Author Topic: Suggestions for Phase 4  (Read 32776 times)

mapurves

  • Shipherd
  • Hero Member
  • *****
  • Posts: 1846
    • View Profile
Re: Suggestions for Phase 4
« Reply #30 on: September 09, 2013, 09:15:00 pm »
You could handle the "don't fill in the blanks" problem by assigning a key for a blank cell, e.g. a "[", or whatever. When the Fill button is clicked, the "[" would be removed. If there was a column with several successive blanks, putting the "[" in the first and last of the consecutive blanks would preserve the string of blank characters when the Fill button was clicked. I really like this idea of vertical entry as long as the log values are right beside the entry boxes. The people on the Concord, at least, have been filling in the log pages by entering the data vertically and, on occasion, they get out of sync so that a temperature and wet bulb don't make sense; i.e. the transcriber might enter five values of 45 when it should have been only four, or maybe six. I imagine the original sheet isn't right beside the log book so it isn't apparent that the wrong number of a singe value have been entered.

For those people using a number pad to enter values, a "." would suffice for a blank. For those using the numbers on the top row of the main keyboard, perhaps a "=", although I think different countries have different characters along the top row. It would be nice to have both characters symbolise a blank so that both types of people would have a character immediately handy.

I have found that the water temperature, especially over the last few log pages, changed hardly at all during the course of 24 hours.

Michael

Randi

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 13191
    • View Profile
Re: Suggestions for Phase 4
« Reply #31 on: September 09, 2013, 09:35:36 pm »
Perhaps a "." or "=" could indicate that value from above should be copied down.

I'm not really comfortable about a general fill for the whole column. There seems to be a lot of room for errors with blank entries, and new transcribers and transcribers that don't visit the forum might not make the distinction between 'no entry' and 'copy the one above'. We need to keep this simple ;) (However, if you want a key for a blank cell, how about a space?)

Craig

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3357
    • View Profile
Re: Suggestions for Phase 4
« Reply #32 on: September 10, 2013, 12:11:50 am »
Good suggestions for the blanks, Michael. Although it might be best to put the symbol in each blank cell given Randi's concern. If the whole block is blank then nothing needs to be done.

Perhaps a fill just for a block (i.e. AM or PM) rather than the whole column, Randi? And, of course, an undo option plus a warning message until you get tired of seeing it. I think if we reduced it to just to copying one cell at a time it would not be worth all the trouble to change the software. I can type most of the values (except compound cloud types) just about as quickly as I could type a period or equal sign. The warning could say "A blank in the log should not be filled with text. Please type a '[' in each cell that should remain a blank. ". (or whatever symbol we agree upon).

I think the biggest challenge will be getting the values in the right cells to indicate the intervals to fill. The capture screen will have the hours marked and it will be important for them to be vis-?-vis the corresponding row of the log. Is the spacing of rows standard for most of the logs? If not, can we adjust he capture screen to adapt to different row heights?  It would be interesting at this point to know if Philip thinks this suggestion is feasible from the software prespective. Perhaps we are wasting our time discussing it.

Randi

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 13191
    • View Profile
Re: Suggestions for Phase 4
« Reply #33 on: September 10, 2013, 07:33:12 am »
Quote
I can type most of the values (except compound cloud types) just about as quickly as I could type a period or equal sign.
You obviously type a lot better than I do! ;D

Craig

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3357
    • View Profile
Re: Suggestions for Phase 4
« Reply #34 on: September 10, 2013, 11:48:22 am »
I can see why you are concerned about blanks, Randi. The Bear, Patterson and the two Ungalas don't have hourly weather reports in the log pages that I could see. The fill option would not be helpful for those logs and the advantage of a vertical capture screen would be very marginal. The others, plus our new ships, have hourly reports as far as I could tell by a cursory look, although there may be some exceptions in certain books. When the Jamestown 1844 is in port there is only 1 report per watch but there is hardly any information to record so the orientation wouldn't matter. Blanks occur even in 24-hour reports, of course but I think one could just refrain from using the fill option, or use Michael's suggestion. Your concern is how to inform newbies or those who don't visit the Forum. I think some modification to the introductory training lesson would be necessary.

Randi

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 13191
    • View Profile
Re: Suggestions for Phase 4
« Reply #35 on: September 10, 2013, 12:16:57 pm »
We also need to be able to handle cases where there is more than one temperature or pressure (early Yukon 1 with mercury and aneroid readings) per hour or there are intermediate readings (e.g., 9.30), and on one ship there was a time change so there were two different 6pm reports :o.

I still like the idea, but I have reservations about its flexibility.
It also has to be fairly intuitive - I have the impression that not everyone pays close attention to the tutorial :'( :'( :'(

Maikel

  • Editor
  • Hero Member
  • *****
  • Posts: 2614
    • View Profile
Re: Suggestions for Phase 4
« Reply #36 on: September 10, 2013, 12:27:35 pm »
I have a slight concern about entering data vertically, in particularly when a transcriber makes a mistake by skipping one field.

Horizontally skipping one field only renders values for one hour incorrect.
Vertically skipping one field will render all hours after it incorrect.
But hopefully that would be negated by the three persons needed to transcribe.

More importantly, it seems that all efforts to attract new transcribers seem to involve technical changes.

Although technical changes can help, and in some cases are needed, I would suggest that more emphasis should be placed on the human or psychological aspect.

Probably new transcribers stop because of the amount of data to enter.
You won't improve that by reducing input time from 10 minutes to 5 minutes.
You'll only get them involved if input is done in a matter of seconds or a couple of clicks, like Galaxy Zoo.
Or have pretty pictures to look at, instead of (sometimes hard to read) boring rows of numbers and codes.

The other point of attraction could be to make it a contest.
Unfortunately that fails because the limited amount of available vessels.
All ships are more or less done in terms of captaincy, not attractive for a starter.

Perhaps the logs should have been made available in many small batches instead of many log-books for a few ships.
Remember how the RN phase had hundreds of ships and you could be captain with a couple of thousand weather reports, there the contest part worked better.
Also, to keep such a contest fair, events should be compulsory or at least should be counted in the statistics.

For transcribers not really interested in weather data, a point of interest could be the story of the ship, what happened on board and the adventures it had.
But again, unfortunately the current vessels are pretty boring in the day to day activities.
Only if you follow a ship long enough you'll understand it had an interesting journey.
During the RN phase the large battleships were the first to be done, because they had a lot of action.
The store ships, river gunboats, and what have you, where the last to be done.
Same sort of weather data, but seemingly less interesting to transcribe.

I also believe that a large number of the British transcribers have stopped transcribing simply because it's now US ships instead of RN ships.
Perhaps there is less bonding with the ships for them.

The places of these lost transcribers should have been filled with US transcribers, but for some reason they haven't.
Perhaps it's one of the reasons I mentioned earlier, or perhaps NOAA hasn't been able to reach out enough.

Craig

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3357
    • View Profile
Re: Suggestions for Phase 4
« Reply #37 on: September 10, 2013, 12:33:46 pm »
We also need to be able to handle cases where there is more than one temperature or pressure (early Yukon 1 with mercury and aneroid readings) per hour or there are intermediate readings (e.g., 9.30), and on one ship there was a time change so there were two different 6pm reports :o.

I still like the idea, but I have reservations about its flexibility.
It also has to be fairly intuitive - I have the impression that not everyone pays close attention to the tutorial :'( :'( :'(

Both orientations have their limitations. The extra temperatures or pressures could be easily handled with a vertical capture screen (with a supplementary column) but they cause problems with the horizontal one. (We had to insert two rows for the same hour, which was annoying). Conversely, the half-hour reading would be a problem with a vertical screen but easily handled with a horizontal one. These are exceptional cases and there can always be work-arounds. We would have to insert two columns for the same variable.

You're right about the tutorials. Many people prefer to get right into transcribing and then look for help if they run into problems. One thing to keep in mind, though, is that people that give up after a few pages only account for very small percentage of the transcriptions. There are a couple of prolific transcribers who don't visit the Forum (unless they have changed their names and I don't recognize them). It would be helpful if the Mods could verify some of their work without having them initiate a request.

Craig

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3357
    • View Profile
Re: Suggestions for Phase 4
« Reply #38 on: September 10, 2013, 12:40:01 pm »
I agree with Mailel on most of his points. Indeed, it would be easy to miss a change in wind direction and have the previous value copied down a number of cells. But this could also happen if there is a mistake made in a transcription that is followed by a number of dittos. You faithfully transcribe the dittos but the wrong value is inferred.

And perhaps he is right that these changes won't make much difference in keeping new members. But I think it would speed up transcibing for those who do stay on and make it a bit less frustrating when you have the feeling of d?j? vu  ;D

Pommy Stuart

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3652
  • A closed mouth gathers no foot.
    • View Profile
Re: Suggestions for Phase 4
« Reply #39 on: September 10, 2013, 09:15:18 pm »
Has anybody considered asking the programmers what they can offer us, instead of proffering what we want?
 

Janet Jaguar

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 10166
  • Smell the sea, feel the sky, & fly into the mystic
    • View Profile
Re: Suggestions for Phase 4
« Reply #40 on: September 10, 2013, 09:24:58 pm »
The programmers started this, with the statement they don't have time to start it for another few months.  This is really throwing out wild ideas to get their more practical imagination started.  Which is needed in its own way.

Craig

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3357
    • View Profile
Re: Suggestions for Phase 4
« Reply #41 on: September 10, 2013, 10:07:37 pm »
In any case, it's normal for the users to make suggestions. We're the ones who have the experience of transcribing. As I mentioned before, if Philip had to enter 10 variables times 24 hours for every log page in 60 log books I bet he would think of ways to make it faster.

What do you think about a vertical capture screen with a Fill option, Stuart?

Pommy Stuart

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3652
  • A closed mouth gathers no foot.
    • View Profile
Re: Suggestions for Phase 4
« Reply #42 on: September 10, 2013, 10:13:19 pm »
I can only speak for the Concord and I think it could work for me.
12 entries at a time.

Craig

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3357
    • View Profile
Re: Suggestions for Phase 4
« Reply #43 on: September 10, 2013, 10:25:59 pm »
Yes, I think 12 hours at a time is optimal because you can see it all at once. You can get a feel for it by opening up Excel, setting the tab direction to down, reducing the window size and moving it next to a column in the log to be transcribed.

I have been doing the Jamestown 1866 for a while now. It's in the Caribbean and the weather conditions have been pretty stable. Lots of opportunity to do automatic fills.

Pommy Stuart

  • Shipherd
  • Hero Member
  • *****
  • Posts: 3652
  • A closed mouth gathers no foot.
    • View Profile
Re: Suggestions for Phase 4
« Reply #44 on: September 12, 2013, 06:02:46 am »
Just thought of a problem with Vertical columns, how would the application know what heading to put on which column and how would the user know what columns to enter?
My only solution would be a list of columns headings which we now have and to click on each heading to drop the capture box.