Let's say, I used two scales that measures - 1. Perception of eco-label and 2. Perception of eco-brand. Each of them have 4 items (questions). Items are measured on 7 point-likert type questions. These scales have been previously used in other papers and their items were grouped under these 2 scales names with Factor Analysis in those papers as well. Therefore, According to previous papers, I know what item belongs to which scale.

My question is the following: Do I still need to do 2 separate Factor Analyses, in my paper, for grouping these items under 2 factors (perception of eco-label and eco-brand)? OR it is enough to know from previous papers which items belong to which scale?

please advise

  • $\begingroup$ Replication is never a bad thing. $\endgroup$ – shadowtalker Dec 20 '15 at 0:21

If your audience is familiar with the questionnaires and your subjects are drawn from a similar population, I don't think you need to redo the factor analysis. Instead, you could write something like

Each subject completed the [Questionnaire Name] and the responses were analyzed as in [citation] to assess subjects' perception of eco-label and eco-brand.

This is pretty common in parts of psychology. One publication describes a Questionnaire/Inventory/Scale and demonstrates its usefulness (e.g., a certain pattern of responses predicts that the subject will be diagnosed with a condition). Subsequent papers typically take the scale as given and show novel applications of it (e.g., selecting treatments), without attempting to re-validate the scale from the original paper.

However, you do have to be careful that you are administering the same scale and to a similar pool of subjects, since even minor changes like moving from a paper questionnaire to a computerized survey or a slightly different subject pool can sometimes affect its validity.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.