If I have understood your issue, you are detailing the inverse relationship, the required increase of an input infers a decrease in the output. For this it is not a simple resolution in SPSS and it will require some coding knowledge.  Please see the linked article to start with, I am not sure if they will necessarily get you all the way to your answer...but I find the more I work these out myself...well the more familiar I am with the software (and its internal runnings), which can never be a bad thing.

https://www.ibm.com/developerworks/community/forums/html/topic?id=77777777-0000-0000-0000-000014865389

If you can see in the attached plot, I need to basically invert the data along the y axis, so that 120 dB is at the bottom and 0 dB is at the top. It simply makes it better fit the format of a typical audiogram. I found that link as well, but haven't been able to incorporate the code. I may have to get a bit fitter at coding.

Perhaps yes, it might be simple also. A great source of information for SPSS is linked below. I assume you have this already, but if not please refer to it. It claims there is a inversion translation option (obviously), but also an option on the specified access to invert the data. Hope this helps.

http://www-01.ibm.com/support/knowledgecenter/SSLVMB_22.0.0/com.ibm.spss.statistics.help/vis_workbench/graphboard_editing_axes.htm?lang=en

Thanks so much for your suggestion! I had read that and was very excited to see it. Unfortunately, it uses the graphboard template system, which can apparently only work with one variable at a time, meaning I could only show one frequency on the graph. Maybe there is some way I haven't yet found to get it to use multiple variables.

You only have to add in the Syntax file "COORD: reflect (dim (2))" in a line before "END GPL."

Hi Antonia, thanks a lot for your idea. It works to a degree. I was able to invert the axis as you can see in the attached pic. However, I am unable to add the other frquencies to the chart. I can only add one frequency at a time. If I use the legacy chart builder, I can add all the frequencies, but I am somehow unable to get the coding correct for the GPL to flip the axis. I have attached the code for the the legacy builder, and then the code for the inverted y axis with only one freuqency. Any idea how to get the GPL code line into the code for the graph with all the frequencies?

Code for Graph with all frequency variables (dosn't seem to use GPL):

EXAMINE VARIABLES=Hz_125 Hz_250 Hz_500 Hz_1000 Hz_2000 Hz_4000 Hz_6000 Hz_8000 

/COMPARE VARIABLE

/PLOT=BOXPLOT

/STATISTICS=NONE

/NOTOTAL

/MISSING=LISTWISE

Code for the attached graph with the inverted axis, which can, however, only show one frequency.

DATASET ACTIVATE DataSet0.

* Chart Builder.

GGRAPH

/GRAPHDATASET NAME="graphdataset" VARIABLES=Hz_125 MISSING=LISTWISE REPORTMISSING=NO

/GRAPHSPEC SOURCE=INLINE.

BEGIN GPL

SOURCE: s=userSource(id("graphdataset"))

DATA: Hz_125=col(source(s), name("Hz_125"))

DATA: id=col(source(s), name("$CASENUM"), unit.category())

GUIDE: axis(dim(2), label("125 Hz"))

ELEMENT: schema(position(bin.quantile.letter(1*Hz_125)), label(id))

COORD: reflect (dim (2))

END GPL.

I am not sure, whrat your problem is. The different Hz are different variables in your data set? Then you have to change the structure first. I have the German version of SPSS, but the path should be data -> restructure -> ... At the end you should have one variable with the result (between 0 and 120) and another variable with the group (125 Hz, 250 Hz, ...). Then you can create the boxplot you are looking for (as I understood). Ich did it for 125, 250 and 500 Hz and with some supposed numbers. The Syntax for the boxplot is:

GGRAPH

/GRAPHDATASET NAME="graphdataset" VARIABLES=Index1 trans1 MISSING=LISTWISE REPORTMISSING=NO

/GRAPHSPEC SOURCE=INLINE.

BEGIN GPL

SOURCE: s=userSource(id("graphdataset"))

DATA: Index1=col(source(s), name("Index1"), unit.category())

DATA: trans1=col(source(s), name("trans1"))

DATA: id=col(source(s), name("$CASENUM"), unit.category())

GUIDE: axis(dim(1), label("Index1"))

GUIDE: axis(dim(2), label("trans1"))

SCALE: linear(dim(2), include(0))

ELEMENT: schema(position(bin.quantile.letter(Index1*trans1)), label(id))

COORD: reflect (dim (2))

END GPL.

The result is attached

Does this help you?

Best wishes,

Antonia

Hi Antonia, thanks so much. It did help and it works great now. It was necessary to move the data into a vertical format. It's a bit more work, as my data set is set up horizontally (excel), but it's not too much work to change it. You saved my life! Thank you!!!