Hypothesis testing Flashcards
Inferential statistics
Used to decide about the population, based on observations of the sample
Characteristics of the population
paramaters
Characteristics of the sample
statistics
mean and sd of the population
μ and σ
mean and sd of the sample
x(bar) and s
Standard error of the mean
the degree to which means differ from one sample to another
If σ is large…
there is a lot of variability between sample means
Null hypothesis
the treatment has no effects and the observed mean is drawn from the original population
the alternative hypothesis
the treatment has an effect and the observed mean is drawn from a different population
Significance level
is also known as the alpha level.
it is the probability value that defines the boundary between rejecting and retaining the null hypothesis
if p < alpha….
we reject the null hypothesis
Region of rejection
the proportion of area in a sampling distribution that represents the sample means that are improbable if the null hypothesis is true
what our critical value for the region of rejection if p=0.05?
z=1.64
When is a one-tailed test used?
When there is evidence to suggest that the treatment will an effect in a particular direction
- must be decided before the experiment!
when is a two-tailed test used?
when there is no reason to predict the direction of the effect
Type I error
rejecting the null hypothesis when is it true
what is the probability of making a type I error?
alpha = region of rejection
Type II error
retaining the null hypothesis when it is false
what is the probability of making a type II error?
beta
true or false - you can both type II and type I errors?
FALSE - they are mutually exclusive
What does minimising type II errors mean?
reducing beta and increasing power (1- beta)
How can we increase power? (four ways)
- increase alpha
- increase n (sample size)
- use powerful statistical tests
- have a good experimental design
How can we get the standard error of the mean when we don’t know the population parameters?
by using an estimate based on the sample
Simple sample t-test
a parametric procedure used to test the null hypothesis for a single-sample experiment when the sd of the population must be estimated
Student’s t-test
a parametric procedure for small samples
Degrees of freedom
n-1
how many scores in the sample are free to vary (all the scores except the last one)
Three assumptions we make when using a single-sample t-test
- the random sample comprises interval or ratio scores
- the distribution of individual scores is normal
- the sd of the mean is estimated using the sample
two-sample t-test
a parametric procedure in which we compare the difference between two sample means