Lecture 3- Experimental Design 2

Question 1

What is an independent variable?

Answer 1

The factor in an experiment that distinguishes groups

The thing that the experimenter is changing/ manipulating

Question 2

How many levels does there need to be of the independent variable?

Answer 2

At least two.

Otherwise there is no point as you can’t make a comparison

Question 3

What’s a manipulated variable and what experiment type does it give rise to?

Answer 3

A manipulated variable is one that the experimenter is directly changing
This gives rise to a true experiment

Question 4

What’s a subject variable and what experiment type does it give rise to?

Answer 4

A subject variable is when a factor is not directly manipulated by the experimenter
there is still a difference between groups but the participants in a way assign their own groups through characteristics/ life experiences.
This is used in a quasi experiment

Question 5

What’s the difference between what you can find out with a true experiment versus a quasi experiment? (why is the later less ideal)

Answer 5

True experiment= prediction and explanation
Quasi experiment= prediction but no explanation. This is because you haven’t directly manipulated the IV so can’t prove a causal relationship

Question 6

If you have an independent variable x and a result y… what are the four possible explanations?

Answer 6

1) X causes Y (there is a direct casual relationship= true experiment)
2) Y causes X (the relationship has gone the other way where the observed difference in the DV is determining what level of the IV the patients are in)
3) A third factor is effecting both the IV and DV
4) Chance where there is not relationship between the variables

Question 7

Not using random assignment to determine IV levels/groups results in…

Answer 7

Confounding variables, if this occurs it is not possible to determine what variable (if it is the IV) is causing a change in the measure/ DV

Question 8

What happened in the Woolfolk, Castellan and Brooks (1983) study?

Answer 8

The Pepsi challenge
They labelled cups S for Pepsi and L for coke
But both cups contained the same drink (either all Pepsi or all Coke)
85% liked the cup labelled S more despite there being no difference between what was in the cups showing how expectation/ bias can effect us

Question 9

What is a control group and why is it important?

Answer 9

Its a comparison group that differs from the others because there is an absence of experimental treatment

Question 10

What is the placebo effect?

Answer 10

What we expect to happen often influences us e.g. if we in a drug trial the knowledge we are receiving a drug even if it is fake is enough to make us feel better/ do better. That’s why often experiments will have a group that believes it is receiving treatment when it is in fact not so we can measure the size of the placebo.

Question 11

Explain with an example why choosing levels of the independent variable correctly is important…

Answer 11

Two experiments are measuring performance of the same task under different stress conditions (high or low). The problem is the two experiments use different stress manipulations meaning the level of stress varies. Because of the Yerkes- Dodson curve this means that opposite trends emerge.

Question 12

What is a single-blind experiment?

Answer 12

Where the participant doesn’t know what experimental group they have been placed in

Question 13

What is a double- blind experiment? Why is this important?

Answer 13

Where both the participant and the experimenter don’t know what experimental group the participant is in. This is important due to placebo effect for the patient and also due to the fact that the experimenter can subconscious alter the patients performance by their own actions.

Question 14

What is a demand characteristic?

Answer 14

It cues in a new situation that people determine as demands for a particular behaviour.
e.g. participants can bring in attitudes that influence behaviour, or the environment + experimenter can influence.

(a subtle cue that makes participants aware of how they are meant to act/ behave in the experiment)

Question 15

What is a between subject design?

Answer 15

Each participant is tested in only one level of the independent variable

Question 16

What do you need to be aware of when using a between-subjects design in terms of subject variables? How do you solve this?

Answer 16

For subject variables when participants essentially choose their level based on personal characteristics (you don’t have direct influence). A between subjects design means you could have commonalities clustered in the same group. e.g. level A of the IV may be all young adults and level B of the IV might be all elderly. To overcome this you need to randomly assign conditions to the participants.

Question 17

What are other ways to control extraneous subject variables that may be confounds?

Answer 17

Matching - this is where important subject variables are matched in the treatment conditions
or you could make the variable a controlled variable by for example testing only females. The cost of this is you reduce the ability to generalize your results to a wider audience.

Question 18

Subject variability leads to an increase in….. which leads to a decrease in….

Answer 18

error variance

in the likelihood of detecting difference in the DV related to the manipulation of the IV

Question 19

What is a within-subjects design? Why it is good?

Answer 19

Every subject is tested in every level of the independent variable
This is good as each participant serves as a control for extraneous variables, we can directly compare how they do in both levels (with IV change) without individual differs clouding our judgement

Question 20

What is a problem with within subjects design?

Answer 20

The order in which participants do tasks is an issue because of the practice effect (whatever they do second they are likely to do better no matter the IV change)

Question 21

What can provide protection against the practice effect?

Answer 21

If half participants do task 1 first then task 2 and the other half of participants do the opposite. This is called counterbalancing.

Question 22

What are control variables?

Answer 22

Any extraneous variables that are help constant throughout the experiment. This is important as uncontrolled variables result in confounding variables.