Reducing Blood Pressure

Step 1: Graphs
Graph > Boxplot > with Groups, Graph variable=Difference, Categorical variable=Gender
Graph > Boxplot > with Groups, Graph variable=Difference, Categorical variable=Treatment

Everything fine with Gender (upper left panel) and Treatment (upper right panel).
Note that Treatment uses numerical ordering, which here is no more meaningful than alphabetic ordering. It is better to change to some other ordering, say by means.
Stat > Basic Statistics > Display Descriptive Statistics, Variable=Difference, By variables= Treatment
shows the order by means to be 1-2-4-3. Trying to change the order by
Select column Schedule by clicking c1, Editor > Column > Value Order
does not work because c1 is a numeric column. Turn it into a text column with
Data > Change data type > Numeric to text. Now change ordering and redraw graph (lower right panel)

Step 2: Summary Statistics

Stat > Basic Statistics > Display Descriptive Statistics, Variable= Difference, By variables=Gender
Gender
Group Sample Size Mean Std
Female 16 4.313 1.352
Male 16 5.250 1.483

Stat > Basic Statistics > Display Descriptive Statistics, Variable= Difference, By variables=Treatment
Treatment
Group Sample Size Mean Std
1 8 3.875 1.356
2 8 4.375 1.188
4 8 4.625 1.061
3 8 6.250 1.282

Step 3: Interaction
Stat > ANOVA > Interactions Plot, Response= Difference, Factors= Gender Treatment

there does not appear to be any interaction. Let's confirm this with the hypothesis test:
Test for Interaction:

Stat > ANOVA > Twoway, Response=Difference, Row Factor=Gender, Column factor=Treatment

1) a=0.05
2) H0: g11=g12=...=g24=0 (no interaction)
3) H0: gij≠0 (some interaction)
4) p-value= 0.724 > a
5) We fail to reject H0, there is no interaction

So we will fit an additive model

Step 4: Hypothesis tests
Stat > ANOVA > Twoway, Response=Difference, Row Factor=Gender, Column factor=Treatment , check box additive model, Graphs > Residual vs. Fits and Normal plot

everything appears fine

Test for Gender:
1) a=0.05
2) H0: a1 = =a2=0 (no difference in the mean for Men and Women)
3) Ha: ai≠0 for some i (some differences in the means of Men and Women) 4) p-value= 0.028 < a
5) We reject H0, there are some differences in the means of Men and Women

Test for Treatment:
1) a=0.05
2) H0: b1 = .. =b4=0 (no difference in the means of different Treatments )
3) Ha: ai≠0 for some i (some differences in the means of different Treatments) 4) p-value= 0.002 < a
5) We reject H0, there are some differences in the means of different Treatments)

Step 5: Multiple Comparison
Which treatment is best? Because there is no interaction this will be the same for both genders. We can therefore do this by ignoring Gender and running a one-way ANOVA of Differences by Treatment, with Tukey's multiple comparison. We find
1 2 4 3
_____  
    _____

and so treatments 3 and 4 are best, with no statistically significant difference between them.