FW: need to calculate a single weighted total, off frequencies/counts, repeatedly [Sec: UNCLASSIFIED]
Locked
FW: need to calculate a single weighted total, off frequencies/counts, repeatedly [Sec: UNCLASSIFIED]
 
|
Greetings all, I have a dataset of iron intakes for children of varying
ages. I have already gone through and computed iron intake categories for
each age group, for each sex separately. There are a total of 14 iron intake
categories, which correspond to the descending probability that of iron
insufficiency in the diet. There are different numbers of children in each category,
and for some children/age combinations there are “missing”
categories as no children fell into the relevant iron intake range. What I need to do, 1.
for each age-sex group, and 2.
for each iron intake category within that age-sex
group, is to multiple the frequency/count of children by the
probability for that iron intake category, to produce a single figure.
Then I need to sum all those single figures to end up with one (weighted)
total. This can be print to the output window. I don’t need any of the probabilities or frequencies
data saved, just that end weighted total for each age-sex combination in the
data. I was thinking that I need to use the matrix/end matrix with
some type of loop inside. For example, here is the frequency output table for females
aged 2-3 years:
OBSERVED IRON 2-3 YEARS FEMALE
Frequency
Percent
Valid Percent
Cumulative Percent Valid
3966
88.4
88.4
88.4
Category 10
2
.0
.0
88.4
Category 11
30
.7
.7
89.1
Category 12
92
2.1
2.1
91.2
Category 13
120
2.7
2.7
93.8
Category 14
277
6.2
6.2
100.0
Total
4487
100.0
100.0 As you can see, there are no females aged 2-3 years who fall
into the first 9 iron intake categories. Basically, how can I calculate
those frequency numbers (in the example above, for Categories 10 through 14),
which I can then weight each one by its relevant probability, then sum them all
to get the weighted total (and print that weighted total to the output window)? Note 1: the “Valid” count relates to other children
in the dataset, who are different age/sex combination to that shown above. Note 2: the string variable “IRON2TO3F” contains
the “Category 10”…”Category 14” values used in
the frequency table above. In case it helps, the syntax used to create the table
was: FREQUENCIES VARIABLES=IRONGROUP /ORDER=ANALYSIS. Each age/sex combination has a different variable storing
the frequencies so I could get around DO statements. For example, males aged
2-3 years have their category data stored in the variable “IRON2TO3M”.
And I need to calculate their single weighted total as well, along with another
8 age-sex combinations. Cheers Michelle ********************************************************************** This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager.
This footnote also confirms that this email message has been swept by MIMEsweeper for the presence of computer viruses.
www.clearswift.com **********************************************************************
|
Re: need to calculate a single weighted total of frequencies/counts, repeatedly [Sec: UNCLASSIFIED]
|
|
MIchelle,
I'm not sure i understand your problem but, to the extent that
i do understand it, i think that you might use the Aggregate command. You
define your age-sex combinations and iron consumption categories. Then (i assume
your file is unweighted)
aggregate outfile=*/break=age_sex
iron_category/cases=nu.
Then divde each count value (cases variable) by 4487, which
appears to be your total N. I don't know that this does everything that you want
done but i don't understand all the pieces.
Gene Maguin
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Gosse, Michelle Sent: Thursday, November 11, 2010 9:37 PM To: [hidden email] Subject: FW: need to calculate a single weighted total, off frequencies/counts, repeatedly [Sec: UNCLASSIFIED] Greetings all, I have a dataset of iron intakes for children of varying
ages. I have already gone through and computed iron intake categories for
each age group, for each sex separately. There are a total of 14 iron
intake categories, which correspond to the descending probability that of iron
insufficiency in the diet. There are different numbers of children in each category, and
for some children/age combinations there are “missing” categories as no children
fell into the relevant iron intake range. What I need to do, 1.
for each age-sex group, and 2.
for each iron intake category within that age-sex group,
is to multiple the frequency/count of children by the
probability for that iron intake category, to produce a single figure.
Then I need to sum all those single figures to end up with one (weighted)
total. This can be print to the output window. I don’t need any of the probabilities or frequencies data
saved, just that end weighted total for each age-sex combination in the
data. I was thinking that I need to use the matrix/end matrix with
some type of loop inside. For example, here is the frequency output table for females
aged 2-3 years:
OBSERVED IRON 2-3 YEARS FEMALE
Frequency
Percent
Valid Percent
Cumulative Percent Valid
3966
88.4
88.4
88.4
Category 10
2
.0
.0
88.4
Category 11
30
.7
.7
89.1
Category 12
92
2.1
2.1
91.2
Category 13
120
2.7
2.7
93.8
Category 14
277
6.2
6.2
100.0
Total
4487
100.0
100.0 As you can see, there are no females aged 2-3 years who fall
into the first 9 iron intake categories. Basically, how can I calculate
those frequency numbers (in the example above, for Categories 10 through 14),
which I can then weight each one by its relevant probability, then sum them all
to get the weighted total (and print that weighted total to the output
window)? Note 1: the “Valid” count relates to other children in the
dataset, who are different age/sex combination to that shown
above. Note 2: the string variable “IRON2TO3F” contains the
“Category 10”…”Category 14” values used in the frequency table above. In case it
helps, the syntax used to create the table was: FREQUENCIES VARIABLES=IRONGROUP /ORDER=ANALYSIS. Each age/sex combination has a different variable storing the
frequencies so I could get around DO statements. For example, males aged 2-3
years have their category data stored in the variable “IRON2TO3M”. And I
need to calculate their single weighted total as well, along with another 8
age-sex combinations. Cheers Michelle ********************************************************************** This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager.
This footnote also confirms that this email message has been swept by MIMEsweeper for the presence of computer viruses.
www.clearswift.com **********************************************************************
|
Re: need to calculate a single weighted total of frequencies/counts, repeatedly [Sec: UNCLASSIFIED]
|
|
Hi Gene, Thanks to your tip, I have read
up on the Aggregate command and it seems to be exactly what I need, using the NU
function as you suggest below. This will be so much easier than
the Matrix/End matrix iteration I thought I was going to be doing. Thanks for your help. J Michelle From: SPSSX(r) Discussion
[mailto:[hidden email]] On Behalf Of Gene Maguin MIchelle, I'm not sure i understand your problem but, to the extent that i do
understand it, i think that you might use the Aggregate command. You
define your age-sex combinations and iron consumption categories. Then (i
assume your file is unweighted) aggregate outfile=*/break=age_sex iron_category/cases=nu. Then divde each count value (cases variable) by 4487, which appears
to be your total N. I don't know that this does everything that you want done
but i don't understand all the pieces. Gene Maguin From: SPSSX(r)
Discussion [mailto:[hidden email]] On Behalf Of Gosse,
Michelle Greetings all, I have a dataset of iron intakes for children of varying
ages. I have already gone through and computed iron intake categories for
each age group, for each sex separately. There are a total of 14 iron
intake categories, which correspond to the descending probability that of iron
insufficiency in the diet. There are different numbers of children in each category,
and for some children/age combinations there are “missing”
categories as no children fell into the relevant iron intake range. What I need to do, 1.
for each age-sex group, and 2.
for each iron intake category within that age-sex group, is to multiple the frequency/count of children by the
probability for that iron intake category, to produce a single figure.
Then I need to sum all those single figures to end up with one (weighted)
total. This can be print to the output window. I don’t need any of the probabilities or frequencies
data saved, just that end weighted total for each age-sex combination in the
data. I was thinking that I need to use the matrix/end matrix with
some type of loop inside. For example, here is the frequency output table for females
aged 2-3 years:
OBSERVED IRON 2-3 YEARS FEMALE
Frequency
Percent
Valid Percent
Cumulative Percent Valid
3966
88.4
88.4
88.4
Category 10
2
.0
.0
88.4
Category 11
30
.7
.7
89.1
Category 12
92
2.1
2.1
91.2
Category 13
120
2.7
2.7
93.8
Category 14
277
6.2
6.2
100.0
Total
4487
100.0
100.0 As you can see, there are no females aged 2-3 years who fall
into the first 9 iron intake categories. Basically, how can I calculate
those frequency numbers (in the example above, for Categories 10 through 14),
which I can then weight each one by its relevant probability, then sum them all
to get the weighted total (and print that weighted total to the output window)? Note 1: the “Valid” count relates to other
children in the dataset, who are different age/sex combination to that shown
above. Note 2: the string variable “IRON2TO3F” contains
the “Category 10”…”Category 14” values used in
the frequency table above. In case it helps, the syntax used to create the
table was: FREQUENCIES VARIABLES=IRONGROUP /ORDER=ANALYSIS. Each age/sex combination has a different variable storing
the frequencies so I could get around DO statements. For example, males aged
2-3 years have their category data stored in the variable
“IRON2TO3M”. And I need to calculate their single weighted
total as well, along with another 8 age-sex combinations. Cheers Michelle **********************************************************************
This email and any files transmitted with it
are confidential and intended solely for the use of the individual
or entity to whom they are addressed. If you have received this
email in error please notify the system manager. This footnote also confirms that this email
message has been swept by MIMEsweeper for the presence of computer
viruses. www.clearswift.com **********************************************************************
|
| Free forum by Nabble | Edit this page |