Original subtitles

post hoc tests you can ask for more more

comparisons here but not in a one way I

know but you have to have between I

believe you have to have a between

subjects effect to actually get these

and when you combine the repeated

measures and between subjects effect in

the same ANOVA you're talking about a

split plot ANOVA or what some people

unfortunately call a mixed design and

OVA which is confusing because sometimes

people confuse it with a mixed effects

ANOVA which is actually a totally

different analysis so those are the the

main things I'm going to choose for this

basic one-way repeated measures ANOVA

and I'm going to click OK to run the

analysis so SPSS is going to give me the

first box here which is this first box

here is a warning and it's just telling

me that it couldn't do one of the

homogeneity of variance tests which is

the Levine's test of homogeneity of

variance in fact you know I'm not even

too sure if I had to click that

homogeneity button because I know it

might actually give the test that I'm

going to take a look at in a second just

automatically so within subjects table

here is just telling me what my my

factor looks like it's got three levels

time one time two times three so people

took the same IQ test three times in a

row let's just say a month apart and we

can see the means and the standard

deviation see and we can see the mean is

increasing linearly so time one to time

two an increase of about one point three

five a one point one point in the means

and then it goes up again by about about

one and we can see the standard

deviations here and we can also see that

the standard deviations are increasing

across and usually in repeated measures

designs in the real world the standard

deviation does increase over time people

at the last time period will usually

have the largest standard deviation and

it's larger usually sometimes

substantially larger than the time one

and one of the assumptions of the

repeated measures ANOVA is that there is

homogeneity of variance and it tests it

with a test called much less Tessa's

veracity which I'm going to get to in a

minute this is the multivariate

estimates and it's outputted by SPSS

automatically it is a legitimate

analysis that something people often do

it

PETA measures design and I'm going to

follow that up in a future video I'm

going to talk about repeated measures

and over in the multivariate manova

context but i'm not going to talk about

it in this analysis except to say when

you violate the assumption of modulus

tests for its diversity some people

often go to the multivariate test

because it doesn't assume that what is

much least tessa sphericity is something

that typically you do not want to reject

the null hypothesis for marshalese test

of sphericity is indirectly testing the

assumption that your variances or

standard deviations are actually going

to be the same as well as the

covariances between time 1 time 2 and

time 3 now typically I'm going to talk a

little bit more about this right now to

get an understanding what's going on

unfortunately spss doesn't give you the

correlation between time 1 time 2 and

times 3 so people who scored high at

time 1 on this IQ test most likely also

scored high at time 2 and also most

likely scored time scored high at time 3

so there's a correlation between the

dependent variable scores from time one

time to two times three the repeated

measures ANOVA assumes that that

correlation between time one and time

two and between time 1 and time three

and between time to and times three is

going to be equal so let's actually take

a look at this let's look at the

correlation so we get the correlations

between time 1 time 2 and time 3 and we

can see that there is some deviations

time 1 the time two are about the same 8

point 8 5 in time 2 and time three point

eight between time - and times 3 and

then time one time 0.85 roughly the same

but then only 0.65 between time 1 and

time 3 so the correlation is actually

going down as we get further and further

away from time 1 which is what usually

you see in practice and much less esses

erisa t indirectly tests that these

correlations are the same

I won't go into too much detail about

exactly what monsters Texas versity is

testing except to say that it's not

actually testing directly that the

variances and covariances or

correlations are the same you can get a

variance covariance matrix in SPSS by

going into the reliability analysis

utility and click on statistics and

getting covariances here interitum will

present pretend that our variables are

items they're not the dist variables but

we can get the variance covariance

matrix and what repeated measures ANOVA

assumes is that these variances here

time one's variance is 7.85 that's

simply this