Hi Egon,

It is interesting if four items of the GSE scale is strongly correlated with the full 10 items version. My own studies (listed on my RG page) using that scale have shown alpha coefficients > 0.90, this being in support of your idea that fewer items may represent the GSE construct quite well. However, I would say that by adapting four of the GSE items to learning (as you describe) it is hard to know what you would be measuring by that modified scale. One way of moving forward might be to use the four modified items within a larger pool of items, where you also could include the full GSE scale and an established scale for measuring academic self-efficacy, and then compare?

Best wishes,

Tore

Hello Egon. As Tore Bonsaksen suggests it is difficult to figure out what kind of scale is that you will be creating by modifying general self-efficacy scale items to learning. Additionally, you have to consider what aspects of learning you intend to measure. If you wish to create a learning self-efficacy scale my suggestion is to use another empirically tested and well-established learning self-efficacy scale or academic self-efficacy scale (see some reports by Usher and Pajares, 2007 and Usher, 2009 using relative scales) or otherwise make a new learning self-efficacy scale applying ideas drawn from Schwartzer and Jerusalem's scale by but not using the same items. Note that general self-efficacy differs from task-specific self-efficacy which aims at evaluating individuals' beliefs about carrying out specific roles, tasks and aspects in a designated level of performance (Bandura, 1997).

Moreover, learning includes deep understanding of a complex system of numerous tasks which could be measured either in a general level or a certain level of accomplishing tasks.In my opinion is not feasible to measure learning issues with a tool of just few items (4 are very few to evaluate any construct). To conclude, If I were you I would rather do the following: (a) decide what do I exactly want to evaluate about learning, (b) make many new items after taking interviews by experts in the field, at least 40 or 50, (c) draw on well-established research in learning self-efficacy or learning processes in general, (d) run a pilot study in which you will be testing the psychometric properties of the items and finally, retain a portion of them, perhaps 15-20.

Best wishes for your research.

Kind regards,

Andronikos

Thank you for your answers.

You helped me do decide what to do.

The time that I have to find a scale gives me little choice, I have to search for an already validated scale.

Kind regards

Egon

Egon I think this is the best available choice due to lack of time..However, try to adapt it the culture and population with which you are going to work. Back translation would help you to do this. I would avoid just taking the scale, translating it and use it instantly. You will find attach two papers focused on learning and academic achievement exploring the sources of academic self-efficacy. As you will notice some of the measures described in the "Instrument" section refer to academic learning or self-regulated learning. Take a look and check out whether you find them appropriate for your study.

All the best,

Andronikos

here they are

Dear Egon,

though my answer might be a little bit too late: It always depends what you want to investigate (big surprise), but I always try to apply (domain) specific instruments or measurments as these are more sensitive for changes an usually have a better exploratory and predictive value. I hope this reply still helps you.

Kind regards

Julian

 It's a long time since I postet this question.

I'd like to communicate, what I finally did.

I took the database of the general self-effecy expectation scale of Schwarzer and Jerusalem (http://userpage.fu-berlin.de/%7Ehealth/world_24nations_25nov2006.sav) and searched for the four or fife best items. These items, I adapted for the domain of [academic] studies and distributed the short questionnaire to students. It resulted in good values (e.g. alpha=.74).