WEBVTT Kind: captions; language: en-us NOTE Treffsikkerhet: 91% (H?Y) 00:00:00.600 --> 00:00:08.000 In this video, we will talk about measurement scales. Measurement scales are very important because 00:00:08.000 --> 00:00:14.500 they determine what we can do with our data. What operations are permissible on the values of our 00:00:14.500 --> 00:00:22.350 variables. So what kinds of operations are we talking about? Well, the first one has to do with 00:00:22.350 --> 00:00:29.900 equality. Can we say that two values are the same or different? or identity n other words NOTE Treffsikkerhet: 91% (H?Y) 00:00:29.900 --> 00:00:38.100 This is something we can always do. The second possible operation is ranking. Can we say that 00:00:38.100 --> 00:00:45.500 some values are greater than others? Or better than others? Or in some sense there is a natural 00:00:45.500 --> 00:00:54.800 ordering among possible values of our variable? The third operation is essentially subtraction so 00:00:54.800 --> 00:00:59.900 that we can find out if there are distances between values. Can we say NOTE Treffsikkerhet: 91% (H?Y) 00:00:59.900 --> 00:01:05.150 that there are meaningful differences between values of our variables. NOTE Treffsikkerhet: 90% (H?Y) 00:01:05.150 --> 00:01:13.000 And the final operation is ratio or division and that has to do with whether we can say that there 00:01:13.000 --> 00:01:17.750 is a natural zero in this scale of our variable. NOTE Treffsikkerhet: 85% (H?Y) 00:01:17.750 --> 00:01:25.050 Now, let us look at some examples in order to understand how this works and what it actually means. 00:01:25.050 --> 00:01:31.600 Let's start with the simplest example, which is the sex variable we talked about before which admits 00:01:31.600 --> 00:01:34.900 to values female or male. NOTE Treffsikkerhet: 91% (H?Y) 00:01:34.900 --> 00:01:42.200 So if you have a bunch of participants, you can measure them for this variable and assign each one, 00:01:42.200 --> 00:01:44.949 a value of either male or female. NOTE Treffsikkerhet: 78% (H?Y) 00:01:44.949 --> 00:01:53.550 So the first question is, can we talk about equality? Can we save the two people have the same sex? 00:01:53.550 --> 00:02:00.900 And, of course yes, you can say that that's always possible. Every pair of values can be checked 00:02:00.900 --> 00:02:08.100 for being the same or different. What about rank? Can we say that one value is better than the 00:02:08.100 --> 00:02:14.700 other, or that is more than the other? I know, some of you might have an opinion about that, but the 00:02:14.700 --> 00:02:15.350 answer is that NOTE Treffsikkerhet: 91% (H?Y) 00:02:15.350 --> 00:02:21.400 there is no natural order in these values. They're just two different labels. So there is no rank 00:02:21.400 --> 00:02:28.000 here. There is no possibility of ranking which also implies that there cannot be any distances or 00:02:28.000 --> 00:02:35.050 ratios. The only thing we can do with sex values is check if they're the same or different and this 00:02:35.050 --> 00:02:39.300 is called a nominal scale of measurement. NOTE Treffsikkerhet: 88% (H?Y) 00:02:40.900 --> 00:02:50.300 What about a variable of education, actually, level of education. Admitting these values is they're 00:02:50.300 --> 00:02:56.900 used in the Norwegian system. So every person would have one of these values in their educational 00:02:56.900 --> 00:03:05.700 level variable measurement. Can we say that two people are equal in education level? That would 00:03:05.700 --> 00:03:08.950 amount to checking if they have the same value or not. NOTE Treffsikkerhet: 91% (H?Y) 00:03:08.950 --> 00:03:11.550 Of course, we can do that. NOTE Treffsikkerhet: 89% (H?Y) 00:03:11.550 --> 00:03:19.400 Can we say that there is ranking? Is there a natural order in these labels such that we can always 00:03:19.400 --> 00:03:27.700 say, for two people if one has more or less education or maybe higher or lower level of education, 00:03:27.700 --> 00:03:35.300 more appropriately. Is there a natural order in these levels? The answer is yes, there is one from 00:03:35.300 --> 00:03:40.900 lowest to highest level of education. So the answer is yes for rank. NOTE Treffsikkerhet: 91% (H?Y) 00:03:40.900 --> 00:03:43.500 What about distance? NOTE Treffsikkerhet: 83% (H?Y) 00:03:43.500 --> 00:03:53.000 Can we say that someone who has for example ungdomskole educational level is as different from 00:03:53.000 --> 00:03:58.200 someone who has only barneskole as from someone who has videreg?ende? Does this make any 00:03:58.200 --> 00:04:05.900 sense? Can we subtract levels of education to find their distance? We can't do that. That would make 00:04:05.900 --> 00:04:12.500 no sense. We can only say if someone has more or less than another. So if there's no distance, there 00:04:12.500 --> 00:04:13.800 is no possibility NOTE Treffsikkerhet: 77% (H?Y) 00:04:13.800 --> 00:04:21.100 for ratios either and this is called an ordinal level scale of measurement. NOTE Treffsikkerhet: 90% (H?Y) 00:04:23.700 --> 00:04:32.100 Our next example is birth year. So this is the variable in which we measure someone and assign the 00:04:32.100 --> 00:04:35.750 value by the year they were born. NOTE Treffsikkerhet: 91% (H?Y) 00:04:35.750 --> 00:04:42.700 Okay, can we say that two people were born in the same year, of course we can. NOTE Treffsikkerhet: 91% (H?Y) 00:04:42.700 --> 00:04:52.900 Can we say that someone was born before or after another previous or later earlier or later than 00:04:52.900 --> 00:04:59.500 another? Yes, we can always say that and birth years are actually ranked. There is an order in birth 00:04:59.500 --> 00:05:00.900 years. NOTE Treffsikkerhet: 91% (H?Y) 00:05:00.900 --> 00:05:04.100 Okay, how about distance? NOTE Treffsikkerhet: 91% (H?Y) 00:05:04.100 --> 00:05:11.300 Can we talk about differences in birth year? Actually we can subtract the years and the 00:05:11.300 --> 00:05:17.700 difference is how many years someone is older or younger than another. So there is meaningful 00:05:17.700 --> 00:05:22.900 distance and we can derive that by subtracting, that makes a lot of sense. NOTE Treffsikkerhet: 89% (H?Y) 00:05:22.900 --> 00:05:33.000 What about ratio? Can we say that someone born in the year 2000 is born in twice the year of someone 00:05:33.000 --> 00:05:34.500 born in 1000? NOTE Treffsikkerhet: 83% (H?Y) 00:05:34.600 --> 00:05:44.100 Can we say that someone born in 2100 is born in five percent more birth year than someone born in 00:05:44.100 --> 00:05:52.400 year 2000? Did that sounds fairly stupid? Yes. It did. This makes no sense. We cannot compute 00:05:52.400 --> 00:06:00.300 ratios of birth years because this makes no sense. This kind of variable is called an interval level 00:06:00.300 --> 00:06:01.900 variable. NOTE Treffsikkerhet: 91% (H?Y) 00:06:03.500 --> 00:06:11.800 And what about height? We can measure height in centimeters for our participants. Can we say that 00:06:11.800 --> 00:06:17.150 two people are equally, tall have the same height? Yes, of course, we can. NOTE Treffsikkerhet: 88% (H?Y) 00:06:17.150 --> 00:06:24.200 Is there a natural order? Can we always say if someone is taller or shorter than someone else? Yes. 00:06:24.200 --> 00:06:26.200 This does make sense. NOTE Treffsikkerhet: 90% (H?Y) 00:06:26.200 --> 00:06:34.850 Can we compute distances? Yes, we can always subtract two heights and say how many centimeters 00:06:34.850 --> 00:06:39.000 someone is taller or shorter than someone else. NOTE Treffsikkerhet: 91% (H?Y) 00:06:39.400 --> 00:06:48.250 Can we also divide to compute ratios? Does it make sense to say that someone who is 2 meters tall 00:06:48.250 --> 00:06:52.250 is twice as tall as someone who is 1 meter tall. NOTE Treffsikkerhet: 86% (H?Y) 00:06:52.250 --> 00:06:59.600 Actually, that sounds reasonable. Can we say that someone who is 1 meter on a hundred and ten 00:06:59.600 --> 00:07:08.400 centimeters is 10% taller than someone who is only 100 centimeters that makes sense as well. We can 00:07:08.400 --> 00:07:16.150 do that height, So there are ratios and therefore, this is a ratio level scale of measurement. NOTE Treffsikkerhet: 91% (H?Y) 00:07:16.150 --> 00:07:22.799 Let's look at this last bit, a little more carefully to understand the difference between interval 00:07:22.799 --> 00:07:24.400 and ratio. NOTE Treffsikkerhet: 76% (H?Y) 00:07:24.400 --> 00:07:33.350 And it all has to do with the meaning of zero. So essentially, when you have variables that are 00:07:33.350 --> 00:07:41.700 measured with numbers, there are two possibilities. One possibility is to be like, birth year, or 00:07:41.700 --> 00:07:49.500 temperature and the other possibility is to be like height or rainfall. What is the difference 00:07:49.500 --> 00:07:53.900 between these. Well in height, and NOTE Treffsikkerhet: 86% (H?Y) 00:07:53.900 --> 00:08:03.300 Rainfall 0 means something very specific. It means there is no quantity of that variable. So zero, 00:08:03.300 --> 00:08:14.200 rainfall means it didn't rain. There was no water in the measuring vessel. Zero height means at the 00:08:14.200 --> 00:08:21.350 ground. Although, you can't expect this to ever happen for a person because nobody has zero height. NOTE Treffsikkerhet: 91% (H?Y) 00:08:21.350 --> 00:08:28.500 Zero height does mean something. So it's not about whether it happens. It's about whether it's 00:08:28.500 --> 00:08:35.299 meaningful. So you can, zero,height means something, zero weight means something, zero rainfall 00:08:35.299 --> 00:08:37.049 means something. NOTE Treffsikkerhet: 89% (H?Y) 00:08:37.049 --> 00:08:44.200 And it means that there is no quantity of the sort we are measuring. That's what a 00:08:44.200 --> 00:08:52.800 meaningful zero refers to. What about temperature or year. Well, zero there isn't 00:08:52.800 --> 00:09:01.500 meaningful. It is an arbitrary point that someone just decided to use so we can use the differences 00:09:01.500 --> 00:09:04.300 from it, for the measurement. NOTE Treffsikkerhet: 76% (H?Y) 00:09:04.300 --> 00:09:11.600 And that's why you can have different ones. So for year of birth, the zero is not the same for every 00:09:11.600 --> 00:09:19.800 culture. We have our 0 that is now called Common Era and others have different starting points for 00:09:19.800 --> 00:09:26.900 their calendars. And in temperature, the Celsius scale of temperature that we're all used to here in 00:09:26.900 --> 00:09:31.750 Europe, has this 0 set at the freezing point of water. NOTE Treffsikkerhet: 86% (H?Y) 00:09:31.750 --> 00:09:38.200 But you know, in the United States, they still use the fahrenheit which has the zero somewhere else. 00:09:38.200 --> 00:09:47.700 I don't know what made them put zero there. But the point is if you're supposed to say it's twice as 00:09:47.700 --> 00:09:55.900 temperature as another temperature that should hold regardless of the scale, right? Like in height, 00:09:55.900 --> 00:09:59.200 whether you measure it in centimeters or inches. NOTE Treffsikkerhet: 91% (H?Y) 00:09:59.200 --> 00:10:03.400 The person will be twice as tall as someone else. NOTE Treffsikkerhet: 89% (H?Y) 00:10:03.400 --> 00:10:09.600 If you're twice as tall in centimeters, you're going to be twice. Is the only inches. But if your 00:10:09.600 --> 00:10:14.600 device is hot in Celsius, you're not going to be twice as hard in Fahrenheit, which means that 00:10:14.600 --> 00:10:22.800 saying twice as hot makes no sense. So the scale itself doesn't have a meaningful zero because zero 00:10:22.800 --> 00:10:30.300 doesn't mean no temperature. It is just one point along this line. So that's what makes this an 00:10:30.300 --> 00:10:34.600 interval scale and this is a ratio scale. NOTE Treffsikkerhet: 85% (H?Y) 00:10:34.900 --> 00:10:42.600 Any doesn't have anything to do with having or not having negative values because in the previous 00:10:42.600 --> 00:10:51.000 examples that are used height and rainfall can only go up but temperature and birth year can 00:10:51.000 --> 00:10:58.900 go below zero, but that's not what's important. So you can have, for example, negative floor. If you 00:10:58.900 --> 00:11:01.800 have a variable on floor. NOTE Treffsikkerhet: 79% (H?Y) 00:11:01.800 --> 00:11:09.400 Like what lures your office on. In a system where you call ground floor with zero one floor above 00:11:09.400 --> 00:11:17.099 ground is one to above is two. And first baseman being - one second, baseman, be minus 2 and so on 00:11:17.099 --> 00:11:26.200 which is used in most of Europe, then there's a meaningful natural zero there. It means zero floors 00:11:26.200 --> 00:11:27.800 from the ground. NOTE Treffsikkerhet: 90% (H?Y) 00:11:27.800 --> 00:11:35.000 Zero means ground floor. So this would be measured on a ratio scale. But if you're using the 00:11:35.000 --> 00:11:42.400 Norwegian system where ground floor is 1, then there is no natural zero and that cannot be a ratio 00:11:42.400 --> 00:11:48.000 level measurement. Depending on what their distances are well-defined. You could have an interval or 00:11:48.000 --> 00:11:50.500 the ordinal level scale. NOTE Treffsikkerhet: 85% (H?Y) 00:11:51.100 --> 00:11:58.300 And, of course, you can have negative values on a lot of other kinds of measures. For example, you 00:11:58.300 --> 00:12:04.700 can have negative pressure, negative force, and lots of other things. So it's not about having or 00:12:04.700 --> 00:12:11.500 not having negative values. It's about what zero means. If zero means no quantity of what you're 00:12:11.500 --> 00:12:14.900 measuring, then that's a natural zero. NOTE Treffsikkerhet: 86% (H?Y) 00:12:15.300 --> 00:12:23.800 Okay, let's now go to another tricky case that actually trips up a lot of people and it has to do 00:12:23.800 --> 00:12:30.800 with the results of tests like IQ and other psychometric scales, which actually provides 00:12:30.800 --> 00:12:36.600 standardized scores. We will talk much more about standardized scores later in this course, and we 00:12:36.600 --> 00:12:40.350 will learn how to compute them and exactly what they mean. NOTE Treffsikkerhet: 91% (H?Y) 00:12:40.350 --> 00:12:47.100 But since you're now on the topic of measuring scales, we might as well, clear up to be able to 00:12:47.100 --> 00:12:50.150 classify this kind of variable. NOTE Treffsikkerhet: 90% (H?Y) 00:12:50.150 --> 00:12:58.900 So, of course, you can say, if two people have the same IQ or not. Okay, we have equality or non 00:12:58.900 --> 00:13:06.800 equality. And of course, you can say that whether someone has a higher or lower IQ. You can rank 00:13:06.800 --> 00:13:14.450 all the IQ values from lowest to highest. So there is a natural order in this measurement scale. NOTE Treffsikkerhet: 91% (H?Y) 00:13:14.450 --> 00:13:22.600 What about other things? Let's start from the most demanding. Can you say that a person with an IQ 00:13:22.600 --> 00:13:28.250 of a hundred and fifty is twice as smart as someone with an IQ of 75? NOTE Treffsikkerhet: 88% (H?Y) 00:13:28.250 --> 00:13:35.400 or does that sound a little silly to say? This actually makes no sense because IQ doesn't have a 00:13:35.400 --> 00:13:37.300 zero at all. NOTE Treffsikkerhet: 88% (H?Y) 00:13:37.300 --> 00:13:48.300 So there cannot be any ratios there? Okay. What about something a little less demanding. Let's say 00:13:48.300 --> 00:13:52.800 we have someone who has an IQ of a hundred and ten. NOTE Treffsikkerhet: 81% (H?Y) 00:13:52.800 --> 00:14:01.200 And let's compare this person to someone with an IQ of 100 and someone else with an IQ of 120. Can 00:14:01.200 --> 00:14:08.200 you say that the person with a hundred and ten is as smarter from the person with an IQ of a 00:14:08.200 --> 00:14:14.600 hundred, as there are less smart than the person with an IQ of a hundred twenty. So, are these 00:14:14.600 --> 00:14:22.400 distances interpretable as distances? Can we talk about a 10-point distance in IQ? NOTE Treffsikkerhet: 91% (H?Y) 00:14:23.100 --> 00:14:25.500 Such that NOTE Treffsikkerhet: 81% (H?Y) 00:14:25.500 --> 00:14:33.600 different 10-point distances at different points in the scale are actually equal distances. That's 00:14:33.600 --> 00:14:41.800 the important thing. If their distances, then there are 10 everywhere. So the distance between 70 00:14:41.800 --> 00:14:43.300 and 80, NOTE Treffsikkerhet: 86% (H?Y) 00:14:43.300 --> 00:14:51.600 is the same as the distance between 120 and 130. Does this make any sense to say? And the 00:14:51.600 --> 00:15:01.200 answer is no, because standardized scores don't mean something on a linear scale. Standardized 00:15:01.200 --> 00:15:11.700 scores are standardized referred to proportions of the population. So an IQ of 100 means that that 00:15:11.700 --> 00:15:13.500 you have an IQ NOTE Treffsikkerhet: 80% (H?Y) 00:15:13.500 --> 00:15:22.800 that is equal to or higher than half of the population. So you're in the middle, you are at 50%. If you 00:15:22.800 --> 00:15:28.800 have an IQ of a hundred and ten, it means you're smarter than three quarters of the population. So 00:15:28.800 --> 00:15:36.500 you are at 75%, If you have an IQ of a hundred and twenty, it means that you have a higher IQ than 00:15:36.500 --> 00:15:43.550 91 percent of the population and that's what these numbers are computed to mean. NOTE Treffsikkerhet: 69% (MEDIUM) 00:15:43.550 --> 00:15:51.700 They're set to to mean exactly that. So because they refer to proportions of the population and 00:15:51.700 --> 00:16:00.700 aren't themselves scale on a some kind of linear measuring tape of intelligence. These kinds of 00:16:00.700 --> 00:16:08.700 psychometric skills are in fact, skilled on an ordinal level of measurement and they're not 00:16:08.700 --> 00:16:13.750 interval. And they're not certainly not ratio and that limits NOTE Treffsikkerhet: 91% (H?Y) 00:16:13.750 --> 00:16:20.100 what we are allowed to do with these scores and how we can formally treat them. NOTE Treffsikkerhet: 84% (H?Y) 00:16:20.200 --> 00:16:28.700 Okay, so, now let's sum this up. What do the different scales of measurement imply for the 00:16:28.700 --> 00:16:39.000 variables that we derive? If we start from the bottom; for a variable for that is measured on a ratio 00:16:39.000 --> 00:16:46.700 scale. Then we can actually divide and compute ratios, that gives us a quantitative variable or 00:16:46.700 --> 00:16:50.100 otherwise known as numeric variable. That means NOTE Treffsikkerhet: 91% (H?Y) 00:16:50.100 --> 00:16:52.000 it's a number. NOTE Treffsikkerhet: 87% (H?Y) 00:16:52.500 --> 00:17:01.200 For variables measured on an interval scale, which means that we can compute distances but not 00:17:01.200 --> 00:17:09.000 ratios. That's also a quantitative or numeric variable. It's a number but it's a number we are 00:17:09.000 --> 00:17:18.200 allowed to subtract. Not a number we're allowed to divide. For variables measured on an ordinal 00:17:18.200 --> 00:17:22.349 scale where you can put values in some order. NOTE Treffsikkerhet: 91% (H?Y) 00:17:22.349 --> 00:17:31.000 But cannot subtract or divide them, then. What do we have there is a qualitative variable otherwise 00:17:31.000 --> 00:17:38.200 known as a factor variable or categorical variable. It means, this variable isn't really composed of 00:17:38.200 --> 00:17:43.449 numbers. It's composed of labels or codes or words. NOTE Treffsikkerhet: 91% (H?Y) 00:17:43.449 --> 00:17:45.900 These aren't numbers. NOTE Treffsikkerhet: 90% (H?Y) 00:17:45.900 --> 00:17:52.700 And for a variable that's measured on a nominal scale were the only thing we can do is to say, 00:17:52.700 --> 00:18:00.250 whether two values are the same or different. Of course we also have a qualitative variable 00:18:00.250 --> 00:18:04.600 or also known as factor or categorical variable. NOTE Treffsikkerhet: 78% (H?Y) 00:18:05.500 --> 00:18:16.300 Now, it's important to realize that if you forfill the conditions for being a ratio variable. So if 00:18:16.300 --> 00:18:23.700 you can, if there is a natural zero that permits meaningful division, then all the conditions above 00:18:23.700 --> 00:18:31.700 are automatically satisfied. It is not possible to have a variable where you can apply ratios, but 00:18:31.700 --> 00:18:34.450 you cannot for example, do distances or ranking. NOTE Treffsikkerhet: 84% (H?Y) 00:18:34.450 --> 00:18:44.200 Doesn't happen. So if you start checking from the bottom then wherever you stop the rest is 00:18:44.200 --> 00:18:53.600 necessarily also going to apply. So if they're if you can rank then you can check identity. If there 00:18:53.600 --> 00:18:56.700 is distance there is also ranked and so on. NOTE Treffsikkerhet: 89% (H?Y) 00:18:57.900 --> 00:19:06.500 Let me clarify one terminological thing here, which sometimes confuses people here because of maybe 00:19:06.500 --> 00:19:15.200 an unfortunate translation issue. So we said that there is quantitative and qualitative variables. 00:19:15.200 --> 00:19:22.300 And what this means is quantitative is quantities is numbers and qualitative with qualities is 00:19:22.300 --> 00:19:28.650 words. And if we look at what these words mean and why this can cause some NOTE Treffsikkerhet: 82% (H?Y) 00:19:28.650 --> 00:19:34.699 confusion. Quality in the English language were qualitative comes from. NOTE Treffsikkerhet: 90% (H?Y) 00:19:34.699 --> 00:19:43.800 means first peculiar and essential character and inherent feature, these kinds of things which in 00:19:43.800 --> 00:19:47.100 norwegian are "egenskaper" NOTE Treffsikkerhet: 82% (H?Y) 00:19:47.100 --> 00:19:54.200 That's the kind of quality or talking about. Of course quality also means degree of excellence or 00:19:54.200 --> 00:20:03.000 superiority which in norwegian is "kvalitet". Now, unfortunately qualitative variables, and qualitative 00:20:03.000 --> 00:20:10.800 research are translated in norwegian as "kvalitativ". Which makes some people think that we mean high 00:20:10.800 --> 00:20:17.550 quality variables or high quality research. That's not what we mean. Not at all. It has nothing to NOTE Treffsikkerhet: 81% (H?Y) 00:20:17.550 --> 00:20:23.600 with it. It doesn't mean higher or lower quality because that's not what they're talking about. 00:20:23.600 --> 00:20:26.550 So qualitative means NOTE Treffsikkerhet: 91% (H?Y) 00:20:26.550 --> 00:20:34.400 composed of words rather than of numbers. So quantitative variables are numbers. Qualitative 00:20:34.400 --> 00:20:36.150 variables are words. NOTE Treffsikkerhet: 91% (H?Y) 00:20:36.150 --> 00:20:42.900 Quantitative research is research in which we measure things with quantitative variables and mostly 00:20:42.900 --> 00:20:49.600 deal with numbers, count, compare and things like that. And qualitative research is research in 00:20:49.600 --> 00:20:56.100 which we have qualitative variables. We deal with words, with descriptions with qualities of that 00:20:56.100 --> 00:21:06.300 sort. Okay. I hope that clears that and let's also clear up one other point. That is, it's NOTE Treffsikkerhet: 72% (MEDIUM) 00:21:06.300 --> 00:21:11.000 to get confused with because we often use numbers. NOTE Treffsikkerhet: 87% (H?Y) 00:21:11.000 --> 00:21:18.600 to denote qualitative variables and that's why you have to think about what the variable means and 00:21:18.600 --> 00:21:26.500 not what it looks like. So, for example, it's possible and in fact, it's frequent that when you 00:21:26.500 --> 00:21:33.000 have a variable like sex, you might choose to code and use the number one, for example, for females 00:21:33.000 --> 00:21:37.800 and two for males, or you can choose the other way around. It doesn't make any difference. You just 00:21:37.800 --> 00:21:39.850 need to be consistent. NOTE Treffsikkerhet: 88% (H?Y) 00:21:39.850 --> 00:21:45.150 So then you're going to have a sex variable that has numbers in it, one and two. NOTE Treffsikkerhet: 91% (H?Y) 00:21:45.150 --> 00:21:53.200 But obviously you still cannot subtract those numbers and derive anything meaningful. You still 00:21:53.200 --> 00:21:59.000 cannot rank those numbers and derive anything meaningful and you certainly cannot divide them. So 00:21:59.000 --> 00:22:06.100 they look like numbers but they aren't really number. They're just labels. Labels for sex in the form 00:22:06.100 --> 00:22:08.300 of a digit. NOTE Treffsikkerhet: 84% (H?Y) 00:22:08.300 --> 00:22:15.200 You can do the same with education level. So you can have zero for no education. One for grunnskole 00:22:15.200 --> 00:22:19.800 and so on. This look like numbers. You're going to get a variable with something that looks 00:22:19.800 --> 00:22:26.600 like numbers but they're not numbers. They are labels for the categories that have the form of a 00:22:26.600 --> 00:22:30.500 digit. You'd still cannot subtract or divide these things. NOTE Treffsikkerhet: 91% (H?Y) 00:22:30.500 --> 00:22:39.100 And also this is not only about a distinction between qualitative and quantitative because some 00:22:39.100 --> 00:22:44.200 things that look like numbers may be partially numbers. So in temperature, you actually have 00:22:44.200 --> 00:22:50.950 numbers, it's a quantitative variable. So you can have 10 degrees and 20 degrees and you're allowed 00:22:50.950 --> 00:22:56.700 to subtract them and talk about 10-degree difference. This is meaningful in a temperature 00:22:56.700 --> 00:23:00.000 measurement, but you still cannot divide them. NOTE Treffsikkerhet: 81% (H?Y) 00:23:00.000 --> 00:23:08.300 So what you can do with your values isn't determine but what the value look like, it's 00:23:08.300 --> 00:23:14.700 determined by what the values mean. This is why you need to think about the nature of your 00:23:14.700 --> 00:23:21.000 variable. So there's a lot of things that look like numbers. In this sense. They aren't really 00:23:21.000 --> 00:23:24.600 numbers in this sense of what you can do with them. NOTE Treffsikkerhet: 89% (H?Y) 00:23:24.600 --> 00:23:32.600 Okay, so I hope I've cleared up and number of things that people sometimes are confused by. And now 00:23:32.600 --> 00:23:41.900 I would like to ask you to try a whole bunch of examples for yourself or maybe in groups and apply 00:23:41.900 --> 00:23:49.000 the criteria of the different operations and classify all these potential variables in more examples 00:23:49.000 --> 00:23:52.700 of your own into the different measurement scales.