Comparing Means - T test

Question 1

State some examples of comparing means and say why it’s relevant

Answer 1

-Treatment (e.g. cognitive therapy) vs no treatment, for depression
-Effect of mood (happy vs sad) on reaction time to emotional images
-Comparing allows us to see if a change in an IV leads to a change in a DV
-Essential for experiments
-See whether data is from the same distributions (equal means) or different distributions (different means)

Question 2

Why do we use a sample of a population?

Answer 2

-We want to know about the population, but we can only measure a sample
-A problem that can occur is that data from 2 samples will give different means even if they are from the same population

Question 3

What is the P value?

Answer 3

-How well the data from the sample fits the null hypothesis
-Probability of seeing the difference in the means in the sample if there was no difference in means of population
-Once we know the T-statistic and the DF, we can figure out the p-value
-Found in statistical table

Question 4

What is a standard error?

Answer 4

-The standard deviation of a sampling distribution depends on; sample size and standard deviation
-The more standard errors the means are way from each other, the more confidence that the means came from different distributions

Question 5

What 2 things can a T test tell you?

Answer 5

-Student’s T statistics - how many standard errors the means are away from each other
-P value - how confident we can be that the sample means came from different distributions
-GOOGLE DOCS FOR EQUATION

Question 6

Describe the degrees of freedom

Answer 6

-Before calculating p-value, we need to calculate degrees of freedom (df)
-The number of parameters which are free to vary
-The amount of info we know, the more info we known the more confident we can be in the results
-df = n - 1

Question 7

What is the one type of T test that is most commonly used?

Answer 7

-Independent samples T-test (also known as Unpaired means T test/Unrelated sample T test/Between groups T test)
-Seeing if 2 groups have significantly different means

Question 8

What assumptions do we have to make?

Answer 8

-Interval/ratio data
-Independent data
-Variables are normally distributed
-Both groups have similar variances

Question 9

How should results be reported?

Answer 9

-p and t letters should be italics
-Report exact same p-value where possible (unless SPSS give a very low number)
-Don’t write p-value with leading 0