1
Lab 10
Chisquare
2
χ
Learning Objectives
State the difference between the
parametric and nonparametric statistical
tests
Compute and interpret correctly chi
square tests by hand for
The goodnessoffit test
The test of independence
Compute and interpret correctly both
types of test using SPSS
Nonparametric Tests
The tests we have considered (t
test, ANOVA, correlation) are based
on estimated parameters (estimates
of
and
).
Nonparametric tests do not estimate
parameters in order to compute
probabilities.
The null doesn’t need
a parameter to find probabilities.
µ
ρ
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
2
The
Goodnessoffit Test
2
χ
Use this (GoF) when you have a single categorical
variable, as in the above examples.
Chisquare is used for testing hypotheses about
frequencies.
Are there equal numbers of men
and women students at USF?
Do certain color
M&Ms appear more often than other colors?
Example:
A sports psychologist wants to know if
a starting lane results in more wins for a horse
race.
He collects data from a track.
GoodnessofFit Example
(1)
Do the number of wins vary by starting lane?
N=
30
2
4
6
5
13
Wins
5
4
3
2
1
Lane
Alpha =.05.
The null is that the frequencies are
equal across lanes in the population. Alt:
some
lanes different.
To calculate the test, we have to
estimate frequencies under the null.
GoF Example (2)
N=
30
2
4
6
5
13
Wins
5
4
3
2
1
Lane
If the null is true, we expect equal frequencies in
each lane.
We have 30 winners and 5 lanes.
We
expect 30/5 = 6 winners per lane.
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '08
 Staff
 Harshad number, Nonparametric statistics

Click to edit the document details