 #### Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

# Different results for linear regression in SPSS vs JASP

Hi all,

Are you familiar with the fact that spss gives different results for regression than regression in JASP?
I did Bayes regression and found good results, and when checking with 'normal' regression in JASP I found similar results. However, the same analysis in SPSS resulted in non-sign findings and no correlation. The opposite of what was found using JASP.

Do you have an explanation? Of course I hope JASP outcomes are correct Thanks!
Sanne

• Hi Sanne,

We use code from R packages, and most of the time SPSS is wrong and R is right, but who knows, this may be an exception. Can you provide more specifics so we can look into it?

Cheers,
E.J.

• edited May 2017

Hi E.J. and Sanne,

I have encountered the same discrepancy (example below).
JASP does indeed provide the same answer as R but it's different from SPSS and I was wondering why? Given that the differences are quite substantial.

Susanne

``````#In R
lm(formula = CPSST_8rank ~ sert_tri_bin, data = DataInc)

Deviance Residuals:
Min       1Q   Median       3Q      Max
-69.622  -52.454    0.046   40.546   82.046

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept)    57.787     10.921   5.291 4.08e-07 ***
sert_tri_bin   17.167      8.359   2.054   0.0417 *
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for gaussian family taken to be 1975.892)

Null deviance: 314598  on 156  degrees of freedom
Residual deviance: 306263  on 155  degrees of freedom
AIC: 1641
``````

In JASP:

Same as R

In SPPS (sorry for the formating):

``````Model Summary
Model   1 R (.288a) R Square (.083) Adjusted R Square (.077)    Std. Error of the Estimate   (35.40)    R Square Change (.083)  F Change (14.017) df1 (1)       df2 (155)       Sig. F Change (.000)

a Predictors: (Constant), sert_tri_bin

ANOVAa
Model       Sum of Squares      df          Mean Square         F           Sig.
1
Regression  17569.313           1           17569.313           14.017      .000b
Residual        194276.068          155         1253.394
Total   211845.381          156
a Dependent Variable: CPSST_8rank
b Predictors: (Constant), sert_tri_bin
``````
• I am not sure what outputs to compare, exactly, but I googled this problem and it has been noted a few times before. From what I see, most of the time the issue is rounding (!). For instance see https://stats.stackexchange.com/questions/113314/spss-and-stata-output-different
It's great that you've also posted this issue on our GitHub page. We might take a while to get to this, but we will at some point.
E.J.

• Susanne, on GitHub you said that you resolved the issue. What was the cause of the discrepancy?
Cheers,
E.J.