I am looking to replicate the findings of Leder, forster, and Gerger (2011) in their publication of the glasses stereotype revisited.

I intend to examine the same conditions of no-glasses, rimmed glasses, and rimless glasses, whilst adding an extra condition of sunglasses.

Currently my struggle is finding a facial database that has identical pre-rated images for each condition. I am hoping to avoid taking my own images, as I would then have to score and test them for neutraL facial expressions as well as make sure that I limit the possibility of previous/ future exposure between models and participants, as they will be rated on percieved intelligence, attractiveness, trustworthiness, and success. If anyone is able to point out some possibilities that would be great. I have found somewith either, glasses present vs glasses not present, and some with sunglasses present vs sunglasses not present, but no unified database with identical models for each condition.


Leder, H., Forster, M., & Gerger, G. (2011). The glasses stereotype revisited: Effects of glasses on perception, recognition and impressions of faces. Swiss Journal of Psychology, 70: pp. 211—222.
DOI: 10.1024/1421-0185/a000059

  • $\begingroup$ Just curious, have you asked the authors of the original study whether they would allow you to use their materials? $\endgroup$ – Jeromy Anglim Dec 17 '14 at 0:23
  • $\begingroup$ I have not contacted them yet no. The although it would help with replicating their results, it would unfortunately does not include the sunglasses condition. The sunglasses condition is actually the primary reason behind this, as I hypothesize that they could result in opposite results due to masking the eyes. $\endgroup$ – David Goddard Dec 17 '14 at 2:44
  • $\begingroup$ I think it's unlikely you'll find a single database that has all the conditions you need. Would it be sufficient to have multiple databases that cover the conditions? $\endgroup$ – Josh de Leeuw Dec 17 '14 at 3:35
  • $\begingroup$ I belive that may have to be the option then.I was hoping that somewhere there was a good database for all conditions such as a Michael Tarr database. I am thinking that perhaps in stead of running the same model in each condition then, if it would be fine to have different models but a larger stimulus pool. I was planing to do a repeated measures, exposing each participant to the entire pool of stimuli. Do you think that this would be minimal, in how much power I lose, by having different models in all the different conditions? $\endgroup$ – David Goddard Dec 17 '14 at 20:21
  • $\begingroup$ could anyone please provide some advice? In stead of having same model face represented in each condition, such as in the original study, would it still be acceptable to compare different faces, if i use 30 stimuli per condition. I understand i will lose some power by doing this but i hope with 30 per condition this will compensate for that. $\endgroup$ – David Goddard Jan 3 '15 at 17:31

Your Answer

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

Browse other questions tagged or ask your own question.