Literacy Project/2012-06-27: Difference between revisions

From OLPC
Jump to navigation Jump to search
No edit summary
No edit summary
 
(16 intermediate revisions by the same user not shown)
Line 1: Line 1:
== Agenda ==
1. Application Status - New applications reviewed, to be added, obstacles / status
2. Application Organization - Updated app presentation on the tablet
3. Third Site Status - MoE discussion, location scouting, tablet preparation
4. Data-collection / funf framework - Meeting report, David's updates
5. Current data collection status (Richard)
6. Assessment data observations (Chris, Richard)

== Attendance ==
== Attendance ==
Maryanne
Maryanne
Line 16: Line 8:
David
David
Max (new from Maryanne's group.)
Max (new from Maryanne's group.)
Eric (" ")
Eric (" ")
Denise
++
++
++


== Notes ==
== Agenda ==
1. Application Status - New applications reviewed, to be added, obstacles / status
Apps and activities by conceptual and linguistic development
2. Application Organization - Updated app presentation on the tablet
: early childhood concepts & words
3. Third Site Status - MoE discussion, location scouting, tablet preparation
4. Data-collection / funf framework - Meeting report, David's updates
5. Current data collection status (Richard)
6. Assessment data observations (Chris, Richard)
7. Schedule


=== App status ===
A categorization developed by Maryanne, Merit, over the past week.


Apps and activities by conceptual and linguistic development : A categorization developed by Maryanne, Merit, over the past week.
; Categorization:
: Early Primer Words
: early childhood concepts & words; Categorization: 6-10 categories, such as 'Early Primer Words'
: Related doc: [https://docs.google.com/document/d/1zIh-N8qjdE0UVDz9_eA-nzq4AfENf1ZVCoCJuQU5kZI/edit on Google Docs]


; New apps: There are some apps planned to be added over the past week.
=== New apps ===
There are some apps planned to be added over the past week.
: Not put into colored squares yet; very visible.
: Not put into colored squares yet; very visible.
: Oceanhouse media aren't yet in the group, but we are hoping we will manage it. Let's avoid running into that problem. e.g., the apps look great, can't connect to a license server when offline, work great for a week. We need to try any apps in a no-network env for extended time.

:: "there was no way we would ahve known that problem unless we tried that for a week w/o network" [well, maybe ''one'' way.]
Oceanhouse media aren't yet in the group, but we are hoping we will manage it. Let's avoid running into that problem. e.g., the apps look great, can't connect to a license server when offline, work great for a week. We need to try any apps in a no-network env for extended time.
: "there was no way we would ahve known that problem unless we tried that for a week w/o network" [well, maybe ''one'' way.]


Pruning a fair amount. Adding some new apps. How many films are reallt instructing in language? How much is just fun? [turned on, leflt on]
Pruning a fair amount. Adding some new apps. How many films are reallt instructing in language? How much is just fun? [turned on, leflt on]
Line 43: Line 39:
Q: on sustained attention. on lions and other vids w small segments within them:
Q: on sustained attention. on lions and other vids w small segments within them:


Alex - is this like memory games where you have to look at one image and find another that's similar.
=== Funf meeting ===
: C - but in one sense we have dull things for our experiments so that we're only measuring something very basic; when you add something fun it makes it hard to interpret the results.
They can lose the launcher; kids can access apps any way they want. we can unify on a smaller # of views for that. and get much better / cleaner data collection.
: E - 'fun' by definition adds somethig they like that skews behavior... [S - not by definition! ] C - I'm not so fatalistic about it. we could add better rewards and still be rigorous. but when you add multiple types of stimulus...]

=== App organization ===
From the funf meeting: We can lose the launcher with the next funf update; kids can access apps any way they want. we can unify on a smaller # of views for that. and get much better / cleaner data collection.

Q: on consolidating skills. can we group things into 'letters' , 'colors' -- editing things into smaller chunks?
: Maryanne to David -it sounds like you're saying we can do this.


=== Third site status, and pre-testing ===
We have two more weeks now for a final image?
;pre-testing
Chris - to do proper pretesting, you need to ship an image that only has matching and balloons and none of the concepts; and then add it later.

M - that might be a soln in several ways. we would prefer doing these segments, but perhaps should ahve a phase 1/2/3 in which the first is just testing, 2d is what we have friday this week, 3rd is that plus segments, hich could be added after 2-3 weeks.

C - we want to give them more than a week for matching and balloons to see how they learn the task. at wonchit hey only gradually learned what they were being asked to do.

E - also wonchi/wolonchete kids had had usage experience. at the third site they won't hav ethat. we almost want to give the kids as et of apps that have nothingt o do with literacy, have them play angry birds for 2 weeks and then replace everything with the literacy image.

M - but lenring to use apps isn't ? learning, it is technical learning...

C -so we could do non-lit tests for a while?
Merrit - well at least "content-free" tests. ones that are not literacy oriented at all.

C - what about only roviding the color and matching tests, and then adding the literacy tests after 2 weeks.

Maryanne -better to start with something that is engaging, then move to matching, then more to the content we want. that wodl give us time to integrate esegments. [time qt 2hqt point in the schedule?]

=== Data-collection / funf framework ===
Meeting report, David's updates.

From David: "I made some major cleanup and funf overhaul fixes that I just pushed to the repo and updated apk's in xoom-update..." (see email)

We're still getting picture data as we always have, but evern :10 rather than every :5 . We now also have info on when the photo was taken. So you could use a very manual process to match a photo with roughly what apps were running at that time. Cf Chris's recent email, we could look at data from one week to see if it's always the same kid in front of the tablet. [Max and Eric can do this; need to be added to the literacy list]


==== Data collection status: Probes and custom activities ====

Can you make a tinkrbook probe for concepts in each of these categories? say, 3 animals, colors, etc throughout.

C - we've done this already, with the popular things they've been exposed to in tinkrbook.
What we found so far is they haven't learned things like baby/duck outsid ea specific context.

M - we expect and hope that having a categorization with things like a cluster of different ducks in a file, will help consolidate ideas.
It offers information on receptive-vocabulary growth.


Perhaps we could have a sequence: from a specific instance of an image with sound. then add a variety of simialr images.
then add a photograph of a real duck, something not in the original cluster, but something that the cluster has been specifically training them to be able to id.

D - if we can confirm they can associate written word with images... the more they know about a wrod the more likely they are to be able to learn other things about it when you see it in written form. background knowledg about each word itself adds scaffolding.
: M - with more time we would efinitely add that scaffolding explicitly. and it's thanks to chris that rather than a specifically scaffolded tablet, this version is a compromise towards a future version.
: E - good that we went through :50 before first mentioning scaffolding! <aside on sth that rhymed with 'mocratic ladflies'.>

=== Assessment data observations ===
* When kids can't get beyond one assessment, can we find another assessment that works better for them?
* [We may need another set of video sounds?]


=== Schedule ===
We shouldn't be adding more content to tablets being taken for testing now.

We are in some sense already out of time (for this build).

We should think about to what degree we want to use and modifyin matching and balloons apps for pre-testing, to get initial data before starting.

M: can we do this by Friday? Merrit: we should consider if we need all these segments ready.

: M - we should be all on the same page. our group is really... to ahve as good as we can, a conceptualization of these caetegories, every segment we can and think is important put in there. BUT: for each category, this full set will be gotten to you within 2 weeks (3 weeks?). And by that point we will be able to ascertain whether we are making similar mistakes with these apps as last time.
: Let's keep thinking in terms of waves, rather than having everything at the start.

Max and Eric: around through most of August. Merrit: away soon; but available then?

: UK Aid will tell us Friday about the initial statement, whether invited to submit the full app.
: USAid will make awards thru Sept; be patient. (we were excluded on one, not the other)

== Other Q's for OLPC from others ==
Something that happened other than this glitch connecting to the G apps store.
# problem opening videos, diff from one tablet to another. a loading problem? 'this file/video cannot be opened/displayed'
# David will chat with Richard after the meeting.

Latest revision as of 20:10, 27 June 2012

Attendance

Maryanne
Ed
Merrit
Richard
Chris
SJ
David
Max (new from Maryanne's group.)
Eric (" ") 
Denise
++

Agenda

1. Application Status - New applications reviewed, to be added, obstacles / status
2. Application Organization - Updated app presentation on the tablet 
3. Third Site Status - MoE discussion, location scouting, tablet preparation
4. Data-collection / funf framework - Meeting report, David's updates
5. Current data collection status (Richard)
6. Assessment data observations (Chris, Richard)
7. Schedule

App status

Apps and activities by conceptual and linguistic development : A categorization developed by Maryanne, Merit, over the past week.

early childhood concepts & words; Categorization: 6-10 categories, such as 'Early Primer Words'
Related doc: on Google Docs
New apps
There are some apps planned to be added over the past week.
Not put into colored squares yet; very visible.
Oceanhouse media aren't yet in the group, but we are hoping we will manage it. Let's avoid running into that problem. e.g., the apps look great, can't connect to a license server when offline, work great for a week. We need to try any apps in a no-network env for extended time.
"there was no way we would ahve known that problem unless we tried that for a week w/o network" [well, maybe one way.]

Pruning a fair amount. Adding some new apps. How many films are reallt instructing in language? How much is just fun? [turned on, leflt on]

(but isn't this a good way to learn too?) Is there real engagement w a very long movie?
"it's a big investmemnt, time and space; if they're not being used..."
Richard: certainly some kids had onet cuked under their arms, walking around with a film blaring. I watched them take an interest in videos of all types and sizes; and they'd jump around a lot in them. Kids now almost never use the launcher. they go to 'apps' and the gallery views; both give them a nicer view to page through.

Q: on sustained attention. on lions and other vids w small segments within them:

Alex - is this like memory games where you have to look at one image and find another that's similar.

C - but in one sense we have dull things for our experiments so that we're only measuring something very basic; when you add something fun it makes it hard to interpret the results.
E - 'fun' by definition adds somethig they like that skews behavior... [S - not by definition! ] C - I'm not so fatalistic about it. we could add better rewards and still be rigorous. but when you add multiple types of stimulus...]

App organization

From the funf meeting: We can lose the launcher with the next funf update; kids can access apps any way they want. we can unify on a smaller # of views for that. and get much better / cleaner data collection.

Q: on consolidating skills. can we group things into 'letters' , 'colors' -- editing things into smaller chunks?

Maryanne to David -it sounds like you're saying we can do this.


Third site status, and pre-testing

We have two more weeks now for a final image?

pre-testing

Chris - to do proper pretesting, you need to ship an image that only has matching and balloons and none of the concepts; and then add it later.

M - that might be a soln in several ways. we would prefer doing these segments, but perhaps should ahve a phase 1/2/3 in which the first is just testing, 2d is what we have friday this week, 3rd is that plus segments, hich could be added after 2-3 weeks.

C - we want to give them more than a week for matching and balloons to see how they learn the task. at wonchit hey only gradually learned what they were being asked to do.

E - also wonchi/wolonchete kids had had usage experience. at the third site they won't hav ethat. we almost want to give the kids as et of apps that have nothingt o do with literacy, have them play angry birds for 2 weeks and then replace everything with the literacy image.

M - but lenring to use apps isn't ? learning, it is technical learning...

C -so we could do non-lit tests for a while? Merrit - well at least "content-free" tests. ones that are not literacy oriented at all.

C - what about only roviding the color and matching tests, and then adding the literacy tests after 2 weeks.

Maryanne -better to start with something that is engaging, then move to matching, then more to the content we want. that wodl give us time to integrate esegments. [time qt 2hqt point in the schedule?]

Data-collection / funf framework

Meeting report, David's updates.

From David: "I made some major cleanup and funf overhaul fixes that I just pushed to the repo and updated apk's in xoom-update..." (see email)

We're still getting picture data as we always have, but evern :10 rather than every :5 . We now also have info on when the photo was taken. So you could use a very manual process to match a photo with roughly what apps were running at that time. Cf Chris's recent email, we could look at data from one week to see if it's always the same kid in front of the tablet. [Max and Eric can do this; need to be added to the literacy list]


Data collection status: Probes and custom activities

Can you make a tinkrbook probe for concepts in each of these categories? say, 3 animals, colors, etc throughout.

C - we've done this already, with the popular things they've been exposed to in tinkrbook. What we found so far is they haven't learned things like baby/duck outsid ea specific context.

M - we expect and hope that having a categorization with things like a cluster of different ducks in a file, will help consolidate ideas. It offers information on receptive-vocabulary growth.


Perhaps we could have a sequence: from a specific instance of an image with sound. then add a variety of simialr images. then add a photograph of a real duck, something not in the original cluster, but something that the cluster has been specifically training them to be able to id.

D - if we can confirm they can associate written word with images... the more they know about a wrod the more likely they are to be able to learn other things about it when you see it in written form. background knowledg about each word itself adds scaffolding.

M - with more time we would efinitely add that scaffolding explicitly. and it's thanks to chris that rather than a specifically scaffolded tablet, this version is a compromise towards a future version.
E - good that we went through :50 before first mentioning scaffolding! <aside on sth that rhymed with 'mocratic ladflies'.>

Assessment data observations

  • When kids can't get beyond one assessment, can we find another assessment that works better for them?
  • [We may need another set of video sounds?]


Schedule

We shouldn't be adding more content to tablets being taken for testing now.

We are in some sense already out of time (for this build).

We should think about to what degree we want to use and modifyin matching and balloons apps for pre-testing, to get initial data before starting.

M: can we do this by Friday? Merrit: we should consider if we need all these segments ready.

M - we should be all on the same page. our group is really... to ahve as good as we can, a conceptualization of these caetegories, every segment we can and think is important put in there. BUT: for each category, this full set will be gotten to you within 2 weeks (3 weeks?). And by that point we will be able to ascertain whether we are making similar mistakes with these apps as last time.
Let's keep thinking in terms of waves, rather than having everything at the start.

Max and Eric: around through most of August. Merrit: away soon; but available then?

UK Aid will tell us Friday about the initial statement, whether invited to submit the full app.
USAid will make awards thru Sept; be patient. (we were excluded on one, not the other)

Other Q's for OLPC from others

Something that happened other than this glitch connecting to the G apps store.

  1. problem opening videos, diff from one tablet to another. a loading problem? 'this file/video cannot be opened/displayed'
  2. David will chat with Richard after the meeting.