help with testing whether there is significance difference between the proportions or not

I realy need help. Let say I have data as below, where each row denotes a company(case). I want to test whether there is significance difference between the proportions of gender or not according to characteristic of companies. How can I test this with SPSS? Which test should I use and how? If you help me, I will be very glad. Thanks a lot. company "characteristic" "no of female" "no of male" -------------------------------------------------------------------------------------------------- 1 a 25 35 2 a 32 54 3 b 25 45 4 b 36 52 5 b 78 65 6 b 45 62 7 c 15 18 8 c 78 92

On 21/02/2011 6:14 PM, oercim wrote: > I realy need help. Let say I have data as below, where each row > denotes a company(case). > I want to test whether there is significance difference between the > proportions of gender or not according to characteristic of > companies. > How can I test this with SPSS? Which test should I use and how? If you > help me, I will be very glad. Thanks a lot. > > company "characteristic" "no of female" "no of > male" > -------------------------------------------------------------------------------------------------- > 1 a > 25 35 > 2 a > 32 54 > 3 b > 25 45 > 4 b > 36 52 > 5 b > 78 65 > 6 b > 45 62 > 7 c > 15 18 > 8 c > 78 92 > > Line-wrap is making your table hard to read. I think you meant it to look like this (use fixed font like Courier to make it line up): Co Char F M ---------------- 1 a 25 35 2 a 32 54 3 b 25 45 4 b 36 52 5 b 78 65 6 b 45 62 7 c 15 18 8 c 78 92 I *think* that what you you're asking is how to crosstabulate Characteristic and Sex. If so, you would need to have two rows per Company*, with variables indicating Characteristic, Sex, and the count for that combination. Like this: Co Char Sex Kount 1 1 1 25 1 1 2 35 2 1 1 32 2 1 2 54 etc 8 3 1 78 8 3 2 82 (Notice that I used 1-3 for the characteristic codes rather than a-c. Generally, it's better to use numeric variables where possible--this gives you more flexibility in how you can use the variable.) Then use the WEIGHT command to "weight by kount" before running the Crosstab. weight by Kount. crosstabs char by sex / stat = chisq. * Actually, you would need two rows for each Characteristic; but then you would need the counts of males and females across all companies with that characteristic. If this is not what you wanted, please clarify your question. -- Bruce Weaver bweaver@lakeheadu.ca http://sites.google.com/a/lakeheadu.ca/bweaver/Home "When all else fails, RTFM."

On Feb 22, 1:41=A0am, Bruce Weaver <bwea...@lakeheadu.ca> wrote: > On 21/02/2011 6:14 PM, oercim wrote: > > > > > > > > > > > I realy need help. Let say I have data as below, =A0where each row > > denotes a company(case). > > I want to test whether there is significance difference between the > > proportions of gender or not according to characteristic of > > companies. > > How can I test this with SPSS? Which test should I use and how? If you > > help me, I will be very glad. Thanks a lot. > > > company =A0 =A0 =A0 =A0"characteristic" =A0 =A0 =A0"no of female" =A0 = =A0 =A0 =A0 "no of > > male" > > -----------------------------------------------------------------------= ---- ----------------------- > > 1 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0a > > 25 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A035 > > 2 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0a > > 32 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A054 > > 3 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0b > > 25 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A045 > > 4 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0b > > 36 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A052 > > 5 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0b > > 78 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A065 > > 6 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0b > > 45 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A062 > > 7 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0c > > 15 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A018 > > 8 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0c > > 78 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A092 > > Line-wrap is making your table hard to read. =A0I think you meant it to > look like this (use fixed font like Courier to make it line up): > > Co Char F =A0 =A0M > ---------------- > 1 =A0 a =A0 25 =A0 35 > 2 =A0 a =A0 32 =A0 54 > 3 =A0 b =A0 25 =A0 45 > 4 =A0 b =A0 36 =A0 52 > 5 =A0 b =A0 78 =A0 65 > 6 =A0 b =A0 45 =A0 62 > 7 =A0 c =A0 15 =A0 18 > 8 =A0 c =A0 78 =A0 92 > > I *think* that what you you're asking is how to crosstabulate > Characteristic and Sex. =A0If so, you would need to have two rows per > Company*, with variables indicating Characteristic, Sex, and the count > for that combination. =A0Like this: > > Co Char Sex Kount > 1 =A0 1 =A0 =A01 =A0 25 > 1 =A0 1 =A0 =A02 =A0 35 > 2 =A0 1 =A0 =A01 =A0 32 > 2 =A0 1 =A0 =A02 =A0 54 > etc > 8 =A0 3 =A0 =A01 =A0 78 > 8 =A0 3 =A0 =A02 =A0 82 > > (Notice that I used 1-3 for the characteristic codes rather than a-c. > Generally, it's better to use numeric variables where possible--this > gives you more flexibility in how you can use the variable.) > > Then use the WEIGHT command to "weight by kount" before running the > Crosstab. > > weight by Kount. > crosstabs char by sex / stat =3D chisq. > > * Actually, you would need two rows for each Characteristic; but then > you would need the counts of males and females across all companies with > that characteristic. > > If this is not what you wanted, please clarify your question. > > -- > Bruce Weaver > bwea...@lakeheadu.cahttp://sites.google.com/a/lakeheadu.ca/bweaver/Home > "When all else fails, RTFM." Yes, that is what I want. But I have the data inputed. And it is a large data set(hundreds of cases-I won't do that just for gender, also for education level, etc). So it is very difficult to make two rows manually for each chrasteristic. Is there a short way to make prepare this data? Thanks a lot .

On Feb 22, 1:54=A0am, oercim <oer...@yahoo.com> wrote: > On Feb 22, 1:41=A0am, Bruce Weaver <bwea...@lakeheadu.ca> wrote: > > > > > > > > > > > On 21/02/2011 6:14 PM, oercim wrote: > > > > I realy need help. Let say I have data as below, =A0where each row > > > denotes a company(case). > > > I want to test whether there is significance difference between the > > > proportions of gender or not according to characteristic of > > > companies. > > > How can I test this with SPSS? Which test should I use and how? If yo= u > > > help me, I will be very glad. Thanks a lot. > > > > company =A0 =A0 =A0 =A0"characteristic" =A0 =A0 =A0"no of female" =A0= =A0 =A0 =A0 "no of > > > male" > > > ---------------------------------------------------------------------= ------ ----------------------- > > > 1 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0a > > > 25 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A035 > > > 2 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0a > > > 32 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A054 > > > 3 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0b > > > 25 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A045 > > > 4 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0b > > > 36 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A052 > > > 5 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0b > > > 78 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A065 > > > 6 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0b > > > 45 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A062 > > > 7 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0c > > > 15 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A018 > > > 8 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0c > > > 78 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A092 > > > Line-wrap is making your table hard to read. =A0I think you meant it to > > look like this (use fixed font like Courier to make it line up): > > > Co Char F =A0 =A0M > > ---------------- > > 1 =A0 a =A0 25 =A0 35 > > 2 =A0 a =A0 32 =A0 54 > > 3 =A0 b =A0 25 =A0 45 > > 4 =A0 b =A0 36 =A0 52 > > 5 =A0 b =A0 78 =A0 65 > > 6 =A0 b =A0 45 =A0 62 > > 7 =A0 c =A0 15 =A0 18 > > 8 =A0 c =A0 78 =A0 92 > > > I *think* that what you you're asking is how to crosstabulate > > Characteristic and Sex. =A0If so, you would need to have two rows per > > Company*, with variables indicating Characteristic, Sex, and the count > > for that combination. =A0Like this: > > > Co Char Sex Kount > > 1 =A0 1 =A0 =A01 =A0 25 > > 1 =A0 1 =A0 =A02 =A0 35 > > 2 =A0 1 =A0 =A01 =A0 32 > > 2 =A0 1 =A0 =A02 =A0 54 > > etc > > 8 =A0 3 =A0 =A01 =A0 78 > > 8 =A0 3 =A0 =A02 =A0 82 > > > (Notice that I used 1-3 for the characteristic codes rather than a-c. > > Generally, it's better to use numeric variables where possible--this > > gives you more flexibility in how you can use the variable.) > > > Then use the WEIGHT command to "weight by kount" before running the > > Crosstab. > > > weight by Kount. > > crosstabs char by sex / stat =3D chisq. > > > * Actually, you would need two rows for each Characteristic; but then > > you would need the counts of males and females across all companies wit= h > > that characteristic. > > > If this is not what you wanted, please clarify your question. > > > -- > > Bruce Weaver > > bwea...@lakeheadu.cahttp://sites.google.com/a/lakeheadu.ca/bweaver/Home > > "When all else fails, RTFM." > > Yes, that is what I want. But I have the data inputed. And it is a > large data set(hundreds of cases-I won't do that just for gender, also > for education level, etc). So it is very difficult to make two rows > manually for each chrasteristic. Is there a short way to make prepare > this data? Thanks a lot . Thanks a lot. I found the solution. Restructure is done my work. Again thanks alot. It was very useful for me

Similiar Articles:

Which Assembler? - comp.lang.asm.x86
I still have not received a definitive answer as to whether Rosasm ... Ultimately yes, but there is a large and significant relative difference between ... All, I need some help ...

How best to detect duplicate values in a column? - comp.databases ...
... as to the meaning ... to know whether there is or is not ... for help - you were looking for free consulting. There is a huge difference. The former I provide. But not the ...

mixed models with repeated measurements - comp.soft-sys.stat.spss ...
... to find out whether CS is nested in AR1. I think it is not, i ... model to test if there is a statistically significant linear relationship between ... lot for your help ...

Packet timestamps when using Windows-7/Vista - comp.protocols.time ...
However, I'm not sure whether this routine is in effect ... is in user space here, so > there may not only be ... The major difference between the two systems is that one ...

1-wire to USB converter that can use 1820s directly - comp.home ...
> A significant proportion of folks wiring ... > For software on the PC there is absolutely no difference between > a ... softdev/owdocs/Docs/1-Wire_SDK_Help.html Whether or not ...

Freescale's Idea of Open Source JTAG - comp.arch.embedded ...
I'm not sure if there is any significance to that or not. ... guess > > there is some sort of difference there ... gcc toolchain support. =A0I am not sure whether ...

Sampling: What Nyquist Didn't Say, and What to Do About It - comp ...
I know there's a few ... challenge, but not impossible. Whether I'm ... Just the difference between a paperback and a "hard-bound" tome alone is significant (though I ...

code speed moving from fortran 77 compiler to f2003 compiler ...
... of the code might help this process even more. There are ... long term, I don't know whether converting to C/C++ is better or not ... Without it, the meaning of the claim ...

MC: 4th power with no small digits - comp.sys.hp48
The difference for me lies in whether or not we are supposed to (a) do a ... I created this little program to help me ... Note that the key thing about whether there are a ...

Need a FORTRAN compiler for Win7 (or XP) - comp.lang.fortran ...
I'm not commenting on whether it is likely to happen ... he's been through a fairly significant one w/ BMDP suite not ... Also in TB, there's a difference between "read next" and ...

Hypothesis Test: Difference Between Proportions
Help ... whether the difference between two proportions is significant. ... that there is no difference between the two population proportions ...

T-test online. Compare two means, two proportions or counts online.
... significant’, the difference is not caused by chance. The t-test is basically not valid for testing the difference between two proportions. ... testing. It is there ... Help ...

7/23/2012 6:04:20 AM