Stats test help?

[kjglen87]

Hi all,


I was wondering if anyone could help me? I am not very good at stats and for my PGCE I have to do a stats test with the data I collected but don't even know where to start


Basically I tested my pupils (I have emitted their names) and got their results (%) tested them after a week (after I implemented memory tasks) and got their percentage scores and then again after 2 weeks.


I now have to do a stats test with this data so I can analyse? What test can I even do? I believe I need to show if it's significant or just show I can analyse?


Any help is greatly appreciated.


Kind regards

 

[Aladin Akyurek]

Are you expecting that they get better across the weeks? That is:

W3 > W2 > W1

If so, you might to consider L test (Page, 1963; Siegel & Castellan (1980).

 

[Stiuart_W]

Or a T-Test for equal means. H0 is Mena 1 = Mean 2 so if you have a p-value less than, say, 0,05 the means are NOT the same

 

[Aladin Akyurek]

Or a T-Test for equal means. H0 is Mena 1 = Mean 2 so if you have a p-value less than, say, 0,05 the means are NOT the same

That should be a paired-T test. Having data aon 3 different weeks would require testing separately for W2 > W1 and for W3 > W2.

 

[Stiuart_W]

That should be a paired-T test. Having data aon 3 different weeks would require testing separately for W2 > W1 and for W3 > W2.

I've never heard of an L-Test. But I would think the OP would do a paired t-test on W1 Vs W2, W1 Vs W3 and W2 Vs W3