Not the most rigorous algorithm to obtain pooled standard erors, but ..

POOLED STANDARD ERRORS:

=SQRT(A1^2+A2^2)

where A1 and A2 are Excel cells that contain two standard errors.

 

Z will be:

Difference in means / pooled standard errors

 

Probability will be:

=1-NORMSDIST(O4)

where O4 here should reference to the Excel cell that contains Z.

 

 

=IF(ABS($A1)>=3.29,"***",IF(ABS(A1)>=2.57,"**", IF(ABS(A1)>=1.96, "*", IF(ABS(A1)>=1.645,"~", IF(ABS(A1)<1.645," ")))))

 

~ p <.10  * p <.05          ** p <.01        *** p <.001

 

Showing significant levels with *s based on P-value (1 tail or 2 tail does not matter)

 

=IF(ABS($A1)<0.001,"***",IF(ABS($A1)<0.01,"**",IF(ABS($A1)<0.05,"*",IF(ABS($A1)<0.1,"~"))))

 

Effect Size Calculation (Jump to Centre for Evaluation & Monitoring)

http://www.cemcentre.org/renderpage.asp?linkID=30325017

I used their Excel sheet to get a pooled standard deviation.  I modified their table to show you how it can be done.  CLICK HERE for the actual excel sheet.

Group A			Group B			pooled standard deviation
mean	n	SD	mean	n	SD
1.6	54	0.72	1.6	56	0.66	=SQRT((C3^2*(B3-1)+F3^2*(E3-1))/(B3+E3-2))

To go from percentages to logit (A32 is a hypothetical Excel cell):

=LN(A32/(1-A32))

 

To go from logit to percentage:

=EXP(B28)/(1+EXP(B28))

 

Count the occurance of certain values

=countif(b1:b5, "kaz")