Hi everyone,

I am trying to graph the coefficients I generated from a multilevel linear regression analysis. Unfortunately, the overall constant is significantly larger than the coefficients of my independent variables. I have plotted the graph without the inclusion of the constant (see below), however, I would like to include it to make it easier to understand the results. I've tried including a graph break using the advice given here: https://www.stata.com/support/faqs/g.../scale-breaks/ However, I wasn't successful.

That's the code I currently use for the graph including the constant

Code:
coefplot,  xline(0) ysize (15) xsize (20) omitted baselevels headings (0.Gender ="{bf:Gender}" 0.samePharmacy="{bf:Same pharmacy}" 0.FrequencyPharmacy="{bf:Frequency of pharmacy use}" 0.freeRx="{bf: Free prescription}" 1.AgeGroups ="{bf:Age group}", nogap)
The corresponding code without the constant is:

Code:
coefplot,  xline(0) ysize (15) xsize (20) drop (_cons) omitted baselevels headings (0.Gender ="{bf:Gender}" 0.samePharmacy="{bf:Same pharmacy}" 0.FrequencyPharmacy="{bf:Frequency of pharmacy use}" 0.freeRx="{bf: Free prescription}" 1.AgeGroups ="{bf:Age group}", nogap)
Can anyone help me to include a scale break as part of the coefplot command (preferably I would like to exclude 2-6 on the x-axis) or how to merge both files into a useful graph (which I have tried unsuccessfully using the advice given in the link above).

Many thanks
Franziska

Array Array