Chat log for LAK11 Weekly Check in
Feb 4, 2011
00:29 - Stephen Downes
So if you polish up your presentation for this course, is it lak lustre?
01:18 - dave cormier
ha
02:11 - dave cormier
tanya?
03:15 - Kelly Edmonds
use in content/courses: marketing research, stats...
04:34 - Sylvia Currie
http://www.itap.purdue.edu/tlt/signals/
05:10 - Jon Dron
hi!
05:24 - dave cormier
perfect
07:15 - dave cormier
next on the mic?
07:23 - dave cormier
come on... someone else
07:33 - Kelly Edmonds
start with (research) questions
08:29 - Francisco Reis
Anyone saw Peter Kruse videos on the link George sent us today?
08:46 - mike.cummings2009@gmail.com
Fantastic video
09:26 - Francisco Reis
Later on i'd like to discuss those videos.
09:38 - Jon Dron
it's not that it's not possible - the risk is that we place too much credence in the results
09:46 - Kelly Edmonds
we need to ensure we start with what learning is and how it occurs so not to get overwhelmed by the technologies and data
10:28 - Jon Dron
I worry, for instnace, that the demotivating aspects of red lights are only worth risking if we are close to 100% sure we are right
10:31 - Sylvia Currie
Yes, anybody else know knowledge forum?
10:32 - Stephen Downes
Maybe a good question is, what sort of information is visualizable using analytics? Just counting numbers of posts? Doesn't sound that useful. But what other types of information...?
10:43 - Shaun
what was the name of it, made by OISIE?
10:54 - Sarah Haavind
knowledge forum
11:03 - Shaun
thx Sarah
11:24 - Kelly Edmonds
right and questionable data and constructs
11:33 - Sarah Haavind
Exactly exactly
11:34 - Stephen Downes
partial data, erronious assumptions - yes, exactly, major risks
11:43 - Adam Weisblatt
One problem is that the scale of the domains of what we are talking about are so vast: predicting an individual's desired next course vs adoption of pedagogies across entire populations
12:02 - Sarah Haavind
I think that's what is holding us all back
12:13 - Stephen Downes
the supposition that x,y,x data = 'improvement' is in itself suspect
12:19 - Sylvia Currie
Especially in open courses, where data is everywhere, people aren't logging in, etc
12:20 - Sarah Haavind
yes yes
12:24 - Adam Weisblatt
Each domain that we measure has its own considerations
12:34 - Sarah Haavind
It's already happening in online evaluation programs for course evals
13:45 - Adam Weisblatt
also, what needs to be captured in order to measure and how much resource does that take
13:52 - Adam Weisblatt
Is the jice worth the squeeze
14:06 - Adam Weisblatt
Is the juice worth the squeeze
14:18 - Stephen Downes
what it means to have a hit on a website: success! $$$$!
14:44 - Stephen Downes
yeah I'm a big sell-out
15:03 - dave cormier
go ahead sarah
15:12 - dave cormier
ur just fine
15:16 - Adam Weisblatt
everything eventually comes down to investment of resources to produce value aka money
16:23 - Sylvia Currie
haha - give 24 hr feedback to students please
16:53 - Stephen Downes
I'm not available 24 hours - I take 3:00 a.m. - 4:00 a.m. for persopnal time
16:59 - dave cormier
yes
17:14 - dave cormier
i can
17:27 - Sylvia Currie
Here's the signals link again http://www.itap.purdue.edu/tlt/signals/
17:40 - dave cormier
I'm nodding my head
18:15 - Sarah Haavind
yes exactly
18:29 - Mary McEwen
that would also depend on class - size, wouldn't it?
18:29 - Sarah Haavind
I'm nodding MY head :)
18:42 - Jon Dron
assuming a formal context, of course
18:57 - Jon Dron
the potentially biggest value might be for self-directed, networked learning
19:29 - Sylvia Currie
class size a factor.. as a way to notice who is doing what
20:05 - Tanya Elias
in training we often have no idea how what we've developed goes over with learners
20:16 - Sylvia Currie
go ahead Kelly
21:11 - dave cormier
back now :)
21:15 - dave cormier
sorry folks
21:32 - Sylvia Currie
Interesting...almost like making the process more transparent
21:38 - Tanya Elias
performance based learning :)
22:07 - Francisco Reis
We go back to putting on the student the responsibilty to what they are doing/following.
22:27 - Jon Dron
excellent points Kelly - it has *always* been true that what we teach bears only a passing resemblance to what and how people learn
22:54 - Jon Dron
at best, we might be modelling a process and offering snippets. The serious learning is outside our field of view
23:01 - Therese Weel
http://www.sonar6.com/ <-- performance analytics
23:50 - Kelly Edmonds
yes, difficult to follow
24:05 - Sylvia Currie
Kelly's comments really hit home for work environment too -- how to make your daily work log available to others, how to include people in communication without bombarding them
24:07 - Therese Weel
sonar6 visuals http://bit.ly/hC666D
24:31 - Mary McEwen
yes!! analytics as it informs instructional design!
24:41 - Kelly Edmonds
Thanks, Therese
25:42 - Stephen Downes
link between analytics and recommendations?
25:53 - Therese Weel
Your welcome - I'll tweet those links under #lak11
26:31 - Jon Dron
@Stephen - yes, definitely - at least in George's view! I think this is still in adaptive systems when we are looking at using results directly to change what happens next
28:24 - Sarah Haavind
Linked In now visualizes connections which you can then label by professional group/association -- I thought it was "neat" but haven't thought of a "use"
28:36 - dave cormier
we're just having a friday chat :)
29:05 - Kelly Edmonds
I used it and it was okay. Not really sure how LinkedIn connections were linked
29:41 - Sylvia Currie
@Sarah, I've played with Linked In too. another one of those... this is neat, now what situations :-)
30:06 - Stephen Downes
the problem is, the paradox of measurement: any measurement depends on our preidentification of what is measured & what the units will be, and hence any measurement begins as inherently subjective
30:25 - Stephen Downes
thus, no measurement is an accurate and objective representation of the domain
31:04 - Kelly Edmonds
I agree Stephen - determining the construct/context of our units of measurement. These have to be thoughtfully selected and not 'slapped' together
31:16 - Sylvia Currie
@Mary, would be interesting to list out measurements that would inform instructional design
31:18 - Stephen Downes
this is especially a problem in education
31:36 - Kelly Edmonds
I think we are
31:57 - Tanya Elias
we also go in a lot of wrong directions when we don't measure
32:05 - Stephen Downes
well, no, Dave - the core of 'how the internet works' is a 'hit' - but in a eworld of search agents and Akamai, what is a 'hit'?
32:21 - dave cormier
:)
33:03 - dave cormier
@stephen i was thinking the nature of the data... html etc...
33:42 - Mary McEwen
@sylvia I think that the most helpful measurements to inform instructional design will be more more specific to a design, than broadly such as at the CMS level.
33:45 - Tanya Elias
what should/ do they want to be able to do at the end of learning?
35:19 - Sarah Haavind
it depends...
35:24 - Jon Dron
I guess the interesting possibility in this kind of thing is that we can ask these questions about learner X, not just 'learners'
35:26 - dave cormier
@sarah lol
35:49 - Sarah Haavind
it depends both on the learners and what ever you want your students to understand
36:10 - dave cormier
and then you have to figure better...
36:14 - Mary McEwen
or even something as simple as are the students viewing the content, and for how long? Is there a correlation between that and other performances?
36:55 - dave cormier
i don't hear you stephen
36:56 - Mike Cummings
Mary that really is a key simple question :-)
37:21 - dave cormier
nice
37:30 - dave cormier
oh sure.
37:40 - Mike Cummings
@Mary teachers, course designers have intuitions but have no way of validating the significance of these thoughts
37:55 - Jon Dron
@Mike, Mary - it's a key question if we are interested in the efectiveness of our instructional design. May have little value to a learner
38:25 - dave cormier
@Mary my question then would be 'what other performances?'
38:43 - dave cormier
i r trying to facilitate and not comment. bad facilitator
38:54 - Mike Cummings
Its all about connectiveness :-) effective instructional design, measured against effective learning
39:15 - dave cormier
does colour affect the length of time people look at a picture?
39:33 - Sylvia Currie
From instructional design perspective, I think would be interesting to see if students are using collaborative processes you expect them to be using.
39:43 - ljp
You could measure what I do; but, it wouldn't tell you much about the course or it's design. What I am doing is driven by the available time and my other activities. Further, most video doesn't work on my computer. My not using the video resources is a technical not a personal choice. How will my lack of activity be interpreted?
39:49 - Tanya Elias
back to need to be performance based
39:49 - Stephen Downes
My 'time on task' data is frequently altered by my having a cat in my lap
40:59 - Jon Dron
@Stephen - I sometimes spend hours on a single 5 minute task (along with dozens of other 5 minute tasks at the same time, from the point of view of a tool that might be measuring such things)
41:10 - dave cormier
anyone else wanna weigh in no this discussion?
41:10 - Stephen Downes
exactly
41:31 - Mary McEwen
exactly - Tanya!
41:40 - mike.cummings2009@gmail.com
But if everyone is spending hours on a 5 minute task, this is an important pattern
41:41 - Stephen Downes
people take courses for some purpose (versus) I don't know why I'm sitting in this lak course
41:56 - Kelly Edmonds
Formal education defines what is to be learned.. good and bad point
42:20 - Stephen Downes
I'm always an outlier
42:28 - Mary McEwen
or that stephen has a cat in his lap distracting him
42:28 - mike.cummings2009@gmail.com
blame the cat
42:35 - Tanya Elias
I think analytics can also be used to personlaize
42:46 - Tanya Elias
among big systems
42:53 - Mary McEwen
yes--tanya, another great potential for use.
43:01 - Jon Dron
@Tany - yes - this gets back to my point about why we are asking the questions, not just what kind of questions we can ask
43:12 - Kelly Edmonds
where is the attention of students?
43:18 - Francisco Reis
It's me!
43:21 - Jon Dron
to improve design, to monitor progress, to know what to do next
43:42 - dave cormier
hi :)
43:49 - dave cormier
that's not a problem
43:54 - Therese Weel
A Questioning Toolkit http://www.fno.org/nov97/toolkit.html
43:55 - Tanya Elias
I really like those 3 jon, going to think more on those...
44:26 - Stephen Downes
Peter Kruse videos: http://www.google.ca/search?q=peter+kruse&tbs=vid:1
46:28 - Mary McEwen
is it possible to calibrate instructional interactions in a similar same way that we calibrate assessment items?
47:35 - dave cormier
@francisco i think so too
47:37 - Stephen Downes
'I link it, I like it' - Peter Kruse (text, English) - http://www.goethe.de/wis/med/idm/fin/en6578052.htm
47:52 - Tanya Elias
@ Mary: Do you student-intructor interactions?
47:53 - ljp
The problem with analytics is I'm always excluded. Normally, I am an outlier and don't cluster well.
48:26 - Mary McEwen
no, though that is a possibility, what I had in mind was interacting with online instructional interactions.
48:36 - Jon Dron
@ljp - mass customisation vs personalisation - analytic methods can personalise for outliers too :-)
49:03 - Francisco Reis
Seeing those topics and their relations evolve over time must be interesting!
49:15 - Kelly Edmonds
like a communiity of practice
49:25 - Francisco Reis
YES!. REAL-TIME!
49:41 - Kelly Edmonds
emerging CoP
49:42 - Stephen Downes
clustering creates tipping points - interesting insight
49:52 - Sylvia Currie
design that system please
49:55 - Jon Dron
interesting that the segmented hierarchical threaded discussion process works in total opposition to this kind of clustering -we are fighting the tools
50:06 - Niklas
Is it possiblie to analyze information that tell the student and teachers about their socio-cultural context. How they communicate with, places, time...?
50:08 - Kelly Edmonds
Yes, Jon!
50:51 - Jon Dron
but again - if we ask the wrong questions we may get clusters of no great value to learners
51:35 - Jon Dron
but choice of algorithm makes enormous difference
52:09 - Jon Dron
content-addressable results vary massively depending on algorithms we use to mine
52:24 - Jon Dron
and that implies categorisation of a sort
52:24 - Francisco Reis
Isn't it defining connections more powerful than finding connections?
52:48 - Jon Dron
funnily enough that's what I was writing about yesterday - ened for adjustable alogrithms
52:52 - Jon Dron
need even
52:57 - Kelly Edmonds
we need some type of frame, Stephen
54:58 - dave cormier
anyone else for last words?
55:20 - Sylvia Currie
Been a great session!
55:32 - Francisco Reis
Congratulations, Dave!
55:40 - Therese Weel
glad to catch this session today - Thank you!
55:42 - Sylvia Currie
That's okay -- remember I'm the asynchronous creature
55:48 - Jon Dron
analytics :-)
55:49 - merce galan
Thank everybody
55:56 - Claudia Guerrero
thanks
56:01 - Tanya Elias
Thanks!
56:02 - Kelly Edmonds
Thanks so much!
56:04 - Niklas
Thanks
56:06 - Jon Dron
bye everyone - thanks for coming!
56:12 - Andreas Link
thx & bye. have a nice weekend!
56:16 - Sylvia Currie
Bye for now!
56:21 - dave cormier
tah tah