Archived

This topic is now archived and is closed to further replies.

rhea_dreams

HL
How do we test the significance level in the Psychology IA

Hello guys, :P

I am new to this forum and I have decided to join because I found some of the discussions really helpful. Thanks to all the people who are a part of the forum.
I am in my second year of IB and I have taken Psychology HL. I am doing my IA on the serial position effect. My report is due in two days. For the statistical -analysis I have done two tests: non-parametric (Mann-Whitney U test) and parametric test (t-test). I wanted some help with the deciding the significance values or the p-values. i really do not understand what they mean. My results are not significant at p < 0.05 and at p < 0.01 for both the test. Does this mean I have to accept my null hypothessis?? If yes, at what level of significance or at what should be the p value?? :D

I will really appreciate some help this.

Share this post


Link to post
Share on other sites
When something is found not to be significant, that's what you say.

You can say p > 0.05 but there is no need to go into more detail than that. Results are either significant to certain degrees, or they are not significant at all. And yes, you have to essentially accept your null hypothesis.

Share this post


Link to post
Share on other sites
[quote name='Deus' post='14632' date='Apr 7 2008, 05:49 PM']When something is found not to be significant, that's what you say.

You can say p > 0.05 but there is no need to go into more detail than that. Results are either significant to certain degrees, or they are not significant at all. And yes, you have to essentially accept your null hypothesis.[/quote]

Is is alright if I just say that the results from both the test are not significant without mentioning any p value??

rhea

Share this post


Link to post
Share on other sites
Yeah, just say no significance was found using the tests.. It's really that simple if you get no significant result. :P

Share this post


Link to post
Share on other sites
[quote name='Deus' post='14683' date='Apr 8 2008, 07:53 PM']Yeah, just say no significance was found using the tests.. It's really that simple if you get no significant result. :P [/quote]

hey deus,

thanx a lot for ur help. i have finally submitted my report.

let me know if i can be of any help

rhea

Share this post


Link to post
Share on other sites

my p value is .0914, which i don't think is that significant in the long run. Should i still say it's significant? :dash::help:

Share this post


Link to post
Share on other sites

my p value is .0914, which i don't think is that significant in the long run. Should i still say it's significant? :dash::help:

According to APA conventions your results are not significant. Did you do a one or two-tailed test? If you did a two-tailed test but your hypothesis was actually a one way then your results are actually significant because you divide the p value by 2 (.0916/2=.048).

Share this post


Link to post
Share on other sites