APU SOCI332 2019 March Week 3 Forum Latest

SOCI332 Statistics for Social Science
Week 3 Forum
Week 3 Forum (Optional-not for credit)
View Full Description
Because you have an assignment due this week, you will have no required forum response. Utilize this forum to ask questions about Assignment 1 and/or your research topic/project. You can also utilize this forum to practice SPSS functions, such as recoding described below.
You will have completed most of the steps for Assignment 1 in Forums 1 and 2. The exception is recoding. Recoding allows you to take numeric variables and group the responses into categories or take the responses to categorical variables and group them into fewer categories. For example, take the variable “age” (age of respondent) for which participants answer with their numeric age. Let’s say you want to group ages into categories.
Here’s one way age is typically categorized:
•    18-24 years old
•    25-34 years old
•    35-44 years old
•    45-54 years old
•    55-64 years old
•    65-74 years old
•    75 years or older
To perform a recode in SPSS, after you have imported in the GSS data file, you use the TRANFORM function. Click on Transform – Recode into different variable.  On the left-hand side, click on the variable “age of respondent [age]” and move it over to the right. Under Output Variable, you will give your new recoded variable a name and label, and click Change. (I would call it “ager” and label it “recode of age
into categories.”) When you recode a numeric (interval/ratio) variable, you are telling it what the original numbers are and how you want to code (assign numbers to) the new categories.
Click on Old and New Values. I would recode it like this:
Old value (Range): 18-24     New value: 1
Old value: 25-34     New Value: 2
Old value: 35-44     New Value: 3
Old value: 45-54     New Value: 4
Old value: 55-64   New Value: 5
Old value: 65-74     New Value: 6
Old value: 75-89     New Value: 7
Old value: 98, 99      New Value: system missing
Continue – OK
Your new recoded variable will then be at the bottom of the variable list. In variable view, you can go in an assign labels to the numbers (i.e., 1 is labeled 18-24, 2 is labeled 25-34, etc.).
Another example might be a categorical variable that has too many categories and you would like to combine some of the categories that are similar to create a variable with fewer categories. Let’s take a look at the variable “denom16” – denomination in which r was raised. With a categorical variable (nominal or ordinal), the first step of recoding is to look at how the variable is coded (numbers assigned to the categories) to begin with. I like to go back to the GSS data explorer site to see this, but in SPSS you can go to ANALYZE – REPORTS – CODEBOOK and look up your variable. Denom 16 is coded like this:
 
Code    Label
10    Am baptist asso
11    Am bapt ch in usa
12    Nat bapt conv of am
13    Nat bapt conv usa
14    Southern baptist
15    Other baptists
18    Baptist-dk which
20    Afr meth episcopal
21    Afr meth ep zion
22    United methodist
23    Other methodist
28    Methodist-dk which
30    Am lutheran
31    Luth ch in america
32    Lutheran-mo synod
33    Wi evan luth synod
34    Other lutheran
35    Evangelical luth
38    Lutheran-dk which
40    Presbyterian c in us
41    United pres ch in us
42    Other presbyterian
43    Presbyterian, merged
48    Presbyterian-dk wh
50    Episcopal
60    Other
70    No denomination
98    Don’t know
99    No answer
0    Not applicable
I think it’s always best to code starting with 0 or 1 (depending on what you’re doing), so here I would start with one. In the recoding process, we will take each of the old categories and combine them into new categories and give them new codes (numbers). It’s not hard, it just needs to be precise.
In SPSS, after you have imported the GSS data file in, click on Transform – Recode into different variable. On the left-hand side, click on the variable “Denomination in which r was raised [denom16]” and move it over to the right. Under Output Variable, give it a new name and label, and click Change. (I would call it “rdenom16” and label it “recode of denom16.”)
Click on Old and New Values. I would recode it like this:
Old value: (Range) 10-18     New value: 1
Old value: 20-28     New Value: 2
Old value: 30-38     New Value: 3
Old value: 40-48     New Value: 4
Old value: 50   New Value: 5
Old value: 60     New Value: 6
Old value: 70     New Value: 0
Old value: 0      New Value: system missing
Old value: 98-99      New Value: system missing
Old value: system or user missing   New Value: system missing
Then click CONTINUE and then OK. (It will look like this…)
The new variable will be placed at the very bottom of the variable list. (You can go into the variable view in SPSS and label the new codes (numbers) with 0=no denomination, 1=baptist, 2=methodist, 3=lutheran, 4=presbyterian, 5=episcopal, 6=other.) When you run future analyses, remember that the recoded variable is at the bottom. The original variable is still there in its original place.