Actually, I have another question. I am seeing some controversy about whether to use Pearson r or Spreaman rho for Likert-type questions. What would you use?
Hello! I have a quick question.
I have made a survey for a sample program evaluation. Many of the questions I have composed are Likert scale type questions. I would like to see if the responses to a few of these items correlate. I'll give you an example. I want to see if the perceived success of volunteers is associated with how supportive the staff is in their opinion. I used the "strongly agree, agree, neutral, disagree, strong disagree" with the coding "1, 2, 3, 4, 5."
Because it is ordinal, I figured I should use Spearman rho, but I am slightly confused about ranking. Do I need to rank this data? This is what I have for the raw data:
S1: X: 5 Y: 2
S2: X: 5 Y: 3
S3: X: 1 Y: 1
S4: X: 3 Y: 2
S5: X: 1 Y: 3
S6: X: 1 Y: 1
S7: X: 1 Y: 4
S8: X: 1 Y: 1
S9: X: 4 Y: 1
S10: X: 3 Y: 3
So the question I have is does this data need to be ranked? Thank you in advance.