how to
plaatje
if one or more of the first eight fields can predict the 9th.
how can spss 16 help me? what menu options i have to choose to import
the data and spss (16) can analyse the data?
thanks
anton
Mark Ebermann
What are those fields?
plaatje
> On Wed, 14 May 2008 09:01:25 -0700 (PDT), plaatje
>
> >i have gained some data. one row has 9 fields/koloms. i like to know
> >if one or more of the first eight fields can predict the 9th.
>
> >how can spss 16 help me? what menu options i have to choose to import
> >the data and spss (16) can analyse the data?
>
> >thanks
> >anton
>
> What are those fields?
thanks for responding.
the fields contains data about people such as age, smooking, type of
car
anton
Mark Ebermann
The most important thing we need to know is the nature of the 9th
field you're trying to predict. Is it a scale variable, ordinal or
nominal?
plaatje
>
> - Tekst uit oorspronkelijk bericht weergeven -
i find out what that (= scale variable, ordinal or
nominal? ) so i can answer your question.
Bruce Weaver
> On 15 mei, 11:33, Mark Ebermann <cfisc...@aachener.de> wrote:
>
> > The most important thing we need to know is the nature of the 9th
> > field you're trying to predict. Is it a scale variable, ordinal or
> > nominal?- Tekst uit oorspronkelijk bericht niet weergeven -
>
>
> nominal? ) so i can answer your question.
Scaled variable = quantitative variable (e.g., height, weight, heart
rate, any kind of count)
Ordinal = ordered categories (e.g., stage of cancer)
Nominal = categories without any inherent order (e.g., type of beer,
brand of running shoe)
What is the 9th variable in your data set?
--
Bruce Weaver
bwe...@lakeheadu.ca
www.angelfire.com/wv/bwhomedir
"When all else fails, RTFM."
plaatje
plaatje
> "When all else fails, RTFM."
Because i am not sure. more info.
My data looks like this:
24 201016 3 1 4 4 10 9 9 323 122 122 183 12
57 301017 3 3 4 4 6 9 8 323 122 122 183 -18
the first colom (with 24 and 57) is age. the last colom (12 and -18) a
value. the data between those are a code for a specifik elements of
that person or his situation. So the are 12 codes. Each code is unique
in that colom.
i hope this helps to answer my question.
anton
Bruce Weaver
> On 16 mei, 21:04, Bruce Weaver <bwea...@lakeheadu.ca> wrote:
>
>
>
> > On May 16, 1:03 pm, plaatje <spssvr...@fastmail.fm> wrote:
>
> > > On 15 mei, 11:33, Mark Ebermann <cfisc...@aachener.de> wrote:
>
> > > > The most important thing we need to know is the nature of the 9th
> > > > field you're trying to predict. Is it a scale variable, ordinal or
> > > > nominal?- Tekst uit oorspronkelijk bericht niet weergeven -
>
> > > i find out what that (= scale variable, ordinal or
> > > nominal? ) so i can answer your question.
>
> > Scaled variable = quantitative variable (e.g., height, weight, heart
> > rate, any kind of count)
> > Ordinal = ordered categories (e.g., stage of cancer)
> > Nominal = categories without any inherent order (e.g., type of beer,
> > brand of running shoe)
>
> > What is the 9th variable in your data set?
>
>
> My data looks like this:
>
24 201016 3 1 4 4 10 9 9 323 122 122 183 12
57 301017 3 3 4 4 6 9 8 323 122 122 183 -18
> the first colom (with 24 and 57) is age. the last colom (12 and -18) a
> value. the data between those are a code for a specifik elements of
> that person or his situation. So the are 12 codes. Each code is unique
> in that colom.
> i hope this helps to answer my question.
> anton
You say the last column is a "value". That suggests it is a
quantitative (numeric) variable. So you're probably looking at a
linear regression model. But in order to carry out the analysis
properly, you have to know what the numbers in the other columns
represent. E.g., in the 4th column, is it a numeric variable, or do 1
and 3 stand for two different categories of a qualitative variable
(e.g., 1 = Nike, 2 = Adidas, 3 = New Balance, etc)?
By the way, the second column looks like an ID number to me. If it
is, it would not enter into the analysis (assuming there is only one
row per ID number).
--
Bruce Weaver
bwe...@lakeheadu.ca
www.angelfire.com/wv/bwhomedir
plaatje
> "When all else fails, RTFM."- Tekst uit oorspronkelijk bericht niet weergeven -
>
> - Tekst uit oorspronkelijk bericht weergeven -
thanks for the explaintion and your example.
colom 2 is nu id cod. it is a code with a kind of system within the
code
colom 2 -13 are code such as (e.g., 1 = Nike, 2 = Adidas, 3 = New
Balance, etc)?
i hope to find out if colom 1 -13 (or some combination) predict
(influence) the value in colom 14
thanks
Bruce Weaver
If the explanatory variables are all codes, the easiest way for you to
proceed would be treating them as fixed "factors" (rather than
"covariates") in GLM-UNIANOVA. In the pull-down menus, it's Analyze --
> GLM --> Univariate. I suggest that you also find someone local to
advise you.
--
Bruce Weaver
bwe...@lakeheadu.ca
www.angelfire.com/wv/bwhomedir
plaatje
found. Because I am unemployed this is once more extra troublesome. I
have nothing to offer for any help (unless this becomes a succes). In
the meantime I continued, but that does not turn out better. If I
attempt all variable together, I get an error message. I have done now
two efforts, but cannot judge the output. I mean, how to read such a
table. What mean the F and Sig values? What do Parial Eta squared
mean? How important is the error value in the column mean square?
This is the output of my 2 attempts
UNIANOVA
DependendOne BY F1 F2 F3
/METHOD = SSTYPE(3)
/INTERCEPT = INCLUDE
/PLOT = PROFILE( F1*F2*F3 )
/PRINT = DESCRIPTIVE ETASQ
/CRITERIA = ALPHA(.05)
/DESIGN = F1 F2 F3 F1
*F2 F1*F3 F2
*F3 F1*F2*F3 .
Tests of Between-Subjects Effects
Dependent Variable: DependendOne
Source Type III Sum df Mean F Sig. Partial Eta
of Squares Suare Sqyared
CM 64918,157(a) 449 144,584 1,909 0,000 0,216
Int.. 338,461 1 338,461 4,469 0,035 0,001
F1 5856,613 27 216,912 2,864 0,000 0,024
F2 4332,569 11 393,87 5,201 0,000 0,018
F3 2305,002 11 209,546 2,767 0,001 0,01
F1*F2 40573,407 212 191,384 2,527 0,000 0,147
F1*F3 2239,358 11 203,578 2,688 0,002 0,009
F2*F3 11223,247 119 94,313 1,245 0,039 0,046
F1*F2
* F3 7509,408 54 139,063 1,836 0,000 0,031
Error 235078,513 3104 75,734
Total 299999 3554
C. Total 299996,67 3553
a R Squared = ,216 (Adjusted R Squared = ,103)
UNIANOVA
DependendOne BY F1 F2 F4
/METHOD = SSTYPE(3)
/INTERCEPT = INCLUDE
/PRINT = DESCRIPTIVE ETASQ
/CRITERIA = ALPHA(.05)
/DESIGN = F1 F2 F4 F1*F2
F1*F4 F2*F4 F1*F2
*F4 .
Tests of Between-Subjects Effects
Dependent Variable: DependendOne
Source Type III Sum df Mean F Sig. Partial Eta
of Squares Suare Sqyared
CM 185033,196(a) 1744 106,097 1,669 0,000 0,617
Intercept 262,577 1 262,577 4,132 0,042 0,002
F1 4897,379 27 181,384 2,854 0,000 0,041
F2 4737,429 11 430,675 6,777 0,000 0,04
F4 541,165 11 49,197 0,774 0,666 0,005
F1*F2 36445,238 205 177,782 2,797 0,000 0,241
F1*F4 23382,42 213 109,777 1,727 0,000 0,169
F2*F4 10214,992 121 84,421 1,328 0,012 0,082
F1*F2
* F4 107041,324 1145 93,486 1,471 0,000 0,482
Error 114963,474 1809 63,551
Total 299999 3554
C. Total 299996,67 3553
a R Squared = ,617 (Adjusted R Squared = ,247)
I hope the output is correct on your screen. For anyone wants to help
i i can send the output buy email. please send your email to
spssvraag@{NOMAIL]fastmail.fm (but ofcourse without the [NOAMAIL].
thanks
Anton