Alejandro Cedeno Posted April 12, 2016 Report Share Posted April 12, 2016 For my Psychology IA, I conducted a study in which I told participants to read out loud a list of words, the words were color names printed in black (this is condition one) and wrote down the time they took to complete the task. Then, I gave them another list of words, but printed in colors different than what their names pronounced (for example: red, green, blue), this was condition two. There are no control groups or experimental groups since every participant went through both conditions. However, I'm having trouble in the inferential statistics criteria for my paper. I don't know what test to use. Please help. Reply Link to post Share on other sites More sharing options...
blindpet Posted April 12, 2016 Report Share Posted April 12, 2016 Your teacher should explain to you that this is a within subjects design, also called a repeated measures design. Your data are interval/ratio since they are times and have an absolute 0 (meaning 0 actually means no time taken). You can use a related t-test for your data to get the p-value. If you have a small sample size you can use the Wilcoxon t-test, even though it is a non-parametric test (for ordinal data like rating the attractiveness of a person out of 10), it can be more robust for small sample sizes with interval/ratio data. You can justify the use of either test, there are online calculators that will return the right values. Remember to report the mean and standard deviation when making your tables and graphs and the p value from the inferential test. Good luck! Reply Link to post Share on other sites More sharing options...
Alejandro Cedeno Posted April 12, 2016 Author Report Share Posted April 12, 2016 I thought the t-test was for independent measures only. Reply Link to post Share on other sites More sharing options...
blindpet Posted April 12, 2016 Report Share Posted April 12, 2016 There are multiple versions of the t-test, there is one for independent measures and one for repeated measures. Reply Link to post Share on other sites More sharing options...
Alejandro Cedeno Posted April 12, 2016 Author Report Share Posted April 12, 2016 (edited) Sorry to bother, but do you think you could do a brief explanation for it? Also my sample consists of 37 participants. Edited April 12, 2016 by Alejandro Cedeno Reply Link to post Share on other sites More sharing options...
blindpet Posted April 12, 2016 Report Share Posted April 12, 2016 There is not much to explain, the repeated measures t-test examines the variance between the means between the first trial and the second trial and looks up the value it returns in a table to calculate the probability the results occurred from chance. If the p value is small enough you conclude the results did not occur from chance and instead from the treatment manipulation you used in your experiment. 1 Reply Link to post Share on other sites More sharing options...
blindpet Posted April 12, 2016 Report Share Posted April 12, 2016 37 participants sounds large enough to use a t-test instead of the wilcoxon non-parametric version. Reply Link to post Share on other sites More sharing options...
blindpet Posted April 12, 2016 Report Share Posted April 12, 2016 You're welcome, good luck Reply Link to post Share on other sites More sharing options...
Alejandro Cedeno Posted April 12, 2016 Author Report Share Posted April 12, 2016 I'm sorry SO much to keep insisting. But is it two-tailed, or one-tailed. Why? I seriously don't understand I've looked for the answer but I don't get it. Reply Link to post Share on other sites More sharing options...
blindpet Posted April 12, 2016 Report Share Posted April 12, 2016 Tails refer to the direction of the hypothesis. When you have clear theoretical reasons to predict one condition will be greater than another then you use a one-tailed test. When you have no theoretical grounds and just assume there will be a difference then you use a two-tailed test. http://www.ats.ucla.edu/stat/mult_pkg/faq/general/tail_tests.htm 1 Reply Link to post Share on other sites More sharing options...
Alejandro Cedeno Posted April 12, 2016 Author Report Share Posted April 12, 2016 Thank you it worked but now the calculator says that the means are different. And that t is higher than the critical value. Does this mean that the null hypothesis is rejected? Reply Link to post Share on other sites More sharing options...
blindpet Posted April 13, 2016 Report Share Posted April 13, 2016 It has been a while since I did it manually but that sounds right, if t is higher than critical value then the null hypothesis is rejected (for some statistical tests the critical value needs to be less than). Reply Link to post Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.