Page 1 of 5
Analysis of Variance
One-Way ANOVA
What it does
The one-way analysis of variance (ANOVA) is used to determine whether or not the dependent variable has
the same population mean in each of several different groups. The independent variable is used to indicate
the group to which each participant belongs.
Using SPSS to obtain the descriptive statistics for the one-way ANOVA
1)
From the
Analyze
menu, select
Compare Means
, and then
Means…
.
2)
Move the dependent variable into the
Dependent List
box.
3)
Move the independent variable (the variable that indicates in which group each participant belongs) into
the
Independent List
box, then click on
OK
.
Running the one-way ANOVA in SPSS
4)
The independent variable (IV) is used to indicate group membership.
Skip to Step 8 if the IV is a
numeric variable
. Otherwise, make the data grid the active window.
5)
Select the
Automatic Recode
procedure from the
Transform
menu. This procedure is used to
automatically convert string variables (which SPSS cannot use in ANOVAs) into numeric variables.
6)
Move the independent variable into the
Variable -> New Name
box.
7)
In the small box to the right of
New Name
, enter the name that SPSS should use for the numeric
variable it will create for the IV. Then, click on
New Name
, and click on
OK
.
8)
From the
Analyze
menu, select
Compare Means
, and then
One-Way ANOVA…
.
9)
Move the dependent variable into the
Dependent List
box.
10)
Move the IV into the
Factor
box, then click on
OK
.
Interpreting the SPSS output
First, examine the descriptive statistics (M, n
, and SD)
for each group, as output by the
Means
procedure.
If the samples means for each group are fairly similar to each other, the corresponding population means
are probably identical to each other. However, if the sample means differ markedly from each other, this
could signify that the population means also differ from each other.
To determine which conclusion to draw, examine the output from the
One-Way ANOVA
procedure, and
in particular the p-value for the ANOVA, which is labeled “Sig.” If p
is greater than .05, the result is not
statistically significant, and you should conclude that all groups share the same population mean for the
dependent variable. However, if p is less than or equal to .05, you have a significant result and the
population mean of at least one group differs from the population mean of the other groups.
If you have a significant result, you must determine its direction. Look again at the sample means for each
group. Groups that have markedly different sample means are also likely to have different population
means, while groups with similar sample means probably have identical population means. Try to determine
which groups have the largest means, and which groups have the smallest means.