Comparing columns -> Grouping

[Rijnsent]

Hi there,
I've built a Chi square test and a column proportion test in Excel and am now trying to build a grouping of the results. I've started off with this example:
https://www.ibm.com/support/knowled...ig_tests_colprop_ex.html#sig_tests_colprop_ex

The data:

ROW	A	B	C	D	E
full time	57.8%	15.5%	66.1%	62.4%	59.1%
part time	10.3%	7.1%	7.8%	9.7%	15.4%
temp not	1.7%	0.7%	2.0%	1.1%	1.7%
unempl	1.0%	1.1%	2.2%	0.0%	4.8%
retired	12.5%	53.0%	11.9%	6.5%	2.6%
school	0.7%	0.4%	1.6%	2.2%	9.0%
house	14.9%	19.4%	5.6%	14.0%	5.3%
other	1.2%	2.8%	2.7%	4.3%	2.1%
	100.0%	100.0%	100.0%	100.0%	100.0%
NR_TOT	1345	283	446	93	663

From that I can calculate the significantly different columns, to get to the first result (as a table). I get this by a massive table to calculate for every pair (e.g. full time - A vs B (57.8% vs 15.5%)) whether the difference is significantly bigger. I concatenate the results of that table, so I get (which matches the second table of the table in the link):

ROW	A	B	C	D	E
full time	B		AB	B	B
part time					ABC
temp not
unempl					AB
retired	E	ACDE	E
school					ABC
house	CE	CE		CE
other

For the (first) full time row: the value AB implies that C is significantly bigger than both A and B. Also, the B in column A implies that A is significantly bigger than B. From that, you can basically make a couple of groups that are internally not significanlty different: Group "B" (smallest) - group "ADE" (bigger) - group "CDE" (biggest). This is shown in the last table (small letters from the SPSS link), but how do I calculate the individual cell values? My sample file is here (desired outcome starts in Q105) https://www.dropbox.com/s/ey0ivf2owqpzqns/chi_bonf_col_tests.xlsx?dl=0:

ROW	A	B	C	D	E
full time	a	b	c	ac	ac
part time	a	a	a	ab	b
temp not	a	a	a	a	a
unempl	a	a	ab		b
retired	a	b	a	ac	c
school	a	a	a	ab	a
house	a	a	b	a	b
other	a	a	a	a	a

Thanks in advance for any pointers!