Significant differences in the means?

[Hans-Christian Stahl]

Dear all



I know these are basics - but I can't help and really need some help or
advice.  Any help is greatly appreciated!



I have the following data:






Mean duration Chemotherapy

Mean duration Targeted Therapy


1st Line

5,72 (A)

6,57 (B)


2nd Line

5,11 (C)

5,68 (D)



For A I have: 236 (Valid CASES)

For B: 128

For C: 141

For D: 71



Moreover I have the following table:






Mean duration Chemotherapy


1st Line

5,46 (A)


2nd Line

5,46 (B)



For A: 763 (Valid CASES)

For B: 289



How can I proceed to determine if there are significant differences in the
groups?



Christian

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Richard Ristow]

At 08:21 AM 10/17/2007, Hans-Christian Stahl wrote:

>I have the following data:
>
>Mean duration Chemotherapy
>Mean duration Targeted Therapy
>
>1st Line 5,72 (A) 6,57 (B)
>2nd Line 5,11 (C) 5,68 (D)
>
>For A I have: 236 (Valid CASES)
>For B: 128
>For C: 141
>For D:  71
>
>How can I proceed to determine if there are significant differences in
>the groups?

I'm very sorry, but you probably can't. You have counts and group means
for your groups; but ANOVA, the standard assessment of significant
differences in means, requires the group standard deviations as well.
Nonparametric tests require various other measures of group dispersion,
sometimes including all the values observed.

If you have the original data from which these means were computed, you
can do a lot. If not, I can't think of a solution.

-Better luck,
  Richard

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Richard Ristow]

Postscript:

At 03:28 AM 10/18/2007, Hans-Christian Stahl wrote:

>I have the original data from which the means were computed

If you have SPSS (I assume you do), I'd enter the original data into an
SPSS file, and analyze from that.

As I wrote before, you probably should find a statistical methodologist
in your area to work with, who can give you more detailed explanations
of the methods than we'll be able to, on the list.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[DEMUNA Chorrillos-Peru]

Dear colleagues.
Please, is there some macro or sintax for compute confidence intervals to partial correlation?

Thanks.

Cesar Merino

_________________________________________________________________
Explore the seven wonders of the world
http://search.msn.com/results.aspx?q=7+wonders+world&mkt=en-US&form=QBRE

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD