the paired samples t test Flashcards
When might we use a paired samples T-test? Examples
Is there a significant difference in heart-rate when watching a horror movie compared to watching an action movie?
Is yoga an effective method for weight management?
Is there a difference in adherence to the speed-limit when driving in silence versus singing in the car?
Features of the data for a paired samples t-test
One independent variable with two dependent groups
One dependent variable measured using continuous and normally distributed data
What is a paired samples T-test?
A paired samples (or dependent) t-test is used to establish whether the mean difference between two sets of observations is zero. Data are collected from the same participant for both sets of observations, this results in ‘paired’ observations.
Average size of change for each participant
score change/variance of that change = t
Independent samples
Difference between means of two independent groups
difference between groups/variance within groups = t
t-value
This is our test statistic
This shows us the size of the difference between our two groups of data – relative to the variance of that difference
Degrees of freedom (df)
How much information the data provides
p-value
What is the probability of our t-value, given that the null hypothesis is true?
Why use a between groups design
It is important to think critically about our research design and whether it is appropriate for our research question
e.g., Is there a significant difference in heart-rate when watching a horror movie compared to watching an action movie?
● We could test this using a between group design. But what would the potential confounding variable be?
Quick summary: paired samples t-test
Tests whether there is a significant difference between groups of continuous data in a within-subject design
Examines the score change between the two groups of data, whilst also accounting for the variance of that change.
r studio
Paired_T_test(dataset, dataset, data = data)
