I am very interested in a survey employing a Likert type scale in my research. I have come across the Rosenberg Self-Esteem Scale. I would like to use this scale to get a sense of how subjects feel about their self-esteem during their trial in my research.

The questions will stay consistent with Rosenberg’s test model keeping five of the items positively worded and keeping five negatively worded using the four-point scale from strongly agree to strongly disagree.

  • Can the scale questions be "changed"/presented without losing validity in the self-esteem measurement?
  • If not, then how do I let participants answer questions in the context of my experiment and what I am trying to improve? For example, I would like to know how they feel about learning a new language (e.g., Japanese) using a traditional method. Then I present a new method and ask again how they feel about learning the same new language? I would like to capture feedback from participants that would let me know if they feel better/more confident about learning a new language by utilizing a new method/approach?

The Rosenberg questions are extremely general (I believe this is the beauty of it) which means it can yield stronger more general results. However, from my limited knowledge, I believe a generally confident person would answer those questions positively most of the time which means filling-out this scale before and after an experiment wouldn't conclude drastic result/change. However, if those questions were directed towards an issue that the subjects in the experiment all find difficult, then I believe their subjective answer to the questions would show considerable change if a *new method promises to improve their perceived intuition over a traditional method?

The following are the questions from Rosenberg’s self-esteem scale:

  1. I feel that I am a person of worth, at least on an equal plane with others.
  2. I feel that I have a number of good qualities.
  3. All in all, I am inclined to feel that I am a failure.
  4. I am able to do things as well as most other people.
  5. I feel I do not have much to be proud of.
  6. I take a positive attitude toward myself.
  7. On the whole, I am satisfied with myself.
  8. I wish I could have more respect for myself.
  9. I certainly feel useless at times.
  10. At times I think I am no good at all.

If a Rosenberg scale shouldn't be used in this case, then what would be an appropriate scale?

I was originally looking at The General Self-Efficacy Scale (GSE). However, I would still ask the same qustions of GSE that I have of Rosenberg's Scale. How can I "retro-fit" GSE to make it in the context of say learning a language & self-efficacy?

The following are the questions from the GSE scale:

  1. I can always manage to solve difficult problems if I try hard enough.
  2. If someone opposes me, I can find the means and ways to get what I want.
  3. It is easy for me to stick to my aims and accomplish my goals.
  4. I am confident that I could deal efficiently with unexpected events.
  5. Thanks to my resourcefulness, I know how to handle unforeseen situations.
  6. I can solve most problems if I invest the necessary effort.
  7. I can remain calm when facing difficulties because I can rely on my coping abilities.
  8. When I am confronted with a problem, I can usually find several solutions.
  9. If I am in trouble, I can usually think of a solution.
  10. I can usually handle whatever comes my way.
  • 1
    $\begingroup$ It's hard to tell what you mean by "context of my experiment" and "what I am trying to improve." I suspect it would take an expert in psychology to give you a good, well-supported answer. Is there some statistical aspect of this question we could help you with? $\endgroup$
    – whuber
    Mar 23 '15 at 21:17
  • $\begingroup$ @whuber: I have revised my question a bit, thanks for your response! $\endgroup$ Mar 24 '15 at 17:11
  • $\begingroup$ Thank you for updating for your post, Susu. I'm afraid it would probably better fit on cognitive.SE as it is more about psychology than statistics or psychometrics per se. We can migrate this question for you if you like. Just let us know (flag your question for moderator attention). $\endgroup$
    – chl
    Mar 24 '15 at 20:19
  • $\begingroup$ @chl: Thank you! I have flagged my question. $\endgroup$ Mar 24 '15 at 20:33
  • 1
    $\begingroup$ It sounds like, based on your language example, that you're not trying to measure self-esteem (a global rating of self-worth) but rather self-efficacy (how confident am I in my ability to do a specific task). Is this accurate? $\endgroup$
    – mrt
    Mar 24 '15 at 20:49

It depends on the types of changes you are looking to make. In general, my experience has been that changes to a response scale can be done with minimal threats to validity (e.g. looking to change the 4 item Likert type scale to a 6 item Likert type scale). On the other hand, changing question wording itself usually can't be done "without losing validity" (Goodman, Iervolino, Collishaw, Pickles, & Maughan, 2007).

To answer the "what now" question, you can either use the survey as is, or modify it to better fit your needs (and "revalidate" as needed). This decision is ultimately up to you and should be based on what you think would give you the best information about what you are trying to measure. Based purely off of your brief description and the questions provided, you may prefer to modify the survey. It doesn't seem like the survey, as is, would give you the information you are looking to measure.


Goodman, R., Iervolino, A. C., Collishaw, S., Pickles, A., & Maughan, B. (2007). Seemingly minor changes to a questionnaire can make a big difference to mean scores: a cautionary tale. Social Psychiatry and Psychiatric Epidemiology, 42(4), 322–327. doi:10.1007/s00127-007-0169-0


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.