Experimental designs

Question 1

what is an experimental method

Answer 1

method that looks at cause and effect

Question 2

what is a variable

Answer 2

anything in the environment that can change between people and across situations

Question 3

if we want to determine reading speed, what 3 things can be a variable

Answer 3

difficulty of word
distractions
lighting

Question 4

we cannot control all variables, which 3 popular ones can we not control

Answer 4

personality
consent
brain

Question 5

what are 2 ways we an control variables

Answer 5

standardisation- stopping any other factors from influencing the outcome
ensuring the process is kept as constant across conditions and participants

Question 6

what is between subjects sample

Answer 6

participants take part in one condition

Question 7

what is within subjects sample

Answer 7

participants take part in all conditions

Question 8

what is an advantage of between subjects sample

Answer 8

reduced chance of fatigue/practise effects

Question 9

what is a disadvantage of between subjects sample

Answer 9

requires more participants which all vary

Question 10

what is an advantage of within subjects sample

Answer 10

reduced participants variability

Question 11

what is a disadvantage of within subjects sample

Answer 11

fatigue/practise effects

Question 12

what is a way to ensure allocation is random

Answer 12

based on test scores and paired with that person

Question 13

which statistical test you use depends on what 3 things

Answer 13

whether data is parametric or not
number of experimental conditions
whether the design was within subjects or between

Question 14

when you use an independent t-test, what 3 things are required

Answer 14

parametric data
between subjects
two conditions

Question 15

what does an independent t-test examine

Answer 15

the difference between the means of the two conditions

Question 16

what are the 5 stages you go through when looking at making a statisitical decision

Answer 16

what type of data is it 
parametric or non?
what is the study looking at
what type of design was used
what was the hypothesis

Question 17

if youre testing a relationship between 2 variables, the data is parametric and youve got within subjects, what test do you use

Answer 17

Related t test

Question 18

if youre testing a relationship between 2 variables, the data is parametric and youve got between subjects, what test do you use

Answer 18

independent t test

Question 19

if youre testing a relationship between 2 variables, the data is non-parametric and youve got within subjects, what test do you use

Answer 19

wilcoxon

Question 20

if youre testing a relationship between 2 variables, the data is parametric and youve got between subjects, what test do you use

Answer 20

mann whitely

Question 21

what 3 types of data can non parametric tests be used with

Answer 21

interval, ration and ordinal

Question 22

what does non parametric data do

Answer 22

rank the data in order

Question 23

does non parametric test create more or less reliable results

Answer 23

less

Question 24

what is a wilcoxon test used for

Answer 24

testing scores before and after an observation e.g. before and after CBT

Question 25

what is mann whitely used for

Answer 25

test the difference between 2 conditions

Question 26

is mann whitely used with parametric data or non

Answer 26

non parametric

Question 27

summary of wilcoxon

Answer 27

calculate the difference between conditions and rank the differences

Question 28

summary of mann whitely

Answer 28

rank scores individually but across all conditions, add up ranked scores separately

Question 29

what is ANOVA

Answer 29

analysis of variance between 3+groups

Question 30

what does ANOVA do

Answer 30

compares 3 or more groups to see if there is a statistical difference between them

Question 31

what are the assumptions when using ANOVA

Answer 31

interval data
normal distribution
homogeneity of variance

Question 32

instead of ANOVA why dont we do multiple t -tests

Answer 32

3x more chance of type or error