1
00:00:02,970 --> 00:00:05,340
[Instructor] This is the
second and final part

2
00:00:05,340 --> 00:00:08,913
of the qualitative research
lecture for this week.

3
00:00:12,840 --> 00:00:15,840
I wanna talk to you
now about three factors

4
00:00:15,840 --> 00:00:20,520
of how you design your interview guide.

5
00:00:20,520 --> 00:00:23,103
What kinds of questions do you ask?

6
00:00:26,910 --> 00:00:29,880
The first one is the structure,

7
00:00:29,880 --> 00:00:34,200
and the first option of these
is a structured interview.

8
00:00:34,200 --> 00:00:37,200
In here you have a precise script.

9
00:00:37,200 --> 00:00:40,920
You read the questions word for word

10
00:00:40,920 --> 00:00:43,920
always in the same order.

11
00:00:43,920 --> 00:00:48,540
You have no deviations,
and because of this,

12
00:00:48,540 --> 00:00:53,013
you'll get more predictable
and therefore reliable answers.

13
00:00:59,670 --> 00:01:04,620
At the other end of the
spectrum is unstructured,

14
00:01:04,620 --> 00:01:06,540
where it's much more freeform.

15
00:01:06,540 --> 00:01:09,870
You just cover the key points.

16
00:01:09,870 --> 00:01:13,110
You can drop and add questions.

17
00:01:13,110 --> 00:01:16,500
And in this way because you're adapting it

18
00:01:16,500 --> 00:01:19,080
to what the respondent is saying,

19
00:01:19,080 --> 00:01:24,000
and sort of allowing the
interview to follow the thread

20
00:01:24,000 --> 00:01:25,500
of the conversation,

21
00:01:25,500 --> 00:01:27,993
you'll probably get more valid answers.

22
00:01:32,220 --> 00:01:33,750
As a rule of thumb,

23
00:01:33,750 --> 00:01:37,770
we generally tend to use
a structured interview

24
00:01:37,770 --> 00:01:42,090
if there are multiple interviewers.

25
00:01:42,090 --> 00:01:45,090
So an example is this
class when you'll be,

26
00:01:45,090 --> 00:01:48,480
each of you will be going
out and doing interviews,

27
00:01:48,480 --> 00:01:51,000
and we will use a structured interview,

28
00:01:51,000 --> 00:01:53,370
not an unstructured interview.

29
00:01:53,370 --> 00:01:55,233
And why do you think that is?

30
00:02:01,440 --> 00:02:03,190
The reason is that

31
00:02:04,950 --> 00:02:08,670
with lots and lots of
different interviewers

32
00:02:08,670 --> 00:02:12,030
we can't be sure that the
key questions are asked

33
00:02:12,030 --> 00:02:15,120
that they were asked in the same way.

34
00:02:15,120 --> 00:02:20,120
So we tend to stick to a stricter script

35
00:02:21,390 --> 00:02:23,790
when there are multiple interviewers.

36
00:02:23,790 --> 00:02:25,980
Whereas if there's only one,

37
00:02:25,980 --> 00:02:30,980
you have a little bit more
leeway to use less structure.

38
00:02:35,940 --> 00:02:38,610
In some ways, the best of both worlds

39
00:02:38,610 --> 00:02:40,770
is the in-between state,

40
00:02:40,770 --> 00:02:43,620
where you used semi-structured interviews.

41
00:02:43,620 --> 00:02:47,220
And these are probably
the most commonly type

42
00:02:47,220 --> 00:02:51,180
of interviews in social science research.

43
00:02:51,180 --> 00:02:54,750
And again, this is sort
of an in-between case

44
00:02:54,750 --> 00:02:57,480
where there's some structure

45
00:02:57,480 --> 00:03:01,170
which ensures that some, you know,

46
00:03:01,170 --> 00:03:04,710
the key points and key questions are hit,

47
00:03:04,710 --> 00:03:06,990
so you get reliable data.

48
00:03:06,990 --> 00:03:09,900
Yet there also allows for some flexibility

49
00:03:09,900 --> 00:03:12,480
and may allow for some new information

50
00:03:12,480 --> 00:03:17,430
that the structured
questions didn't elicit

51
00:03:17,430 --> 00:03:21,513
and therefore also get valid information.

52
00:03:25,800 --> 00:03:27,990
Here are some basic tips.

53
00:03:27,990 --> 00:03:31,440
Try to set an easy conversational tone.

54
00:03:31,440 --> 00:03:34,217
Set the subject at ease.

55
00:03:37,740 --> 00:03:42,390
Let the responses shape
the next questions.

56
00:03:42,390 --> 00:03:46,957
Have lots of probes like,
"How does that work?"

57
00:03:46,957 --> 00:03:48,817
"Tell me more."

58
00:03:48,817 --> 00:03:50,280
"What's an example?"

59
00:03:50,280 --> 00:03:52,890
And like almost anything else in life,

60
00:03:52,890 --> 00:03:57,363
the best way to get better at
doing interviews is practice.

61
00:04:02,670 --> 00:04:06,480
The next consideration

62
00:04:06,480 --> 00:04:10,200
is whether you're having
open-end questions

63
00:04:10,200 --> 00:04:12,390
or closed-ended questions.

64
00:04:12,390 --> 00:04:16,173
Usually in interviews we
asked open-ended questions.

65
00:04:17,130 --> 00:04:19,980
Closed-ended question is like a yes-no,

66
00:04:19,980 --> 00:04:22,620
or where you're asking them to choose

67
00:04:22,620 --> 00:04:27,450
from a set of answers that you provide.

68
00:04:27,450 --> 00:04:30,810
So that's more like a survey.

69
00:04:30,810 --> 00:04:31,920
And an open-ended

70
00:04:31,920 --> 00:04:35,520
tends to just let them
answer in their own words

71
00:04:35,520 --> 00:04:38,940
and give the answer that they think of.

72
00:04:38,940 --> 00:04:42,870
So think about what will
give you more reliable

73
00:04:42,870 --> 00:04:46,023
and what will give you more valid answers.

74
00:04:50,220 --> 00:04:54,330
Here is the way it works.

75
00:04:54,330 --> 00:04:57,210
Closed ended has more reliable

76
00:04:57,210 --> 00:05:01,080
and open-ended has more valid.

77
00:05:01,080 --> 00:05:03,750
And if you don't
understand why that is so,

78
00:05:03,750 --> 00:05:07,983
let's discuss it in the class
meeting, the discussions.

79
00:05:12,390 --> 00:05:16,890
The next consideration is
the order of questions.

80
00:05:16,890 --> 00:05:20,700
Now note that these may be in conflict,

81
00:05:20,700 --> 00:05:24,360
but here are three really
important principles.

82
00:05:24,360 --> 00:05:27,600
First general to specific,

83
00:05:27,600 --> 00:05:30,000
most important to least important,

84
00:05:30,000 --> 00:05:32,910
and safest to riskiest.

85
00:05:32,910 --> 00:05:36,990
And we'll talk in class,
in our class meeting,

86
00:05:36,990 --> 00:05:40,383
how that works and why
that is the right order.

87
00:05:41,370 --> 00:05:44,943
But feel free to think
about why that is so.

88
00:05:49,710 --> 00:05:51,810
As we learned very early on

89
00:05:51,810 --> 00:05:55,740
in our human inquiry and science session,

90
00:05:55,740 --> 00:05:57,960
taking notes is important.

91
00:05:57,960 --> 00:06:00,150
That you need to record it.

92
00:06:00,150 --> 00:06:03,450
That you can't just trust your memory.

93
00:06:03,450 --> 00:06:07,770
And it's very important to
take notes in real time.

94
00:06:07,770 --> 00:06:10,650
And then once you are
done with the interview,

95
00:06:10,650 --> 00:06:15,330
go back and sort of rewrite,
revise, reorganize them,

96
00:06:15,330 --> 00:06:18,960
just like you would during a class.

97
00:06:18,960 --> 00:06:21,750
So yeah, I think that's
a good thing to to do

98
00:06:21,750 --> 00:06:24,150
when you take class notes

99
00:06:24,150 --> 00:06:27,060
is like write down things on the fly

100
00:06:27,060 --> 00:06:32,060
and then go back and
revise and reorganize them.

101
00:06:32,100 --> 00:06:34,470
Again, it's important
to take lots of notes.

102
00:06:34,470 --> 00:06:36,750
And as with anything else,

103
00:06:36,750 --> 00:06:39,453
the best way to get better
at it is to practice.

104
00:06:44,940 --> 00:06:46,890
So what are the strengths and weaknesses

105
00:06:46,890 --> 00:06:50,223
of qualitative methods?

106
00:06:51,360 --> 00:06:53,400
The strengths is it's flexible.

107
00:06:53,400 --> 00:06:55,800
So these, you can ask different questions

108
00:06:55,800 --> 00:06:57,510
to different people.

109
00:06:57,510 --> 00:06:59,520
You get great depth of answer,

110
00:06:59,520 --> 00:07:02,400
that they can go into as much
depth as they would want to.

111
00:07:02,400 --> 00:07:04,440
And you get very valid responses

112
00:07:04,440 --> 00:07:07,350
because they are putting
it in their own words.

113
00:07:07,350 --> 00:07:11,190
The weaknesses is in most cases,

114
00:07:11,190 --> 00:07:16,190
no sort of statistical analysis
or inference can be done.

115
00:07:16,380 --> 00:07:20,010
There may be a lack of
breadth that because of time

116
00:07:20,010 --> 00:07:25,010
you may not be able to sort of
cover all possible questions.

117
00:07:25,050 --> 00:07:27,540
And it tends to be not as reliable

118
00:07:27,540 --> 00:07:32,100
because you're not just
having them pick answers

119
00:07:32,100 --> 00:07:34,383
off of a sheet of paper.

120
00:07:38,310 --> 00:07:39,910
So again, they're valid

121
00:07:40,950 --> 00:07:44,223
because they're firsthand direct accounts.

122
00:07:45,420 --> 00:07:47,010
They'll probably tell you things

123
00:07:47,010 --> 00:07:49,023
that you would not have thought of.

124
00:07:49,023 --> 00:07:51,870
That if you had a more reliable approach

125
00:07:51,870 --> 00:07:55,050
that you might have missed
something important.

126
00:07:55,050 --> 00:07:59,100
And as we learned in
mixed methods discussions

127
00:07:59,100 --> 00:08:01,473
it can inform future inquiry.

128
00:08:07,470 --> 00:08:12,163
There are ways to overcome
the lack of reliability.

129
00:08:14,040 --> 00:08:18,090
So because it's very personal,

130
00:08:18,090 --> 00:08:22,020
it's subject to the researcher's bias

131
00:08:22,020 --> 00:08:24,990
and to the respondent's biases,

132
00:08:24,990 --> 00:08:29,617
and it very much just reflect
their own perspectives.

133
00:08:32,070 --> 00:08:33,810
So you can address this first

134
00:08:33,810 --> 00:08:36,450
by being aware of your own bias,

135
00:08:36,450 --> 00:08:41,450
by being reflexive as we
learned about very early on.

136
00:08:41,460 --> 00:08:46,460
And it can, very often it
helps to use comparisons

137
00:08:48,120 --> 00:08:49,710
with specificity.

138
00:08:49,710 --> 00:08:54,710
So it's much more, much better account

139
00:08:55,170 --> 00:08:59,220
to name the things that the person does.

140
00:08:59,220 --> 00:09:03,287
So I talk about here that
Interviewee A, a senior,

141
00:09:05,190 --> 00:09:10,190
was able to name key energies
saving steps but B could not.

142
00:09:11,034 --> 00:09:14,130
And A was familiar.

143
00:09:14,130 --> 00:09:17,310
And by being specific
like this, and especially,

144
00:09:17,310 --> 00:09:22,310
as we'll learn, adding
some key quotations,

145
00:09:22,950 --> 00:09:25,860
it sort of avoids the,
"Take my word for it."

146
00:09:25,860 --> 00:09:29,400
That that's a criticism
of qualitative work

147
00:09:29,400 --> 00:09:34,400
is it's, "Well, that's just
your opinion of what they said."

148
00:09:40,530 --> 00:09:42,330
So speaking of which,

149
00:09:42,330 --> 00:09:46,650
these are some of the
steps that you can take

150
00:09:46,650 --> 00:09:51,650
to provide evidence of rigor
of qualitative methods.

151
00:09:54,870 --> 00:09:59,870
First, did you do a good
job on the lit review?

152
00:10:00,120 --> 00:10:02,910
Did you build upon prior studies

153
00:10:02,910 --> 00:10:07,910
and show how your work addresses
a gap in the existing work?

154
00:10:08,370 --> 00:10:11,670
Next, does your sample make sense?

155
00:10:11,670 --> 00:10:15,390
Did you collect data from a broad array?

156
00:10:15,390 --> 00:10:20,010
Are you clear about what
your sampling method was

157
00:10:20,010 --> 00:10:24,900
and why you did what
you did in the analysis?

158
00:10:24,900 --> 00:10:29,220
Are you transparent about your codes?

159
00:10:29,220 --> 00:10:32,340
So when we're gonna
learn how to code data,

160
00:10:32,340 --> 00:10:34,950
are you upfront and
transparent about that?

161
00:10:34,950 --> 00:10:38,850
And in many cases, it's good
to have multiple coders,

162
00:10:38,850 --> 00:10:43,850
so several researchers
looking at the same data set,

163
00:10:43,950 --> 00:10:48,753
and see if you're seeing the same thing.

164
00:10:49,650 --> 00:10:53,880
Again, really addressing this reliability

165
00:10:53,880 --> 00:10:56,910
by having multiple coders
and are they saying,

166
00:10:56,910 --> 00:10:59,820
I mean, seeing the same thing?

167
00:10:59,820 --> 00:11:04,020
And finally, are you
transparent in the reporting?

168
00:11:04,020 --> 00:11:08,640
And does the voice of the
respondent come through?

169
00:11:08,640 --> 00:11:13,440
Do you have good quotations
that really sum up a theme?

170
00:11:13,440 --> 00:11:17,820
Can you sort of hear,
maybe in a sense, the voice

171
00:11:17,820 --> 00:11:21,633
of the respondent when
you read the results?

172
00:11:24,300 --> 00:11:29,300
So here are the key takeaways
for qualitative research.

173
00:11:33,330 --> 00:11:36,810
That it's part of mixed methods.

174
00:11:36,810 --> 00:11:38,730
We talked about the balance

175
00:11:38,730 --> 00:11:41,130
between the insider and outsider view.

176
00:11:41,130 --> 00:11:43,140
Talked a bit about grounded theory.

177
00:11:43,140 --> 00:11:46,080
How do we structure our
question and options?

178
00:11:46,080 --> 00:11:47,910
And what are the strengths and weaknesses

179
00:11:47,910 --> 00:11:50,253
of valid versus reliable?

180
00:11:53,520 --> 00:11:56,610
This is the end of the
second and final part

181
00:11:56,610 --> 00:11:59,970
of qualitative research.

182
00:11:59,970 --> 00:12:02,490
And so make sure that
you do the assessment

183
00:12:02,490 --> 00:12:05,160
before we discuss it in class.

184
00:12:05,160 --> 00:12:05,993
Thank you.