Thanks! Meta-analysis just keeps on getting more and more important, in this age of the replication crisis and rise of Open Science. Bring on more replication, more meta-analysis, and open data and materials. And may all your confidence intervals be short!
Some questions: Up front you mentioned Estimation being good... but it is not mentioned again. Does this mean confidence intervals? And... your presentation seems to assume known statistics. How would you handle drug testing, where only the null hypothesis statistics are known (only random effects), because the positive hypothesis (it improves outcomes) is of unknown strength and cannot be modelled. That is why drug trials use only the null hypothesis as set p = .05 in advance.
Thanks very much Claire! With Open Science practices marching ahead, meta-analysis only becomes more important, more central. Even beginning students should learn about at least the basics of meta-analysis, and the forest plot makes it all visual and simple. Bob and I are working at the moment on the second edition of our intro text 'Introduction to the new statistics: Estimation, Open Science, and beyond'. May all your confidence intervals be short! Geoff
Dear Professor , downloaded your "ESCI-for-Excel-07-10" software. BEAUTIFUL! I am not a research student of Statistics (but a life long learner) so I am facing difficulties to use it. Can you suggest which excel macro to use for Trend Analysis of Stock Market.
Thank you Geoff. That was a nice summary
Thanks! Meta-analysis just keeps on getting more and more important, in this age of the replication crisis and rise of Open Science. Bring on more replication, more meta-analysis, and open data and materials. And may all your confidence intervals be short!
Some questions: Up front you mentioned Estimation being good... but it is not mentioned again. Does this mean confidence intervals?
And... your presentation seems to assume known statistics. How would you handle drug testing, where only the null hypothesis statistics are known (only random effects), because the positive hypothesis (it improves outcomes) is of unknown strength and cannot be modelled. That is why drug trials use only the null hypothesis as set p = .05 in advance.
Dr. Cumming, this is a job well done. :)
Thanks very much Claire! With Open Science practices marching ahead, meta-analysis only becomes more important, more central. Even beginning students should learn about at least the basics of meta-analysis, and the forest plot makes it all visual and simple. Bob and I are working at the moment on the second edition of our intro text 'Introduction to the new statistics: Estimation, Open Science, and beyond'. May all your confidence intervals be short! Geoff
Dear Professor , downloaded your "ESCI-for-Excel-07-10" software. BEAUTIFUL! I am not a research student of Statistics (but a life long learner) so I am facing difficulties to use it. Can you suggest which excel macro to use for Trend Analysis of Stock Market.