# Graphs type for likert scaled questionnaire?

My questionnaire is on Likert Scale and is divided into 7 sections say A, B, C, D, E, F, and G. In each section there are a number of questions which the respondent has to fill in accordance with the likert scale (0=No to 5=very high).

To make it simpler my questionnaire looks like:

• Q1
• Q2

• Q1
• Q2
• Q3
• Q4

And so on...

## Queries

1. I want to make graphs of each section (i.e. A, B, C, D, E, F, and G) in such a manner that for section "A" a combined graph is formed in which all the questions i.e. Q1, Q2, and Q3 are represented.
2. I am a civil engineer by profession therefore any guide regarding the statistics required for a research based project, will be highly appreciated.
3. Is it possible that I could upload my questionnaire here and, after seeing it any one might help me in identifying the type of statistics to be used on it?
• About question 3, you can provide a minimal reproducible example, or a small sample of your data (10 records as plain text in your original post would be ok). Anyone could help provided it is known what the data looks like (e.g., are there external demographic variables that are of interest) and what question(s) you are trying to answer with your summary/inferential statistics. – chl Aug 13 '12 at 10:00
• Besides Visualizing Likert Item Response Data (which chl mentions in the comment to Blain's answer) also see Visualizing Likert responses using R or SPSS. There are alot of questions about likert items on the site, I would suggest you peruse that tag for more help on your query 3. – Andy W Aug 13 '12 at 12:01

For each section you can visualize pairwise that how many have answered a particular combination of the possible answers. A matrix plot of the questions may reveal that. At first create a data frame whose columns will be the responses of each question. Then simply plot this. You get something exactly like the scatter plot. Better answers to the visualization problem may exist. If you use R, please see the following codes to understand what I meant.

> Q1<-c(1,2,0,2,0,1,1,0,2)
> Q2<-c(3,2,1,2,3,2,3,0,1)
> Q3<-c(0,1,1,1,1,0,0,1,0)
> Sec_A<-data.frame(Q1,Q2,Q3)
> plot(Sec_A)


In response to your third question I would suggest you to upload your data on any site that allows you to upload your files and then simply post the link here. Experts may suggest you about the procedure. This site certainly provides you with great suggestions.

• About visualization techniques, there are some discussions lurking around there: Visualizing Likert Item Response Data. For raw data, I believe having a short overview should be enough, although it is possible to rely on external web hosting systems. Now, the question is what the graphical displays and numerical quantities aim to demonstrate/illustrate. – chl Aug 13 '12 at 10:40
• Wow, great reference, this is far better than my answer. – Blain Waan Aug 13 '12 at 10:53