Help -- Method For Measuring Prejudice Against Stereotypes

[x-Alex-x]

Hello --- Any Response to Any Questions Asked In This Post Will Be Greatly Appreciated

I am looking at people's reaction to certain stereotypes portrayed through the fashion worn.
I plan to go out to a public area dressed as several stereotypes, either a * Young Lady/ Girly Girl*, * Tomboy/ Casual* and * Anti-social/ Emo* (which will remain constant) and approach people, who i will then ask for either the time or some change to make a call.

I Need ideas of ways of effectivly measuring the reactions and a possible scale to use, because people may have several unrelated reasons for why they can't help... And I can't film them without consent, for later reference because it is against the law :[

So far, I Have written sheets which I ( the 1st hand observer) and an assistant, who will sit 3-10m away can fill in, responding to observations made on: body language, time of interaction and dictance away from me. Any Other Suggestions For possible meduims to measure?

Also, I was contemplating recording them, and then afterwards having my assistant and i aproach them and explain what just occurred, then asking for consent to use the recording and also to possibly compile more data -- Any suggestions?

Thanks a Ton!

x-Alex-x

P.s Ignore My Last Post, Its A Bad Version of The Above.

[Lise Byrd]

Alex,

The Experts in this forum check posts frequently, but not constantly, so you may have to wait a little while for responses.

Your question is an interesting one. I would suggest having your tomboy/ young lady ask for the time rather than change; I think more people would be likely to comply with the time than with money.

You will certainly need to obtain consent from your subjects to include their data in your research. Approaching the subject after your assistant asks for the time and explaining what just occurred should not be a problem. You just won't be able to use any data from subjects who do not agree. Also, since you will need a few minutes to explain your project, I would suggest choosing people who do not look as if they are in a hurry when you are choosing whom to film.

Keep as many variables as possible constant. For example, use the same person as the young lady/ tomboy and have her say the same words when she asks for the time. You can also narrow variables by choosing to ask people of the same age or gender. I would suggest keeping the distance between yourself and the person you are filming constant as well. Try to place yourself somewhere so you are not immediately obvious to your subject.

The data you currently have on your sheets are a good start. Other data you can consider taking are
-tone of voice (gentle/ sharp),
-distance from the tomboy/ young lady,
-whether or not the person makes eye contact with the tomboy/ young lady,
-the wording used by the person giving the time.

If you are not within earshot when you are recording the interactions, you can ask your tomboy/ young lady to give you the tone of voice/ words as objectively as possible.

Since you are using two people to record the interaction (which is a great idea), you might consider not exchanging notes until all the studies are completed. This way you and your assistant will not color each other's opinions of the interactions. You can also consider having another person view the recordings of the interactions. If you black out the young lady/ tomboy (so the third person does not know who is asking), you can get another, more impartial view of the data.

I hope this helps you. Post back if anything is unclear or if you have more questions.
Sonia

[x-Alex-x]

Wow -- Thank You So Much For That Response. It Has Heped Greatly.
Do you have suggestions to alternatives to filming, because I am concerned that some people may not be pleased with me filming them, even if ido ask for consent.

I was wondering if you knew of an effective method of analyzing data to do with people, which have LOTS of variables, so i can effectivly portray my results?

Thanks So Much!

x-Alex-x

P.s Ayone ese with any suggestions, hey wouldb greatly appeciated

[Lise Byrd]

Alex,

Being able to film people would allow you to go back and re-check your data, especially for things such as the amount of time the interaction takes. It would also allow you to do a double-blind study (having a third person view the video without knowing if the interaction is with the tomboy or the young lady), if that interests you. I think most people would not mind too much about the filming, especially if you explain that you are doing this for a science fair project. However, if the filming does concern you, you could also just fill out survey sheets and not use the ones for people who don't give you permission.

For analyzing your data, I would suggest bar graphs or pie graphs, especially since much of your data will be qualitative rather than quantitative. Pie graphs are good for showing percentages (for instance, how many people did/ didn't give the young lady/ tomboy the time); bar graphs are good for comparing multiple variables (did certain kinds of people take more time with the young lady/ tomboy, respond gently or sharply, make eye contact?). I would try playing around with your data in different combinations; that's probably how you'll make the most interesting conclusions. A table might also be an effective way to present your data.

Hope this helps!
Sonia

[x-Alex-x]

Does Anyone Know If I Ca Get In Contact With Other People From This Website Who Did a similar experiment To Me? I would rweaoly like to discuss the results with them.

Also What Do You Thin IS A Suitable Sample Size?

x-Alex-x

[davidkallman]

Hi alex,

What one is trying to do with sampling is trying to calculate a value for an entire population by sampling a population, and running the test on the only the sample subset of the population. By selecting a sample, rather than the entire population, one introduces an error between the value calculated on the sample, vs. the true value, which is running the test on the entire population.

The wikipedia article is at: http://en.wikipedia.org/wiki/Margin_of_error

Other articles can be found by submitting: "sample size error" to google.com.

[campbtj]

I agree with Lise. In addition to her suggestion, I would add a stop-watch to your third person evaluation. This would you to allow you to measure something called "latency time" which is the time that it takes for the judgment to be evoked.