Inferring differences over time Flashcards
(19 cards)
when should a repeated measures t-test be used?
when looking at a difference over two time points with continuous variables that are normally distributed
if you are looking at a difference between two time points with continuous variables that are normally distributed, which statistical test should you use?
Repeated measures t-test
when should a Wilcoxon Z test be used?
when looking at a difference over two time points with continuous variables that are not normally distributed
if you are looking at a difference between two time points with continuous variables that are not normally distributed, which statistical test should you use?
Wilcoxon Z test
if you are looking at a difference between two time points with ordinal variables, which statistical test should you use?
Wilcoxon Z test
if you are looking at a difference between two time points with nominal binary variables, which statistical test should you use?
Wilcoxon Z test
if you are looking at a difference between two time points with nominal non-binary variables, which statistical test should you use?
Chi square
what is the repeated measures t-test?
Compares the means of these 2 variables and tests the null hypothesis (p<0.05): ‘there has been no change in the mean value over time in the population’
what is the Wilcoxon Z test?
Turns the data into ranks - discards some information so less desirable
Compares mean ranks of these 2 measures and tests the null hypothesis (p<0.05): ‘there has been no change in the mean rank value over time in the population’
if p > 0.05 then…
the result is not significant so you report p=n.s. meaning we cannot reject our null hypothesis
if p < 0.05 then…
the result is significant so you report p<0.05 meaning we can reject our null hypothesis
if p < 0.01 then…
the result is significant so you report p<0.01 meaning we can reject our null hypothesis
if p < 0.001 then…
the result is significant so you report p<0.001 meaning we can reject our null hypothesis
how do you carry out SPSS output & report for repeated measures t-test?
Which time point has the higher mean?
Look at significance value:
i. If value of Sig is more than 0.05 –> not a significant difference –> accept null hypothesis
ii. If value of Sig is less than 0.05 –> was a significant difference –> reject null hypothesis
What elements from the SPSS table is needed when reporting the results of a repeated measures t-test?
Mean, standard deviation, degrees of freedom, t-value, p-value
Example write-up of the results from a repeated measures t-test:
“the average maths knowledge of pupils at the beginning of term (M=0.08, SD=12.72) was not significantly different to that at the end of this term (M=0.40, SD=12.06); f(461)=0.61, p=n.s.”
how do you carry out SPSS output & report for Wilcoxon Z test?
Which time point had the higher mean rank?
Look at significance value:
i. If value of Sig is more than 0.05 –> not a significant difference –> accept null hypothesis
ii. If value of Sig is less than 0.05 –> was a significant difference –> reject null hypothesis
What elements from the SPSS table is needed when reporting the results of a Wilcoxon Z test?
Mean rank, Z-value, p-value
Example write-up of the results from a Wilcoxon Z test:
“The average maths knowledge of pupils at the beginning of term was not significantly different to that at the end of this term; Z=0.73, p=n.s.”
