Why is there no back button?

• by lukesmith moderator

  This has come up a few times today. We need to 'close' the page once you have submitted, so that we know when to stop presenting it for tagging. We then to match up each person's tags statistically. If you want to check something in a page you have already tagged, you can view your recently tagged pages on your PROFILE at:
  http://www.operationwardiary.org/#/profile

  p.s. Don't worry if you missed something or made a mistake. We use a statistical matching process to weed them out. Any pages where 'agreement' has not been reached will be put forward for tagging again at a later stage.

  Many thanks.

  Posted January 14, 2014 4:10 PM

• by richardlight

  That's clever, but it doesn't let you correct mistakes you subsequently realise you have made.

  Posted January 14, 2014 4:52 PM

• by richardlight

  Also, if you just click on "Profile" from within the Talk context, you go to http://talk.operationwardiary.org/#/profile, which /doesn't/ give you this information. So there will lots of puzzled people trying to follow your advice from the Talk context and failing. Bug?

  Posted January 14, 2014 4:58 PM

• by BizzieLizzie

  How do you get back into a page that is not finished? I hit something and was moved off the page, when I went back it took me to the next page - how do we get back to previous pages to check, or correct something you later realise was wrong.

  Posted January 14, 2014 6:03 PM

• by HeatherC moderator

  You can't go back once you move off the page for any reason it appears. Frustrating but we seem to be stuck with it.

  Posted January 14, 2014 6:21 PM

• by DannySpud

  I had a page where the King was visiting the troops. I clicked to go to the talk page to see if there was anything about how exactly to tag the King and now I can't get back to it. Incredibly frustrating.

  Posted January 14, 2014 7:08 PM

• by ianscrim

  Frustrating the inability to correct previous pages! As one proceeds through you start to understand the terminology and also the place names for an area, an edit facility would be useful, Thanks

  Posted January 15, 2014 12:43 AM

• by phillowes

  We need the ability to go back and add more tags. I know I have moved on from pages that I could have added more to 😦

  Posted January 15, 2014 1:16 PM

• by lukesmith moderator

  @richardlight My apologies, there are indeed 2 profiles. Talk and the classifier itself. I should have pointed you to the later profile :
  http://www.operationwardiary.org/#/profile

  Thanks

  Posted January 15, 2014 1:45 PM

• by lukesmith moderator

  p.s. Don't worry if you missed something or made a mistake. We use a statistical matching process to weed them out. Any pages where 'agreement' has not been reached will be put forward for tagging again at a later stage.

  Posted January 15, 2014 1:48 PM

• by cpe

  I find that the responses from the moderators here entirely miss the point that is being made (see my comments in another thread). Both the point about a back button and correcting mistakes are simply elements of the poor intrinsic structure of this process. The assumption is being made that tagging on a page is a one time, one person activity. I deduce - though it does not appear to be said anywhere - that at some stage after one person has finished with a diary it is opened up for others; only when several people have tagged can the "statistical matching" have any meaning. We need a clear explanation of whether and how it is possible to arrange multiple visits to a page, either by the same individual or several individuals. If we have missed the instructions then make them clearer.

  Posted January 15, 2014 2:12 PM

• by BizzieLizzie

  The response to the 'no back button' ,misses the point. Why put a page 'forward for tagging again at a later stage' when the problem could be dealt with immediately, as we gain more experience with the handwriting etc we may realise mistakes have been made or information missed. The system should allow for a good job to be done at the outset, I have only just started this but am frustrated already. There is also the question of not enough options for the tags which have been highlighted, is this going to be addressed or is 'other' going to be the most used option????

  Posted January 15, 2014 4:42 PM

• by hodgsonp

  I agree with cpe. Yes the system is all very clever (too clever?) but please try to see it from the point of view of the volunteers doing the work. One of the reasons we do this is that we would like to get some satisfaction from the process too! Its very frustrating not being able to go back and correct mistakes you know you have made or to check something. I take pride in what I do. I want to do the best job possible and I also want to follow the diary of the unit I have picked. To come back and find, after having a cup of coffee, that the next page is now jumped 6 pages further on (presumably because someone else has tagged the pages?) and that I have lost the thread and context of what was happening, to find the unit moved miles from where they were and having to effectively start again (and know it will happen again) is really,really, really maddening!

  Posted January 15, 2014 4:48 PM

• by thedisraeli

  I just popped into a diary page that was continuing from the pages before, so the place, for example, was not re-mentioned just as "the village" or something like that. I had no chance to look up the preceding page so I could not tag the place. Is there a chance to look up certain pages of a diary (maybe with tagging function deactivated?)

  Thanks

  Posted January 15, 2014 5:02 PM

• by beadlady

  Is there any good reason why this was designed without a "BackButton"

  Posted January 15, 2014 5:08 PM

• by ttfnrob admin

  Hi - I'm Rob, one the Zooniverse developers and I have experience with data reduction on many of our projects. The use of a back button presents a tricky situation for the data analysis that occurs after you have finished each page. If we allow volunteers to 'edit' their transcriptions then:

  1. We have created a sort of Wiki, which is not the idea. We want structure, independent transcription of these documents.
  2. We never know when a page is finished - and perhaps they never will be.

  We ask multiple people to look at each page and when they have been seen 5 times we use the collective tags, and correspondence between them, to create a final set of tags from the group. Now that 5 might turn out to be 3 or even 10 over time - but once it's reached, the page is done and we remove that page from circulation (but not from Talk). Editing pages means that we no longer know when they are finished. We have 1.5 million to get through so we need to keep things moving. However, pages that produce uncertain results could re-enter the system later.

  All of that being said - I can see that it would be extremely useful to be able to refer back the pages you have recently tagged (e.g. the last 3?). That is a feature I believe we may be able to implement, and I will look in to seeing if it can be done within the licensing agreements Zooniverse have with IWM and the National Archives.

  Hopefully that helps?

  Posted January 15, 2014 5:37 PM

• by cpe

  Rob

  I do not agree with the points you are making and they do not seem to have specifically answered the comment I made earlier. In particular it is not clear how the "multiple people" approach is being implemented. As others have said it appears that there is no lock on a diary for any individual who starts tagging. If they stop and restart then they get back in at an arbitrary place after someone else's pages. Can you please explain how the multiple people approach is implemented and managed.

  As far as I can see any tagger can "finish" a page with an arbitrary amount of information, perhaps no more than a classification. Is that right?

  Posted January 15, 2014 5:53 PM

• by BarbPetagna

  I went into my profile page to see how I was doing, and a lot of my tags don't seem to have registered. It's showing that I've tagged no dates at all, for example, when I know I've tagged several. Also, when I click on the timeline button, the dropdown is always empty, even though I've entered tags on the page.

  Posted January 15, 2014 8:38 PM

• by ttfnrob admin

  cpe,

  I'm not sure what you don't agree with, as I was only trying to state the facts of the matter. We (the Zooniverse) run a great many projects like this and thus is just how they work. Though I did say that think the idea of being able to return to a diary (even if not to mark on it) would be useful, and I'll see what can be done.

  Diary pages that have not been seen by 5 people are assigned sequentially to anyone that choses to follow a specific diary, or they are randomly given to people that do not pick a specific diary. There is no 'lock' of any sort, and multiple people can tag the same page at once without knowing it. Once a page has been tagged by 5 different people it will no longer be displayed unless the subsequent analysis finds a problem.

  The analysis is be performed offline, by researchers working at the Zooniverse, IWM and National Archives. We are starting with a repeat of 5, and if this doesn't appear to be enough then we'll increase it and many of the pages will become available again.

  Rob

  Posted January 15, 2014 8:38 PM

• by ttfnrob admin in response to BarbPetagna's comment.

  This is a bug and I will look into it. Can you let me know what browser and operating system you are using?

  Posted January 15, 2014 8:40 PM

• by KJL1911

  I was half way through classifying a long list of the wounded... suddenly the page disappeared and a note came on the screen thanking me for finishing that particular section.
  So disappointing as I had followed this particular regiment right through - so wanted to finish it 😦

  Posted January 15, 2014 8:55 PM

• by Andrew_Stevenson in response to DannySpud's comment.

  On a windows computer if you press the CTRL key while clicking a link, it will open the link in a new browser tab so you can check discussions, while the page you're working on remains open.

  Posted January 15, 2014 10:07 PM

• by JudithLlangarron

  I'm sure this has come up somewhere else, but ... I have been presented with a new page on my visit today, but although the date is noted there, by the original author, the month and year is not, and I cannot remember from yesterday what they were. HELP, do I just move on or what?

  Posted January 16, 2014 2:41 PM

• by Mike100

  It is frustrating there is no back button!

  Posted January 16, 2014 3:11 PM

• by cpe in response to ttfnrob's comment.

  Rob

  Thanks for responding specifically to my comments. Just to be clear I was disagreeing with the process as described rather than you! However the phrase "thus is just how they work" is a little worrying in a project seeking to attract volunteers and asking for improvements to the process.

  Apologies for the relative length of what follows.
  The process
  From what you have said here and others elsewhere I assume that the description of the process is as follows, with queries where I remain unclear:

  When we "start a diary" we will be assigned a start page (or is it a range of pages?). We can then proceed page by page.

  We can tag an arbitrary range of information before finishing each page. The minimum is simply to classify the page. (Is that right? I can certainly do so on ones I have tried)

  Neither we (by going back) nor other people can see the tags we have applied to any page. Although the pages can be seen in our profile the tags are not revealed and the subheading "tags" in that page means something different.

  On some occasions when we "start a diary" we will be assigned pages that have already been tagged by others but we will not see the tags they have applied.

  Once five people (or whatever number you select) have processed a page it will be declared finished. Do you apply any further test for example if there is only minimal content in all five sets of tags?

  If this is approximately right then I think it would be very helpful if it were explained more clearly somewhere: it does not come across in the tutorial for example.

  The implementation

  Whether we think that is the right process or not there can then be some discussion about the implementation. Hence this debate about the back button. I think that there are clearly improvements. If the absence of review or edit is intrinsic to the statistical integrity of the process then at least one option would be to lock the data but allow commenting to highlight mistakes or alternatives, at least for our own entries. This would be similar to the approach on the main TNA site about correcting existing index information.

  Similarly most tags do not need to be located on the page image unless there is some expectation that the eventual unit of viewing will be part pages. They could equally well be on on parallel page (akin to the talk element of a wiki but controlled rather than free form). This would make for much easier handling.

  The process

  I do have some views about whether this statistical approach is right. There is potentially highly structured data within these document; the project appears to be creating a cross between a transcription and an index. One of the common experiences of transcriptions in genealogical work is that the focus is too often on text and too little on creating structure that others can then manipulate or interrogate. So I support the aim of imposing structure but I would expect the creation of a visible index with challengeable corrections would have been at least as good. However this is presumably not up for debate now so I will retreat quietly.

  Conclusion

  More explanation will help a lot, including how this approach will enable an innovative way of analysing the war.

  CPE

  Posted January 16, 2014 6:49 PM

• by lukesmith moderator in response to cpe's comment.

  Hi @CPE,

  Thank you for your thoughful post. A few quick comments, in advance of more considered responses:

  • I agree that we need to do a bit more to explain the method of creating an agreed set of data for each page. We will do this -- please bear with us.
  • If there is not good agreement between the classifications and tags for a page, then it will be reviewed manually or resubmitted for tagging.
  • I agree that we could do a bit more to show you what you have already done -- the 'back button' question. We are considering this along with other enhancements.
  • This is an innovative approach that, we believe gets us through the mammoth task more efficiently.
  • You are right, this is about getting structured data, rather than text, hence the tagging approach.
  • In terms of innovation, we will be able to use the 'big data' to gain a better view on how much time was spent doing the various activities. There is stil considerable debate about this among historians. Plotting unit movements on maps, over time, is likely to produce new knowledge and will certainly aid public understanding -- the front may have been fairly static, but the people certainly were not. By using the data in Lives of the First World War, we can find out the journeys and activities of individuals whose service records were destroyed -- of value to descendants. The intersection with Lives of the First World War will allow correlations between, for example, birthplace, age, casualty, locations, weather conditions, divisions, etc. New hypotheses can be formed by historians and tested against data -- a valuable new tool to complement existing case study based approaches.

  More on all these topics soon,

  many thanks,

  Luke.

  Posted January 16, 2014 7:38 PM

• by rogerlwhite in response to ttfnrob's comment.

  'We ask multiple people to look at each page and when they have been seen 5 times we use the collective tags...' Well that's interesting and as far as I can is information that has not been presented anywhere else on the site. It's a bit late that this comes out after you've gone live.

  You may have experience with 'data reduction on many of our projects' but that is a tecchie speaking and I don't have a clue what it means.

  I'm afraid that in the light of all the other comments made on many threads the statement that 'We (the Zooniverse) run a great many projects like this and thus (sic) is just how they work 'comes over as arrogant. Haven't you read those other threads?

  Posted January 16, 2014 9:09 PM

• by lukesmith moderator in response to rogerlwhite's comment.

  Hi,

  I agree that we need to do a bit more to explain the method of creating an agreed set of data for each page. We will do this -- please bear with us.

  This is a long term project -- we know that there are things that can be improved.

  many thanks.

  Posted January 16, 2014 9:33 PM

• by Milliest

  Are we supposed to add up all the wounded (or dead) on a page or tag them separately.

  Posted January 16, 2014 11:03 PM

• by Milliest

  Please can you provide a link to a list of abbreviations etc so we can understand this better.

  Posted January 16, 2014 11:15 PM

• by Cidermonkey

  When I last logged off The last date that I had was Sept 7th 1914 , when I logged on today the first date was Nov 5th ( year unknown ) , so I have lost at least 2 months , possibly years . Is there a way to go back ? The reason I ask is because I am very interested in the subject and the lesser known regiments and it seems unfair to ask people to give their time to help you if the reward of finding their own treasures from this project is denied them ,thank you

  Posted January 17, 2014 10:52 AM

• by Cidermonkey in response to ttfnrob's comment.

  It is not really just a question of being useful for you ,all of the people doing this are giving their time , and many , like me are hoping to find out about the lives of their relatives. I feel that it is unfair to let people put in the time and then , because they can't sit at their computer all day ,every day they miss relevant pages in the diaries.

  Posted January 21, 2014 8:30 PM

• by spof moderator

  The indexing being done here will link into the IWM project Lives of the First World War so the information should be readily available when that launches late next month.

  Posted January 21, 2014 8:45 PM

• by bonnypearl

  I have discovered that i have been misreading his signature and so giving the wrong name to the diary author - terrible! Now I can't go back to correct. Wish I could.

  Posted April 1, 2014 9:18 AM

• by looker1940

  The reasons given for lack of back button are very poor. There are very good reasons for needing to go back, not least that on some occasions everyone will have had the same problem so the information will never be complete. My example, I just had a page where the scrawled signature at the bottom of a report page was totally in-decipherable, the 2nd and final page of the rport had the same signature over a typed name, I could have gone back and completed the first page with a back button.

  Posted April 7, 2014 6:45 PM

• by lukesmith moderator

  Hi There,

  Thank you for your contribution to date.

  Operation War Diary has 3 aims:

  1.to provide evidence about the experience of named individuals in IWM’s Lives of the First World War project – you are contributing to their permanent digital memorial

  2.to present academics with large amounts of accurate data to help them gain a better understanding of how the war was fought

  3.to enrich The National Archives’ catalogue descriptions for the unit war diaries – this will enable you and others to find what they want in those 1.5 million pages, in future

  Citizen History is a two way street –volunteers must get at least as much out as the effort they put in. Our own experience of doing Operation War Diary tells us that this work gives us a fine-grained, close-up view of the Western Front. It makes us think about the war in ways we never did before and leads us on to learn more from other sources. Chris Lintott, Principal Investigator at Zooniverse has a wonderful phrase for this, he describes it as an “engine of motivation.”

  Operation War Diary is built using the powerful Zooniverse crowdsourcing platform. This is designed to present a series of images for classification and tagging by volunteers and to optimise that process. However, it is not designed to be a diary browsing tool. We appreciate that some people wish to view the whole diary . You can download the full diaries from The National Archives. There are no charges for visitors to their reading room at Kew and there is also free access in many Universities and libraries.

  many thanks,

  Luke (Smith)
  IWM

  Posted April 8, 2014 4:47 PM

• by lukesmith moderator

  p.s. There really is no reason to be concerned about errors. Every page is tagged by several people and the results cross referenced.

  Posted April 8, 2014 4:48 PM

• by arleena

  hi luke. i know, no back button. but can i tag a page i left no tagged at all? i clicked twice

  Posted August 5, 2014 9:05 AM

• by HeatherC moderator

  Luke isn't likely to answer since this thread is 4 months old! No you can't go back. Don't worry as each page is tagged by at least 5 people and a "best fit" of tags from all 5 is used.

  Posted August 5, 2014 9:21 AM

• by anon2643

  It is frustrating that you can't go back and fix errors. Once you've done a few pages you realise that you made mistakes on some of the earlier ones. I appreciate that you do that statistical matching but surely being able to fix errors would speed the matching process up - even if it was perhaps one 24 hour period to do corrections before making a final submission.

  Posted August 5, 2014 6:32 PM

• by wildcatjenny moderator in response to anon2643's comment.

  Hello annon2643!
  You are quite correct, it is frustrating! I think we all sympathize with your sentiments. But for now, that is just how things work I'm afraid. We will be sure to bring it up again if there is a new round of development scheduled.
  wildcatjenny

  Posted August 5, 2014 7:06 PM

• by MitreSquareMurder

  It really would be useful if we could go back and add things to pages we worked on in the past. Getting used to a person's handwriting takes time and there are often 'mystery' words on the first few pages that I figure out after I've immersed myself in the way the author forms his letters and seen them written a few more times. The same is true of place names. As something of a perfectionist, it galls me that on the first page I was given of a particular diary, the author neglected to put the year, so I had to guess. As soon as I moved onto the next page, I found that I had put the wrong year but had no way to go back and change such a simple, but important, thing. That's just one example, but it's frustrating.

  Posted August 9, 2014 6:18 PM

• by HeatherC moderator in response to MitreSquareMurder's comment.

  Yes I agree, but that's why you don't guess. You tag only what you see. If it's not there don't add it. You can click in the top text box of the date tag and delete the bits you don't know. Because the system requires each page to be tagged by at least 5 people and then takes a "best fit" of their tags, someone else who didn't get the middle page of the month as their first one, but started at the first page will know the year and so their tag will be used for the year and yours will confirm theirs for the day and month. That's the beauty of how crowd-sourcing works, but it relies on us all only tagging what we actually see and not what we may work out from elsewhere or even guess....

  Posted August 9, 2014 9:01 PM

• by coopernicola

  Having been tagging the same diary from April 1914, when I started, I'm now up to Sept 1915. I became frustrated at being unable to return to add extra information on earlier pages as I got used to things. However I now make a note of the date I am up to, as this always defaults to 1914 and can get confusing), interesting items e.g. names of officers who often take turns at being the diary author (identified only by initials), and other useful things such as the veterinary officers. This has proved useful as I fit in tagging when I have a spare hour.
  I often need to use 'other' in the category when tagging as 'my' diary is for an ammunition supply group and those available do not always apply. Generally I add unusual activities e.g. "cutting red clover for horses" in the comments.
  It's worth noting that if the years are not written you can double check if you have a church parade on the page as these will be on a Sunday.

  Posted August 18, 2014 10:57 AM

• by wildcatjenny moderator in response to coopernicola's comment.

  Hi coppernicola,

  Well of course, if you know the day of the week you can figure out what year you are working on! But, I had not thought of that myself. Thanks for bringing it up.

  wildcatjenny

  Posted August 18, 2014 6:59 PM

• by wildcatjenny moderator

  Please see the FAQ - General Topics post for more information.

  Posted November 11, 2014 1:23 AM