is SPSS ignoring empty cells?

[TomZ]

Hi,
I was wondering if SPSS ignires empty cells. That is if it ignores
empty cells when making a regression or any compoutations
If not how can I make it so?
More information below.

Regards,
tomZ
Maastricht University

I have a table that is organised in the following way:

coutry ddm1 ddm2 ddm3
1 44 45 51
1 34 55 44
1 44 34
1 22 23
2 88 54 44
2 99 12 11
2 100 101 120
3
.
.
.
20

Now, I would like to make regressions and hostograms using that table.
Each colum represents a different model (valuation of property
companies).
I would like not to have to reorganise that table for each model.
Deleting some empty cells here and there would make the country
collumn not appcicable to the datat any more as not an equal amount of
data is available for each model.
Then the data in one collum would be shifted upwards when deleting
some cells and it would not fit the county index as the data of one
country would shift upwards to another.
I hope it is clear what I mean.

Using 9.0.1

[Rich Ulrich]

On Mon, 11 Dec 2000 14:01:38 GMT, ba...@junglebook.disney (TomZ) wrote:

> Hi,
> I was wondering if SPSS ignires empty cells. That is if it ignores
> empty cells when making a regression or any compoutations
> If not how can I make it so?
> More information below.

[ ... ]

> Then the data in one collum would be shifted upwards when deleting
> some cells and it would not fit the county index as the data of one
> country would shift upwards to another.
> I hope it is clear what I mean.
>
> Using 9.0.1

SPSS holds data as a "database."
If a "case" is left out of an analyses because of missing data,
that says the ROW (that you see in the display) is omitted
and the total N is reduced by 1.

There is never anything like "the data in one column"
being "shifted upwards", but I have trouble imagining what
convention someone is using that would give them that as
a model. (You can get in trouble with "free-formatting" when
you have a blank in the line, because a VARIABLE will use the
next value encountered. But that is a "shifting over" instead of
"shifting up.")

SO - I *think* that the person asking the question
should not be worried, but the question is so unusual
to my frame of reference that I could be missing it entirely.

--
Rich Ulrich, wpi...@pitt.edu
http://www.pitt.edu/~wpilib/index.html