Loading ...
Sorry, an error occurred while loading the content.

Crossover Designs

Expand Messages
  • srmillis@mindspring.com
    I m not certain whether this is an appropriate question for this list, but I ll take a chance: I ve just started to study corssover designs in medical
    Message 1 of 6 , Mar 30, 2000
    • 0 Attachment
      I'm not certain whether this is an appropriate question for this list,
      but I'll take a chance:

      I've just started to study corssover designs in medical research, ie,
      both groups of patients/subjects get treatment and placebo but at
      different time periods.

      My question involves how to analyze the data: the crossover design seems
      to be a 1-within, 1-between ANOVA. Am I correct? Any recommendations for
      books/articles? What software do you prefer for an analysis like this?

      Thanks,
      SR Millis
    • Alex Gray
      Hi SR Millis, The crossover design can be view as such depending upon the actual design. Sometimes however, (again depending upon the design) I view the
      Message 2 of 6 , Mar 30, 2000
      • 0 Attachment
        Hi SR Millis,

        The crossover design can be view as such depending upon the actual design.
        Sometimes however, (again depending upon the design) I view the cross-over
        as a latin square and analyse it that way. In terms of software I would
        have to say SAS. Since release 6.12 SAS has beefed up the Proc Mixed, which
        allows one to analyse repeated measures designs in a univariate set-up with
        appropriate modelling for covariance structure.

        I would be happy to help you with the code for this type of analysis in SAS
        or run the analysis for you if you need help.

        Regards.

        Alex.

        srmillis@... wrote:

        > I'm not certain whether this is an appropriate question for this list,
        > but I'll take a chance:
        >
        > I've just started to study corssover designs in medical research, ie,
        > both groups of patients/subjects get treatment and placebo but at
        > different time periods.
        >
        > My question involves how to analyze the data: the crossover design seems
        > to be a 1-within, 1-between ANOVA. Am I correct? Any recommendations for
        > books/articles? What software do you prefer for an analysis like this?
        >
        > Thanks,
        > SR Millis
        >
        > ------------------------------------------------------------------------
        > You have a voice mail message waiting for you at iHello.com:
        > http://click.egroups.com/1/2377/6/_/_/_/954424662/
        > ------------------------------------------------------------------------
        >
        > To unsubscribe from this group, send an email to:
        > EcFinMatStGradSchool-unsubscribe@onelist.com
      • srmillis@mindspring.com
        Thanks, Alex. We re in the process of getting SAS as well as collecting the data now. I ll be in touch. Scott Millis
        Message 3 of 6 , Mar 30, 2000
        • 0 Attachment
          Thanks, Alex. We're in the process of getting SAS as well as collecting
          the data now. I'll be in touch.

          Scott Millis


          Alex Gray wrote:
          >
          > Hi SR Millis,
          >
          > The crossover design can be view as such depending upon the actual design.
          > Sometimes however, (again depending upon the design) I view the cross-over
          > as a latin square and analyse it that way. In terms of software I would
          > have to say SAS. Since release 6.12 SAS has beefed up the Proc Mixed, which
          > allows one to analyse repeated measures designs in a univariate set-up with
          > appropriate modelling for covariance structure.
          >
          > I would be happy to help you with the code for this type of analysis in SAS
          > or run the analysis for you if you need help.
          >
          > Regards.
          >
          > Alex.
          >
          > srmillis@... wrote:
          >
          > > I'm not certain whether this is an appropriate question for this list,
          > > but I'll take a chance:
          > >
          > > I've just started to study corssover designs in medical research, ie,
          > > both groups of patients/subjects get treatment and placebo but at
          > > different time periods.
          > >
          > > My question involves how to analyze the data: the crossover design seems
          > > to be a 1-within, 1-between ANOVA. Am I correct? Any recommendations for
          > > books/articles? What software do you prefer for an analysis like this?
          > >
          > > Thanks,
          > > SR Millis
          > >
          > > ------------------------------------------------------------------------
          > > You have a voice mail message waiting for you at iHello.com:
          > > http://click.egroups.com/1/2377/6/_/_/_/954424662/
          > > ------------------------------------------------------------------------
          > >
          > > To unsubscribe from this group, send an email to:
          > > EcFinMatStGradSchool-unsubscribe@onelist.com
          >
          > ------------------------------------------------------------------------
          > @Backup- Protect and Access your data any time, any where on the net.
          > Try @Backup FREE and recieve 300 points from mypoints.com Install now:
          > http://click.egroups.com/1/2345/6/_/_/_/954426254/
          > ------------------------------------------------------------------------
          >
          > To unsubscribe from this group, send an email to:
          > EcFinMatStGradSchool-unsubscribe@onelist.com
        • Alex Gray
          I look forward to it. Cheers. Alex.
          Message 4 of 6 , Mar 30, 2000
          • 0 Attachment
            I look forward to it.

            Cheers.

            Alex.

            srmillis@... wrote:

            > Thanks, Alex. We're in the process of getting SAS as well as collecting
            > the data now. I'll be in touch.
            >
            > Scott Millis
            >
            > Alex Gray wrote:
            > >
            > > Hi SR Millis,
            > >
            > > The crossover design can be view as such depending upon the actual design.
            > > Sometimes however, (again depending upon the design) I view the cross-over
            > > as a latin square and analyse it that way. In terms of software I would
            > > have to say SAS. Since release 6.12 SAS has beefed up the Proc Mixed, which
            > > allows one to analyse repeated measures designs in a univariate set-up with
            > > appropriate modelling for covariance structure.
            > >
            > > I would be happy to help you with the code for this type of analysis in SAS
            > > or run the analysis for you if you need help.
            > >
            > > Regards.
            > >
            > > Alex.
            > >
            > > srmillis@... wrote:
            > >
            > > > I'm not certain whether this is an appropriate question for this list,
            > > > but I'll take a chance:
            > > >
            > > > I've just started to study corssover designs in medical research, ie,
            > > > both groups of patients/subjects get treatment and placebo but at
            > > > different time periods.
            > > >
            > > > My question involves how to analyze the data: the crossover design seems
            > > > to be a 1-within, 1-between ANOVA. Am I correct? Any recommendations for
            > > > books/articles? What software do you prefer for an analysis like this?
            > > >
            > > > Thanks,
            > > > SR Millis
            > > >
            > > > ------------------------------------------------------------------------
            > > > You have a voice mail message waiting for you at iHello.com:
            > > > http://click.egroups.com/1/2377/6/_/_/_/954424662/
            > > > ------------------------------------------------------------------------
            > > >
            > > > To unsubscribe from this group, send an email to:
            > > > EcFinMatStGradSchool-unsubscribe@onelist.com
            > >
            > > ------------------------------------------------------------------------
            > > @Backup- Protect and Access your data any time, any where on the net.
            > > Try @Backup FREE and recieve 300 points from mypoints.com Install now:
            > > http://click.egroups.com/1/2345/6/_/_/_/954426254/
            > > ------------------------------------------------------------------------
            > >
            > > To unsubscribe from this group, send an email to:
            > > EcFinMatStGradSchool-unsubscribe@onelist.com
            >
            > ------------------------------------------------------------------------
            > Good friends, school spirit, hair-dos you'd like to forget.
            > Classmates.com has them all. And with 4.4 million alumni already
            > registered, there's a good chance you'll find your friends here:
            > http://click.egroups.com/1/2622/6/_/_/_/954426466/
            > ------------------------------------------------------------------------
            >
            > To unsubscribe from this group, send an email to:
            > EcFinMatStGradSchool-unsubscribe@onelist.com
          • srmillis@mindspring.com
            Alex: Background: I m a neuropsychologist who has recently gone back to school in applied stats. My study: determining efficacy of a particular drug in
            Message 5 of 6 , Mar 31, 2000
            • 0 Attachment
              Alex:
              Background: I'm a neuropsychologist who has recently gone back to school
              in applied stats. My study: determining efficacy of a particular drug in
              improving various aspects of cognition in person with multiple
              sclerosis. Patients have been give cognitive tests that measure stuff
              like learning and memory, rate of information processing, and capacity
              to initiate problem solving. The scores obtained from tese tests can be
              considered interval.

              The study was double-blind, placebo controlled. If A=drug and B=placebo,
              about half the subjects got ABAB while the other half got BABA. We had
              drug "wash-out" periods in between. Subjects were given the cognitive
              tests after each period. (The tests had alternate forms).

              My first question is: now that we have the data collected, what is the
              best way to set up the database "spreadsheet"? When I've used SPSS for
              repeated measures analysis, subjects were in "rows" with each subject
              having 1 row of data. And TestA--time1 was a column variable and
              TestA-time2 was a second column variable, etc.

              However, when I've done hierarchical linear modeling/growth curve
              analysis, each time period was a separate row of data.

              I also need to mention that this is a small pilot study with less than
              15 subjects. Would you recommend using an exact test rather than its
              parametric counterpart?

              Thanks for your help!

              Scott Millis

              --
              Scott R. Millis, PhD, ABPP
              Kessler Medical Rehabilitation Research & Education Corp
              1199 Pleasant Valley Way
              West Orange, New Jersey 07052

              Tel: 973.243.6976
              Fax: 973.243.6990
              Emails: smillis@...
              srmillis@...

              Web: www.rehabtrials.org
            • srmillis@mindspring.com
              I apologize to the list for my last posting. It was intended to be sent to an individual on the list rather than the entire lsit. Sorry! Scott Millis
              Message 6 of 6 , Mar 31, 2000
              • 0 Attachment
                I apologize to the list for my last posting. It was intended to be sent
                to an individual on the list rather than the entire lsit.

                Sorry!

                Scott Millis
              Your message has been successfully submitted and would be delivered to recipients shortly.