Standard Deviation vs Standard Error in Scatter Dot Plot
Erik.Gylfe
Wed, 10/18/2017 - 06:23 am
When make a Scatter Dot Plot together with mean values and Standard Deviation it looks fine. However, if I change to Standard Error the error bars actualy become much smaller than the true Standard Error. Is this a bug or am I doing something wrong. I enclose an illustration of the problem. The data in wave A6 has a Mean Value of 1.485, a Standard Deviation of 0.447 and a Standard Error of 0.075. The Mean Values and Standard Deviation are shown correctly in the left graph but the right graph indicates a Standard Error of 0.0128 (measured with Show Info) rather than 0.075.
John Weeks
WaveMetrics, Inc.
support@wavemetrics.com
October 18, 2017 at 10:06 am - Permalink
I think that's a bug. I use scatter dot plot a lot but I always use s.d. for the error bars, but I can reproduce what you see in IP7.05 on macOS.
John, s.e.m. is stdev/sqrt(n). But even so, the standard error calculated by the scatter dot plot package is out.
•wavestats dd
V_npnts= 20; V_numNaNs= 0; V_numINFs= 0; V_avg= 10.2673;
V_Sum= 205.346; V_sdev= 1.08918; V_sem= 0.243549; V_rms= 10.322;
V_adev= 0.82628; V_skew= -0.0352432; V_kurt= -0.77727;
V_minloc= 15; V_maxloc= 12; V_min= 8.30005; V_max= 12.1819;
V_minRowLoc= 15; V_maxRowLoc= 12; V_startRow= 0; V_endRow= 19;
// make a scatter dot plot using the GUI
// print the magnitude of error plotted by the package
•print root:Packages:WMScatterDotPlot:Graph0:counts_ddSy[2]-root:Packages:WMScatterDotPlot:Graph0:counts_ddSy[3]
0.0573254
October 18, 2017 at 12:17 pm - Permalink
Oops. Of course. I'll look at the Scatter Dot Plot package.
John Weeks
WaveMetrics, Inc.
support@wavemetrics.com
October 18, 2017 at 12:36 pm - Permalink
John Weeks
WaveMetrics, Inc.
support@wavemetrics.com
October 18, 2017 at 12:54 pm - Permalink
Erik
October 20, 2017 at 02:15 am - Permalink