Digits of accuracy in Minimum and Maximum

[Bob Schacht-3]

I want to segment "Age at application" into age groups.
I define "AgeGroup = 1" if Age at Application < 22, and other age groups
similarly.
Afterwards, to check, I do a "Compare Means" with Age at Application as the
dependent variable, and AgeGroup as the independent variable, asking for
the minimum and maximum of each group.

I am then informed that the "maximum" of Age Group 1 is "22.00", which
should not be correct because I specified Age Group = 1 if Age at
Application is LESS THAN 22.

I then ran FREQUENCIES on Age group 1, which only provides 2 places of
accuracy as well, and find that the "maximum" = 22.00, there being one
example of this value.

It occurs to me that the maximum age of that group might be, say,
21.999999, which would round off to 22.00 if only two places of accuracy
are specified.

How can I find out the *actual* (as opposed to rounded) maximum value of
this group?
Is there a way to specify the number of places of accuracy in the maximum
and minimum reports from these outputs?

BTW, I tried this with SPSS ver. 16 as well as SPSS ver. 12, with the same
results.

Bob Schacht

Robert M. Schacht, Ph.D. <[hidden email]>
Pacific Basin Rehabilitation Research & Training Center
1268 Young Street, Suite #204
Research Center, University of Hawaii
Honolulu, HI 96814

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Garry Gelade]

Bob, if you right click on the table you will get the option to open the
table in Pivot Table view. In Pivot table view you can double click on the
cell and see the unrounded value. In Pivot Table view You can also click on
a row or column data label --> Select Data Cells--> Cell Properties  and
from there set the number of decimal places for the row or column.

Garry Gealde

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Bob
Schacht
Sent: 17 September 2008 22:49
To: [hidden email]
Subject: Digits of accuracy in Minimum and Maximum

I want to segment "Age at application" into age groups.
I define "AgeGroup = 1" if Age at Application < 22, and other age groups
similarly.
Afterwards, to check, I do a "Compare Means" with Age at Application as the
dependent variable, and AgeGroup as the independent variable, asking for the
minimum and maximum of each group.

I am then informed that the "maximum" of Age Group 1 is "22.00", which
should not be correct because I specified Age Group = 1 if Age at
Application is LESS THAN 22.

I then ran FREQUENCIES on Age group 1, which only provides 2 places of
accuracy as well, and find that the "maximum" = 22.00, there being one
example of this value.

It occurs to me that the maximum age of that group might be, say, 21.999999,
which would round off to 22.00 if only two places of accuracy are specified.

How can I find out the *actual* (as opposed to rounded) maximum value of
this group?
Is there a way to specify the number of places of accuracy in the maximum
and minimum reports from these outputs?

BTW, I tried this with SPSS ver. 16 as well as SPSS ver. 12, with the same
results.

Bob Schacht

Robert M. Schacht, Ph.D. <[hidden email]> Pacific Basin Rehabilitation
Research & Training Center
1268 Young Street, Suite #204
Research Center, University of Hawaii
Honolulu, HI 96814

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list of
commands to manage subscriptions, send the command INFO REFCARD

__________ NOD32 3449 (20080917) Information __________

This message was checked by NOD32 antivirus system.
http://www.eset.com

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Muir Houston-2]

I think it also provides you with the unrounded value if you simply
double click on the value in question - without opening a pivot table?

Muir

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Garry Gelade
Sent: 18 September 2008 11:03
To: [hidden email]
Subject: Re: Digits of accuracy in Minimum and Maximum

Bob, if you right click on the table you will get the option to open the
table in Pivot Table view. In Pivot table view you can double click on
the cell and see the unrounded value. In Pivot Table view You can also
click on a row or column data label --> Select Data Cells--> Cell
Properties  and from there set the number of decimal places for the row
or column.

Garry Gealde

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Bob Schacht
Sent: 17 September 2008 22:49
To: [hidden email]
Subject: Digits of accuracy in Minimum and Maximum

I want to segment "Age at application" into age groups.
I define "AgeGroup = 1" if Age at Application < 22, and other age groups
similarly.
Afterwards, to check, I do a "Compare Means" with Age at Application as
the dependent variable, and AgeGroup as the independent variable, asking
for the minimum and maximum of each group.

I am then informed that the "maximum" of Age Group 1 is "22.00", which
should not be correct because I specified Age Group = 1 if Age at
Application is LESS THAN 22.

I then ran FREQUENCIES on Age group 1, which only provides 2 places of
accuracy as well, and find that the "maximum" = 22.00, there being one
example of this value.

It occurs to me that the maximum age of that group might be, say,
21.999999, which would round off to 22.00 if only two places of accuracy
are specified.

How can I find out the *actual* (as opposed to rounded) maximum value of
this group?
Is there a way to specify the number of places of accuracy in the
maximum and minimum reports from these outputs?

BTW, I tried this with SPSS ver. 16 as well as SPSS ver. 12, with the
same results.

Bob Schacht

Robert M. Schacht, Ph.D. <[hidden email]> Pacific Basin
Rehabilitation Research & Training Center
1268 Young Street, Suite #204
Research Center, University of Hawaii
Honolulu, HI 96814

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list
of commands to manage subscriptions, send the command INFO REFCARD

__________ NOD32 3449 (20080917) Information __________

This message was checked by NOD32 antivirus system.
http://www.eset.com

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list
of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Peck, Jon]

All pivot table cell displays are rounded using a format that depends on the type of statistic and the variables it stems from.  The full precision value is always available as discussed below.  Of course you can override a cell format in the pivot table editor.

For statistics whose values lie in the domain of the variable, such as min, max, and mean, the default format is generally based on the format of the underlying variable although it may add or subtract extra decimal places.  So to get more precision in a frequency table or a descriptives table, you can just increase the number of decimals in the variable's format in the Data Editor Variable View or with the FORMAT command.  Or you could change the format to scientific notation.

Since computed values or variables with fractional values not exactly representable in standard floating point can easily arise, a value may well have more nonzero decimals than you might expect.

HTH,
Jon Peck

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Muir Houston
Sent: Thursday, September 18, 2008 4:18 AM
To: [hidden email]
Subject: Re: [SPSSX-L] Digits of accuracy in Minimum and Maximum

I think it also provides you with the unrounded value if you simply
double click on the value in question - without opening a pivot table?

Muir

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Garry Gelade
Sent: 18 September 2008 11:03
To: [hidden email]
Subject: Re: Digits of accuracy in Minimum and Maximum

Bob, if you right click on the table you will get the option to open the
table in Pivot Table view. In Pivot table view you can double click on
the cell and see the unrounded value. In Pivot Table view You can also
click on a row or column data label --> Select Data Cells--> Cell
Properties  and from there set the number of decimal places for the row
or column.

Garry Gealde

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Bob Schacht
Sent: 17 September 2008 22:49
To: [hidden email]
Subject: Digits of accuracy in Minimum and Maximum

I want to segment "Age at application" into age groups.
I define "AgeGroup = 1" if Age at Application < 22, and other age groups
similarly.
Afterwards, to check, I do a "Compare Means" with Age at Application as
the dependent variable, and AgeGroup as the independent variable, asking
for the minimum and maximum of each group.

I am then informed that the "maximum" of Age Group 1 is "22.00", which
should not be correct because I specified Age Group = 1 if Age at
Application is LESS THAN 22.

I then ran FREQUENCIES on Age group 1, which only provides 2 places of
accuracy as well, and find that the "maximum" = 22.00, there being one
example of this value.

It occurs to me that the maximum age of that group might be, say,
21.999999, which would round off to 22.00 if only two places of accuracy
are specified.

How can I find out the *actual* (as opposed to rounded) maximum value of
this group?
Is there a way to specify the number of places of accuracy in the
maximum and minimum reports from these outputs?

BTW, I tried this with SPSS ver. 16 as well as SPSS ver. 12, with the
same results.

Bob Schacht

Robert M. Schacht, Ph.D. <[hidden email]> Pacific Basin
Rehabilitation Research & Training Center
1268 Young Street, Suite #204
Research Center, University of Hawaii
Honolulu, HI 96814

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list
of commands to manage subscriptions, send the command INFO REFCARD

__________ NOD32 3449 (20080917) Information __________

This message was checked by NOD32 antivirus system.
http://www.eset.com

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list
of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[ViAnn Beadle]

Double-clicking opens the pivot table.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Muir Houston
Sent: Thursday, September 18, 2008 4:18 AM
To: [hidden email]
Subject: Re: Digits of accuracy in Minimum and Maximum

I think it also provides you with the unrounded value if you simply
double click on the value in question - without opening a pivot table?

Muir

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Garry Gelade
Sent: 18 September 2008 11:03
To: [hidden email]
Subject: Re: Digits of accuracy in Minimum and Maximum

Bob, if you right click on the table you will get the option to open the
table in Pivot Table view. In Pivot table view you can double click on
the cell and see the unrounded value. In Pivot Table view You can also
click on a row or column data label --> Select Data Cells--> Cell
Properties  and from there set the number of decimal places for the row
or column.

Garry Gealde

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Bob Schacht
Sent: 17 September 2008 22:49
To: [hidden email]
Subject: Digits of accuracy in Minimum and Maximum

I want to segment "Age at application" into age groups.
I define "AgeGroup = 1" if Age at Application < 22, and other age groups
similarly.
Afterwards, to check, I do a "Compare Means" with Age at Application as
the dependent variable, and AgeGroup as the independent variable, asking
for the minimum and maximum of each group.

I am then informed that the "maximum" of Age Group 1 is "22.00", which
should not be correct because I specified Age Group = 1 if Age at
Application is LESS THAN 22.

I then ran FREQUENCIES on Age group 1, which only provides 2 places of
accuracy as well, and find that the "maximum" = 22.00, there being one
example of this value.

It occurs to me that the maximum age of that group might be, say,
21.999999, which would round off to 22.00 if only two places of accuracy
are specified.

How can I find out the *actual* (as opposed to rounded) maximum value of
this group?
Is there a way to specify the number of places of accuracy in the
maximum and minimum reports from these outputs?

BTW, I tried this with SPSS ver. 16 as well as SPSS ver. 12, with the
same results.

Bob Schacht

Robert M. Schacht, Ph.D. <[hidden email]> Pacific Basin
Rehabilitation Research & Training Center
1268 Young Street, Suite #204
Research Center, University of Hawaii
Honolulu, HI 96814

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list
of commands to manage subscriptions, send the command INFO REFCARD

__________ NOD32 3449 (20080917) Information __________

This message was checked by NOD32 antivirus system.
http://www.eset.com

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list
of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Bob Schacht-3]

At 03:46 AM 9/18/2008, ViAnn Beadle wrote:
>Double-clicking opens the pivot table.

Thanks to one and all who pointed this feature out to me. Works like a charm!

Bob Schacht

>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
>Muir Houston
>Sent: Thursday, September 18, 2008 4:18 AM
>To: [hidden email]
>Subject: Re: Digits of accuracy in Minimum and Maximum
>
>I think it also provides you with the unrounded value if you simply
>double click on the value in question - without opening a pivot table?
>
>Muir
>
>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
>Garry Gelade
>Sent: 18 September 2008 11:03
>To: [hidden email]
>Subject: Re: Digits of accuracy in Minimum and Maximum
>
>Bob, if you right click on the table you will get the option to open the
>table in Pivot Table view. In Pivot table view you can double click on
>the cell and see the unrounded value. In Pivot Table view You can also
>click on a row or column data label --> Select Data Cells--> Cell
>Properties  and from there set the number of decimal places for the row
>or column.
>
>Garry Gealde
>
>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
>Bob Schacht
>Sent: 17 September 2008 22:49
>To: [hidden email]
>Subject: Digits of accuracy in Minimum and Maximum
>
>I want to segment "Age at application" into age groups.
>I define "AgeGroup = 1" if Age at Application < 22, and other age groups
>similarly.
>Afterwards, to check, I do a "Compare Means" with Age at Application as
>the dependent variable, and AgeGroup as the independent variable, asking
>for the minimum and maximum of each group.
>
>I am then informed that the "maximum" of Age Group 1 is "22.00", which
>should not be correct because I specified Age Group = 1 if Age at
>Application is LESS THAN 22.
>
>I then ran FREQUENCIES on Age group 1, which only provides 2 places of
>accuracy as well, and find that the "maximum" = 22.00, there being one
>example of this value.
>
>It occurs to me that the maximum age of that group might be, say,
>21.999999, which would round off to 22.00 if only two places of accuracy
>are specified.
>
>How can I find out the *actual* (as opposed to rounded) maximum value of
>this group?
>Is there a way to specify the number of places of accuracy in the
>maximum and minimum reports from these outputs?
>
>BTW, I tried this with SPSS ver. 16 as well as SPSS ver. 12, with the
>same results.
>
>Bob Schacht
>
>Robert M. Schacht, Ph.D. <[hidden email]> Pacific Basin
>Rehabilitation Research & Training Center
>1268 Young Street, Suite #204
>Research Center, University of Hawaii
>Honolulu, HI 96814
>
>=====================
>To manage your subscription to SPSSX-L, send a message to
>[hidden email] (not to SPSSX-L), with no body text except the
>command. To leave the list, send the command SIGNOFF SPSSX-L For a list
>of commands to manage subscriptions, send the command INFO REFCARD
>
>__________ NOD32 3449 (20080917) Information __________
>
>This message was checked by NOD32 antivirus system.
>http://www.eset.com
>
>=====================
>To manage your subscription to SPSSX-L, send a message to
>[hidden email] (not to SPSSX-L), with no body text except the
>command. To leave the list, send the command SIGNOFF SPSSX-L For a list
>of commands to manage subscriptions, send the command INFO REFCARD
>
>=====================
>To manage your subscription to SPSSX-L, send a message to
>[hidden email] (not to SPSSX-L), with no body text except the
>command. To leave the list, send the command
>SIGNOFF SPSSX-L
>For a list of commands to manage subscriptions, send the command
>INFO REFCARD
>
>=====================
>To manage your subscription to SPSSX-L, send a message to
>[hidden email] (not to SPSSX-L), with no body text except the
>command. To leave the list, send the command
>SIGNOFF SPSSX-L
>For a list of commands to manage subscriptions, send the command
>INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Maguin, Eugene]

All,

I've just had 16 installed and I patched it to 16.0.2, but have not added
the hotfix because people here don't seem to be absolutely sure whether it
is the academic version. I just used it to do some litte tasks and several
odd--meaning different from 15--things have shown up and I wonder if others
have seen this or if this could be due to idiosyncracies of our network
system.

1) The default syntax font was some unknown to me monospaced font that I
changed to new courier, which have used consistently in the past. I now
notice that odd 'square boxes' display in the output view between words.
Instead of a blank space there is this square box. These boxes do not
print--fortunately. Changing back to the 'monotype' font does not correct
the problem. Anybody else seen this?

2) Printing. On 15 the delay between pressing the print button and having
the print box come up was maybe 5 seconds. I print to a network printer (HP
LJ4200/4300 series printer). Now the delay is more like 10-15 seconds. I
thought that maybe there were network changes that would account for this
but our network person says No. Does this happen to anyone else?

3) Printing (more). Previously (and on every other windows program) the
clicking on the print command (file -> print) brought up a printer control
box in which you could select the printer, change properties from portrait
to landscape, one page per sheet to two pages per sheet, printing front and
back, etc (The HP LJ4200/4300 series printer supports those options.)
However, those options are now missing on the 16 printer control box. Is
this true for others?

4) I find that the Utilities --> variables dropdown does not work from
either syntax window or output window unless it has been first activated in
the data window. Is this true for others?

5) I find that on one computer the insert/overtype toggle does not work but
on another computer it does. Is this true for others?

6) Using a dataset created in 15, I find that a frequencies command for
variables with a defined format of f8.2 yields an as printed in the
frequency table format of f18.16. Curiously, opening the cell format shows a
format of #.#, i.e., f3.1. Have others noticed this also? How can it be
fixed--permanently--so that it matches the defined format.

In order of importance to me is 3), 6), 4), 1), 2), 5). Can anyone offer
solutions or workarounds?

I can only guess that there are many, many other issues. I'd be interested
to hear of them OFF-LIST along with work-arounds found, if any. I'd also be
curious to know if they are documented--not that it matters any more. I'd
also be curious to know if beta testers found these problems. I'd like to
know what I'll be facing for the next year. Perhaps they are idiosyncratic
to our computers, e.g., a Dell GX270 with a pentium 4 chip and XP
professional 2002 with service pack 3.

Gene Maguin

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Oliver, Richard]

If you typically open data files via syntax (e.g., get file='filename.sav'), problem #4 has been fixed in version 17. The best workaround in version 16 is the one you already discovered: activate the Data Editor window. This problem affects any dialogs that contain variable lists. The workaround in command syntax in version 16 is to explicitly name and activate the dataset, as in:

get file='filename.sav'.
dataset name somename window=front.

But if you typically have syntax jobs that open (and subsequently automatically close) multiple data files in a sequential manner, you may not want to have explicitly named datasets.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Gene Maguin
Sent: Friday, September 19, 2008 2:24 PM
To: [hidden email]
Subject: V16 questions

All,

I've just had 16 installed and I patched it to 16.0.2, but have not added
the hotfix because people here don't seem to be absolutely sure whether it
is the academic version. I just used it to do some litte tasks and several
odd--meaning different from 15--things have shown up and I wonder if others
have seen this or if this could be due to idiosyncracies of our network
system.

1) The default syntax font was some unknown to me monospaced font that I
changed to new courier, which have used consistently in the past. I now
notice that odd 'square boxes' display in the output view between words.
Instead of a blank space there is this square box. These boxes do not
print--fortunately. Changing back to the 'monotype' font does not correct
the problem. Anybody else seen this?

2) Printing. On 15 the delay between pressing the print button and having
the print box come up was maybe 5 seconds. I print to a network printer (HP
LJ4200/4300 series printer). Now the delay is more like 10-15 seconds. I
thought that maybe there were network changes that would account for this
but our network person says No. Does this happen to anyone else?

3) Printing (more). Previously (and on every other windows program) the
clicking on the print command (file -> print) brought up a printer control
box in which you could select the printer, change properties from portrait
to landscape, one page per sheet to two pages per sheet, printing front and
back, etc (The HP LJ4200/4300 series printer supports those options.)
However, those options are now missing on the 16 printer control box. Is
this true for others?

4) I find that the Utilities --> variables dropdown does not work from
either syntax window or output window unless it has been first activated in
the data window. Is this true for others?

5) I find that on one computer the insert/overtype toggle does not work but
on another computer it does. Is this true for others?

6) Using a dataset created in 15, I find that a frequencies command for
variables with a defined format of f8.2 yields an as printed in the
frequency table format of f18.16. Curiously, opening the cell format shows a
format of #.#, i.e., f3.1. Have others noticed this also? How can it be
fixed--permanently--so that it matches the defined format.

In order of importance to me is 3), 6), 4), 1), 2), 5). Can anyone offer
solutions or workarounds?

I can only guess that there are many, many other issues. I'd be interested
to hear of them OFF-LIST along with work-arounds found, if any. I'd also be
curious to know if they are documented--not that it matters any more. I'd
also be curious to know if beta testers found these problems. I'd like to
know what I'll be facing for the next year. Perhaps they are idiosyncratic
to our computers, e.g., a Dell GX270 with a pentium 4 chip and XP
professional 2002 with service pack 3.

Gene Maguin

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD