which might take I know it's ready we started coding and so that the, to, to, to have
16:11:15 as much clarity at the moment as we can. The most urgent task I understand is to finish the problems.
16:11:25 And, in the sense, make sure that all the new exes are done.
16:11:30 And in order to do that, the guys, there are those requests I showed you with some merging and forking, which is, you know, most efficient is when, you know, we assign it to those of you who were the authors of those codes most, most cases.
16:11:53 So, we also assigned a few to ourselves. So, in the Polish corpus there are some things you're some codes of, you know, like to give you another example of something that we have to sort out very quickly.
16:12:04 The code polarization is extremely important, obviously when you talk about populism so we have a way code polarization and this morning we were thinking about, you know, new access for it and I think we will propose a game, very simple.
16:12:24 The political ideological polarization one x, social, cultural polarization the other one.
16:12:31 And, and, you know, again, being pragmatic because when we started looking at the text, it's some it's very difficult to, to say this cultural and social but social cultural if it works, and it is separate from political.
16:12:46 So I think the decisions like that, but they seem. I mean those decisions we've noticed are coming to us faster now before we had a lot of back and forth and a lot of kind of maybe this maybe that, and I guess we I don't know we've met might have acquired
16:13:05 some skill or, you know, we are more practiced and, you know, it's easier to make those decisions.
16:13:13 But we will keep you posted. Again, we will try to be done by the end of the weekend.
16:13:19 We'll see you Erica.
16:13:24 Yeah, I'm just like, I would like to ask is it forward to what you just said about the forking in merging decisions isn't something that should technically mean that we really recalled the old codes, with the new codes straight ahead, instead of writing
16:13:44 it back up in the emerging yes merging you're absolutely right, I should. Yeah, you're right, we need to because we realized this this also the morning.
16:13:57 I don't, I just maybe a practical example is in my mind if there's like kind of like okay, there is this polarization thing, and we had called polarization there is some seven courts in check.
16:14:10 That I inserted somewhere like technically to do it through the Google, I would have to list out the individual annotations by their numbers and say, Where do I proposed the replacement with the new code that appears possibly to be done with like maybe
16:14:28 10 codes if there are like one decisions to be made it's maybe easier to just use your new codes and do it straight to head, I don't know.
16:14:39 No. Well, when you because we started doing that in the Polish carpet. So, if, if you are forking maybe like a sense that, too much is pack, right.
16:14:51 You have to go in and for each annotation decide what the new code should be. And you're right it should be the, The new code should be the new coach
16:15:05 right now.
16:15:07 So, it means that, technically, you need to wait for us to finish proposing new access. Right.
16:15:15 Okay, I see that. Yes, okay. You're right. because we're creating another loop that will come back to you, and for you in a way that's a good point fat.
16:15:26 Absolutely and and, yeah, it really helps that you articulated that now. Okay, you're right.
16:15:33 We didn't think it through all the way to the end. That was just a little Lido anyway we have to we have to react to all these kind of like under the table coats, as I call them that are still hanging on board on the table calls Yes, under the table.
16:15:49 Yeah, there's a bit there but it's most of it is out. Yeah, manage it, you know, that the only thing would be useful in merge you would see a few suggest the name of the new code.
16:16:04 Yeah, sure, because you don't need to worry about merging one to another, in the old categorization of access, but just say this, if it says merge.
16:16:16 It has some old code. What new code should be there now.
16:16:23 Yep.
16:16:24 We don't need to kind of like finish doing what we're doing, because those new axes are already there suggested.
16:16:34 that can be used. And by used by being used it will be created in the system. So, if you already for example have those say four categories of prioritization and say, social, cultural, political, economic, or whatever, economic, then they can really be
16:16:53 used.
16:16:55 Yeah. And then you unmute Yes, yes. But let me show you one more thing that they also asking for your advice very quickly, because I that we have to but then you are still in the Google Sheets, they are not in the back end.
16:17:08 Yeah. When it comes to problem. Yes.
16:17:11 Yeah, but if you put them, you know, so he has for a while they will be the multiplication of codes, but it will be very quickly resolved, once we start changing old names to new names.
16:17:25 and you know we tried it really goes very fast because, you know, despite of sometimes those numbers look kind of scary like 4550 or 60.
16:17:35 As you may recall, in most cases, most of them actually quite a huge majority are not our coach, but from other projects so for us this is less work than we may see in those numbers clearly less work than we've seen those numbers, but here is if you see
16:17:57 the screen, you know environment like. So P environments which is why it's a problem environment.
16:18:05 And, and those are old x's.
16:18:08 And we, we don't have new exes as you can see.
16:18:14 But in this case, it may be the same strategy, as in other Z's, which means that we will just repeat those new codes, just add env in front of them because they seem to be very sensible.
16:18:31 You know I've been thinking about them yesterday lol whatever two days ago and I. It's like, you know, environmental problems maybe should be forked right so I have to do that because that I started that.
16:18:44 But then climate change pollution human intervention emissions decarbonisation, you know i i don't know how to measure them they are all important and separate issues right and then water.
16:18:58 Yeah, there should be a
16:19:04 an emissions. So emissions we have twice that those are easy issues.
16:19:09 So something like that, that will be quickly resolved. if you go
16:19:17 one more question for you guys.
16:19:20 We have social relations and social anime.
16:19:25 And as you can see here in line to 39 merge with anime we are going back and forth on that with voice.
16:19:43 One more thing. It also pertains to many of the comments especially by your car.
16:19:51 This category consists of magic between resource needs and problems and since we cannot kind of like, based on on the definition of the problem and the need resolve the
16:20:08 difficulty in distinguishing between the two, we merge those two categories and this category basically is problems and needs because they kind of like are the
16:20:21 pretty strongly connected.
16:20:24 Yes. Yeah.
16:20:26 So the fact that somebody has something probably means that somebody else doesn't have it or they want something which is their need means that the lack of it is is a problem in fact and so on.
16:20:37 So, um, there are some residual codes in this category that come from the contemporary of needs like for example adequate pensions benches salaries, or like decent salaries, I think, is to code.
16:20:54 The fact that somebody. I don't know, says that they're happy with with the salary and Poland and we know that like, people don't learn that much in here is a problem slash need so so just bear that in mind when it goes through those codes.
16:21:07 There's not not just and then Erica thank you for the suggestion I think that's a great solution with as a support deficit, and it really merges the whole bunch of things and you know from nine to 72.
16:21:23 Basically 279.
16:21:25 So yeah, that if you, if you have a moment guys and you can have a look I mean that those suggestions are fantastic because it saves our time, you know, basically we, we will just sort of maybe just to reply to what you're asking, I you know I think that,
16:21:41 in terms of the, let's say this kind of like high definition look at coats, that we started up with, and the one that we're heading to I I don't personally see that much of a functional differentiation on the episode level between the enemy.
16:22:00 And the relations.
16:22:02 Since, the social relations here are more or less cut like a on the negative side of social relations.
16:22:12 Totally fit into the world. That is kind of like generated by the enemy codes. And also, if we look at the new access the enemy codes should be reduced greatly by the new access suitable also because it seems like there's too much in enemy to just kind
16:22:30 of like bunched together with the relations which technically are problematic social relations, right. So I think this is really, like, come together in one episode.
16:22:45 Okay so good. I, this was our growing sense that, you know, maybe it's also already indicated that we colored in the same way.
16:22:57 Ok.
16:22:57 Ok so I guess that's the decision then we will merge that.
16:23:04 And then this will be considerably reduced.
16:23:10 As this example of the year cause support deficit code suggests that it will really shrink it into this the few codes.
16:23:32 But we have to think about it a little bit because you know here, the codes in social relations which indeed a kind of negative social relations that the codes are pretty important right so we will we will think about this for a moment.
16:23:41 Yeah, man I just.
16:23:44 Actually I had a kind of like a slightly more important thing to talk about before I reacted to this.
16:23:53 And I don't know if you already took some decisions but I was thinking a lot about the emotions about how to actually, you know, do it in the same way that you are doing it.
16:24:06 And if I don't know if you have it open but looking at all the categories. I just came to the feeling that maybe with one exception.
16:24:18 the excess and epsilon and emotions school all technically become access. It's all together, 28 codes that somehow, again with one or two exceptions are quite distinguishable.
16:24:32 And the problem that I was still kind of like juggling with was like how to epsilon eyes, the whole thing right.
16:24:40 What should be the new episodes, but to be honest, by take a look at it. They all seem more or less like x codes and the only episodes that really come to my mind are like positive or negative, which, again, is maybe even useless so when I was thinking
16:24:57 is where the emotions really, they, they, if they really need any episodes, you know, since it's quite small compared to the other sets, of course. So I would say, what's your what's your take on it.
16:25:12 Do you think there should be some kind of differentiation more.
16:25:16 Should I just try to divide them into positive and negative emotional response content, reflection something along that line, or should they just be like, without any episodes.
16:25:39 I'm trying to open emotions and for some strange reason doesn't want to open.
16:25:45 Well,
16:25:49 what's your feeling guys about what's your emotion about emotions.
16:25:57 It should we try to come up with.
16:26:03 I don't know, to two wise positive negative, or it's too silly to simple yeah this is just like the thing that it doesn't, it would be if we would like to have, which is still my believe that we should be able to do it and it might be interesting to have
16:26:22 the epsilon view.
16:26:26 But look, we have this if you open that spreadsheet, you have two pages one is called full list and the other is called final, final is the revised version, right, that it has new Greg's right.
16:26:40 Yeah, but like the problem was that actually as we were, we were saying, the last time, and it was the whole thing that I should have should have considered whether those because I have the epsilon codes contain codes, that's the problem the epsilon,
16:27:00 were excess in the,
16:27:04 in the old methods. So I was thinking about how to abstract, a bit more to like a higher level.
16:27:14 In order to encapsulate both the epsilon and the excess in the current final view and divided into some categorization of new epsilon, with no annotations that would make sense.
16:27:27 But to be honest, I struggled to find a consistent higher level categorization of things like hate and laugh.
16:27:39 Yeah. And at the same time.
16:27:41 I think they're quite operable operational I mean they're they're easy to utilize in my opinion because the right now is 28 codes I think I could reduce it down to like 25.
16:27:52 I do believe that that's a solid code book for according emotions and they would still work individually reasonably and we will not have to record the hundreds of annotations because that's, that was the task that we said we would do, right, because I
16:28:04 said, Yeah, I had a problem that my episodes contain annotations so we said okay let's put it all down to x and create new episodes but I simply just, you know, I signed up my mind can't generate in between level that would be like feeling good and feeling
16:28:31 bad and all the things just split along that line. I am convinced.
16:28:34 It's like, don't see any problem with having no regrets and to whatever 25 to 28, x is.
16:28:41 Since this is done and it's just like we've spent so much time thinking about emotions and, you know, this is a pretty exhaustive list.
16:28:50 It catches all the emotions that played a role in theorizing about right wing populist.
16:28:58 As we know, we've been so I suggest we accept you because suggestion.
16:29:08 Okay, so I will just identified it nowadays, the but the emotions that could be like divided on positive negative definitely they could be like emotions divided based on if they are kind of like light, or like strong like strong strong reaction or then
16:29:25 they are, I think the type ology is that they're like reactive that it's an reaction on something, or it is a state of kind of mind.
16:29:35 And then maybe they're like relational like love and hate are kind of like relational then state is I feel bad I feel good or something or I'm stressed, maybe or stress know, and then their reaction or that they're like reaction on something.
16:29:53 So like laugh like something is for example it's either reactive or relational it's reactive I think, so maybe there could be a table and you know there's like some reactions something, or just like state of the people that they have this emotion.
16:30:12 Like, they're more like stable in a in a in a time. That's a good point but that's a good point I'm afraid from the theoretical point of view when you go down to the annotations, they turn to be more eclectic, you know that they might contain text which
16:30:26 is both reacted reactive and both.
16:30:31 Somehow I, there was a life feeling of the people, and again from the analytical interpretive state. Well, the point of view, I it's hard sometimes to assume what is a reaction to something that might be like you know missing out of the texts are missing
16:30:48 out of the context of the little piece of things I don't know, that's a that's a good tip. On the other hand, if I would think about the list that I see right now.
16:30:58 I kind of.
16:31:00 Yeah, it's, it might be the right techniques to run. Yeah, I have to go to another meeting, but here's, here's what I think very quickly.
16:31:15 The point raised by Zdenek is of course very important and we could come up with, you know, as you suggested very several possibilities of creating that.
16:31:27 But think about it from the point of view of what it will, how would it help.
16:31:33 It would help at all in our
16:31:51 study of right wing populism right to know that some emotion is as you call it.
16:31:46 Reactive or relational versus not relational.
16:31:54 Yeah, I don't know, it may be right but at this point, it seems to me that the most important thing. I mean given you know we've all read some of this material on other people studying emotions and populism.
16:32:08 Basically this is the discussion, using the words, anger, fear, hate, empathy, constantly people talk in this literature about those emotions so naming them.
16:32:22 Precisely and and having the analysis at that level, which is consistent with the vocabulary with what other people are saying about populism, it will allow us very quickly to contribute something to this debate.
16:32:39 You say, you know, that anger, versus, whatever the debate is an anger versus fear for example.
16:32:46 Okay, here's contextually. Here's the context within which people talk about anger versus fear.
16:32:56 If I made it's precisely, just quickly that's that's my point there when I was thinking about it again, I imagined the epsilon structure, and you see the epsilon codes which will be selecting the right to me when you can.
16:33:09 When we can talk. Okay. All right. Cheers guys. So, I like imagine the network where you see social problem. And next to it a relational emotional response, you know that doesn't say much and then again like social problems and that negative emotion,
16:33:26 definitely. But that doesn't tell us nothing because we know that social problems produce negative emotions right. We are more interested in the nature of the whole thing.
16:33:37 So the second the second thing that might be done is actually taking emotions as a special category and just create void epsilon, with the same labels above it just for the sake of the visualization and seeing it on both levels right that they will be
16:33:57 kind of like universal operating, both on epsilon and RX explains. Very good point. very good beer so we could do that easily.
16:34:02 Just copy paste right and.
16:34:05 Or I could know, whatever, that's sort of hold you back. This is not that important, is just like a, but you know it's great that that you sort of thought this through and realize those things.
16:34:18 It's. Yeah, let's think about it a little bit more and we'll make a decision but right now. Let's go with this idea that we have access because creating excellence, that's very easy task.
16:34:30 That will take no time. Should we make some decision. The key thing is that we capture things the way that is sort of broadly makes sense to us.
16:34:51 No, make sense, or they can be like the episode can be also emotions.
16:35:09 Emotions its own emotions and then there will be these emotions. So you know, to which topic emotion is attached any emotion and yeah and then I hope to which topics no emotions are attached Yeah, yeah, individualization at the level of a wise.