SPSSX Discussion

Different result for SPSS ANOVA and UNIANOVA

Classic

List

Threaded

4 messages Options

kd1

Different result for SPSS ANOVA and UNIANOVA

I am not sure if the below provides enough information, but why are the values different if I test the same variables using ANOVA or UNIANOVA (e.g. with ANOVA Bonferroni the difference between variable H and S is not significant, but is using UNIANOVA)?

ANOVA
Multiple Comparisons
Dependent Variable: Max Length (mm)
(I) Impactor Mean Difference (I-J) Std. Error Sig. 95% Confidence Interval
Lower Bound Upper Bound
Tukey HSD A S 7.1743* 1.2667 .000 4.184 10.165
H 4.1871* 1.2667 .003 1.197 7.177
S A -7.1743* 1.2667 .000 -10.165 -4.184
H -2.9871 1.2667 .050 -5.977 .003
H A -4.1871* 1.2667 .003 -7.177 -1.197
S 2.9871 1.2667 .050 -.003 5.977
Bonferroni A S 7.1743* 1.2667 .000 4.117 10.232
H 4.1871* 1.2667 .003 1.130 7.244
S A -7.1743* 1.2667 .000 -10.232 -4.117
H -2.9871 1.2667 .058 -6.044 .070
H A -4.1871* 1.2667 .003 -7.244 -1.130
S 2.9871 1.2667 .058 -.070 6.044
*. The mean difference is significant at the 0.05 level.

UNIANOVA
Multiple Comparisons
Dependent Variable: Max Length (mm)
(I) Impactor Mean Difference (I-J) Std. Error Sig. 95% Confidence Interval
Lower Bound Upper Bound
Tukey HSD A S 7.174* 1.2226 .000 4.288 10.061
H 4.187* 1.2226 .002 1.301 7.074
S A -7.174* 1.2226 .000 -10.061 -4.288
H -2.987* 1.2226 .041 -5.874 -.101
H A -4.187* 1.2226 .002 -7.074 -1.301
S 2.987* 1.2226 .041 .101 5.874
Bonferroni A S 7.174* 1.2226 .000 4.223 10.125
H 4.187* 1.2226 .002 1.236 7.138
S A -7.174* 1.2226 .000 -10.125 -4.223
H -2.987* 1.2226 .046 -5.938 -.036
H A -4.187* 1.2226 .002 -7.138 -1.236
S 2.987* 1.2226 .046 .036 5.938
"Based on observed means.
The error term is Mean Square(Error) = 52.316."
*. The mean difference is significant at the .05 level.

kd1

Re: Different result for SPSS ANOVA and UNIANOVA

Sorry,
the output is very garbled, here is a screenshot

test.png

Bruce Weaver

Re: Different result for SPSS ANOVA and UNIANOVA

Administrator

In reply to this post by kd1

Please post your syntax.

kd1 wrote

I am not sure if the below provides enough information, but why are the values different if I test the same variables using ANOVA or UNIANOVA (e.g. with ANOVA Bonferroni the difference between variable H and S is not significant, but is using UNIANOVA)?

ANOVA
Multiple Comparisons
Dependent Variable: Max Length (mm)
(I) Impactor Mean Difference (I-J) Std. Error Sig. 95% Confidence Interval
Lower Bound Upper Bound
Tukey HSD A S 7.1743* 1.2667 .000 4.184 10.165
H 4.1871* 1.2667 .003 1.197 7.177
S A -7.1743* 1.2667 .000 -10.165 -4.184
H -2.9871 1.2667 .050 -5.977 .003
H A -4.1871* 1.2667 .003 -7.177 -1.197
S 2.9871 1.2667 .050 -.003 5.977
Bonferroni A S 7.1743* 1.2667 .000 4.117 10.232
H 4.1871* 1.2667 .003 1.130 7.244
S A -7.1743* 1.2667 .000 -10.232 -4.117
H -2.9871 1.2667 .058 -6.044 .070
H A -4.1871* 1.2667 .003 -7.244 -1.130
S 2.9871 1.2667 .058 -.070 6.044
*. The mean difference is significant at the 0.05 level.

UNIANOVA
Multiple Comparisons
Dependent Variable: Max Length (mm)
(I) Impactor Mean Difference (I-J) Std. Error Sig. 95% Confidence Interval
Lower Bound Upper Bound
Tukey HSD A S 7.174* 1.2226 .000 4.288 10.061
H 4.187* 1.2226 .002 1.301 7.074
S A -7.174* 1.2226 .000 -10.061 -4.288
H -2.987* 1.2226 .041 -5.874 -.101
H A -4.187* 1.2226 .002 -7.074 -1.301
S 2.987* 1.2226 .041 .101 5.874
Bonferroni A S 7.174* 1.2226 .000 4.223 10.125
H 4.187* 1.2226 .002 1.236 7.138
S A -7.174* 1.2226 .000 -10.125 -4.223
H -2.987* 1.2226 .046 -5.938 -.036
H A -4.187* 1.2226 .002 -7.138 -1.236
S 2.987* 1.2226 .046 .036 5.938
"Based on observed means.
The error term is Mean Square(Error) = 52.316."
*. The mean difference is significant at the .05 level.

--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

PLEASE NOTE THE FOLLOWING:
1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above.
2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/).

Rich Ulrich

Re: Different result for SPSS ANOVA and UNIANOVA

In reply to this post by kd1

Why do the tests show different p values?
- The tables do show different standard errors being
used for the tests, so they will create different t-tests.
The mean differences are the same so it seems it should
be the same data.

Why do they have different standard errors?
- I hope that is explained by different specifications of
the models.

--
Rich Ulrich

> Date: Sun, 10 Mar 2013 06:18:39 -0700

> From: [hidden email]
> Subject: Re: Different result for SPSS ANOVA and UNIANOVA
> To: [hidden email]
>
> Sorry,
> the output is very garbled, here is a screenshot
> <http://spssx-discussion.1045642.n5.nabble.com/file/n5718515/test.png>
> test.png
> <http://spssx-discussion.1045642.n5.nabble.com/file/n5718515/test.png>
>
> ...