You are not currently logged in. Please create an account or log in to view the full course.
Measurement Levels
- About
- Transcript
- Cite
Research Methods - Being in Control
In this course, Dr Ashok Jansari (Goldsmiths, University of London) explores points of consideration and control when designing and undertaking a scientific study. In the first lecture, we think about measurement levels by clearly defining nominal, ordinal, and interval data. In the second lecture, we think about how to construct a questionnaire, utilising open and closed questioning. In the third lecture, we think about variables by closely reviewing the role of independent, dependent, and extraneous variables within a study. Next, we think about experimental controls, going beyond just extraneous variables to think about randomisation and counterbalancing. In the fifth and final lecture, we think about demand characteristics and their impact, both as those which originate in the researcher as well as in the participant.
Measurement Levels
In this lecture, we think about measurement levels, focusing in particular on: (i) understanding the importance of measuring data and defining variables in a scientific study, as certain statistical analyses are only valid with certain types of data; (ii) nominal data, the least sophisticated of the three measurement levels discussed, with relatively incomparable categories and limited scope for statistical analysis; (iii) ordinal data, which differs from nominal data by the fact that it can be put in order, with the scale representing a measured difference between the data points; (iv) interval data, the most sophisticated of the three measurement levels, which differs from ordinal data by the fact that the differences between each point, e.g. on a scale of how tall or how heavy a person is, are equal and mean something.
My name is Dr Asher Chancery,
00:00:07and I'm a cognitive neuropsychologist at Goldsmiths University of London.
00:00:09Basically, what my work entailed is to study people with brain injury.
00:00:13We've got problems in things like memory,
00:00:18face recognition or organisational abilities to
00:00:20try to understand how we remember how
00:00:24we recognise faces and how we plan and organise things so effectively.
00:00:26What I do is to look at the damage system to try
00:00:31to understand how you and I function on an everyday basis.
00:00:35When we're conducting experiments in psychology,
00:00:39we have to measure the things that we're looking at
00:00:42and because it's a scientific process,
00:00:45we have to be quite precise in how we define our variables because of
00:00:47the fact that we're going to do statistical analysis on some of this data.
00:00:52So knowing this is important because some statistics
00:00:57are only valid for certain types of data.
00:01:01Now there are three different types of
00:01:05measurements that we generally use in psychology.
00:01:06There are nominal, orginal and interval.
00:01:10Now each of those has got different characteristics.
00:01:14The first of them, which is nominal, is basically categories so male, female,
00:01:18Asian, African, Caucasian.
00:01:26Now the thing about this type of data is that there is no hierarchy there,
00:01:30so putting male and then female is exactly the same as male than female.
00:01:35You can't add or subtract male from female or Asian from white etcetera.
00:01:42So with nominal type of data, you're effectively putting people into boxes,
00:01:48and you're working out how many people of a
00:01:54certain type fall into each of those nominal categories.
00:01:56There are certain types of statistics that you can do,
00:02:01but they're relatively limited.
00:02:04The next type of data is known as ordinary.
00:02:07An ordinary data can be put into an order so you can have a range
00:02:11classic.
00:02:18Example would be How much do you like this music on a scale of 1 to 10,
00:02:19a six is higher than a three.
00:02:24It's lower than a nine, so the order
00:02:27of the numbers in your data set does make a difference.
00:02:31However,
00:02:36it's not quite as precise as we'd like to be,
00:02:38because ultimately it's all subjective.
00:02:41What you call a sex might be different from what I call it sticks,
00:02:43so the important thing there is that you can put them in order, unlike the nominal.
00:02:47But there is subjectivity you can do more analysis with this type of data,
00:02:54but there's still a subjectivity involved.
00:02:59Then the final type that we tend to use is known as Interval.
00:03:02An interval allows you to put things into order. 123456, et cetera.
00:03:07But very importantly,
00:03:13the difference between successive stages does
00:03:14make a difference and they're equal.
00:03:19So the difference between five and six is the
00:03:21same as a difference between eight and nine.
00:03:24So this is about something that actually exists in the outside world.
00:03:27So something like, say, temperature there is a difference between 23
00:03:3024 C
00:03:36and between 32
00:03:3733 C.
00:03:39Or we could look at height,
00:03:40height has got interval data and the differences
00:03:42between each of the different values is meaningful.
00:03:47And so what we have is that we've got three different types of
00:03:52data that we can use the nominal data which is just categories,
00:03:57and we can't subtract or do any calculations with those.
00:04:01Then we have orginal where we can put things into order.
00:04:06But there's still a bit of subjectivity,
00:04:10and then finally, we've got
00:04:13interval data where you can put things into order
00:04:16and the differences between successive stages does have an impact,
00:04:20and the result of all of that is that you end up with different
00:04:25types of data which you can use in different ways for statistical analysis.
00:04:29
Cite this Lecture
APA style
Jansari, A. (2022, March 03). Research Methods - Being in Control - Measurement Levels [Video]. MASSOLIT. https://massolit.io/courses/research-methods-being-in-control
MLA style
Jansari, A. "Research Methods - Being in Control – Measurement Levels." MASSOLIT, uploaded by MASSOLIT, 03 Mar 2022, https://massolit.io/courses/research-methods-being-in-control