R p-value half to SPSS rANOVA I have done rANOVA in SPSS and then in R and got two different p-values from the same model. While SPSS gives 0.032, R gives 0.0162, which when rounded is a half. Other data are exactly the same.
1. Why?
Intuitively, one of them does a one-tailed, the other a two-tailed test. Other values are: Sum Sq 19.071 (SPSS) v. 24.14 (R) or F value 4.863 (SPSS) v. 6.156 (R).
Other parts of the table are the same though, both IV1 * IV2 and residuals. 
2. Why? Which is “more correct”?
3. How to get the SPSS-reported values from R?
SPSS approach:
Analyze –> General Linear Model –> Repeated Measures…

R approach:
summary(aov(DV ~ IV1 * IV2 + Error(subject/(IV1 * IV2)), data))

Update to answer @PeterFlom and give additional details:
I am evaluating the effect of an intervention (treatment: a lecture) on the ability to spot behavioral clues of emotions using visual stimuli using pretest-posttest plus control group experimental design.
The experimental group is given the first set of stimuli, then there is intervention and the second set of stimuli follows. Naturally, the same without the intervention occurs in the case of the control group. Two groups, two measurements each.
The hypothesis is that the intervention significantly raises the ability to spot behavioral clues of emotions.
SPSS calculates the same interaction too.
The data:
    subject   group   phase     value
1        A1     exp     pre        13
2        A2     exp     pre         7
.
35       B1    cont     pre         9
36       B2    cont     pre        14
.
57       A1     exp    post        11
58       A2     exp    post        12
.
91       B1    cont    post        13
92       B2    cont    post        12

Honestly, I am using it as is because a tutorial said so. I have got a very shallow understanding of the correct process in R.
Update in response to @ttnphns:
SPSS data:
group    pre    post
  exp     13      11
  exp      7      12
 cont      9      13
 cont     14      12

SPSS commands:
GET DATA 
  /TYPE=TXT 
  /FILE="F:\file.csv" 
  /DELCASE=LINE 
  /DELIMITERS=";" 
  /ARRANGEMENT=DELIMITED 
  /FIRSTCASE=2 
  /IMPORTCASE=ALL 
  /VARIABLES= 
  group A4 
  pre F2.0 
  post F2.0. 
CACHE. 
EXECUTE. 
DATASET NAME DataSet1 WINDOW=FRONT. 
GLM pre post BY group 
  /WSFACTOR=phase 2 Polynomial 
  /METHOD=SSTYPE(3) 
  /PLOT=PROFILE(phase*group) 
  /PRINT=DESCRIPTIVE ETASQ 
  /CRITERIA=ALPHA(.05) 
  /WSDESIGN=phase 
  /DESIGN=group.

 A: As you indicated in the comments, SS type I (aov()) vs. III (SPSS) is indeed one reason for the different results. The other reason is that your formula for aov() is incorrect, as the Error() term you used is the for two crossed within factors - whereas you have one between (group) and one within (phase) factor. Here's a reproducible example on how to replicate the results from SPSS in R. First SS type I with aov():
set.seed(23)         # make reproducible
Nj  <- c(22, 34)     # number of subjects in control and experimental group (unbalanced)

# data frame long format -> make sure subject, group and phase are factors
dfL <- data.frame(subject=factor(rep(1:sum(Nj), times=2)),
                  group=factor(rep(rep(c("A", "B"), times=Nj), time=2)),
                  phase=factor(rep(1:2, each=sum(Nj))),
                  value=round(rnorm(2*sum(Nj), rep(c(-0.5, 0.5), each=sum(Nj)), 2), 3))

summary(aov(value ~ group*phase + Error(subject/phase), data=dfL))

results in:
Error: subject
          Df Sum Sq Mean Sq F value Pr(>F)
group      1   0.81   0.812   0.198  0.658
Residuals 54 221.72   4.106               

Error: subject:phase
            Df Sum Sq Mean Sq F value  Pr(>F)   
phase        1  29.04  29.040   8.747 0.00459 **
group:phase  1   5.80   5.799   1.747 0.19188   
Residuals   54 179.28   3.320

Now SS type III using Anova() from package car. Make sure to use effect coding instead of treatment coding as SS type III requires this.
# effect coding
options(contrasts=c(unordered="contr.sum", ordered="contr.poly"))

# data frame in wide format
dfW <- reshape(dfL, direction="wide", v.names="value", timevar="phase",
               idvar=c("subject", "group"))

library(car)   # for Anova()
fitSPFpq   <- lm(cbind(value.1, value.2) ~ group, data=dfW)
inSPFpq    <- data.frame(phase=gl(2, 1))
AnovaSPFpq <- Anova(fitSPFpq, idata=inSPFpq, idesign=~phase, type="III")
summary(AnovaSPFpq, multivariate=FALSE, univariate=TRUE)

gives
Univariate Type III Repeated-Measures ANOVA Assuming Sphericity

                 SS num Df Error SS den Df      F  Pr(>F)  
(Intercept)  1.7844      1   221.72     54 0.4346 0.51256  
group        0.8118      1   221.72     54 0.1977 0.65834  
phase       22.5410      1   179.28     54 6.7896 0.01182 *
group:phase  5.7985      1   179.28     54 1.7466 0.19188

... the same as SPSS.
A: Harold,  Not sure if this is why the difference, but I note that in SPSS you specify METHOD=SSTYPE(3).  In base R, the default is a type 1. You will need to load a different package than base in order to estimate type 3. 
