Section 2.3: Independent variables
From Research Methods in Psychology
Contents |
[edit] WHAT ARE YOU STUDYING?
IN THIS SECTION WE WILL LOOK AT SOME OF THE DIFFERENT WAYS THAT INDEPDENT VARIABLES (IVS) CAN BE OPERATIONALISED. WE WILL FOCUS ON MOOD AND NOISE AS TWO DIFFERENT INDEPENDENT VARIABLES THAT CAN BE OPERATIONALISED IN MANY DIFFERENT WAYS. WE WILL ALSO SHOW HOW INDEPENDENT VARIABLES CAN BE CLASSIFIED INTO DIFFERENT TYPES.
[edit] Independent Variables
An independent variable in an experiment is one that is manipulated by the experimenter. The manipulation is carried out to determine the effect this variable has on the dependent variables. It is independent because it is manipulated independently from any changes in other variables. The independent variable is the cause of something and the dependent variable is the effect. As we saw in Chapter 1, the independent variable will be specified by our theory, but the theory will talk about the variable in psychological terms. The theory will use words like ‘noise,’ or ‘learning,’ or ‘scores on a maths test.’ We need to make those psychological terms real. We need to specify what we will do, to make those psychological terms real - we need to say what operations we will carry out when we do this. Because we need to specify the operations that we will carry out, this process is sometimes called ‘operationalising.’ A definition that tells you what the process was is called an operational definition. We can operationalise independent variables in very many ways, and a different approach might have different implications.
- When you are designing a study, you need to consider the different ways in which you could operationalise the independent variable.
- When you are reading about a study, you must consider whether they have operationalised the independent in the best way, or whether, if the researchers had done it differently, they would have a different result.
Any independent variable can be operationalised in many ways. We will consider one independent variable: mood, in some detail, and look at some different ways in which it could be operationalised. It is possible to go through a similar process for almost every independent variable. Psychologists are often interested in the effects of mood on different tasks. To experimentally investigate what effect mood has, we need to be able to change people’s moods. Usually we want to make one group of people happy, and one group sad. There are many ways on inducing different moods in participants (for a review, see Gerrards-Hesse, Spies and Hesse, 1994). Some of them include:
- Weather. Wait until a rainy day for sad participants, wait until a sunny day for happy participants. (Or as Rind, 1996 did, put participants in a position where they cannot see the weather, and then tell them that it is raining, or sunny.)
- Gifts – give participants a gift to make them happy. Sweets which remind them of childhood can work well – Love Hearts, for example. (You might be aware that Love Hearts have recently been updated, and included requests such as ‘Email me’. These usually raise a smile, which is what we want.)
- False feedback – ask people to fill out a questionnaire (intelligence or personality) and then provide false feedback about the results. (“You have done less well than average on this test. Do you usually find tests difficult?” or “The test suggests that you may have some insecurities. Do you have problems making friends?”)
- Music – play happy music (Beatles: “Yellow Submarine”) or play sad music for example Prokofiev: “Russia under the Mongolian yoke” at half speed.)
- Give people a drug that will induce the appropriate mood.
Each of these methods of mood induction would be appropriate for the hypothesis we wish to test, and each would be likely to give slightly different results, although each is a perfectly valid representation of the independent variable that we are interested in.
| Crucial Tip: In a psychology experiment, we want to create in the participant a particular psychological state.
We are not interested in the independent variable, for itself, we are interested in the psychological state created by the independent variable, and it is important to distinguish between these two things. Example: A psychologist who is interested in stereotyping introduces two people to a group. The psychologist tells the groups that one person is from Queen’s, New York, and the other is from Manhattan, New York. The psychologist is not interested in Queen’s and Manhattan, they are interested in the psychological states that are induced by the beliefs. If you replicate this study in the UK, it will make no sense, because Queen’s and Manhattan, New York, don’t mean anything to people in the UK. You should always consider whether the IV is appropriate for the cultural group that you are studying, and the historical period that you are in. |
[edit] Levels of Independent Variables
IVs are usually described as having levels. If your experiment comprises one IV, which has two forms, it is described as having two levels. Examples of this would be noise, and no noise, or happy mood and sad mood. Sometimes one of those conditions can be considered a neutral condition, in which case the group that is experiencing that neutral condition would be called the control group, whilst the group experiencing one of the other conditions would be an experimental group. Thus in the example we considered earlier on in this chapter, the noise group would be considered to be the experimental group, and the no noise group would be considered to be the control group. Note: The above is a second meaning of the word ‘control.’ We also used ‘control’ to mean ways of controlling extraneous variables. In some experiments, there may be more than two groups. We might have a no-noise experimental condition and the group tested in this condition would be considered the control group. There might also be a quiet noise condition and a loud noise condition and the two groups experiencing these respective noise conditions would be called the experimental groups. So there would be two experimental groups, and one control group. A further complication is that there may be more than one IV. We might have 3 noise groups (none, quiet, loud) combined with 2 mood groups (happy, sad). In effect, we then have six groups. In this case, we would probably not talk of a control group. This experiment would be described as a 3 × 2 design (spoken as “three by two”). Table 1: A 3 x 2 design, incorporating 2 independent variables, Noise (which has 3 levels) and mood (which has 2 levels)
[edit] Section Summary
This section has focussed on operationalising independent variables – that is turning them the variable from a psychological concept into a physical reality. We looked at noise and mood as examples of variables that can be defined in many different ways. We classified IVs into three different types: instructions, stimuli and events. Finally, we looked at levels of independent variable, considering what was mean by ‘experimental’ and ‘control’ groups, and looking at some occasions where there may be more than two levels of the independent variable.
