 So what's next for American Fact Finder? American Fact Finder, I think most people have had issues with it. So I think this is perhaps good news for most people that it's going to just be going away entirely. So I'm going to talk today about the overview of what all is happening and the timeline for that migration to the new platform. The content that is included in the preview version as you're calling the data test and the limitations, some of the limitations that it has. I'll show you the new interface and explain how the Census Bureau is asking for feedback in terms of the questions they would like you to answer and the address to send it to. And then at the end I'll do some example searches. The content today is based on a webinar the Census Bureau did back in December and also on my own testing. I would say that my own testing is not nearly comprehensive and in the webinar the Census Bureau actually restrained from doing any example searches because they didn't want to prejudice people's experience with it. So I would encourage you to come up with your own examples. There is a link at the end of my presentation to that Census Bureau webinar. So if you want to see it from the horses mouth you will see that I copied a lot of information from their content. They also said that there would be another webinar in March but that did not materialize. So this is the most current information that seems to be publicly available. I aim to have a 30-minute presentation. So we should have plenty of time for sample searches, questions. Go ahead and chat in questions as they come up for you. Because I'm in presenter mode I won't see the chat but Linda will prompt me when there are questions. Also use the chat to send any searches you would like to see do. There are some slides about the specific pieces of data that are included in the preview. So there might be some searches that I can't actually do that won't actually work in this version. But go ahead and send in any examples that you would like to see and we can at least try them. All right, so all of the Census Bureau's data products are intended to move into this new interface. And it really is time for a new interface. All of the Census Bureau's tools really have at least a few years of exposure. The most recent one, the API debuted in 2012. And so it's really time for a reworking of this. The unit within the Census Bureau that's leading the development is SEDCI as they're pronouncing the acronym. The Center for Enterprise Extermination Services. I think it's easy to forget how many tools the Census Bureau makes available. Right now there are 37 tools in all. So this is really a huge effort to get all of this in. And so as you can imagine, it's a multi-year effort. So this is directly from the Census Bureau's presentation. I think it requires a bit of explanation to understand what this all means. They do nicely have this legend at the bottom. But I think the thing that's probably easiest to miss is there are these red. So these red platforms are intended to represent the technology platform on which each tool is built. And as you see where we start out, each tool has its own platform. So this is really kind of a massive, not only time sink, but just resource sink in terms of having to maintain all these different platforms. Each one has its own funding. Each one had its own developments. So efficiency is a big reason for this transfer. So as you can see, here we are in actually stage two, labeled Begin Capability Transitions. All of last year was spent on creating this preview environment at data.census.gov, which is sort of the test environment for how all of this is going to work. Once they had the platform stable, then this year they began moving American FactFinder into this platform. And part of what that involves is loading the data into data.census.gov, and then encouraging people to test it and see where they run into problems. The other tools are beginning to be moved in. So you see Data Share It here and Quick Finder, Quick Fax. Both of those are on the move. Their platforms are going away, and their data are advancing into data.census.gov. It's not all there yet. When I looked at Quick Fax this week, it looked to me like it's still using the original interface. I haven't been historically a big user of Quick Fax, so it might be mistaken about that. If any of you use that a good bit, you may see differences there. But it did look to me like if you select Quick Fax in the data.census.gov interface, it sort of exits out to a Quick Fax interface. So one thing that's important to remember here is that this is a projected timeframe, and these dates could easily change. I would even say you should pretty much expect them to change. So then over time, you see that once we get to stage three, all of those legacy platforms will have gone away, and all of these tools will have moved into the new platform. Some of them will still probably have legacy development, maintenance, and funding aspects, so they're not completely within data.census.gov, but they're moving in that direction. This was something that I didn't clue into for a while. It's not actually going to continue to be called data.census.gov. It's actually going to become the Census Bureau's homepage, just census.gov. So that should happen in the next two years. I think I also forgot to point out data ferret moving in is again why we didn't do that August webinar on data ferret. So there you are. So then finally by 2022, stage four, they're hoping to finally be in full production mode, everything operating on the same platform, and all of those legacy things going away, not just the platforms, but the development, maintenance, and funding aspects as well. And there was one more question just to clarify the name of the tool, or is it just searching Census.gov? Yeah, so the preview version, they're calling data.census.gov, and they're expecting sometime in the next couple of years for that to change just to Census.gov to be the Census Bureau's main homepage. A couple of things that came up as questions in the Census Bureau's webinar, microdata. People were asking if microdata would be available in the tool, and they were saying that it is on the table, but that's further down the road. So it may be that when I talked to the head of the data ferret office, probably this was a couple of months ago now, he thought that within six months data ferret would be in this tool. But it sounded to me from the webinar like they might not be expecting it in the next six months. They might be expecting it like next year sometime. So all of the timeline is pretty much fluid. Somebody asked about access to historical data. I know that with academic researchers in particular, people really want access to all of the years of the Census. My answer to that has always been that's not the Census Bureau's prime objective. Their mission is to collect new data. And so I really don't expect that they're ever going to make all of the historical data available. But they know that people want it, and so I would expect that there would be some historical data available in the final product. It might just be further down the road in terms of once they've tested out the interface ways to do that, that they would get more of the historical data in there. Someone asked about the ability to add their own data so that they could combine it with Census data. And the Census Bureau presenter talked about the fact that it is possible with Census Business Builder and on that now for people to upload their own shapefiles to combine with the data that the Census Bureau presents in those tools. They know that people want other kinds of uploading of their own data to be possible, but that is to be determined. They haven't really gotten to that point yet. They kind of have to deal with their own data first. I'm not going to talk very much today about the technical back end of this. It's both my perception that that is sort of not the main interest of most of the people on this kind of webinar, but also it's not my expertise at all. So all I will say is that they mentioned in the webinar that the new platform is based on their API. And the result of that is that it both makes it easier for the Census Bureau to provide service and deliver data, but it also makes it easier for users because the current American Fact Finder display requires table formatting in order to display the data. And that locks down the data in very specific ways so that it's difficult to export it from the system and use it then. And with the new system being based on the API, they're no longer going to be tied to that kind of table formatting. So the public will have the same access that developers inside the Census Bureau have. So that's good news. Okay, so what's in the test system? You will see that out of 11 elements here, six of them are related to American Community Survey data. So that's really where they're focusing the development in terms of the American Fact Finder is, I would say, primarily used to access the American Community Survey. And so those are the data that they're testing for that to be the first migrated tool. They do have this nice big warning on the bottom. Everything in here is test data and should only be used for purposes of evaluating functionality as you have them. You will see in the next couple of slides some of those limitations and why the data should only be used for evaluating the platform. So as we go on, here are known limitations. First of all, there's no tract level access in this interface yet. They are developing that and some of their tools have mapping capabilities. So as they develop the geospatial tools, then they will also make the tract level geographies available. Again, data fair at cross-tabbing and further down the road. And someone asked in the Census Bureau webinar about whether it would be possible to directly access the Census Bureau data for services like ArcGIS online. The Census Bureau isn't paying attention to that right now, but they know that people want it so it may come later on. And then there are these known limitations. I'll particularly point out pseudo geographies. They actually have false geographies in there just to test the interface. And as you saw from the data content, it's very limited data sets that are available there. So there are going to be lots of queries that people try to do that they're not going to get any answer for because that's not what the system is intended to do. For now, until we hear otherwise, American FactFinder is still available and will still be available until they completely transition it over. And there will be lots of warning about that. So as we turn to look at the new interface itself, think about how you search. Think about how you like to search for things. We're not talking about really American FactFinder at this point because in those cases you don't have all these options in American FactFinder. But essentially what's driving the development of this new platform is lots and lots and lots of new users who've never had experience with the Census Bureau before coming in and finding it difficult to search for what they want. And so they've developed this interface specifically in the answer to those questions. The Census Bureau pretty much drives all of their development based on where they're getting the majority of their questions. And so given that new users are where they're getting the most of their questions, that's what's driving this interface. I recommend you think consciously about how you like to search and try that, try the way you regularly would search, but also try to mix it up. Try other ways of searching if you want to get a feel for how your users might approach this. This webinar series, the Help webinar series is oriented towards librarians, so I do assume that everybody here pretty much is a librarian. I really like using library patron searches, actual things that people have asked for as examples to try to do better searches and understand where people might run into problems. So this is it. This is the interface. And this is what you're going to see when you come to both the preview, data.census.gov, but then eventually, just as the Census Bureau's website, this is what it's going to look like. So it really is, as some people have described it, an approach to looking like Google. They want it to be super simple. There is an advanced search here, but it's kind of hidden. If you click on the search box, you get this option for the advanced search. And then when you click on that, you get to a screen that looks like this. To zoom it in a little bit, you see that at the top is this sort of exposed Boolean search option screen. So people can choose. They want all of these words, this exact word or phrase, at least one of these words, that kind of thing. And then down below that, you get sort of an access to the hierarchies of the Census Bureau. So if people want to look at geographies over years or particular surveys or particular industries, this is where you would access that. So my feeling in having used this a little bit, I will still say a little bit, even though I've been playing around with it for a few months, is that I think this is going to be challenging for new users. The Census Bureau has really focused on making access easier. You don't have to know as much to just throw in a search and find something, but new users are still going to have a great deal of difficulty evaluating what they find. And depending on how they do the search, they might not find something and not realize that something was available, that if they had just used different terms, they would have found more. They're also not going to be able to differentiate very easily between different kinds of information. So if they were looking at search results for sex age and they got a table from the American Community Survey and another table from the Decennial Census, they're not going to understand the difference between 100% data in the Census and calculated estimates from the American Community Survey. This question was asked in the Census Bureau webinar and the presenters answered, well, they're going to be able to tell the difference between those two because the estimates are going to have margins of error and the 100% data aren't. And the sources are going to tell them where the data comes from. And my reaction to that is, well, that's not enough information to really explain to them what the difference is. So I know because this question was asked in the webinar and because I have already sent feedback to them about this issue that they're going to get a lot of feedback. But as I say, the development of their tools is always based on where they get the most questions or where they get the most pushback about something. So I really encourage all of you to submit feedback about that kind of issue where you see ways that the Census Bureau could make it better. So we'll talk about that a little bit more. Another place where new users would have trouble, I actually did search, this is one of the examples for this webinar, a search for race income. And you don't find any tables for that because all of those tables that cross-tabulate race and income actually are for the single race alone categories. So they don't actually have the word race in the table title. So you have to search white income, black income, Native American income in order to find those tables. And then because the hierarchical lists are buried in the advanced search, I think it's going to be very hard for new users to find those. And so they're not going to automatically be exposed to complete lists of geographies or complete lists of topics. And that's going to make it very hard to explore and really discover some of the unusual tables that are actually available. So feedback now is critical because they're doing the development now. The Census Bureau does have a good record of doing development on a continual basis. Just as an example, this Center for Enterprise Dissemination Services and Consumer Innovation, SESI, has been around for quite a while. It's been operating for at least seven years, and I think longer than that, in trying to find a platform that would really work as not an alternative, but a replacement for American Factfinder. So these are the questions that the Census Bureau would like you to answer when you send them feedback. What was it that you actually searched for? What terms did you use? How did you construct it? Did you use and in the middle? Did you use quotation marks? And then what did you find? Was it what you expected to find? Did the search meet your expectation? And if not, how could they improve it? You can send lots of feedback. I would say a ton of feedback, maybe fewer than 10 emails so far, but I'm expecting to send more. One thing you need to be aware of, though, this is not something that you're going to get an answer from. If you actually have a question, then you need to contact the Census Bureau directly to actually get an answer, but they will take all kinds of feedback, whatever comments and suggestions you have. And again, please do send reports or requests, even if you know someone else has already reported or requested something, because they base their work on the number of requests they get about something, the number of questions they get about something. So the more you send, the more likely it is that they will respond to that. I have also pointed out navigation issues. I found it really difficult to find the links for the technical documentation. They are there when you get down to a specific table, but I felt like they were really hard to see and that not many people were going to find them, even if they were looking for them. I think it's hard to find them. The system does the thing where you do a search and it immediately presents you with a highlighted table, a kind of featured table. And the things that they suggest in my experience almost never super relevant to the thing that you actually searched. So I had sent them feedback about the relevance being an issue there. And then featured requests or suggestions. I found that when I searched for a town, it does give me a list of suggested other towns. So I found out that there's a Chapel Hill, Tennessee. I did not know that, but I didn't want Chapel Hill, Tennessee, but that's what it was giving me. And so I clicked on Chapel Hill, North Carolina to switch to that, but even though it's hyperlink, it doesn't return search results when I click on the link. All it does is repopulate the search box with the new geography. I didn't find that very helpful. And then finally, I think I already told you that example about adding race metadata for all the tables so that if you search for race income, you actually find something instead of having to search for white income or black income. Okay, so we're up to the test searches. Michelle, I had a couple of questions you want to wait. Great, yeah, go ahead. One is, is there a date by which we should send feedback? Nope, keep sending it. And then, but agencies looking to preserve historical data collections that might drop off the census site? If you could talk about that maybe. There's actually a lot of places to access historical data. So the first search that I'm going to do is just for race and income, so you can see this. So one of the things that I have reported and complained about is that this is really slow. I am expecting that that is partly because it is a test system, but given that the test system has less data in it than the real system will, I'm a little concerned about this. All right, so this is the results that I get in Chrome. That's actually another point for just your information. When the Census Bureau does development in the past, they have been extremely limited in the number of browsers that they are allowed to do development for. For a very long time, the Census Bureau was only allowed to do development for Internet Explorer. And this extended past 2010. I mean, this is not an old-time thing. This is really pretty recent. So I'm hoping that with the new platform, that kind of restriction will be done away with altogether. So it working in Chrome gives me hope. So you see here I did search for race and income, and I got a race table, which is good, is only about race, not about income. But I think that this UALL Tables link down here is pretty hard to see. It's pretty obvious that you could view the full table here, but if you want more tables, you really have to look for that link. It's also pretty settled to me that you can choose between tables, maps, and web pages here. But the Census Bureau is really expecting that this is going to replace their website. And so it's not just the data that are in here. It's also the maps and web pages. So eventually, when you search for American Housing Survey, you'll be able to get to both the web pages and the tables for that as well. That's not even on the chart in terms of, you know, the timeline chart in terms of when specifically that survey will get transferred over. But that's the intention long term. So if we go to view all tables, now you can see that particular table. You can filter the results if you want to add other terms. As I say, you're just not going to get to race and income tables because the income tables don't have race in their title. But you can browse down through all the other tables that are available. There's another question. I have a question. When will the Census tools be disabled? Is there a protected timeline? There is no specific date type of that. The timeline chart that I showed earlier is the only indication so far of what they're aiming for. And like I say, the legacy tools are going to continue to be available until the new platform is absolutely ready to go. So you don't need to be worried about not being able to get to the old interfaces. I would imagine that there will be a lot of buzz around the legacy tools being taken away. And even before that, a lot of buzz around, hey, data.census.gov is ready to go. Generally, my memory, when they premiered the new American Fact Finder interface, I think that was in 2011, there was at least a six-month period when both interfaces were available so that you could take some time to explore the new one before you actually had to start using it. And I would expect the same thing here. But they haven't published any dates any more specifically than that close to timeline five that I showed. So then that means that the 2020 Census will still be available in the legacy tools. Yes, I would expect that it will be. Yeah. It's hard to say. The 2020 data won't be out until 2021. And by that time, they are expecting American Fact Finder to be completely available, completely transitioned over to the new platform. So they might well decide not to put 2020 data into the old interface. And I wouldn't be surprised if that's part of what's driving the development timeline. They would rather not put it two places. But it depends on how long it takes. If the new interface, if the new platform isn't ready, then they're just going to have to. It's not like they can avoid releasing the data. There's a legal requirement that the census data be released within one year of taking the census. So if I change to black and income, again, it gives me a race table. Okay, I'm going to go back again and change to the advanced search. I don't know if I said this already. I do encourage you to do searches in both the basic and the advanced interfaces, the same search, because you learn different things about the interface by doing in both. I have quite different feedback for the Census Bureau about the advanced search than I do about the basic search. So I think it's a useful exercise. Here we go. I'm putting, I do want all the words, but I want both of the words up there. You could add something here if you wanted to be more specific. You can also go to geography to your surveys or industries to add additional pieces. I found this bit of the interface to be a little confusing because you have this enter text filter. I thought that you had to search in order to add a geography and it doesn't work that way. If you search for North Carolina here, what you get is the congressional districts, and that's the only option that it offers you. Instead, you have to go through this hierarchy and wait. Eventually you will get a list of all of the states and not only the states, but also the District of Columbia and all of the territories so that you can select from those lower level geographies instead of having to search for them. Because like I say, when you search for North Carolina and you don't get an option to select just North Carolina, all you get is the congressional districts. Well, I'm not sure how well it's going to work for us to do any more searching. In my slides, I do have some other example searches listed, so you can look at those and try those searches on your own and see what happens. For instance, I searched for the word rice and I discovered to my surprise that there are actually a lot of communities in the United States called rice or that have rice in their name. But there's not very much economic data in this new system. I got better results searching for NAICS codes for specific things, for specific industries than I did in getting anything with words describing commodities. So that's one thing. I searched for Education North Carolina and I expected to get a variety of different tables. You know, school enrollment, educational attainment, sex by age by educational attainment. But the only thing that I got was a single school enrollment table. So I think that again is about metadata that's attached to those tables. So those were the three examples that I had to show you. They are adding data to the system as they go along. So I would expect that more economic census data will show up in there as they get around to moving census business builder into the system. Those are probably just further down the road because that's not what we're focused on transitioning right now. Do folks have other questions? We could answer any questions that you have or I could turn to the question about historical census data access. I'm not seeing any private ones. So in terms of census data access, I will mention this commercial product even though it is subscription based. You have to pay for it. Some people may have seen social explorer before for a long time it was developed by a university. And so during that time it was freely available, but it has it then quickly moved into a subscriber model. And it's now owned by I believe Oxford University Press. So it's not cheap. I don't offhand know what we're paying for it. But we actually have a subscription through our statewide NCOs collaboration of all public educational, well educational covers abroad. It's pretty much true, but it's both school and institutions of higher education as well as public libraries that pay for it. Our price wouldn't be helpful to you anyway if you're looking to subscribe to this as an individual institution. Social Explorer has all of the census data going back to 1790. That's a little misleading. They don't like situations where the data get complicated. And so there are pieces of the census data that are not available here. But it's really easy to use and it combines the mapping interface along with the tables. And it makes it really easy to export images of the maps. So it's super easy to use with undergraduates and people who aren't very knowledgeable about data. The irony here is that it basically uses American FactFinders original interface as a way to dig into the data. So you first choose the survey that you're working with. And then if we're going to go with the most current American community survey data, you begin a report. You drill down to the specific geography that you want and then choose from list either just one or use shift or control to choose multiples. Add them to your selections and then proceed to tables. This first selection box you get is all of the tables. And Social Explorer presents its own version of the tables in a lot of cases where they do things like calculate percentages for you, which the Census Bureau sort of notoriously does not do. You can switch over to the Census Bureau tables if you prefer. What I find most helpful is to search by keyword because that list of all the tables is really long. And then this is a much smaller list still long because there are so many tables without income that you can choose whatever one you are particularly interested in. I believe, yeah, it keeps the original labels for the Census Bureau's version. So if you're used to seeing the base and compressed tables from the American Community Survey, those will be familiar to you. The ones that are marked just T are Social Explorers tables. And then you again have to add show results. So here's your table. If you want to export it, you can export it easily to Excel. It lets you choose whether you want to have percentages or aggregates or not. And then you can just download it. If you need to use it in a statistical software, you can use the data download and do tag the limited files for data or it will give you set up files for SAS or SDSS. So you can also choose to include the geographic identifiers if people are doing geospatial data. That becomes important. So it's pretty easy to use. Social Explorers has added and authorized data in recent years. If you go to the tables list, you can see the long list. Environmental summaries is new. U.S. religion data from the RCMS is there. Crime data from the Uniform Crime Reports. Health data, cancer data, those are generally county level. Not anything smaller than that. So that reminds me to tell you that the geographies are based on the particular data set. So you'll get quite different things if you're looking at U.S. cancer data versus the UK census or the American Community Survey. So there are differences there among the different data sets that are available. I remember correctly the dictionary links are what will get you to the technical documentation. I still almost never use their technical documentation for Census Bureau products. I will always go directly to the Census Bureau itself, but there is some technical documentation here. Another product that actually was the basis for Social Explorer is something called NHGIS. This is the National Historical GIS System from the Minnesota Population Center. The Minnesota Population Center operates on grant funding. And so they make all of these products available for free. They collect statistics about them so they do ask you to register before you use them. But they are substantially less intuitive than Social Explorer. So I would never send an undergrad to use this product just right off of that. I would always sit down with somebody and give them a careful introduction before setting them loose from this. Even doctoral students or faculty. I think those are the two major ones that I would recommend for historical data. NHGIS is the first geocoding of all of the Census boundaries historically. This also has the Census Data Facts 1790. It's not complete yet, but they are working steadily to add as much as they can get their hands on. So I would expect that eventually this will be a complete collection of all of the Census data. We're pretty much at time. If anybody has remaining questions, my contact information is on the last slide in my presentation. And then we'll put those up on the website, along with a link to the recording. So do feel free to contact me if you have questions or if you just want to get in touch about your use of Census data. I would welcome that. Great. Thank you very much, Michelle. This is great. A little scary. Definitely useful. Many of my co-workers. If there are any final questions, feel free to chat with us. We can't see you typing, so just go ahead and chat and get them in there. Thank you, everybody, for coming, though. And thank you again, Michelle. We'll definitely have you come back and reassure us later on. Sure. Happy to do it.