I'm struggling to figure out how to create a sum of the # of respondents who selected a particular response for a total scale. I've been trying to do this in R but haven't been able to properly figure out the code to get my desired results. For example:
The scale is made up of 4 questions that are indicative of "X". Response options are Strongly agree, agree, disagree, and strongly disagree. How can I generate a table that will tell me the total number of respondents selecting "strongly agree" across all 4 items (e.g., on scale "X", 94% of respondents strongly agreed)?
I'm able to generate frequencies for each individual item using table(), but can't find a way to combine it.
Thank you for your help!
You will have to post some code and/or data to properly answer this question.
Good point. I attached a dummy dataset in .csv format.
The first column "ID" is a dummy respondent ID followed by four columns "X1-X4" serving as the questions, with responses coded from 1 to 4.
Hi,
You might generally get quicker answers on R from stackoverflow or the R help list but I happened to spot your question. There's always more than one way to do things in R, the following should help.
# the dummy data
dat = data.frame(ID = 521:525, X1=c(4,4,1,1,3), X2=c(4,3,1,4,1), X3=c(3,2,3,4,1), X4=c(3,1,2,3,4))
# table from the 4 variables
tab = table(dat$X1, dat$X2, dat$X3, dat$X4, dnn=c("X1","X2","X3","X4"))
tab
# make output into a data frame as easier to read and use
res = data.frame(tab)
res
# one way of picking out a row, answered 4 for all questions
# to see help file use ?"&"
res[res$X1==4 & res$X2==4 & res$X3==4 & res$X4==4, ]
# or just the frequency for that row
res$Freq[res$X1==4 & res$X2==4 & res$X3==4 & res$X4==4]
# answered 4 for any of the questions, i.e. X1 or X2 or X3 or X4
any4 = res[res$X1==4 | res$X2==4 | res$X3==4 | res$X4==4, ]
sum(any4$Freq)
Dear Randal
If I got Your question correctly (what I am not sure) and You do not mind to do it in Excel instead of R, please see attachment. If I did not get it, can You please clarify?
Thank you for your responses. I think you're both interpreting me correctly, but I might be miscommunicating what my end goal is.
1) I do not want to do this in excel, if at all possible. What I've been doing is copying the frequencies from R into excel and hand-calculating total % based on the formula (# of "1" responses/total # of responses for the question).
2) I'm also needing to calculate overall responses, such as how many people selected "1" on items x1, x2, x3, x4 TOTAL, and then determine the % based on the total # of respondents across all items.
In David's code, all makes sense until selecting the # of people selecting 4s. I'm looking for a table that will give me these calculations for all responses (1,2,3,4) on each question separately (x1, x2, x3, x4) and total (x1+x2+x3+x4). Does that help clarify at all?
Update: I have figured out how to create a table with %/item using the command:
total
Have you tried transposing the variables in your table? For example, use the responses as columns and the individual events/responders as rows. That should allow you to conduct the operation you need. From there, you can also create variables that give you the results that you are looking for. For example, a
I've made a little R code for you. I hope I've understood your needs correctly. In any case, the script will show you some efficient ways to process your data. Cheers.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R Statistical Package