ONE-WAY BETWEEN SUBJECTS ANOVA
(Psychological Statistics – Laboratory)
Mathematics Learning Area
De La Salle Lipa
ONE-WAY BETWEEN SUBJECTS ANOVA
• The one –way between subjects analysis of variance (ANOVA) is used when the
means of two or more independent groups are compared on a dependent variable of
interest.
• In the one-way between subjects ANOVA, the independent variable is a between
subjects factor, where each participant receives only one level of factor (i.e each
person is in a single group).
ONE-WAY BETWEEN SUBJECTS ANOVA
Example 1:
A student was interested in investigating whether there was a difference between
three different learning strategies (strategies A, B, and C) on word recall. Thirty (30)
students who agreed to participate in the study were randomly assigned to receive one of
the three strategies (with 10 students in each strategy group) . After learning the strategy
of interest, the student were shown a list of 15 words for five minutes and were asked to
remember as many of the words as possible using the new strategy. After a brief delay,
each student wrote down as many of the words as he or she could remember from the
list, with the researcher recording the number of words correctly recalled for each
participant.
ONE-WAY BETWEEN SUBJECTS ANOVA
Objective and Data Requirements of the Dependent-Samples t Test
Objective Data Requirements Example
One between subjects
factor independent
Learning Strategies
To test whether the means of variable with two or more (Strategies A, B, and C)
two or more groups differ separate groups or
significantly on a dependent categories
variable of interest
One continuous dependent Number of words correctly
variable recalled
ONE-WAY BETWEEN SUBJECTS ANOVA
Research Question:
“Does the number of words recalled depend on the type of learning strategy used?”
Null Hypothesis:
There is no significant difference in the mean number of words recalled between
who received three different learning strategies.
𝜇strategy A = 𝜇strategy B = 𝜇strategy C
Alternative Hypothesis:
There is at least one of the mean is different from the others.
ONE-WAY BETWEEN SUBJECTS ANOVA
The Data
Participant Strategy Word Recall Participant Strategy Word recall Participant Strategy Word recall
1 1 8 11 2 12 21 3 6
2 1 10 12 2 10 22 3 4
3 1 9 13 2 9 23 3 5
4 1 7 14 2 10 24 3 6
5 1 9 15 2 9 25 3 4
6 1 7 16 2 8 26 3 2
7 1 10 17 2 12 27 3 5
8 1 8 18 2 10 28 3 3
9 1 12 19 2 10 29 3 7
10 1 9 20 2 12 30 3 5
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 1: Create the Variables
1. Use strategy and wordrecall.
2. Create value labels for the variable strategy.
1 – Strategy A
2 – Strategy B
3 – Strategy C
Step 2: Enter the Data
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 3: Analyze the Data
1. From the menu bar, select Analyze >
Compare Means > One-Way
ANOVA…
A One-Way ANOVA dialog box appears
with the variables strategy and wordrecall
in the left-hand side of the dialog box.
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 3: Analyze the Data
2. Select the dependent variable,
wordrecall, and click the upper right-
arrow button to move it into the
Dependent List box.
3. Select the independent variable, strategy,
and click the lower right arrow button to
move it into the Factor box.
4. Click Options. The One-Way ANOVA:
Options dialog box opens. Under
Statistics, select Descriptive and
Homogeneity of variance test.
5. Click Continue.
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 3: Analyze the Data
6. Click Post Hoc. The One-Way ANOVA: Post Hoc Multiple Comparisons dialog
box opens. Under Equal Variance Assumed, select Tukey (not Tukey’s b).
7. Click Continue.
8. Click OK.
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 4: Interpret the Results
1. The first table output, Descriptives, displays the descriptive statistics for each of the
groups (and for the total sample). Examining the means for the different learning
strategies, strategy B had the highest average number of words recalled (M = 10.20),
followed by strategy A (M = 8.90), with strategy C (M = 4.70) producing the lowest
number of words recalled, on average, of the three groups.
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 4: Interpret the Results
2. The next table, Test for Homogeneity of Variances, tests if the variances are equal for the three groups, an
assumption of the between subjects ANOVA.
a. If p > 0.05, equal population variances are assumed. Interpret the ANOVA F.
• If the ANOVA is significant (i.e. p 0.05), interpret the results of Tukey’s post hoc procedure.
• If the ANOVA is not significant (i.e. p > 0.05), stop. Write the results stating that there is no
significant difference between the groups.
b. If p 0.05 for Levene’s test, equal population variance are not assumed. Return the analysis
selecting either the Brown –Forsythe or the Welch procedure (by clicking on the Options button) and
select one of the post hoc tests does not assume equal variances between the groups (e.g Dunnet’s T3)
by clicking Post Hoc.
• If the overall test (Brown-Forsythe or Welch) is significant (i.e. p 0.05), interpret the results of
the post hoc procedure
• If the overall test (Brown-Forsythe or Welch) is not significant (i.e. p > 0.05) , stop. Write the
results stating that there is no significant difference between the groups.
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 4: Interpret the Results
2. The next table, Test for Homogeneity of Variances, tests if the variances are equal for the three groups, an
assumption of the between subjects ANOVA.
For our data, Levene’s test produced an F value of
0.021, and a p-value of 0.980. Since 0.980 is greater
than 0.05, equal population variances are assumed
(we will interpret the ANOVA table).
3. The next table, ANOVA, provides the answer to our
research question. ANOVA table produces an F test.
For this example, the p-value is 0.000 (which is read
as “less than 0.001” not “zero”). Since the p-value is
less than 0.05, the null hypothesis is rejected, and it
is concluded that at least one of the strategies is
different from the others.
Since the result is significant, we will interpret the Tukey’s Post Hoc procedure.
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 4: Interpret the Results
4. The next table, Post Hoc Tests, tests if the variances are equal for the three groups, an
assumption of the between subjects ANOVA.
By default, SPSS provides two different tables of
output for Tukey’ test, Multiple Comparisons and
Homogeneous Subsets. While either table can be used
to interpret the results of the pair-wise comparisons, the
results will be discussed using the latter table. In the
Homogeneous Subsets table, two different columns,
labeled “1” and “2”, are presented. When interpreting
the results, groups that share the came column are not
significantly different from each other while groups
that do not share the same column are significantly
different.
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 4: Interpret the Results
4. The next table, Post Hoc Tests, tests if the variances are equal for the three groups, an
assumption of the between subjects ANOVA.
Do the groups share
TEST Result
the same column?
A vs B Yes Not significant; assume A and B are equal
A vs C No Significant; A recalled more words than C
B vs C No Significant; B recalled more words than C
For our example, since strategies A and B share the same column (column 2), they are not
significantly different from each other. However, since strategy C (M = 4.7) does not share a
column with strategies A and B, C is significantly different from A (M = 8.9) and B(M = 10.2).
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
Step 4: Interpret the Results
5. Effect Size
The measure of effect size commonly used with ANOVA is eta-square (η2).
Sum of squares between groups 165.267
η2 = = = 0.74
Sum of squares total 223.867
Cohen’s conventions for small, medium, and large effect sizes fro eta-square
are 0.01, 0.06, and 0.14, respectively. Eta-square can be expressed in terms of
the percentage of variance in the dependent variable that is accounted for by the
independent variable. A value of 0.74 corresponds to a very large effect in
practice and indicates that the learning strategies accounted for 74% of the
variance in word recall.
ONE-WAY BETWEEN SUBJECTS ANOVA
Data Entry and Analysis in SPSS
3. Expression of the Results in APA Format
In writing the results, the conclusion of the hypothesis test, the degrees of
freedom (df), the F value, the p-value, and the effect size are reported along with
the means and standard deviations of the three strategy groups. Below is the brief
write-up in APA format of the presented example.
The number of words recalled varied by the learning strategy used, F(2, 27) =
38.07, p < 0.05, 2 = 0.74. Tukey’s post hoc procedure indicated that those who used
strategy A (M = 8.90, SD = 1.52) and those who used strategy B (M = 10.20, SD =
1.40) recalled significantly more words than those who used strategy C (M = 4.70,
SD = 1.49). There was no significant difference in the number of words recalled
between strategies A and B.
ONE-WAY BETWEEN SUBJECTS ANOVA
Example 2:
A medical researcher wanted to investigate the effects of different pain medications on
people suffering from migraine headaches. Twenty-one (21) people who had recently seen
a doctor for migraine headaches were randomly assigned to receive one of the three pills,
wither drug A, drug B, or a placebo. While taking the appropriate pill, each participant
recorded their pain level three times a day at a regular intervals for one week (pain was
recorded on a 1 to 10 scale, with higher scores indicating greater pain). The average pain
level over the one-week period was calculated for each participant and is reported in the
table below.
Drug Pain
A 5.2 4.1 5.8 6.85 4.75 1.75 4
B 3.05 6.15 5.5 6.15 1.85 6.4 3.1
Placebo 8.15 7.15 6.2 7.85 9.45 9.25 6.3
ONE-WAY BETWEEN SUBJECTS ANOVA
Example 2:
a. State the null and alternative hypotheses:
• Null Hypothesis (Ho): The pain level experienced by people does not differ on
the type of drug used.
𝜇drug A = 𝜇Drug B = 𝜇Placebo
• Alternative Hypothesis (Ha): At least one of the means is different from the
others.
b. State a research question for the data.
• Does the level of pain experienced by people with migraine headaches depend
on the type of drug used?
ONE-WAY BETWEEN SUBJECTS ANOVA
Example 2:
c. Test for the assumption of equal variances. Do the data suggest unequal variances
between the groups?
• The variances are not significantly different, F(2, 18) = 1.267, p > 0.05.
ONE-WAY BETWEEN SUBJECTS ANOVA
Example 2:
d. Is there a significant difference in the reported pain levels between the groups?
• Since the p-value of 0.002 is less than 0.05, null hypothesis is rejected and it is
concluded that there is a significant difference in the reported pain levels
between groups.
e. What is the effect size for overall ANOVA? Would you characterize the effect size as
small, medium, or large?
46.205
η2 = = 0.50
93.057
• This is a large effect.
ONE-WAY BETWEEN SUBJECTS ANOVA
Example 2:
f. If the overall ANOVA is significant,
briefly summarize the results of Tukey’s
post hoc procedure.
• Drug A and Drug B are less than the
placebo. There is not a significant
difference between Drug A and Drug
B.
ONE-WAY BETWEEN SUBJECTS ANOVA
Example 2:
g. Write the results of the study in APA format as appropriate.
The pain levels of migraine sufferers differed by the type of drug used, F(2, 18) = 8.876,
p < 0.05, η2 = 0.50. Tukey’s post hoc procedure revealed that those who took drug A (M
= 4.64, SD = 1.61) and those who took drug B (M = 4.60, SD = 1.87) had significantly
lower reported pain than those who took the placebo (M = 7.76, SD = 1.30). There was
not a significant difference in the pain levels between drug A and drug B.
ONE-WAY BETWEEN SUBJECTS ANOVA
Test Your Understanding 19:
A researcher wanted to know whether the level of frustration over high gas prices differed for
different types of vehicle owners. Following a sharp increase in the price of gas, the researcher asked 30
people to indicate their level of frustration with prices at the pump (frustration scores ranged from 1 to 10,
with higher scores indicative of greater frustration). Ten (10) people were surveyed who owned a
motorcycle as their primary means of transportation (coded a 1), 10 were surveyed who owned a hybrid
vehicle (coded a 2), and 10 were surveyed who owned a nonhybrid vehicle (coded a 3). The data are
shown on the table on the next slide.
a. State the null and alternative hypotheses.
b. State a research question for the data.
c. Test for the assumption of equal variances. Do the data suggest unequal variances between the
groups?
d. Is there a significant difference in the frustration levels of vehicle owners?
e. What is the effect size for the study? Would you characterize the effect size as small, medium, or
large?
f. If the overall ANOVA is significant, briefly summarize the results of Tukey’s post hoc procedure.
g. Write the results of the study in APA format as appropriate.
ONE-WAY BETWEEN SUBJECTS ANOVA
Test Your Understanding 19:
The Data
Participant Vehicle Frustration Participant Vehicle Frustration Participant Vehicle Frustration
1 1 5 11 2 2 21 3 8
2 1 6 12 2 4 22 3 9
3 1 7 13 2 3 23 3 10
4 1 3 14 2 5 24 3 7
5 1 4 15 2 4 25 3 8
6 1 7 16 2 3 26 3 5
7 1 6 17 2 1 27 3 6
8 1 9 18 2 3 28 3 10
9 1 6 19 2 4 29 3 9
10 1 5 20 2 6 30 3 8
