Tip:
Highlight text to annotate it
X
So let me give you a little bit of support and scaffolding
now that you are recomputing this variance here.
For those of you who never got
a very good grasp of variance,
where we covered it earlier in the class,
you must do it, do it sooner rather than later,
you're gonna have to be computing variances
every single week, week in and week out
for the rest of the course
and, for you guys, here is a little extra reminder
about the steps in computing variance within a group.
So you're going to be recomputing this.
Variance in the original example was two,
but now it won't be,
because we've moved these numbers closer together,
there's less spread between them,
and there's less spread, you know,
within this group now,
so the within group variance should be smaller.
Remember, variance is a measure
of spread dispersion.
So just a little extra reminder,
and if this doesn't look familiar to you,
you gotta go back and you have to master it,
so that you'll be in better shape
for the entire rest of this class.
You're gonna take deviations within the group,
not between. Not total. Within!
And I'll give you a hint,
it's not gonna be a group mean
minus the grand, like it was in ANOVA.
It's a different thing.
Hopefully you know how to get a deviation
within a group.
So how far is the score from the middle of the group.
So look up here and figure out what's the score
and what's the middle of the group.
And you're gonna turn the deviations
into squares by multiplying them by themselves.
I'm gonna let you do that here
and then you're gonna add up these numbers
to make a sum of squares within this group.
Label, label, label,
and then you're gonna compute the variance.
What's the word that tells you
how you're computing variance?
The other word for variance? Mean squares.
This is a label for variance that tells you exactly
how you're computing it,
you're making a mean of the squares.
It's gonna be sum of squares
over degrees of freedom.
And degrees of freedom is N minus one.
Where N is the number in this group.
Okay.
So, right now I'd like you to go ahead
and compute the new variance within this new group
based on these two scores here.
By getting deviations, squaring them,
summing them up,
dividing them by degrees of freedom.