Experiments and statistics - L1

Question 1

What is Experimental design?

Answer 1

Formulate a number of research hypotheses

Translate hypotheses into treatment conditions (or levels)

Administer treatments to groups or same participants

Measure performance on a response measure

Question 2

What is an independent variable?

Answer 2

Treatment conditions (the variables being manipulated) are commonly known as independent variables

Drug or placebo

N in the N-back task

Amount of H20

Question 3

What are dependent variables?

Answer 3

Response measures are commonly known as dependent variables

Blood sugar levels

Performance on the AAS*

Question 4

What are the three types of independent variables?

Answer 4

Quantitative
Qualitative
Classification

Question 5

What is a quantiative variable?

Answer 5

Quantitative variables represent variation in amount

Amount of drug, loudness of noise, difficulty of test

Question 6

What is a qualiatitive variable?

Answer 6

Qualitative variables represent variations in kind or type

Teaching strategy, type of psychotherapy

Question 7

What is a classification variable?

Answer 7

Classification variables represent characteristics that are intrinsic to the subjects/participants

Sex, species, age group

Question 8

What is a nuisance variable?

Answer 8

Nuisance variables are potential independent variables which, if left uncontrolled, could exert a systematic influence on the different treatment conditions

Question 9

What are examples of nuisance variables?

Answer 9

Different researchers may produce an “experimenter effect”

Time of day can influence outcomes

Individual subject characteristics can influence outcomes

Question 10

What are nuisance variables also known as?

Answer 10

Uncontrolled nuisance variables are also known as confounding variables – they confound any inference derived from the experiment

Question 11

How do we decide on a dependent variable?

Answer 11

Once we have formulated a hypothesis and designed an experiment, we need to decide exactly what we want to measure (and how we want to measure it)

A good dependent variable should capture the hypothesised differences

Our hypothesis is, essentially, that the observed data will be somehow dependent on the independent variable

Question 12

How does random allocation help you achieve experimental control?

Answer 12

eg if The two labs are practically identical, except that temperature cannot be controlled
Temperature variations may lead to systematic differences in task performance

random allocation of treatment conditions to each lab gives an equally likely “chance” that different random temperatures will be associated with the different treatment conditions

Question 13

What happens in a completely randomised design?

Answer 13

In a completely randomised design, each subject is randomly assigned to one of the treatment conditions

Question 14

How does random assignment help?

Answer 14

helps to prevent non-manipulated systematic differences (confounds) from occurring between treatment groups

Question 15

What is a completely randomised design also known as?

Answer 15

This is also known as a between subjects design, as any observed differences are observed between groups of participants

Question 16

What is randomised block design?

Answer 16

Randomized block design uses blocks of subjects who are matched closely on some relevant characteristic

A common procedure is to treat a subject as a “block”, wherein the subject serves in all the treatment conditions of an independent variable

Question 17

what is randomised block design also known as?

Answer 17

When subjects complete all the treatment conditions this type of design is commonly referred to as a repeated measures design or a within-subjects design

Question 18

What is a research hypothesis?

Answer 18

A research hypothesis is a fairly general statement about the presumed nature of the world that inspires a specific experiment

“Physical exercise decreases dementia symptoms”

Question 19

What is a statistical hypothesis?

Answer 19

A statistical hypothesis is a precise statement about the parameters of distributions for different treatment populations

“Mean dementia scores will be lower for the Exercise group than for the No-Exercise group (more so than what we would expect to observe by chance)”

Question 20

What are the two statistical hypothesis?

Answer 20

The null hypothesis

The alternative hypothesis

These hypotheses are mutually exclusive, or incompatible statements about the treatment parameters

Question 21

What are treatment parameters?

Answer 21

The parameters of the distribution for each treatment population (the mean µ and standard deviation σ)*:

Question 22

What is the equation for a null hypothesis with three levels?

Answer 22

H0:u1=u2=u3 (look on powerpoint)

Question 23

What is a null hypothesis the same thing as?

Answer 23

This is the same as saying that no treatment effects are present in the population

Question 24

What happens if the treatment parameters do not satisfy the null hypothesis?

Answer 24

we reject the null hypothesis in favour of its inverse, the alternative hypothesis (H1 or Ha)

Question 25

What does H1 (the alternative hypothesis) usually state?

Answer 25

the parameters are not all equal between the treatment populations

Question 26

How do we know when to reject or not to reject?

Answer 26

to decide whether to reject H0 or not, we quantify differences between groups as an f value (our test statistic)

Next we comute that the probability (p value) of finding this F value ( or greater), given that H0 is ttue comparing it to the F- distribution

The lower the p value, the less likely H0 is true

Question 27

What is chance?

Answer 27

in experiments we sample form a population, our observed parameters are the sample mean and the sample standard error s - these are approximations of the true population parmeters

We use random sampling to chose a sample that is representative of the underlying population, howevr samples may deviate from the population which is determined bt random variation

Question 28

What is the typical statistical threshold for a?

Answer 28

0.05

Question 29

What happens if the p value is less then a?

Answer 29

it is statistically significant and vice versa, fail to reject

Question 30

What is a type one error?

Answer 30

False positive

they decide to reject H0 and accept H1
even though in realist H0 is true and H1 is false

Question 31

What is a type two error?

Answer 31

False negative

they decide to fail to reject H0 and reject H1

When in reality H0 is false and H1 is true

Question 32

What does hypothesis testing consist of?

Answer 32

deciding wether or not to reject the null hypothesis based on a sample drawn from the population of interest

Question 33

What is the goal with hypothesis testing?

Answer 33

to minimise the chance that the decision is erroneous ( type 1 or type 11 error)