WEBVTT 1 00:00:01.650 --> 00:00:05.370 Hello and welcome to this video lecture on indexes, 2 00:00:05.370 --> 00:00:07.893 scales and typologies. 3 00:00:08.730 --> 00:00:13.020 And these are tools that we can use 4 00:00:13.020 --> 00:00:16.020 to sort of understand attributes 5 00:00:16.020 --> 00:00:19.950 in our research subjects. 6 00:00:19.950 --> 00:00:22.290 Are they high or low? 7 00:00:22.290 --> 00:00:27.290 And sort of how can they compare with other subjects? 8 00:00:31.950 --> 00:00:36.950 So indexes and scales, which is gonna be the majority 9 00:00:39.000 --> 00:00:43.980 of this talk, are both ordinal variables. 10 00:00:43.980 --> 00:00:48.687 That means that we can tell the order of things 11 00:00:54.150 --> 00:00:58.350 that on a scale of one to five, if I like, 12 00:00:58.350 --> 00:01:01.350 if I say this is a five and this one is a two, 13 00:01:01.350 --> 00:01:06.350 it means that I like the one I gave a scale of five better 14 00:01:06.540 --> 00:01:10.500 to two, but they don't have units. 15 00:01:10.500 --> 00:01:15.240 And your five and my five may not be the same thing, 16 00:01:15.240 --> 00:01:18.240 that you can order them from things 17 00:01:18.240 --> 00:01:19.950 that I feel most strongly about 18 00:01:19.950 --> 00:01:23.160 to things I feel least strongly about. 19 00:01:23.160 --> 00:01:27.690 But it's very hard to compare your answers to mine. 20 00:01:27.690 --> 00:01:30.363 And we'll see a few examples of this. 21 00:01:31.350 --> 00:01:34.983 And both indexes and scales are ordinal. 22 00:01:41.250 --> 00:01:46.250 So indexes tend to be more of a composite score of things, 23 00:01:47.700 --> 00:01:50.700 a composite rating. 24 00:01:50.700 --> 00:01:55.700 So one example of if you wanted to sort of measure 25 00:01:55.800 --> 00:02:00.393 how much of an NFL football fan is someone, 26 00:02:01.470 --> 00:02:04.080 well, one of the ways is, so here's an example 27 00:02:04.080 --> 00:02:09.080 of five questions that you can ask. 28 00:02:09.120 --> 00:02:12.990 Have you ever attended a game live? 29 00:02:12.990 --> 00:02:17.883 Do you at least name one per week, watch one game per week? 30 00:02:19.080 --> 00:02:24.080 And then by adding up all the yeses, that might be one way 31 00:02:26.100 --> 00:02:30.240 of saying this person is a very, you know, strong fan 32 00:02:30.240 --> 00:02:34.410 or big fan and this person is not a fan at all, 33 00:02:34.410 --> 00:02:37.743 and a number of points in between. 34 00:02:40.350 --> 00:02:44.610 And they should be sort of equal intensity, 35 00:02:44.610 --> 00:02:49.320 that if you would ask a question like, 36 00:02:49.320 --> 00:02:53.460 can you name every single player on every single team, 37 00:02:53.460 --> 00:02:57.270 hardly anyone will be a yes there. 38 00:02:57.270 --> 00:03:01.503 So that's really not a useful question. 39 00:03:04.123 --> 00:03:07.680 And something like, have you ever heard of the NFL, 40 00:03:07.680 --> 00:03:09.930 almost everybody will be a yes. 41 00:03:09.930 --> 00:03:14.550 So those would not be very useful in sorting people 42 00:03:14.550 --> 00:03:18.153 and sort of rating how much of a fan they are. 43 00:03:20.130 --> 00:03:22.170 Note that in this example, 44 00:03:22.170 --> 00:03:24.240 that everything gets an equal weight, 45 00:03:24.240 --> 00:03:27.480 that you would just add up the number of yeses 46 00:03:27.480 --> 00:03:31.830 and that would be their score, that would be their index. 47 00:03:31.830 --> 00:03:36.830 But it's also possible to assign different weights to them, 48 00:03:38.372 --> 00:03:42.570 that if you think one question does a better job 49 00:03:42.570 --> 00:03:45.090 of answering than another, 50 00:03:45.090 --> 00:03:47.880 you could give that a greater weight 51 00:03:47.880 --> 00:03:50.520 and sort of make it like twice as important 52 00:03:50.520 --> 00:03:53.253 or something like that. 53 00:03:54.120 --> 00:03:59.120 The key is to be clear on why you select those questions 54 00:03:59.910 --> 00:04:02.250 and why you assign those weights, 55 00:04:02.250 --> 00:04:05.340 that we all wouldn't do it exactly the same way. 56 00:04:05.340 --> 00:04:07.380 But it's important that a reader 57 00:04:07.380 --> 00:04:10.770 or someone who's looking at your work understands 58 00:04:10.770 --> 00:04:14.070 why you did it that way, that it's logical, 59 00:04:14.070 --> 00:04:15.303 that it makes sense. 60 00:04:18.120 --> 00:04:21.123 So when you're thinking about these questions, 61 00:04:22.860 --> 00:04:24.720 there's a few factors. 62 00:04:24.720 --> 00:04:28.530 First, is it logical, does it make sense? 63 00:04:28.530 --> 00:04:32.850 Does it tell us something about what we want to measure? 64 00:04:32.850 --> 00:04:36.810 You know, if we're measuring NFL fans 65 00:04:36.810 --> 00:04:39.930 and you ask, are you wearing blue socks, 66 00:04:39.930 --> 00:04:42.810 that doesn't have anything to do with it. 67 00:04:42.810 --> 00:04:44.580 So it wouldn't be logical. 68 00:04:44.580 --> 00:04:47.520 So you wanna ask questions that really do a good job 69 00:04:47.520 --> 00:04:51.390 of understanding it and you want them to vary, 70 00:04:51.390 --> 00:04:54.000 that what you want out of an index 71 00:04:54.000 --> 00:04:59.000 is to sort of divide your subjects into some sort of high, 72 00:04:59.190 --> 00:05:03.390 medium, low, to rate them on some overall number, 73 00:05:03.390 --> 00:05:08.190 maybe zero to 100 is often a way to do it. 74 00:05:08.190 --> 00:05:12.300 And like where the zeros are, they are not that thing at all 75 00:05:12.300 --> 00:05:15.210 and 100s are wow, they are really that thing, 76 00:05:15.210 --> 00:05:19.473 they're like the most, and then many points in between. 77 00:05:22.590 --> 00:05:27.590 So here's another example about job related depression. 78 00:05:28.590 --> 00:05:30.660 You can add all these up 79 00:05:30.660 --> 00:05:35.660 and sort of see how much of an issue is this? 80 00:05:35.670 --> 00:05:40.290 Is this person experiencing job related depression? 81 00:05:40.290 --> 00:05:43.380 That those that would have a low score, 82 00:05:43.380 --> 00:05:48.210 this really is not a big problem for them. 83 00:05:48.210 --> 00:05:51.210 Those that have a high score, this is a big issue 84 00:05:51.210 --> 00:05:53.043 and it needs to be dealt with. 85 00:05:56.790 --> 00:06:00.520 Here's another example of an index 86 00:06:03.300 --> 00:06:08.300 that measures the gender gap in development across nations. 87 00:06:48.360 --> 00:06:51.840 So very often in social science 88 00:06:51.840 --> 00:06:55.890 we use what are called Likert scales. 89 00:06:55.890 --> 00:07:00.180 Maybe the most common thing is a five point scale 90 00:07:00.180 --> 00:07:04.470 where you ask some sort of a statement 91 00:07:04.470 --> 00:07:07.350 and you ask on this five point scale 92 00:07:07.350 --> 00:07:10.110 where one equals strongly disagree, 93 00:07:10.110 --> 00:07:12.540 two equals somewhat agree, three equals neutral, 94 00:07:12.540 --> 00:07:16.500 four equals somewhat agree, five equals strongly agree, 95 00:07:16.500 --> 00:07:21.473 and you let somebody sort of rate themselves on that scale. 96 00:07:23.460 --> 00:07:25.830 They can also ask sort of likelihood, 97 00:07:25.830 --> 00:07:28.200 how likely are you to do this thing? 98 00:07:28.200 --> 00:07:31.200 How often do you do this thing? 99 00:07:31.200 --> 00:07:33.480 You could rate a quality of something 100 00:07:33.480 --> 00:07:36.750 from very poor to very good. 101 00:07:36.750 --> 00:07:38.490 How important is this thing? 102 00:07:38.490 --> 00:07:40.350 There's many, many ways 103 00:07:40.350 --> 00:07:44.760 in which these Likert scales can be used, 104 00:07:44.760 --> 00:07:49.030 and we use them a great deal in social science 105 00:07:49.964 --> 00:07:53.370 to sort of map out where does someone fall on this? 106 00:07:53.370 --> 00:07:57.030 How important is this, how likely are you? 107 00:07:57.030 --> 00:07:58.023 Things like that. 108 00:08:02.040 --> 00:08:05.700 The five star is a five point scale. 109 00:08:05.700 --> 00:08:09.030 How would you rate this Airbnb, 110 00:08:09.030 --> 00:08:14.030 this delivery service, this app, all kinds of stuff. 111 00:08:14.640 --> 00:08:19.200 It's become a very well-known scale of five stars. 112 00:08:19.200 --> 00:08:21.630 One star, you know, one star is very bad, 113 00:08:21.630 --> 00:08:24.690 five star is great, and then points in between. 114 00:08:24.690 --> 00:08:27.930 So I think you're probably already familiar 115 00:08:27.930 --> 00:08:29.133 with these things. 116 00:08:31.530 --> 00:08:35.910 So, so far we've been looking at, at least in scales, 117 00:08:35.910 --> 00:08:38.550 sort of one dimension at a time. 118 00:08:38.550 --> 00:08:40.110 One of the things that we can do 119 00:08:40.110 --> 00:08:45.110 are perceptual maps or typologies. 120 00:08:45.240 --> 00:08:50.240 The term perceptual map is often used in marketing, 121 00:08:50.430 --> 00:08:53.610 whereas typology is used in more 122 00:08:53.610 --> 00:08:56.130 in social science and psychology, 123 00:08:56.130 --> 00:09:00.780 and it sort of measures someone across multiple dimensions. 124 00:09:00.780 --> 00:09:05.780 So here is a two dimensional map of cereal, 125 00:09:07.200 --> 00:09:12.200 where the Y axis is taste and the X axis is nutrition. 126 00:09:14.100 --> 00:09:19.100 And you can tell sort of where do various cereals lie. 127 00:09:19.710 --> 00:09:23.583 And this is also used in social science too. 128 00:09:25.380 --> 00:09:28.800 Here is one of a political map 129 00:09:28.800 --> 00:09:33.800 where the Y axis is authoritarian and libertarian 130 00:09:40.800 --> 00:09:45.800 and the X axis is more sort of left, right in the economic. 131 00:09:54.510 --> 00:09:57.063 And have a look at things like that. 132 00:10:00.180 --> 00:10:04.140 Many of you may have taken a Myers-Briggs test. 133 00:10:04.140 --> 00:10:09.140 So this is a four dimensional, which looks at four sort 134 00:10:09.240 --> 00:10:12.780 of psychological personality, 135 00:10:12.780 --> 00:10:16.440 whether you're an introvert or extrovert 136 00:10:16.440 --> 00:10:19.470 or whether you see the big picture or details, 137 00:10:19.470 --> 00:10:23.160 whether you sort of go by sort 138 00:10:23.160 --> 00:10:25.740 of rational things or emotions 139 00:10:25.740 --> 00:10:29.190 and whether you like to make a plan 140 00:10:29.190 --> 00:10:31.113 or just go with the flow. 141 00:10:32.040 --> 00:10:36.573 And there are a lot of these two that one can use. 142 00:10:38.100 --> 00:10:40.500 So that is what we did. 143 00:10:40.500 --> 00:10:42.360 We talked about indexes, we talked about scales, 144 00:10:42.360 --> 00:10:45.663 and we talked about typologies, thank you.