1
00:00:00,380 --> 00:00:01,690
Accessibility testing.

2
00:00:01,690 --> 00:00:04,110
We all know we need to conduct accessibility

3
00:00:04,120 --> 00:00:05,720
testing of our entire website.

4
00:00:05,770 --> 00:00:08,200
There are two types of accessibility testing

5
00:00:08,220 --> 00:00:09,600
that you’ve probably heard

6
00:00:09,600 --> 00:00:11,600
about: manual accessibility testing

7
00:00:11,600 --> 00:00:13,680
and automated accessibility testing.

8
00:00:13,680 --> 00:00:15,280
But what is the difference and

9
00:00:15,280 --> 00:00:16,960
how can you choose between them?

10
00:00:16,960 --> 00:00:18,410
So let’s talk about manual

11
00:00:18,410 --> 00:00:20,240
accessibility testing – what it

12
00:00:20,280 --> 00:00:21,950
entails and its pros and cons.

13
00:00:21,980 --> 00:00:23,530
Manual testing - pros

14
00:00:23,550 --> 00:00:25,760
When it comes to manual accessibility

15
00:00:25,760 --> 00:00:27,450
testing, it can identify all

16
00:00:27,470 --> 00:00:29,740
accessibility errors in your website,

17
00:00:29,740 --> 00:00:30,760
which is great!

18
00:00:30,760 --> 00:00:32,760
But you need to remember on average,

19
00:00:32,760 --> 00:00:35,150
it takes one full day to manually test

20
00:00:35,150 --> 00:00:37,440
one webpage. That’s just not realistic

21
00:00:37,440 --> 00:00:39,440
when you have a large website.

22
00:00:39,440 --> 00:00:41,440
Manual testing - workarounds

23
00:00:41,440 --> 00:00:43,390
So how people have managed to work

24
00:00:43,390 --> 00:00:45,230
around this is that they select a

25
00:00:45,230 --> 00:00:46,970
sub-set of pages to test, and we’ll

26
00:00:46,970 --> 00:00:48,970
see how that can cause problems.

27
00:00:48,970 --> 00:00:50,650
Manual testing - cons

28
00:00:50,670 --> 00:00:52,750
Manual testing takes a lot of time,

29
00:00:52,760 --> 00:00:55,050
but there is another problem with manual

30
00:00:55,050 --> 00:00:57,440
testing. Without a proper testing framework,

31
00:00:57,440 --> 00:00:59,760
each accessibility tester tends to come up

32
00:00:59,760 --> 00:01:01,840
with different results. This means that

33
00:01:01,840 --> 00:01:03,840
you could test the same page by three

34
00:01:03,840 --> 00:01:06,110
different people and have three different

35
00:01:06,110 --> 00:01:07,150
numbers of errors.

36
00:01:07,150 --> 00:01:09,560
What this means is that manual accessibility

37
00:01:09,560 --> 00:01:11,560
testing isn’t really reliable. Only a

38
00:01:11,560 --> 00:01:13,660
small number of pages are tested and

39
00:01:13,660 --> 00:01:15,660
therefore fixed, and if there isn’t a

40
00:01:15,660 --> 00:01:17,770
testing framework, even these pages are

41
00:01:17,770 --> 00:01:19,770
likely to not meet accessibility

42
00:01:19,770 --> 00:01:21,160
standards consistently.

43
00:01:21,180 --> 00:01:23,690
So let’s talk about automated accessibility

44
00:01:23,720 --> 00:01:26,360
testing – what it entails and its pros and cons.

45
00:01:26,380 --> 00:01:28,520
Automated testing - pros

46
00:01:28,540 --> 00:01:30,890
When it comes to automated accessibility

47
00:01:30,890 --> 00:01:33,980
testing, it can test all pages on your website,

48
00:01:33,980 --> 00:01:35,980
and quickly  – which is great!

49
00:01:35,980 --> 00:01:37,980
And there are no people required! Just

50
00:01:37,980 --> 00:01:39,980
press a big red button and it’s done!

51
00:01:39,980 --> 00:01:41,980
But you need to remember on average,

52
00:01:41,980 --> 00:01:44,190
automated testing tools can only find

53
00:01:44,220 --> 00:01:46,680
30 percent of all accessibility errors.

54
00:01:46,680 --> 00:01:48,910
That means 30 percent of errors can be

55
00:01:48,910 --> 00:01:51,180
identified by automated accessibility

56
00:01:51,210 --> 00:01:53,260
testing tools, but it also means that

57
00:01:53,260 --> 00:01:55,920
70 percent of all accessibility errors are

58
00:01:55,920 --> 00:01:57,320
missed by automated

59
00:01:57,340 --> 00:01:59,310
accessibility testing tools.

60
00:01:59,310 --> 00:02:01,310
Automated testing - cons

61
00:02:01,310 --> 00:02:04,040
Although automated testing doesn’t require

62
00:02:04,040 --> 00:02:06,680
people, there remains a huge amount of testing

63
00:02:06,680 --> 00:02:08,440
that still needs to be done.

64
00:02:08,440 --> 00:02:11,230
What this means is that automated accessibility

65
00:02:11,230 --> 00:02:13,960
testing isn’t really reliable. Only a small

66
00:02:13,960 --> 00:02:15,950
number of errors are tested

67
00:02:15,950 --> 00:02:17,280
and therefore fixed.

68
00:02:17,280 --> 00:02:19,690
The problem is that automated accessibility

69
00:02:19,690 --> 00:02:22,400
testing is fast, but it can only identify

70
00:02:22,400 --> 00:02:24,880
30 percent of all accessibility errors

71
00:02:24,890 --> 00:02:27,600
(for example, images missing ALT attributes).

72
00:02:27,600 --> 00:02:30,410
The remainder of the testing needs to be done

73
00:02:30,410 --> 00:02:33,480
manually, and that takes a lot of people power.

74
00:02:34,160 --> 00:02:37,840
AccessibilityOz has a solution: OzART.

75
00:02:37,840 --> 00:02:40,300
OzART is different. OzART is an

76
00:02:40,300 --> 00:02:41,840
accessibility system.

77
00:02:41,840 --> 00:02:44,400
Yes, OzART has automated testing, and

78
00:02:44,400 --> 00:02:46,700
can identify approximately 30 percent of

79
00:02:46,700 --> 00:02:49,340
errors this way. But through our accessibility

80
00:02:49,340 --> 00:02:51,150
testing framework, 60 percent

81
00:02:51,150 --> 00:02:53,150
of all accessibility testing

82
00:02:53,150 --> 00:02:55,480
can be done within OzART. We call this

83
00:02:55,480 --> 00:02:58,750
Facilitated Automated Testing, as the tool

84
00:02:58,750 --> 00:03:01,240
categorises information so it is easy –

85
00:03:01,260 --> 00:03:03,000
and most importantly, quick! –

86
00:03:03,000 --> 00:03:05,500
for manual testing to be done, and we’ll

87
00:03:05,500 --> 00:03:07,710
look at an example on the next slide.

88
00:03:07,710 --> 00:03:09,980
The remaining 10 percent of the testing

89
00:03:10,000 --> 00:03:12,110
requires a tester to go back to the site,

90
00:03:12,110 --> 00:03:14,860
but OzART provides detailed instructions on

91
00:03:14,880 --> 00:03:17,500
how to complete the testing, which means that

92
00:03:17,500 --> 00:03:20,330
the results will be consistent across testers.

93
00:03:20,350 --> 00:03:23,050
What is OzART facilitated testing?

94
00:03:23,050 --> 00:03:25,630
Basically, the tool categorises information

95
00:03:25,630 --> 00:03:28,190
in a way for a human to quickly determine

96
00:03:28,190 --> 00:03:30,280
whether something passes or fails

97
00:03:30,280 --> 00:03:31,950
 accessibility requirements.

98
00:03:31,950 --> 00:03:34,320
In this example, there are four images.

99
00:03:34,320 --> 00:03:37,530
The first one is of a nuclear explosion that has

100
00:03:37,550 --> 00:03:41,770
the ALT attribute of "08-29-2013nucleartest",

101
00:03:41,790 --> 00:03:44,350
which is not a good description of the image.

102
00:03:44,350 --> 00:03:47,150
Remember that ALT attributes are read aloud by

103
00:03:47,150 --> 00:03:50,240
screen readers, so people with vision impairments

104
00:03:50,300 --> 00:03:52,220
know what the image is displaying.

105
00:03:52,220 --> 00:03:54,640
The second ALT attribute is even worse.

106
00:03:54,640 --> 00:03:57,310
 It reads “31319”.

107
00:03:57,310 --> 00:03:59,310
The next two images have

108
00:03:59,310 --> 00:04:01,310
much better ALT attributes,

109
00:04:01,310 --> 00:04:03,960
the first being “A Mongolian family in Tehran,

110
00:04:03,960 --> 00:04:06,910
Mongolia who cultivates potatoes in their yard”

111
00:04:06,910 --> 00:04:09,530
and “Aerial view of UN Headquarters

112
00:04:09,530 --> 00:04:11,280
in New York City.”

113
00:04:11,280 --> 00:04:13,520
An experienced accessibility tester

114
00:04:13,520 --> 00:04:15,200
can identify which of these

115
00:04:15,200 --> 00:04:17,020
pass and fail within seconds.

116
00:04:17,040 --> 00:04:19,200
If they had to go to the site and test this

117
00:04:19,200 --> 00:04:21,370
manually, it could take up to ten minutes.

118
00:04:21,370 --> 00:04:23,480
The other great thing about OzART

119
00:04:23,480 --> 00:04:25,480
is that it groups images if they

120
00:04:25,480 --> 00:04:27,480
occur on your site more than once.

121
00:04:27,480 --> 00:04:29,480
That way, a tester can review it once

122
00:04:29,480 --> 00:04:31,260
 and determine if it passes

123
00:04:31,260 --> 00:04:33,260
or fails. If this was done via manual

124
00:04:33,260 --> 00:04:35,500
testing, the tester would need to review

125
00:04:35,500 --> 00:04:37,500
every single image on the site.

126
00:04:37,500 --> 00:04:38,910
OzART - pros

127
00:04:38,920 --> 00:04:41,550
When it comes to OzART, it can test all pages

128
00:04:41,560 --> 00:04:43,800
in your website just like an automated

129
00:04:43,800 --> 00:04:46,240
accessibility testing tool, and it provides

130
00:04:46,240 --> 00:04:48,560
facilitated testing to make the remaining

131
00:04:48,560 --> 00:04:50,060
testing more streamlined,

132
00:04:50,060 --> 00:04:52,060
and most importantly, consistent.

133
00:04:52,060 --> 00:04:54,520
How does this translate into the real world?

134
00:04:54,520 --> 00:04:56,520
AccessibilityOz has a contract

135
00:04:56,520 --> 00:04:58,040
 with a Fortune 10 company

136
00:04:58,040 --> 00:05:01,360
to test five 15,000 page websites

137
00:05:01,360 --> 00:05:03,920
(that’s 7,500 pages)

138
00:05:03,920 --> 00:05:06,620
every week in under 30 hours.

139
00:05:06,620 --> 00:05:08,620
That’s four minutes per page.

140
00:05:08,620 --> 00:05:11,180
What this means is that OzART tests for

141
00:05:11,180 --> 00:05:13,760
all errors and all pages, which means that

142
00:05:13,760 --> 00:05:16,800
your entire website will be compliant.

143
00:05:16,816 --> 00:05:19,456
Talk to us about OzART today.

144
00:05:19,472 --> 00:05:23,152
Visit www.accessibilityoz.com to learn more

145
00:05:23,168 --> 00:05:26,750
about our products, services and free resources.

146
00:05:26,750 --> 00:05:28,750
[End of captioning]