1
00:00:01,701 --> 00:00:03,194
Voiceover: Let's look at
the difference between

2
00:00:03,194 --> 00:00:06,212
top-down and bottom-up processing.

3
00:00:06,443 --> 00:00:10,084
So, what is bottom-up processing?

4
00:00:14,757 --> 00:00:19,382
Bottom-up processing basically
begins with the stimulus,

5
00:00:19,382 --> 00:00:21,912
so let's imagine that
we're looking at something,

6
00:00:21,912 --> 00:00:24,360
or let's say I'm looking at a banana.

7
00:00:24,360 --> 00:00:26,645
The banana is sitting there and

8
00:00:26,645 --> 00:00:29,100
it influences what we perceive.

9
00:00:29,470 --> 00:00:38,211
So, stimulus influences what
we perceive, our perception.

10
00:00:39,556 --> 00:00:43,525
So, if I know absolutely
nothing about something,

11
00:00:43,525 --> 00:00:47,305
then the stimulus or whatever
it is I'm looking at,

12
00:00:47,305 --> 00:00:49,691
yet I don't know anything
about, I've never seen it,

13
00:00:49,691 --> 00:00:53,450
I don't have any preconceived
cognitive constructs

14
00:00:53,450 --> 00:00:55,259
about what it is I'm looking at.

15
00:00:55,259 --> 00:00:59,254
The stimulus basically is
influencing my perception.

16
00:00:59,254 --> 00:01:02,115
So, for example let's imagine that

17
00:01:02,115 --> 00:01:05,644
I'm looking at a cockpit of a plane.

18
00:01:05,731 --> 00:01:07,234
I'm not a pilot, so I'm not really

19
00:01:07,234 --> 00:01:08,773
too familiar with everything and

20
00:01:08,773 --> 00:01:12,266
everything kinda looks fairly confusing.

21
00:01:12,661 --> 00:01:15,414
So, basically all the different stimuli,

22
00:01:15,414 --> 00:01:18,581
so this stimulus, a bunch of gauges,

23
00:01:18,581 --> 00:01:20,696
and this rudder-looking thing,

24
00:01:20,696 --> 00:01:23,300
I'm basically looking at all the different

25
00:01:23,300 --> 00:01:27,419
little parts of something
that is new and novel to me,

26
00:01:27,419 --> 00:01:31,937
and trying to kind of comprehend
what it is I'm looking at.

27
00:01:31,937 --> 00:01:32,980
So, this is bottom-up.

28
00:01:32,980 --> 00:01:37,300
This is when you start
with no preconceived idea

29
00:01:37,300 --> 00:01:38,631
of what it is that you're looking at,

30
00:01:38,631 --> 00:01:42,202
and allow the stimulus to influence

31
00:01:42,202 --> 00:01:44,901
your perception of what it
is that you're looking at.

32
00:01:44,901 --> 00:01:47,797
So, bottom-up processing is data-driven,

33
00:01:50,220 --> 00:01:52,408
and your perception of what
it is that you're looking at

34
00:01:52,408 --> 00:01:56,909
directs your cognitive
awareness of the object.

35
00:01:57,498 --> 00:02:04,358
So, in contrast, top-down processing

36
00:02:05,588 --> 00:02:07,577
basically uses your background knowledge,

37
00:02:07,577 --> 00:02:10,220
so uses your background knowledge

38
00:02:13,904 --> 00:02:18,447
to influence perception.

39
00:02:20,508 --> 00:02:23,164
So, let's look at this example over here.

40
00:02:23,883 --> 00:02:27,916
So, what we're actually
seeing are a bunch of circles,

41
00:02:28,391 --> 00:02:29,816
they are just a bunch of circles and then

42
00:02:29,816 --> 00:02:35,994
inside the circle there are
a couple of lines drawn.

43
00:02:35,994 --> 00:02:39,657
So, we are looking at this set of circles,

44
00:02:39,657 --> 00:02:42,490
these white circles with
lines drawn inside of them.

45
00:02:42,490 --> 00:02:45,210
We are creating this cube.

46
00:02:45,210 --> 00:02:47,430
We're basically taking
these lines and then

47
00:02:47,430 --> 00:02:49,447
putting them together in
order to create a cube.

48
00:02:49,447 --> 00:02:51,436
Even though the stimulus itself,

49
00:02:51,436 --> 00:02:53,910
which is the circles with the lines,

50
00:02:53,910 --> 00:02:55,898
actually doesn't draw a cube because

51
00:02:55,898 --> 00:02:57,733
there are these black spaces over here,

52
00:02:57,733 --> 00:03:00,586
and there is absolutely
nothing in the black spaces,

53
00:03:00,586 --> 00:03:03,411
but our brains are basically
taking this information

54
00:03:03,411 --> 00:03:06,534
and using our knowledge of cubes

55
00:03:06,534 --> 00:03:08,163
and what they're supposed to look like,

56
00:03:08,163 --> 00:03:12,436
we're recreating a cube despite a lack of

57
00:03:12,436 --> 00:03:15,307
a cube actually being
present in the image.

58
00:03:15,307 --> 00:03:16,650
So, that's top-down processing.

59
00:03:16,650 --> 00:03:18,790
It's using your background information,

60
00:03:18,790 --> 00:03:20,534
your background knowledge, your learning,

61
00:03:20,534 --> 00:03:22,379
your expectations, in order to influence

62
00:03:22,379 --> 00:03:24,466
what it is that we're perceiving.

63
00:03:24,466 --> 00:03:26,276
So, in other words, it's theory-driven.

64
00:03:26,276 --> 00:03:28,949
We look at this and we assume that

65
00:03:28,949 --> 00:03:30,876
they're trying to represent a cube,

66
00:03:30,876 --> 00:03:33,449
even though one's not
actually drawn there,

67
00:03:33,449 --> 00:03:35,950
and we're using that theory in order to

68
00:03:35,950 --> 00:03:37,723
shape our cognitive understanding

69
00:03:37,723 --> 00:03:40,133
of what it is that we're looking at.

70
00:03:41,205 --> 00:03:43,076
So, our perception, our behavior

71
00:03:43,076 --> 00:03:45,032
is influenced by our expectations,

72
00:03:45,032 --> 00:03:46,900
which is top-down processing.

73
00:03:46,900 --> 00:03:49,546
So, we're using what's
already in our heads

74
00:03:49,546 --> 00:03:52,471
in order to perceive what
it is that we're looking at,

75
00:03:52,471 --> 00:03:55,116
whereas in bottom-up
processing we're using

76
00:03:55,116 --> 00:03:58,857
the stimulus itself in order
to drive our perception.

77
00:03:58,857 --> 00:04:00,424
So, another good example of top-down

78
00:04:00,424 --> 00:04:02,394
processing would be "Where's Waldo?"

79
00:04:02,394 --> 00:04:03,324
So, in "Where's Waldo?", we have a

80
00:04:03,324 --> 00:04:04,694
mental idea of what we're trying to do,

81
00:04:04,694 --> 00:04:07,192
which is to find Waldo amidst this

82
00:04:07,192 --> 00:04:10,295
really jumbled mess of a picture.

83
00:04:10,295 --> 00:04:12,148
So, if we were using bottom-up processing

84
00:04:12,148 --> 00:04:13,957
in order to look at this we would just

85
00:04:13,957 --> 00:04:15,866
be seeing a whole bunch of little people,

86
00:04:15,866 --> 00:04:18,658
and we wouldn't really be goal-driven,

87
00:04:18,658 --> 00:04:21,310
we wouldn't be trying to do anything,

88
00:04:21,310 --> 00:04:23,756
but with top-down
processing we have a goal,

89
00:04:23,756 --> 00:04:26,286
and we're able to look
through here to find Waldo.