RMA Week 3: experimental methods I

what’s an experiment?

1. manipulation of one or more variables

2. to determine the effect of this manipulation on another variable

3. as a test of causality between the variables

describe the experimental/alternative hypothesis

manipulating the IV leads to a change in the DV - treatment has an effect

describe the null hypothesis

treatment (the IV) does not lead to an effect (on the DV)

what is a nuisance variable

an additional factor that affects the dependent variable e.g. time of testing and location on test marks

how do you deal with nuisance variables?

turn it into a control variable - a potential IV that’s held constant, counterbalance, include a control group

when does a nuisance variable become a confounding variable?

if it varies across the IV e.g. - amount of revision varies depending on whether listening to music or not

what should you do if a nuisance variable varies across all conditions of the IV?

hold the variable constant for all participants e.g. - fix how much revision participants are allowed to do

what should you do if a nuisance variable varies across participants e.g. - gender?

randomly assign participants to treatment groups

why is better to include more than one IV in the same experiment instead of making another one?

1. more efficient

2. better control of nuisance variables (you can turn them into IV’s)

3. results often more representative of behaviour

if there is interaction between IV’s, will the lines be non-parallel or parallel on a representative line graph?

not parallel

what’s an example of two dependent variables that measure different (but similar) behaviours?

response errors and reaction times

advantages of experiments

1. relative strong test of causality

2. possibility of a variety of manipulative controls

disadvantages of experiments

1. unnatural setting and tasks

2. reactivity (also in non-exp. research)

3. some phenomena can’t be studied in controlled conditions

4. ethical limitations (e.g. deception, harm)