Posted by
Rohan Lulham-2 on
Aug 28, 2006; 10:33am
URL: http://spssx-discussion.165.s1.nabble.com/significantly-not-significant-tp1070571p1070584.html
Hello
Thank you for the really helpful and interesting suggestions. The paper was
really helpful- CI approach may be relevant.
I haven't got a lot of statistical training, and I think the idea of going
back through the theory is the way to go. It is likely this will need to
stay on hold as need to finish a thesis.
The issue really is a post hoc question as there was not a specific
hypothesis. It does however fit with the theoretical basis of my research (
the research uses control theory to look at influence of physical
environment on staff and detainee perceptions of self and others in
juvenile detention centres) the data is likely also not particularly strong
data for conducting complicated statistical analyses.
Based on perceptual control theory (Powers, 1973) a proposition is that
people control perceptions in the environment - they maintain them at a
reference level and resist disturbances (e.g. maintaining perceptions of
self as tough) ( I think I am providing too much information - apologies).
A result in my research was staff participants in three different simulated
environments all had rating of self (good-bad) which were very similar. The
estimated marginal means are shown below. From the results it appears they
are controlling perceptions of self - although ratings of other objects,
and self on other dimensions (toughness), varied.
Estimates
Dependent Variable: DC You- Evaluation
TreatmentMeanStd. Error95% Confidence Interval
Lower BoundUpper Bound
Unit 1.667.468-.2751.608
Unit 2.778.449-.1251.680
Unit 3.756.454-.1581.669
It is likely may need to do more looking in control theory for analysis
approaches - but due to the qualities of data describing the data as
indicative of the explanation will likely suffice.
Thankyou for time - I better get back to it - as need to get this thesis
finished.
regards rohan
At 05:10 PM 28/08/2006, la volta statistics wrote:
>Hello Rohan
>
>look at the following link:
>
http://www.diahome.org/content/abstract/1999/d3341205.pdf>
>It contains a good description for what you are aiming for.
>
>Good luck, Christian
>
>*******************************
>la volta statistics
>Christian Schmidhauser, Dr.phil.II
>Weinbergstrasse 108
>Ch-8006 Zürich
>Tel: +41 (043) 233 98 01
>Fax: +41 (043) 233 98 02
>mailto:
[hidden email]
>
>
>-----Ursprüngliche Nachricht-----
>Von: SPSSX(r) Discussion [mailto:
[hidden email]]Im Auftrag von
>Rohan Lulham
>Gesendet: Sonntag, 27. August 2006 08:21
>An:
[hidden email]
>Betreff: "significantly" not significant
>
>
>Hello
>
>This maybe be a naive question, but was wondering how to test if the
>difference between to means are significantly similar (e.g. "significantly"
>not significant differences). For example being able to say with 95%
>confidence that the two means are similar or the same based on the results
>of a test. It is very basic question, but rarely talked about were most
>research traditions are predominantly looking for differences.
>
>Can the probability value of say an ANOVA test be used so that a p value
>= .95 would indicate 95% confidence they are similar? I am thinking due to
>some underlying statistical theory to the approach this is not appropriate,
>but am not quite sure.
>
>Is there another way to set up an anlysis that can test for similarity
>between two means, or distributions?
>
>Thanks for time
>
>Rohan
Rohan Lulham
Ph.D. Student
Environment, Behaviour and Society Research Group
Faculty of Architecture, University of Sydney
Australia