I am trying to do an online survey where we are looking at unnatural vs natural hair color over a 4 stimuli presentation design. In terms of stimuli, we have are 3 different hair color conditions (1 natural and 2 unnatural) per each of 4 races (black, white, asian, and hispanic)-- so 12 total. The twist is I want to have natural, natural, unnatural, unnatural presented over the four stimuli (order randomized) with each of the four stimuli being difference races. Additionally, I would like each to have a resume three being highly qualified and one of them being a random validity check with a resume that is low quality… help!!!
You’ve said survey but I assume you mean experiment.
You have four ethnicities but you are only presenting them three times. How about a Latin Square design where you randomise the order of the ethnicities and then have:
hair_colour[ethnicity0] = [0,1,2]
hair_colour[ethnicity1] = [1,2,0]
hair_colour[ethnicity2] = [2,0,1]
hair_colour[ethnicity3] = [0,1,2]
If you always want two natural then I think you need 16 trials
Begin Experiment
ethnicity = ['black', 'white', 'asian', 'hispanic']
shuffle(ethnicity)
hair_colour = ['natural', 'blue', 'pink']
hair_colour[0] = [0,1,0,2] # ethnicity0 0 = natural
hair_colour[1] = [1,0,2,0] # ethnicity1
hair_colour[2] = [0,2,0,1] # ethnicity2
hair_colour[3] = [2,0,1,0] # ethnicity3
I would also recommend one of the two natural hair colours to be the low quality resume
ethnicity = ['black', 'white', 'asian', 'hispanic']
shuffle(ethnicity)
hair_colour = ['natural_unqualified','natural', 'blue', 'pink']
hair_colour[ethnicity[0]] = [0,2,1,3] # 0 = validity check
hair_colour[ethnicity[1]] = [2,0,3,1]
hair_colour[ethnicity[2]] = [1,3,0,2]
hair_colour[ethnicity[3]] = [3,1,2,0]
If you have a spreadsheet with one column called ethnicityIndex containing numbers 0 to 3 and you have four repeats then you would use ethnicity[ethnicityIndex] for the ethnicity and hair_colour[ethnicityIndex][trials.thisRepN] for the hair colour / resumé.