Need design help...not sure how to analyze

#1
Hi, I am writing a proposal in which I am trying to implicitly activate a stereotype. Specifically, I want to see if professional sexy attire (not revealing...i.e., knee length pencil skirt, collared shirt, fitted blazer) will activate the "business woman" stereotype or the "sexy woman" stereotype. What I want to do is present one of 3 photo conditions of the same woman wearing either conservative (blazer, slacks, collared shirt), professional sexy (see above description), or unprofessional sexy (collared shirt with ruffled front and several buttons undone, short skirt, "strappy" high heels) to each participant along with two word lists (within subjects) said to be comprised of words (1) her friends and family use to describe her and (2) her co-workers use to describe her. I will counterbalance the order of presentation so that participants either see the image first followed by the word lists, or the word lists first, followed by the image. Included in the word lists (each with 5-10 words) will be both words shown to be associated with the "business woman" stereotype (e.g., competent) and some associated with the "sexy woman" stereotype (e.g., good figure) as well as some neutral words (e.g., funny). Then, I want to have participants perform a math-related distractor task before administering a Word Recognition Test and a Gist Test (in which participants will be asked to identify words that seem like they went with a theme or concept that was presented in the initial word lists...even if they weren't actually presented). I expect that participants who see the conservatively dressed woman will recognize more "business woman" associated words and gist words. The unprofessional sexy woman is expected to lead participants to recognize more "sexy woman" associated words and gist words. There is no formal hypothesis about the professional sexy woman, as this is the one of interest and there is not much information in the literature about which stereotype would be activated for this type of attire. There are two dependent variables: (1) proportion of "business woman" to "sexy woman" stereotype words recognized on the Word Recognition Test and (2) proportion of "business woman" to "sexy woman" stereotype words identified on the Gist Test.

So, as I see it, I have a between subjects independent variable with three levels [3 (image attire: conservative, professional sexy, unprofessional sexy) crossed with one within subjects variable with two levels [2 (order of presentation: Image first, Word lists first)]. --Would this be a 3X2 ANOVA? Someone in one of my classes suggested that this may be a Split Plot Factorial design...but I am not sure why...maybe because I have one within and one between subjects variable?

I am also interested in whether gender of the participant makes a difference with regard to which stereotype is activated. For instance, in a pilot study male participants rated the model as significantly more sexy than female participants rated her (although both rated her as equally highly sexy across all three image conditions). So I would expect that male participant in the proposed study might recognize more "sexy woman" words than female participants do overall, regardless of the image condition. --Would this make it and ANCOVA? Or do I just run the 3X2 ANOVA with participant gender as a covariate?

Please excuse this if it is a silly question. I am a first year Master's student and I am just learning about all of this stuff. I fear I have been a bit too ambitious here...but any help is appreciated! Thank you in advance!
 
Last edited: