Daphne Koller: What we're learning from online education

668,704 views ・ 2012-08-01

TED


Please double-click on the English subtitles below to play the video.

00:00
Translator: Morton Bast Reviewer: Thu-Huong Ha
0
0
7000
00:15
Like many of you, I'm one of the lucky people.
1
15985
3111
00:19
I was born to a family where education was pervasive.
2
19096
3400
00:22
I'm a third-generation PhD, a daughter of two academics.
3
22496
4238
00:26
In my childhood, I played around in my father's university lab.
4
26734
3794
00:30
So it was taken for granted that I attend some of the best universities,
5
30528
3849
00:34
which in turn opened the door to a world of opportunity.
6
34377
3801
00:38
Unfortunately, most of the people in the world are not so lucky.
7
38178
4120
00:42
In some parts of the world, for example, South Africa,
8
42298
3135
00:45
education is just not readily accessible.
9
45433
2705
00:48
In South Africa, the educational system was constructed
10
48138
2975
00:51
in the days of apartheid for the white minority.
11
51113
2873
00:53
And as a consequence, today there is just not enough spots
12
53986
2700
00:56
for the many more people who want and deserve a high quality education.
13
56686
3852
01:00
That scarcity led to a crisis in January of this year
14
60538
3880
01:04
at the University of Johannesburg.
15
64418
1836
01:06
There were a handful of positions left open
16
66254
2131
01:08
from the standard admissions process, and the night before
17
68385
2969
01:11
they were supposed to open that for registration,
18
71354
2560
01:13
thousands of people lined up outside the gate in a line a mile long,
19
73914
4052
01:17
hoping to be first in line to get one of those positions.
20
77966
3880
01:21
When the gates opened, there was a stampede,
21
81846
2308
01:24
and 20 people were injured and one woman died.
22
84154
3652
01:27
She was a mother who gave her life
23
87806
1940
01:29
trying to get her son a chance at a better life.
24
89746
4063
01:33
But even in parts of the world like the United States
25
93809
3157
01:36
where education is available, it might not be within reach.
26
96966
4356
01:41
There has been much discussed in the last few years
27
101322
2672
01:43
about the rising cost of health care.
28
103994
1989
01:45
What might not be quite as obvious to people
29
105983
2642
01:48
is that during that same period the cost of higher education tuition
30
108625
4022
01:52
has been increasing at almost twice the rate,
31
112647
2480
01:55
for a total of 559 percent since 1985.
32
115127
4280
01:59
This makes education unaffordable for many people.
33
119407
4534
02:03
Finally, even for those who do manage to get the higher education,
34
123941
3801
02:07
the doors of opportunity might not open.
35
127742
2625
02:10
Only a little over half of recent college graduates
36
130367
3207
02:13
in the United States who get a higher education
37
133574
2313
02:15
actually are working in jobs that require that education.
38
135887
3463
02:19
This, of course, is not true for the students
39
139350
1840
02:21
who graduate from the top institutions,
40
141190
1952
02:23
but for many others, they do not get the value
41
143142
2632
02:25
for their time and their effort.
42
145774
3536
02:29
Tom Friedman, in his recent New York Times article,
43
149310
3030
02:32
captured, in the way that no one else could, the spirit behind our effort.
44
152340
4368
02:36
He said the big breakthroughs are what happen
45
156708
3120
02:39
when what is suddenly possible meets what is desperately necessary.
46
159828
3899
02:43
I've talked about what's desperately necessary.
47
163727
2621
02:46
Let's talk about what's suddenly possible.
48
166348
2512
02:48
What's suddenly possible was demonstrated by
49
168860
3119
02:51
three big Stanford classes,
50
171979
1568
02:53
each of which had an enrollment of 100,000 people or more.
51
173547
3880
02:57
So to understand this, let's look at one of those classes,
52
177427
3384
03:00
the Machine Learning class offered by my colleague
53
180811
1920
03:02
and cofounder Andrew Ng.
54
182731
1729
03:04
Andrew teaches one of the bigger Stanford classes.
55
184460
2319
03:06
It's a Machine Learning class,
56
186779
1209
03:07
and it has 400 people enrolled every time it's offered.
57
187988
3518
03:11
When Andrew taught the Machine Learning class to the general public,
58
191506
3265
03:14
it had 100,000 people registered.
59
194771
2616
03:17
So to put that number in perspective,
60
197387
2009
03:19
for Andrew to reach that same size audience
61
199396
2359
03:21
by teaching a Stanford class,
62
201755
1826
03:23
he would have to do that for 250 years.
63
203581
3926
03:27
Of course, he'd get really bored.
64
207507
3486
03:30
So, having seen the impact of this,
65
210993
2737
03:33
Andrew and I decided that we needed to really try and scale this up,
66
213730
3128
03:36
to bring the best quality education to as many people as we could.
67
216858
4120
03:40
So we formed Coursera,
68
220978
1495
03:42
whose goal is to take the best courses
69
222473
3137
03:45
from the best instructors at the best universities
70
225610
3317
03:48
and provide it to everyone around the world for free.
71
228927
4028
03:52
We currently have 43 courses on the platform
72
232955
2600
03:55
from four universities across a range of disciplines,
73
235555
3199
03:58
and let me show you a little bit of an overview
74
238754
1833
04:00
of what that looks like.
75
240587
3278
04:03
(Video) Robert Ghrist: Welcome to Calculus.
76
243865
1213
04:05
Ezekiel Emanuel: Fifty million people are uninsured.
77
245078
1880
04:06
Scott Page: Models help us design more effective institutions and policies.
78
246958
3271
04:10
We get unbelievable segregation.
79
250229
2408
04:12
Scott Klemmer: So Bush imagined that in the future,
80
252637
1792
04:14
you'd wear a camera right in the center of your head.
81
254429
2378
04:16
Mitchell Duneier: Mills wants the student of sociology to develop the quality of mind ...
82
256807
4254
04:21
RG: Hanging cable takes on the form of a hyperbolic cosine.
83
261061
3665
04:24
Nick Parlante: For each pixel in the image, set the red to zero.
84
264726
3071
04:27
Paul Offit: ... Vaccine allowed us to eliminate polio virus.
85
267797
2977
04:30
Dan Jurafsky: Does Lufthansa serve breakfast and San Jose? Well, that sounds funny.
86
270774
3623
04:34
Daphne Koller: So this is which coin you pick, and this is the two tosses.
87
274397
3616
04:38
Andrew Ng: So in large-scale machine learning, we'd like to come up with computational ...
88
278013
3687
04:41
(Applause)
89
281700
5609
04:47
DK: It turns out, maybe not surprisingly,
90
287309
2274
04:49
that students like getting the best content
91
289583
2238
04:51
from the best universities for free.
92
291821
2887
04:54
Since we opened the website in February,
93
294708
2522
04:57
we now have 640,000 students from 190 countries.
94
297230
4358
05:01
We have 1.5 million enrollments,
95
301588
2152
05:03
6 million quizzes in the 15 classes that have launched
96
303740
2850
05:06
so far have been submitted, and 14 million videos have been viewed.
97
306590
4916
05:11
But it's not just about the numbers,
98
311506
2518
05:14
it's also about the people.
99
314024
1641
05:15
Whether it's Akash, who comes from a small town in India
100
315665
2976
05:18
and would never have access in this case
101
318641
2175
05:20
to a Stanford-quality course
102
320816
1489
05:22
and would never be able to afford it.
103
322305
2515
05:24
Or Jenny, who is a single mother of two
104
324820
2038
05:26
and wants to hone her skills
105
326858
1967
05:28
so that she can go back and complete her master's degree.
106
328825
3135
05:31
Or Ryan, who can't go to school,
107
331960
3136
05:35
because his immune deficient daughter
108
335096
1865
05:36
can't be risked to have germs come into the house,
109
336961
3383
05:40
so he couldn't leave the house.
110
340344
1840
05:42
I'm really glad to say --
111
342184
1632
05:43
recently, we've been in correspondence with Ryan --
112
343816
2252
05:46
that this story had a happy ending.
113
346068
1932
05:48
Baby Shannon -- you can see her on the left --
114
348000
1903
05:49
is doing much better now,
115
349903
1351
05:51
and Ryan got a job by taking some of our courses.
116
351254
4198
05:55
So what made these courses so different?
117
355452
2244
05:57
After all, online course content has been available for a while.
118
357696
3720
06:01
What made it different was that this was real course experience.
119
361416
3712
06:05
It started on a given day,
120
365128
1726
06:06
and then the students would watch videos on a weekly basis
121
366854
3634
06:10
and do homework assignments.
122
370488
1855
06:12
And these would be real homework assignments
123
372343
1791
06:14
for a real grade, with a real deadline.
124
374134
3304
06:17
You can see the deadlines and the usage graph.
125
377438
2056
06:19
These are the spikes showing
126
379494
2088
06:21
that procrastination is global phenomenon.
127
381582
3789
06:25
(Laughter)
128
385371
2576
06:27
At the end of the course,
129
387947
1672
06:29
the students got a certificate.
130
389619
1856
06:31
They could present that certificate
131
391475
2160
06:33
to a prospective employer and get a better job,
132
393635
2153
06:35
and we know many students who did.
133
395788
2060
06:37
Some students took their certificate
134
397848
1919
06:39
and presented this to an educational institution at which they were enrolled
135
399767
3122
06:42
for actual college credit.
136
402889
1841
06:44
So these students were really getting something meaningful
137
404730
2214
06:46
for their investment of time and effort.
138
406944
2834
06:49
Let's talk a little bit about some of the components
139
409778
2555
06:52
that go into these courses.
140
412333
1892
06:54
The first component is that when you move away
141
414225
2628
06:56
from the constraints of a physical classroom
142
416853
2297
06:59
and design content explicitly for an online format,
143
419150
2840
07:01
you can break away from, for example,
144
421990
2528
07:04
the monolithic one-hour lecture.
145
424518
2415
07:06
You can break up the material, for example,
146
426933
1785
07:08
into these short, modular units of eight to 12 minutes,
147
428718
3376
07:12
each of which represents a coherent concept.
148
432094
2974
07:15
Students can traverse this material in different ways,
149
435068
2570
07:17
depending on their background, their skills or their interests.
150
437638
3704
07:21
So, for example, some students might benefit
151
441342
2520
07:23
from a little bit of preparatory material
152
443862
2760
07:26
that other students might already have.
153
446622
2071
07:28
Other students might be interested in a particular
154
448693
2440
07:31
enrichment topic that they want to pursue individually.
155
451133
3086
07:34
So this format allows us to break away
156
454219
3235
07:37
from the one-size-fits-all model of education,
157
457454
2824
07:40
and allows students to follow a much more personalized curriculum.
158
460278
3992
07:44
Of course, we all know as educators
159
464270
2343
07:46
that students don't learn by sitting and passively watching videos.
160
466613
3360
07:49
Perhaps one of the biggest components of this effort
161
469973
2945
07:52
is that we need to have students
162
472918
2592
07:55
who practice with the material
163
475510
2409
07:57
in order to really understand it.
164
477919
3156
08:01
There's been a range of studies that demonstrate the importance of this.
165
481075
3268
08:04
This one that appeared in Science last year, for example,
166
484343
2532
08:06
demonstrates that even simple retrieval practice,
167
486875
2832
08:09
where students are just supposed to repeat
168
489707
2792
08:12
what they already learned
169
492499
1400
08:13
gives considerably improved results
170
493899
1920
08:15
on various achievement tests down the line
171
495819
2269
08:18
than many other educational interventions.
172
498088
4304
08:22
We've tried to build in retrieval practice into the platform,
173
502392
2962
08:25
as well as other forms of practice in many ways.
174
505354
2254
08:27
For example, even our videos are not just videos.
175
507608
4144
08:31
Every few minutes, the video pauses
176
511752
2043
08:33
and the students get asked a question.
177
513795
2151
08:35
(Video) SP: ... These four things. Prospect theory, hyperbolic discounting,
178
515946
2221
08:38
status quo bias, base rate bias. They're all well documented.
179
518167
3092
08:41
So they're all well documented deviations from rational behavior.
180
521259
2767
08:44
DK: So here the video pauses,
181
524026
1624
08:45
and the student types in the answer into the box
182
525650
2256
08:47
and submits. Obviously they weren't paying attention.
183
527906
3223
08:51
(Laughter)
184
531144
869
08:52
So they get to try again,
185
532013
2010
08:54
and this time they got it right.
186
534023
2536
08:56
There's an optional explanation if they want.
187
536559
2193
08:58
And now the video moves on to the next part of the lecture.
188
538752
4257
09:03
This is a kind of simple question
189
543009
1878
09:04
that I as an instructor might ask in class,
190
544887
2081
09:06
but when I ask that kind of a question in class,
191
546968
2500
09:09
80 percent of the students
192
549468
1300
09:10
are still scribbling the last thing I said,
193
550768
1866
09:12
15 percent are zoned out on Facebook,
194
552634
3321
09:15
and then there's the smarty pants in the front row
195
555955
2456
09:18
who blurts out the answer
196
558411
1359
09:19
before anyone else has had a chance to think about it,
197
559770
2207
09:21
and I as the instructor am terribly gratified
198
561977
2872
09:24
that somebody actually knew the answer.
199
564849
1648
09:26
And so the lecture moves on before, really,
200
566497
2792
09:29
most of the students have even noticed that a question had been asked.
201
569289
3529
09:32
Here, every single student
202
572818
2607
09:35
has to engage with the material.
203
575425
2784
09:38
And of course these simple retrieval questions
204
578209
1936
09:40
are not the end of the story.
205
580145
1662
09:41
One needs to build in much more meaningful practice questions,
206
581807
2970
09:44
and one also needs to provide the students with feedback
207
584777
2353
09:47
on those questions.
208
587130
1663
09:48
Now, how do you grade the work of 100,000 students
209
588793
2888
09:51
if you do not have 10,000 TAs?
210
591681
3082
09:54
The answer is, you need to use technology
211
594763
2354
09:57
to do it for you.
212
597117
1495
09:58
Now, fortunately, technology has come a long way,
213
598612
2648
10:01
and we can now grade a range of interesting types of homework.
214
601260
3268
10:04
In addition to multiple choice
215
604528
1527
10:06
and the kinds of short answer questions that you saw in the video,
216
606055
3153
10:09
we can also grade math, mathematical expressions
217
609208
3260
10:12
as well as mathematical derivations.
218
612468
1952
10:14
We can grade models, whether it's
219
614420
2874
10:17
financial models in a business class
220
617294
2176
10:19
or physical models in a science or engineering class
221
619470
2984
10:22
and we can grade some pretty sophisticated programming assignments.
222
622454
3744
10:26
Let me show you one that's actually pretty simple
223
626198
1919
10:28
but fairly visual.
224
628117
1480
10:29
This is from Stanford's Computer Science 101 class,
225
629597
2477
10:32
and the students are supposed to color-correct
226
632074
1604
10:33
that blurry red image.
227
633678
1592
10:35
They're typing their program into the browser,
228
635270
2018
10:37
and you can see they didn't get it quite right, Lady Liberty is still seasick.
229
637288
4058
10:41
And so, the student tries again, and now they got it right, and they're told that,
230
641346
3756
10:45
and they can move on to the next assignment.
231
645102
2359
10:47
This ability to interact actively with the material
232
647461
3148
10:50
and be told when you're right or wrong
233
650609
1684
10:52
is really essential to student learning.
234
652293
3126
10:55
Now, of course we cannot yet grade
235
655419
2275
10:57
the range of work that one needs for all courses.
236
657694
2834
11:00
Specifically, what's lacking is the kind of critical thinking work
237
660528
3301
11:03
that is so essential in such disciplines
238
663829
1922
11:05
as the humanities, the social sciences, business and others.
239
665751
3597
11:09
So we tried to convince, for example,
240
669348
2249
11:11
some of our humanities faculty
241
671597
1616
11:13
that multiple choice was not such a bad strategy.
242
673213
2696
11:15
That didn't go over really well.
243
675909
2191
11:18
So we had to come up with a different solution.
244
678100
2433
11:20
And the solution we ended up using is peer grading.
245
680533
3074
11:23
It turns out that previous studies show,
246
683607
2422
11:26
like this one by Saddler and Good,
247
686029
1672
11:27
that peer grading is a surprisingly effective strategy
248
687701
2488
11:30
for providing reproducible grades.
249
690189
3214
11:33
It was tried only in small classes,
250
693403
1770
11:35
but there it showed, for example,
251
695173
1487
11:36
that these student-assigned grades on the y-axis
252
696660
2482
11:39
are actually very well correlated
253
699142
1311
11:40
with the teacher-assigned grade on the x-axis.
254
700453
2296
11:42
What's even more surprising is that self-grades,
255
702749
3160
11:45
where the students grade their own work critically --
256
705909
2311
11:48
so long as you incentivize them properly
257
708220
1737
11:49
so they can't give themselves a perfect score --
258
709957
1938
11:51
are actually even better correlated with the teacher grades.
259
711895
3191
11:55
And so this is an effective strategy
260
715086
1607
11:56
that can be used for grading at scale,
261
716693
2104
11:58
and is also a useful learning strategy for the students,
262
718797
2736
12:01
because they actually learn from the experience.
263
721533
2255
12:03
So we now have the largest peer-grading pipeline ever devised,
264
723788
4649
12:08
where tens of thousands of students
265
728437
2504
12:10
are grading each other's work,
266
730941
1198
12:12
and quite successfully, I have to say.
267
732139
3069
12:15
But this is not just about students
268
735208
2260
12:17
sitting alone in their living room working through problems.
269
737468
3041
12:20
Around each one of our courses,
270
740509
1807
12:22
a community of students had formed,
271
742316
2160
12:24
a global community of people
272
744476
1880
12:26
around a shared intellectual endeavor.
273
746356
2532
12:28
What you see here is a self-generated map
274
748888
2652
12:31
from students in our Princeton Sociology 101 course,
275
751540
2961
12:34
where they have put themselves on a world map,
276
754501
2759
12:37
and you can really see the global reach of this kind of effort.
277
757260
2960
12:40
Students collaborated in these courses in a variety of different ways.
278
760220
4567
12:44
First of all, there was a question and answer forum,
279
764787
2639
12:47
where students would pose questions,
280
767426
2144
12:49
and other students would answer those questions.
281
769570
2424
12:51
And the really amazing thing is,
282
771994
1713
12:53
because there were so many students,
283
773707
1670
12:55
it means that even if a student posed a question
284
775377
2365
12:57
at 3 o'clock in the morning,
285
777742
1632
12:59
somewhere around the world,
286
779374
1582
13:00
there would be somebody who was awake
287
780956
2074
13:03
and working on the same problem.
288
783030
2313
13:05
And so, in many of our courses,
289
785343
1958
13:07
the median response time for a question
290
787301
2329
13:09
on the question and answer forum was 22 minutes.
291
789630
3418
13:13
Which is not a level of service I have ever offered to my Stanford students.
292
793048
4577
13:17
(Laughter)
293
797625
1341
13:18
And you can see from the student testimonials
294
798966
1942
13:20
that students actually find
295
800908
1687
13:22
that because of this large online community,
296
802595
2521
13:25
they got to interact with each other in many ways
297
805116
2599
13:27
that were deeper than they did in the context of the physical classroom.
298
807715
4193
13:31
Students also self-assembled,
299
811908
2344
13:34
without any kind of intervention from us,
300
814252
1863
13:36
into small study groups.
301
816115
1903
13:38
Some of these were physical study groups
302
818018
2362
13:40
along geographical constraints
303
820380
1826
13:42
and met on a weekly basis to work through problem sets.
304
822206
2722
13:44
This is the San Francisco study group,
305
824928
1900
13:46
but there were ones all over the world.
306
826828
2319
13:49
Others were virtual study groups,
307
829147
2032
13:51
sometimes along language lines or along cultural lines,
308
831179
2989
13:54
and on the bottom left there,
309
834168
1444
13:55
you see our multicultural universal study group
310
835612
3796
13:59
where people explicitly wanted to connect
311
839408
1763
14:01
with people from other cultures.
312
841171
3006
14:04
There are some tremendous opportunities
313
844177
2111
14:06
to be had from this kind of framework.
314
846288
3325
14:09
The first is that it has the potential of giving us
315
849613
3654
14:13
a completely unprecedented look
316
853267
2434
14:15
into understanding human learning.
317
855701
2289
14:17
Because the data that we can collect here is unique.
318
857990
3463
14:21
You can collect every click, every homework submission,
319
861453
4009
14:25
every forum post from tens of thousands of students.
320
865462
4363
14:29
So you can turn the study of human learning
321
869825
2343
14:32
from the hypothesis-driven mode
322
872168
1933
14:34
to the data-driven mode, a transformation that,
323
874101
2858
14:36
for example, has revolutionized biology.
324
876959
3041
14:40
You can use these data to understand fundamental questions
325
880000
3424
14:43
like, what are good learning strategies
326
883424
1880
14:45
that are effective versus ones that are not?
327
885304
2696
14:48
And in the context of particular courses,
328
888000
2240
14:50
you can ask questions
329
890240
1537
14:51
like, what are some of the misconceptions that are more common
330
891777
3255
14:55
and how do we help students fix them?
331
895032
2177
14:57
So here's an example of that,
332
897209
1424
14:58
also from Andrew's Machine Learning class.
333
898633
2016
15:00
This is a distribution of wrong answers
334
900649
2208
15:02
to one of Andrew's assignments.
335
902857
1610
15:04
The answers happen to be pairs of numbers,
336
904467
1893
15:06
so you can draw them on this two-dimensional plot.
337
906360
2271
15:08
Each of the little crosses that you see is a different wrong answer.
338
908631
3778
15:12
The big cross at the top left
339
912409
2406
15:14
is where 2,000 students
340
914815
2148
15:16
gave the exact same wrong answer.
341
916963
3045
15:20
Now, if two students in a class of 100
342
920008
2327
15:22
give the same wrong answer,
343
922335
1287
15:23
you would never notice.
344
923622
1351
15:24
But when 2,000 students give the same wrong answer,
345
924973
2560
15:27
it's kind of hard to miss.
346
927533
1697
15:29
So Andrew and his students went in,
347
929230
2192
15:31
looked at some of those assignments,
348
931422
1520
15:32
understood the root cause of the misconception,
349
932942
4088
15:37
and then they produced a targeted error message
350
937030
2520
15:39
that would be provided to every student
351
939550
2249
15:41
whose answer fell into that bucket,
352
941799
2179
15:43
which means that students who made that same mistake
353
943978
2084
15:46
would now get personalized feedback
354
946062
2026
15:48
telling them how to fix their misconception much more effectively.
355
948088
4399
15:52
So this personalization is something that one can then build
356
952487
3811
15:56
by having the virtue of large numbers.
357
956298
3140
15:59
Personalization is perhaps
358
959438
2312
16:01
one of the biggest opportunities here as well,
359
961750
2423
16:04
because it provides us with the potential
360
964173
2345
16:06
of solving a 30-year-old problem.
361
966518
2690
16:09
Educational researcher Benjamin Bloom, in 1984,
362
969208
3349
16:12
posed what's called the 2 sigma problem,
363
972557
2251
16:14
which he observed by studying three populations.
364
974808
3062
16:17
The first is the population that studied in a lecture-based classroom.
365
977870
3608
16:21
The second is a population of students that studied
366
981478
2777
16:24
using a standard lecture-based classroom,
367
984255
1719
16:25
but with a mastery-based approach,
368
985974
2080
16:28
so the students couldn't move on to the next topic
369
988054
1920
16:29
before demonstrating mastery of the previous one.
370
989974
3354
16:33
And finally, there was a population of students
371
993328
2294
16:35
that were taught in a one-on-one instruction using a tutor.
372
995622
4528
16:40
The mastery-based population was a full standard deviation,
373
1000150
3272
16:43
or sigma, in achievement scores better
374
1003422
2288
16:45
than the standard lecture-based class,
375
1005710
2394
16:48
and the individual tutoring gives you 2 sigma
376
1008104
2144
16:50
improvement in performance.
377
1010248
1830
16:52
To understand what that means,
378
1012078
1463
16:53
let's look at the lecture-based classroom,
379
1013541
1833
16:55
and let's pick the median performance as a threshold.
380
1015374
2919
16:58
So in a lecture-based class,
381
1018293
1338
16:59
half the students are above that level and half are below.
382
1019631
3879
17:03
In the individual tutoring instruction,
383
1023510
2098
17:05
98 percent of the students are going to be above that threshold.
384
1025608
4801
17:10
Imagine if we could teach so that 98 percent of our students
385
1030409
3920
17:14
would be above average.
386
1034329
2198
17:16
Hence, the 2 sigma problem.
387
1036527
3423
17:19
Because we cannot afford, as a society,
388
1039950
2399
17:22
to provide every student with an individual human tutor.
389
1042349
3072
17:25
But maybe we can afford to provide each student
390
1045421
2249
17:27
with a computer or a smartphone.
391
1047670
2019
17:29
So the question is, how can we use technology
392
1049689
2189
17:31
to push from the left side of the graph, from the blue curve,
393
1051878
3375
17:35
to the right side with the green curve?
394
1055253
2738
17:37
Mastery is easy to achieve using a computer,
395
1057991
2337
17:40
because a computer doesn't get tired
396
1060328
1405
17:41
of showing you the same video five times.
397
1061733
3073
17:44
And it doesn't even get tired of grading the same work multiple times,
398
1064806
3251
17:48
we've seen that in many of the examples that I've shown you.
399
1068062
3026
17:51
And even personalization
400
1071088
1854
17:52
is something that we're starting to see the beginnings of,
401
1072942
2136
17:55
whether it's via the personalized trajectory through the curriculum
402
1075078
3192
17:58
or some of the personalized feedback that we've shown you.
403
1078270
3264
18:01
So the goal here is to try and push,
404
1081534
2488
18:04
and see how far we can get towards the green curve.
405
1084022
3497
18:07
So, if this is so great, are universities now obsolete?
406
1087519
5359
18:12
Well, Mark Twain certainly thought so.
407
1092878
2992
18:15
He said that, "College is a place where a professor's lecture notes
408
1095870
2545
18:18
go straight to the students' lecture notes,
409
1098415
1703
18:20
without passing through the brains of either."
410
1100118
2376
18:22
(Laughter)
411
1102494
4047
18:26
I beg to differ with Mark Twain, though.
412
1106541
2668
18:29
I think what he was complaining about is not
413
1109209
2665
18:31
universities but rather the lecture-based format
414
1111874
2750
18:34
that so many universities spend so much time on.
415
1114624
2784
18:37
So let's go back even further, to Plutarch,
416
1117408
3159
18:40
who said that, "The mind is not a vessel that needs filling,
417
1120567
2227
18:42
but wood that needs igniting."
418
1122794
2023
18:44
And maybe we should spend less time at universities
419
1124817
2190
18:47
filling our students' minds with content
420
1127007
2571
18:49
by lecturing at them, and more time igniting their creativity,
421
1129578
3800
18:53
their imagination and their problem-solving skills
422
1133378
3255
18:56
by actually talking with them.
423
1136633
2498
18:59
So how do we do that?
424
1139131
1367
19:00
We do that by doing active learning in the classroom.
425
1140498
3431
19:03
So there's been many studies, including this one,
426
1143929
2449
19:06
that show that if you use active learning,
427
1146378
2080
19:08
interacting with your students in the classroom,
428
1148458
2416
19:10
performance improves on every single metric --
429
1150874
2696
19:13
on attendance, on engagement and on learning
430
1153570
2449
19:16
as measured by a standardized test.
431
1156019
2055
19:18
You can see, for example, that the achievement score
432
1158074
1864
19:19
almost doubles in this particular experiment.
433
1159938
2870
19:22
So maybe this is how we should spend our time at universities.
434
1162808
4401
19:27
So to summarize, if we could offer a top quality education
435
1167209
4577
19:31
to everyone around the world for free,
436
1171786
1903
19:33
what would that do? Three things.
437
1173689
2821
19:36
First it would establish education as a fundamental human right,
438
1176510
3421
19:39
where anyone around the world
439
1179931
1366
19:41
with the ability and the motivation
440
1181297
1921
19:43
could get the skills that they need
441
1183218
1951
19:45
to make a better life for themselves,
442
1185169
1585
19:46
their families and their communities.
443
1186754
2017
19:48
Second, it would enable lifelong learning.
444
1188771
2631
19:51
It's a shame that for so many people,
445
1191402
1951
19:53
learning stops when we finish high school or when we finish college.
446
1193353
3312
19:56
By having this amazing content be available,
447
1196665
2481
19:59
we would be able to learn something new
448
1199146
2743
20:01
every time we wanted,
449
1201889
1136
20:03
whether it's just to expand our minds
450
1203025
1329
20:04
or it's to change our lives.
451
1204354
1959
20:06
And finally, this would enable a wave of innovation,
452
1206313
3145
20:09
because amazing talent can be found anywhere.
453
1209458
3072
20:12
Maybe the next Albert Einstein or the next Steve Jobs
454
1212530
3008
20:15
is living somewhere in a remote village in Africa.
455
1215538
2615
20:18
And if we could offer that person an education,
456
1218153
2656
20:20
they would be able to come up with the next big idea
457
1220809
2356
20:23
and make the world a better place for all of us.
458
1223165
2404
20:25
Thank you very much.
459
1225569
1160
20:26
(Applause)
460
1226729
7583
About this website

This site will introduce you to YouTube videos that are useful for learning English. You will see English lessons taught by top-notch teachers from around the world. Double-click on the English subtitles displayed on each video page to play the video from there. The subtitles scroll in sync with the video playback. If you have any comments or requests, please contact us using this contact form.

https://forms.gle/WvT1wiN1qDtmnspy7