Compare odds ratios

[Charlotte-9]

Dear all,

Could someone please tell me the correct test to use to compare whether
two odds ratios are significantly different.  For instance, I want to
compare the odds ratio for group A (odds ratio = 0.487) against the odds
ratio for the reference group (odds = 1).

Thanks,

Lou

[Marta García-Granero]

Hi Lou

Perhaps you havent' explained yourself completely... or I see an error
in your question: you just can't compare and odds-ratio with an odd,
and an OR from a group with the one of its reference group. I think
I'm being confused by the use of the term "reference group". Are you
talking about one qualitative variable that has been dummy coded, with
one group being the reference or is it something different?

L> Could someone please tell me the correct test to use to compare whether
L> two odds ratios are significantly different.  For instance, I want to
L> compare the odds ratio for group A (odds ratio = 0.487) against the odds
L> ratio for the reference group (odds = 1).


--
Regards,
Dr. Marta García-Granero,PhD           mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does
not understand. SPSS syntax guide cannot supply this knowledge, and it
is certainly no substitute for the basic understanding of statistics
and statistical thinking that is essential for the wise choice of
methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

[Hector Maletta]

I suspect what you mean is whether the 0.487 odds ratio is significantly
different from 1. In other words, you want to know whether you can reject
the null hypothesis that a certain risk factor has no additional effect at
all compared with the reference situation. Is that so? I would not suspect
that with 0.487, except if my sample is really too small, but if the figure
is 0.98 the suspicion may be justifiable.
Hector

-----Mensaje original-----
De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de Marta
García-Granero
Enviado el: Tuesday, August 29, 2006 11:25 AM
Para: [hidden email]
Asunto: Re: Compare odds ratios

Hi Lou

Perhaps you havent' explained yourself completely... or I see an error
in your question: you just can't compare and odds-ratio with an odd,
and an OR from a group with the one of its reference group. I think
I'm being confused by the use of the term "reference group". Are you
talking about one qualitative variable that has been dummy coded, with
one group being the reference or is it something different?

L> Could someone please tell me the correct test to use to compare whether
L> two odds ratios are significantly different.  For instance, I want to
L> compare the odds ratio for group A (odds ratio = 0.487) against the odds
L> ratio for the reference group (odds = 1).


--
Regards,
Dr. Marta García-Granero,PhD           mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does
not understand. SPSS syntax guide cannot supply this knowledge, and it
is certainly no substitute for the basic understanding of statistics
and statistical thinking that is essential for the wise choice of
methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

[Charlotte-9]

Hi Marta,

Okay, I'll try to clarify matters.  As usual, I'm trying to immitate work
that has been done previously, so the whole thing may be inappropriate.

I have used logistic regression to obtain adjusted odds ratios.  The
outcome is 'screened' or 'not screened' and the independent variables are
age, ethnicity and deprivation.  I am focusing on the odds ratios for the
ethnicity variable.  When I talk about reference group, I mean that the
odds ratios are all calculated with respect to one group.  In other words,
I have five Asian categories (Hindu, Muslim, Sikh etc) and the odds of
undertaking screening are compared to the odds for the non-Asian group
(the reference group).  I am trying to show that the odds of undertaking
screening are lower for the Asian groups when compared to the non-Asian
group.

In previous analysis, the people who produced the report have stated the
following, "the Muslim group demonstrated a significantly (p<0.05) lower
uptake with adjusted odds ratio 0.37 versus 1.0 for non-Asians".  I'm
basically trying to work out what they did and assumed they somehow
compared the odds.

Hope this clarifies things a bit - any suggestions gratefully received.

Best wishes,

Lou



On Tue, 29 Aug 2006 16:24:52 +0200, =?ISO-8859-15?B?
TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]> wrote:

>Hi Lou
>
>Perhaps you havent' explained yourself completely... or I see an error
>in your question: you just can't compare and odds-ratio with an odd,
>and an OR from a group with the one of its reference group. I think
>I'm being confused by the use of the term "reference group". Are you
>talking about one qualitative variable that has been dummy coded, with
>one group being the reference or is it something different?
>
>L> Could someone please tell me the correct test to use to compare whether
>L> two odds ratios are significantly different.  For instance, I want to
>L> compare the odds ratio for group A (odds ratio = 0.487) against the

odds

>L> ratio for the reference group (odds = 1).
>
>
>--
>Regards,
>Dr. Marta García-Granero,PhD           mailto:[hidden email]
>Statistician
>
>---
>"It is unwise to use a statistical procedure whose use one does
>not understand. SPSS syntax guide cannot supply this knowledge, and it
>is certainly no substitute for the basic understanding of statistics
>and statistical thinking that is essential for the wise choice of
>methods and the correct interpretation of their results".
>
>(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

[Charlotte-9]

....Hmmmmm....having just re-read what the people who produced the last
report wrote, I'm now thinking that they used a chi-squared test to
compare uptake levels and stated the odds ratios as a separate matter
(just to try to strengthen their argument).  Does this make more sense?
If so, it's very ambiguous what they have written.  I would still be
interested to hear if I should be doing anything with the odds ratios to
back up the results.....

Thanks,

Lou

On Tue, 29 Aug 2006 10:35:08 -0400, Lou <[hidden email]> wrote:

>Hi Marta,
>
>Okay, I'll try to clarify matters.  As usual, I'm trying to immitate work
>that has been done previously, so the whole thing may be inappropriate.
>
>I have used logistic regression to obtain adjusted odds ratios.  The
>outcome is 'screened' or 'not screened' and the independent variables are
>age, ethnicity and deprivation.  I am focusing on the odds ratios for the
>ethnicity variable.  When I talk about reference group, I mean that the
>odds ratios are all calculated with respect to one group.  In other words,
>I have five Asian categories (Hindu, Muslim, Sikh etc) and the odds of
>undertaking screening are compared to the odds for the non-Asian group
>(the reference group).  I am trying to show that the odds of undertaking
>screening are lower for the Asian groups when compared to the non-Asian
>group.
>
>In previous analysis, the people who produced the report have stated the
>following, "the Muslim group demonstrated a significantly (p<0.05) lower
>uptake with adjusted odds ratio 0.37 versus 1.0 for non-Asians".  I'm
>basically trying to work out what they did and assumed they somehow
>compared the odds.
>
>Hope this clarifies things a bit - any suggestions gratefully received.
>
>Best wishes,
>
>Lou
>
>
>
>On Tue, 29 Aug 2006 16:24:52 +0200, =?ISO-8859-15?B?
>TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]> wrote:
>
>>Hi Lou
>>
>>Perhaps you havent' explained yourself completely... or I see an error
>>in your question: you just can't compare and odds-ratio with an odd,
>>and an OR from a group with the one of its reference group. I think
>>I'm being confused by the use of the term "reference group". Are you
>>talking about one qualitative variable that has been dummy coded, with
>>one group being the reference or is it something different?
>>
>>L> Could someone please tell me the correct test to use to compare

whether

>>L> two odds ratios are significantly different.  For instance, I want to
>>L> compare the odds ratio for group A (odds ratio = 0.487) against the
>odds
>>L> ratio for the reference group (odds = 1).
>>
>>
>>--
>>Regards,
>>Dr. Marta García-Granero,PhD           mailto:[hidden email]
>>Statistician
>>
>>---
>>"It is unwise to use a statistical procedure whose use one does
>>not understand. SPSS syntax guide cannot supply this knowledge, and it
>>is certainly no substitute for the basic understanding of statistics
>>and statistical thinking that is essential for the wise choice of
>>methods and the correct interpretation of their results".
>>
>>(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

[Adam Troy-2]

Lou,

I've pasted a reference below which discusses comparing odds ratios from two
different samples, but I'm not sure that's what you need.


Altman, D. G., & Bland, J. M. (2003). Statistics notes: Interaction
revisited: the difference between two estimates. BMJ, 326, 219.


_____________________________
Adam B. Troy, Ph.D.
Director of Research
GoalQuest, Inc.




----- Original Message -----
From: Lou <[hidden email]>
Date: Tuesday, August 29, 2006 10:40 am
Subject: Re: Compare odds ratios

> ....Hmmmmm....having just re-read what the people who produced the
> lastreport wrote, I'm now thinking that they used a chi-squared
> test to
> compare uptake levels and stated the odds ratios as a separate matter
> (just to try to strengthen their argument).  Does this make more
> sense?If so, it's very ambiguous what they have written.  I would
> still be
> interested to hear if I should be doing anything with the odds
> ratios to
> back up the results.....
>
> Thanks,
>
> Lou
>
> On Tue, 29 Aug 2006 10:35:08 -0400, Lou <[hidden email]>
> wrote:
> >Hi Marta,
> >
> >Okay, I'll try to clarify matters.  As usual, I'm trying to
> immitate work
> >that has been done previously, so the whole thing may be
> inappropriate.>
> >I have used logistic regression to obtain adjusted odds ratios.  The
> >outcome is 'screened' or 'not screened' and the independent
> variables are
> >age, ethnicity and deprivation.  I am focusing on the odds ratios
> for the
> >ethnicity variable.  When I talk about reference group, I mean
> that the
> >odds ratios are all calculated with respect to one group.  In
> other words,
> >I have five Asian categories (Hindu, Muslim, Sikh etc) and the
> odds of
> >undertaking screening are compared to the odds for the non-Asian
> group>(the reference group).  I am trying to show that the odds of
> undertaking>screening are lower for the Asian groups when compared
> to the non-Asian
> >group.
> >
> >In previous analysis, the people who produced the report have
> stated the
> >following, "the Muslim group demonstrated a significantly
> (p<0.05) lower
> >uptake with adjusted odds ratio 0.37 versus 1.0 for non-Asians".  I'm
> >basically trying to work out what they did and assumed they somehow
> >compared the odds.
> >
> >Hope this clarifies things a bit - any suggestions gratefully
> received.>
> >Best wishes,
> >
> >Lou
> >
> >
> >
> >On Tue, 29 Aug 2006 16:24:52 +0200, =?ISO-8859-15?B?
> >TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]> wrote:
> >
> >>Hi Lou
> >>
> >>Perhaps you havent' explained yourself completely... or I see an
> error>>in your question: you just can't compare and odds-ratio
> with an odd,
> >>and an OR from a group with the one of its reference group. I think
> >>I'm being confused by the use of the term "reference group". Are you
> >>talking about one qualitative variable that has been dummy
> coded, with
> >>one group being the reference or is it something different?
> >>
> >>L> Could someone please tell me the correct test to use to compare
> whether
> >>L> two odds ratios are significantly different.  For instance, I
> want to
> >>L> compare the odds ratio for group A (odds ratio = 0.487)
> against the
> >odds
> >>L> ratio for the reference group (odds = 1).
> >>
> >>
> >>--
> >>Regards,
> >>Dr. Marta García-Granero,PhD           mailto:[hidden email]
> >>Statistician
> >>
> >>---
> >>"It is unwise to use a statistical procedure whose use one does
> >>not understand. SPSS syntax guide cannot supply this knowledge,
> and it
> >>is certainly no substitute for the basic understanding of statistics
> >>and statistical thinking that is essential for the wise choice of
> >>methods and the correct interpretation of their results".
> >>
> >>(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)
>

[Marta García-Granero]

Hi Lou:

If you have used logistic regression, then the answer is right in the
output:

- Check the dummy coding table to find out which dummy variable of the
set reflects "Muslim group" (could be ethnicity(1), or ethnicity(2)...
depending on the order of the categories).
- Check the Wald test significance for that dummy variable: that's the
significance you are looking for (caveats: not very sensitive, but it
can't be helped). Alternatively, the same information can be deduced
from the 95%CI for exp(b), provided you've been farsighted enough to
ask for them (this reminds me I have to write to Kyle Weeks
concerning some ideas and suggestions for SPSS 16, and that's one of
them: 95%CI should be default output in most statistical methods).

If you get lost with this explanation (about locating dummies and Wald
test and so on), just send the logistic regression output (draft) and
I'll point the important results to you

The term "adjusted odds-ratio" the authors used in the report means
that they used multiple logistic regression to control for other
factors (age and deprivation in this case).

L> Okay, I'll try to clarify matters.  As usual, I'm trying to immitate work
L> that has been done previously, so the whole thing may be inappropriate.

The old "imitation theorem": "if it's published, then it's correct" ;)

L> I have used logistic regression to obtain adjusted odds ratios. The
L> outcome is 'screened' or 'not screened' and the independent variables are
L> age, ethnicity and deprivation.  I am focusing on the odds ratios for the
L> ethnicity variable.  When I talk about reference group, I mean that the
L> odds ratios are all calculated with respect to one group.  In other words,
L> I have five Asian categories (Hindu, Muslim, Sikh etc) and the odds of
L> undertaking screening are compared to the odds for the non-Asian group
L> (the reference group).  I am trying to show that the odds of undertaking
L> screening are lower for the Asian groups when compared to the non-Asian
L> group.

L> In previous analysis, the people who produced the report have stated the
L> following, "the Muslim group demonstrated a significantly (p<0.05) lower
L> uptake with adjusted odds ratio 0.37 versus 1.0 for non-Asians".  I'm
L> basically trying to work out what they did and assumed they somehow
L> compared the odds.

L> Hope this clarifies things a bit - any suggestions gratefully received.

L> Best wishes,

L> Lou



L> On Tue, 29 Aug 2006 16:24:52 +0200, =?ISO-8859-15?B?
L> TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]> wrote:

>>Hi Lou
>>
>>Perhaps you havent' explained yourself completely... or I see an error
>>in your question: you just can't compare and odds-ratio with an odd,
>>and an OR from a group with the one of its reference group. I think
>>I'm being confused by the use of the term "reference group". Are you
>>talking about one qualitative variable that has been dummy coded, with
>>one group being the reference or is it something different?
>>
>>L> Could someone please tell me the correct test to use to compare whether
>>L> two odds ratios are significantly different.  For instance, I want to
>>L> compare the odds ratio for group A (odds ratio = 0.487) against the

L> odds

>>L> ratio for the reference group (odds = 1).
>>
>>
>>--
>>Regards,
>>Dr. Marta García-Granero,PhD           mailto:[hidden email]
>>Statistician
>>
>>---
>>"It is unwise to use a statistical procedure whose use one does
>>not understand. SPSS syntax guide cannot supply this knowledge, and it
>>is certainly no substitute for the basic understanding of statistics
>>and statistical thinking that is essential for the wise choice of
>>methods and the correct interpretation of their results".
>>
>>(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

L> __________ Información de NOD32, revisión 1.1724 (20060824) __________

L> Este mensaje ha sido analizado con  NOD32 antivirus system
L> http://www.nod32.com




--
Regards,
Dr. Marta García-Granero,PhD           mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does
not understand. SPSS syntax guide cannot supply this knowledge, and it
is certainly no substitute for the basic understanding of statistics
and statistical thinking that is essential for the wise choice of
methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

[Charlotte-9]

Hi Marta,

Thanks for your response.  I think it's fair to say that I'm having a bad
day (month, actually) and my brain packed up for a while there.  I
realised as soon as I posted the question that it was stupid!  I
understand everything you have said below, so feel happy at interpreting
the results.  I think I'm going to have to adopt a policy of not looking
at this other report because it sends me insane!!

Thanks to you and everyone else who responded for your help.

Lou

On Tue, 29 Aug 2006 16:50:27 +0200, =?ISO-8859-1?B?
TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]> wrote:

>Hi Lou:
>
>If you have used logistic regression, then the answer is right in the
>output:
>
>- Check the dummy coding table to find out which dummy variable of the
>set reflects "Muslim group" (could be ethnicity(1), or ethnicity(2)...
>depending on the order of the categories).
>- Check the Wald test significance for that dummy variable: that's the
>significance you are looking for (caveats: not very sensitive, but it
>can't be helped). Alternatively, the same information can be deduced
>from the 95%CI for exp(b), provided you've been farsighted enough to
>ask for them (this reminds me I have to write to Kyle Weeks
>concerning some ideas and suggestions for SPSS 16, and that's one of
>them: 95%CI should be default output in most statistical methods).
>
>If you get lost with this explanation (about locating dummies and Wald
>test and so on), just send the logistic regression output (draft) and
>I'll point the important results to you
>
>The term "adjusted odds-ratio" the authors used in the report means
>that they used multiple logistic regression to control for other
>factors (age and deprivation in this case).
>
>L> Okay, I'll try to clarify matters.  As usual, I'm trying to immitate

work
>L> that has been done previously, so the whole thing may be inappropriate.
>
>The old "imitation theorem": "if it's published, then it's correct" ;)
>
>L> I have used logistic regression to obtain adjusted odds ratios. The
>L> outcome is 'screened' or 'not screened' and the independent variables
are
>L> age, ethnicity and deprivation.  I am focusing on the odds ratios for
the
>L> ethnicity variable.  When I talk about reference group, I mean that the
>L> odds ratios are all calculated with respect to one group.  In other
words,
>L> I have five Asian categories (Hindu, Muslim, Sikh etc) and the odds of
>L> undertaking screening are compared to the odds for the non-Asian group
>L> (the reference group).  I am trying to show that the odds of
undertaking
>L> screening are lower for the Asian groups when compared to the non-Asian
>L> group.
>
>L> In previous analysis, the people who produced the report have stated
the
>L> following, "the Muslim group demonstrated a significantly (p<0.05)
lower

>L> uptake with adjusted odds ratio 0.37 versus 1.0 for non-Asians".  I'm
>L> basically trying to work out what they did and assumed they somehow
>L> compared the odds.
>
>L> Hope this clarifies things a bit - any suggestions gratefully received.
>
>L> Best wishes,
>
>L> Lou
>
>
>
>L> On Tue, 29 Aug 2006 16:24:52 +0200, =?ISO-8859-15?B?
>L> TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]> wrote:
>
>>>Hi Lou
>>>
>>>Perhaps you havent' explained yourself completely... or I see an error
>>>in your question: you just can't compare and odds-ratio with an odd,
>>>and an OR from a group with the one of its reference group. I think
>>>I'm being confused by the use of the term "reference group". Are you
>>>talking about one qualitative variable that has been dummy coded, with
>>>one group being the reference or is it something different?
>>>
>>>L> Could someone please tell me the correct test to use to compare

whether

>>>L> two odds ratios are significantly different.  For instance, I want to
>>>L> compare the odds ratio for group A (odds ratio = 0.487) against the
>L> odds
>>>L> ratio for the reference group (odds = 1).
>>>
>>>
>>>--
>>>Regards,
>>>Dr. Marta García-Granero,PhD           mailto:[hidden email]
>>>Statistician
>>>
>>>---
>>>"It is unwise to use a statistical procedure whose use one does
>>>not understand. SPSS syntax guide cannot supply this knowledge, and it
>>>is certainly no substitute for the basic understanding of statistics
>>>and statistical thinking that is essential for the wise choice of
>>>methods and the correct interpretation of their results".
>>>
>>>(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)
>
>L> __________ Información de NOD32, revisión 1.1724 (20060824) __________
>
>L> Este mensaje ha sido analizado con  NOD32 antivirus system
>L> http://www.nod32.com
>
>
>
>
>--
>Regards,
>Dr. Marta García-Granero,PhD           mailto:[hidden email]
>Statistician
>
>---
>"It is unwise to use a statistical procedure whose use one does
>not understand. SPSS syntax guide cannot supply this knowledge, and it
>is certainly no substitute for the basic understanding of statistics
>and statistical thinking that is essential for the wise choice of
>methods and the correct interpretation of their results".
>
>(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

[Charlotte-9]

Dear all,

While I'm on the topic of odds ratios, just another quick question.  I
have calculated both unadjusted and adjusted odds ratios for the three
independent variables in my logistic regression model (age, ethnicity,
deprivation).  I also have 95% CIs.  For the age and deprivation
variables, the unadjusted and adjusted odds ratios are practically the
same.  In the case of the ethnicity variable, however, there are some
fairly large differences, with the adjusted odds being larger (but still
less than 1).  Does this mean that the influence of ethnicity on screening
is not as large as the influence of deprivation and age and that adjuting
for age and deprivation shows the reduced effect of ethnicity?

Thanks,

Lou



On Tue, 29 Aug 2006 10:55:34 -0400, Lou <[hidden email]> wrote:

>Hi Marta,
>
>Thanks for your response.  I think it's fair to say that I'm having a bad
>day (month, actually) and my brain packed up for a while there.  I
>realised as soon as I posted the question that it was stupid!  I
>understand everything you have said below, so feel happy at interpreting
>the results.  I think I'm going to have to adopt a policy of not looking
>at this other report because it sends me insane!!
>
>Thanks to you and everyone else who responded for your help.
>
>Lou
>
>On Tue, 29 Aug 2006 16:50:27 +0200, =?ISO-8859-1?B?
>TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]> wrote:
>
>>Hi Lou:
>>
>>If you have used logistic regression, then the answer is right in the
>>output:
>>
>>- Check the dummy coding table to find out which dummy variable of the
>>set reflects "Muslim group" (could be ethnicity(1), or ethnicity(2)...
>>depending on the order of the categories).
>>- Check the Wald test significance for that dummy variable: that's the
>>significance you are looking for (caveats: not very sensitive, but it
>>can't be helped). Alternatively, the same information can be deduced
>>from the 95%CI for exp(b), provided you've been farsighted enough to
>>ask for them (this reminds me I have to write to Kyle Weeks
>>concerning some ideas and suggestions for SPSS 16, and that's one of
>>them: 95%CI should be default output in most statistical methods).
>>
>>If you get lost with this explanation (about locating dummies and Wald
>>test and so on), just send the logistic regression output (draft) and
>>I'll point the important results to you
>>
>>The term "adjusted odds-ratio" the authors used in the report means
>>that they used multiple logistic regression to control for other
>>factors (age and deprivation in this case).
>>
>>L> Okay, I'll try to clarify matters.  As usual, I'm trying to immitate
>work
>>L> that has been done previously, so the whole thing may be

inappropriate.
>>
>>The old "imitation theorem": "if it's published, then it's correct" ;)
>>
>>L> I have used logistic regression to obtain adjusted odds ratios. The
>>L> outcome is 'screened' or 'not screened' and the independent variables
>are
>>L> age, ethnicity and deprivation.  I am focusing on the odds ratios for
>the
>>L> ethnicity variable.  When I talk about reference group, I mean that
the
>>L> odds ratios are all calculated with respect to one group.  In other
>words,
>>L> I have five Asian categories (Hindu, Muslim, Sikh etc) and the odds of
>>L> undertaking screening are compared to the odds for the non-Asian group
>>L> (the reference group).  I am trying to show that the odds of
>undertaking
>>L> screening are lower for the Asian groups when compared to the non-
Asian

>>L> group.
>>
>>L> In previous analysis, the people who produced the report have stated
>the
>>L> following, "the Muslim group demonstrated a significantly (p<0.05)
>lower
>>L> uptake with adjusted odds ratio 0.37 versus 1.0 for non-Asians".  I'm
>>L> basically trying to work out what they did and assumed they somehow
>>L> compared the odds.
>>
>>L> Hope this clarifies things a bit - any suggestions gratefully

received.

>>
>>L> Best wishes,
>>
>>L> Lou
>>
>>
>>
>>L> On Tue, 29 Aug 2006 16:24:52 +0200, =?ISO-8859-15?B?
>>L> TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]> wrote:
>>
>>>>Hi Lou
>>>>
>>>>Perhaps you havent' explained yourself completely... or I see an error
>>>>in your question: you just can't compare and odds-ratio with an odd,
>>>>and an OR from a group with the one of its reference group. I think
>>>>I'm being confused by the use of the term "reference group". Are you
>>>>talking about one qualitative variable that has been dummy coded, with
>>>>one group being the reference or is it something different?
>>>>
>>>>L> Could someone please tell me the correct test to use to compare
>whether
>>>>L> two odds ratios are significantly different.  For instance, I want

to

>>>>L> compare the odds ratio for group A (odds ratio = 0.487) against the
>>L> odds
>>>>L> ratio for the reference group (odds = 1).
>>>>
>>>>
>>>>--
>>>>Regards,
>>>>Dr. Marta García-Granero,PhD           mailto:[hidden email]
>>>>Statistician
>>>>
>>>>---
>>>>"It is unwise to use a statistical procedure whose use one does
>>>>not understand. SPSS syntax guide cannot supply this knowledge, and it
>>>>is certainly no substitute for the basic understanding of statistics
>>>>and statistical thinking that is essential for the wise choice of
>>>>methods and the correct interpretation of their results".
>>>>
>>>>(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)
>>
>>L> __________ Información de NOD32, revisión 1.1724 (20060824) __________
>>
>>L> Este mensaje ha sido analizado con  NOD32 antivirus system
>>L> http://www.nod32.com
>>
>>
>>
>>
>>--
>>Regards,
>>Dr. Marta García-Granero,PhD           mailto:[hidden email]
>>Statistician
>>
>>---
>>"It is unwise to use a statistical procedure whose use one does
>>not understand. SPSS syntax guide cannot supply this knowledge, and it
>>is certainly no substitute for the basic understanding of statistics
>>and statistical thinking that is essential for the wise choice of
>>methods and the correct interpretation of their results".
>>
>>(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

[Marta García-Granero]

Hi Lou

It just reflects the fact that age and deprivation are confounding
factors in the relationship between ethnicity and the outcome, but
that it doesn't happen the other way (since confounding isn't
necessarily a reciprocal phenomenon, while interaction is).

L> While I'm on the topic of odds ratios, just another quick question.  I
L> have calculated both unadjusted and adjusted odds ratios for the three
L> independent variables in my logistic regression model (age, ethnicity,
L> deprivation).  I also have 95% CIs.  For the age and deprivation
L> variables, the unadjusted and adjusted odds ratios are practically the
L> same.  In the case of the ethnicity variable, however, there are some
L> fairly large differences, with the adjusted odds being larger (but still
L> less than 1).  Does this mean that the influence of ethnicity on screening
L> is not as large as the influence of deprivation and age and that adjuting
L> for age and deprivation shows the reduced effect of ethnicity?


--
Regards,
Dr. Marta García-Granero,PhD           mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does
not understand. SPSS syntax guide cannot supply this knowledge, and it
is certainly no substitute for the basic understanding of statistics
and statistical thinking that is essential for the wise choice of
methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)